From dave.bucklin at gmail.com  Mon Oct  1 00:55:27 2018
From: dave.bucklin at gmail.com (Dave Bucklin)
Date: Sun, 30 Sep 2018 19:55:27 -0500
Subject: Blog [was Re: Progress migrating to list]
In-Reply-To: <20180930231616.GC2081@nobelware.com>
References: <812B33A2-F194-4C3A-8024-3B02057D9D55@begriffs.com>
 <20180929152647.GA24123@nobelware.com>
 <20180929200634.efeg6fd7sihd2rri@19a6.tech>
 <20180930141608.GA23003@nobelware.com>
 <20180930161222.topqqqn2qk62ycpp@19a6.tech>
 <20180930174841.GC61103@mysterium.local>
 <20180930205611.5vflholrrqvqxoyq@19a6.tech>
 <20180930231616.GC2081@nobelware.com>
Message-ID: <22CEDD8A-4641-4A4E-8FAA-AC73A8CF5A58@gmail.com>

On September 30, 2018 6:16:16 PM CDT, Ioannis Nompelis <nompelis at nobelware.com> wrote:
>That was a good read Joe.
>
>Dave and Joe, can you post your RSS feeds and aggregators so that I can
>have
>a look -- but also to have those in the email archive, so I can get
>back to
>this anytime later...

Here's my feed: https://davebucklin.com/feed.xml

>Here is that aggregator that I have adopted from a friend and admin at
>my
>department at the U:
>
>http://hokan.org

I love that. Bookmarked!


From dave.bucklin at gmail.com  Mon Oct  1 01:15:38 2018
From: dave.bucklin at gmail.com (Dave Bucklin)
Date: Sun, 30 Sep 2018 20:15:38 -0500
Subject: Progress migrating to list
In-Reply-To: <20180930232409.GE2081@nobelware.com>
References: <812B33A2-F194-4C3A-8024-3B02057D9D55@begriffs.com>
 <20180929152647.GA24123@nobelware.com>
 <20180930172345.GA61103@mysterium.local>
 <20180930232409.GE2081@nobelware.com>
Message-ID: <46BF9ADC-E4B0-448C-B064-9B6C581B68E2@gmail.com>

On September 30, 2018 6:24:09 PM CDT, Ioannis Nompelis <nompelis at nobelware.com> wrote:
>I think we just invite people we know, and rely on our gut-feeling from
>getting
>to know people. Let's just try to be gentlemen about it, all sense of
>the
>word.

I support this.


From nompelis at nobelware.com  Mon Oct  1 03:25:11 2018
From: nompelis at nobelware.com (Ioannis Nompelis)
Date: Sun, 30 Sep 2018 22:25:11 -0500
Subject: Blog [was Re: Progress migrating to list]
In-Reply-To: <22CEDD8A-4641-4A4E-8FAA-AC73A8CF5A58@gmail.com>
References: <812B33A2-F194-4C3A-8024-3B02057D9D55@begriffs.com>
 <20180929152647.GA24123@nobelware.com>
 <20180929200634.efeg6fd7sihd2rri@19a6.tech>
 <20180930141608.GA23003@nobelware.com>
 <20180930161222.topqqqn2qk62ycpp@19a6.tech>
 <20180930174841.GC61103@mysterium.local>
 <20180930205611.5vflholrrqvqxoyq@19a6.tech>
 <20180930231616.GC2081@nobelware.com>
 <22CEDD8A-4641-4A4E-8FAA-AC73A8CF5A58@gmail.com>
Message-ID: <20181001032511.GA4850@nobelware.com>

Dave, I read your first two posts on that blog feed. Cool. I liked the
verbosity of the Postfix configuration; great!

Regarding tools for programming, I am a command-line addict myself, and
hard=headed, and I do use GITK, which is very good. But if you are going
to run any Git thing on Windows or WINE, allow me to introduce you to
SourceTree. It is made by that company that also makes Jira and others.
SourceTree is actually very good for browsing branches and tags, while all
else I recommend be done with CLI Git and some GITK.


From joe at begriffs.com  Tue Oct  2 17:44:31 2018
From: joe at begriffs.com (Joe Nelson)
Date: Tue, 2 Oct 2018 12:44:31 -0500
Subject: Hardcore hacknight discussion group
In-Reply-To: <370b451f-46e5-e743-1b66-89bc05ca1da1@gmx.com>
References: <0101016612ba9e79-41c8df54-203d-428e-9aaf-b5c668579b06-000000@us-west-2.amazonses.com>
 <370b451f-46e5-e743-1b66-89bc05ca1da1@gmx.com>
Message-ID: <20181002174431.GA27551@mysterium.local>

(CC'ing the list in case someone there knows about such things)

pstelzig wrote:
> I'm getting an error when I try to subscribe

Thank you for the error report. Email is tricky to set up because what
can work for some servers doesn't work for others, so it gives the false
sense that everything is working when things may still be broken.

> An error occurred while sending mail. The mail server responded:
> Requested action not taken: mailbox unavailable
> invalid DNS MX or A/AAAA resource record.
>  Please check the message recipient "friends-request at talk.begriffs.com" and
> try again
> 
> Is there a chance your server is setup to deny incoming email form some
> domains?

Hmm, https://postmaster.gmx.com/en/error-messages says:

---------------------------------------------------------------------
5xy Invalid DNS MX or A/AAAA resource record
The sender domain of the email server you are using has not been
configured in agreement with the RFC guidelines in the domain name
service (DNS). The following reasons may be responsible for this:

Neither an A- nor an MX record has been configured in the DNS for the
sender domain.
The MX record points to a domain name that cannot be resolved.
The MX record directly points to an IP address.
The MX server does not have a valid A-record, but only a CNAME record.
---------------------------------------------------------------------

Your domain seems fine.

$ dig gmx.com MX +short
10 mx01.gmx.net.
10 mx00.gmx.net.

Also your mail servers have A records
mx00.gmx.net.		900	IN	A	212.227.15.10
mx01.gmx.net.		896	IN	A	212.227.17.4

HOWEVER, my own MX record points to a bare IP
talk.begriffs.com.	300	IN	MX	1 45.76.26.49.

I'll update that record after sending this message and you can try again
to see if it helps.

From nompelis at nobelware.com  Tue Oct  2 20:09:25 2018
From: nompelis at nobelware.com (Ioannis Nompelis)
Date: Tue, 2 Oct 2018 20:09:25 +0000
Subject: Hardcore hacknight discussion group
In-Reply-To: <20181002174431.GA27551@mysterium.local>
References: <0101016612ba9e79-41c8df54-203d-428e-9aaf-b5c668579b06-000000@us-west-2.amazonses.com>
 <370b451f-46e5-e743-1b66-89bc05ca1da1@gmx.com>
 <20181002174431.GA27551@mysterium.local>
Message-ID: <20181002200925.GA9130@nobelware.com>

Yup! Typical error I have made before. DNS lookups pertaining to email are
very picky when it comes to getting a legitimate host (i.e. on that has a
hostname and reverse-lookup-able IP) for the MX record.

I was connected from work to my home desktop and ran two examples, one showing
that your server, Joe, is show up correctly:

nompelis at bigpapa:~/work/DSMC/devel> host nobelware.com
nobelware.com has address 173.203.202.144
nobelware.com mail is handled by 0 iznogoud.nobelware.com.
nompelis at bigpapa:~/work/DSMC/devel> host talk.begriffs.com
talk.begriffs.com has address 45.76.26.49
talk.begriffs.com mail is handled by 1 talk.begriffs.com.
nompelis at bigpapa:~/work/DSMC/devel> 

I have my VPS hosted someplace that has nameservers. I have directed the NS
records of my domain(s) at my registrar's system as the VPS company's NS. Then,
I manage all the DNS from there, although I could have bootstraped the DNS
from the VPS itself (bad idea). The MX record is pointing to a hostname.

The bad idea part has to do with me losing the VPS to hackers... and then,
for a short time, they control all the email that comes here. That would be
a BIG FAIL from a so-called veteran like myself...


From pstelzig at gmx.com  Thu Oct  4 00:23:55 2018
From: pstelzig at gmx.com (pstelzig)
Date: Wed, 3 Oct 2018 19:23:55 -0500
Subject: Working now
Message-ID: <3b416f92-0241-de42-1310-e2d8dd50012b@gmx.com>

What you did fixed the sign up for me.



From nompelis at nobelware.com  Thu Oct  4 01:56:08 2018
From: nompelis at nobelware.com (Ioannis Nompelis)
Date: Wed, 3 Oct 2018 20:56:08 -0500
Subject: Working now
In-Reply-To: <3b416f92-0241-de42-1310-e2d8dd50012b@gmx.com>
References: <3b416f92-0241-de42-1310-e2d8dd50012b@gmx.com>
Message-ID: <20181004015608.GA2866@nobelware.com>

Welcome!

From nompelis at nobelware.com  Fri Oct  5 14:03:36 2018
From: nompelis at nobelware.com (Ioannis Nompelis)
Date: Fri, 5 Oct 2018 14:03:36 +0000
Subject: list and weblogs
Message-ID: <20181005140336.GA21213@nobelware.com>

Regarding the list, how may of us are here? I know of Dave, Joe, Paul and I.
Also, not that we have the time to get together too often, but maybe we
should have a meeting, and it does not have to be in person. Maybe we want
to try the mourmor and some screen-sharing. It will function as a meeting and
as a sort of test of what we have in place on hte VPS.

There is a $50 in my wallet that wants to go into Dave's pocket too, but I
will turn that into an excuse for an in-person meeting.

Also, Dave and Joe regularly maintain weblogs, which I follow. What software,
if any, do you gys use to write your weblog? I have a "weblog" to add random
content on a website of mine, but it is ad-hoc PHP written by the idiot typing
this message.

IN


From joe at begriffs.com  Fri Oct  5 17:32:01 2018
From: joe at begriffs.com (Joe Nelson)
Date: Fri, 5 Oct 2018 12:32:01 -0500
Subject: list and weblogs
In-Reply-To: <20181005140336.GA21213@nobelware.com>
References: <20181005140336.GA21213@nobelware.com>
Message-ID: <20181005173201.GA91555@mysterium.local>

Ioannis Nompelis wrote:
> Regarding the list, how may of us are here? I know of Dave, Joe, Paul and I.

There are ten subscribers now. If any of you want to jump in and
introduce yourselves, feel free. :)

> Maybe we want to try the mourmor and some screen-sharing. It will
> function as a meeting and as a sort of test of what we have in place
> on hte VPS.

Sounds great. Anyone want to pair with me on frostbyte.cc to finish the
exercises in K&R? I'm currently at exercise 6-2.

> Also, Dave and Joe regularly maintain weblogs, which I follow. What software,
> if any, do you gys use to write your weblog? I have a "weblog" to add random
> content on a website of mine, but it is ad-hoc PHP written by the idiot typing
> this message.

Mine is a static site generated by https://jaspervdj.be/hakyll/
(with several customizations). That plus http rules in nginx provide
everything I need. I would *not* recommend this setup to other people
though, there are simpler ways to make a static site. Check out SSG:
https://www.romanzolotarev.com/ssg.html

From salo at saloits.com  Fri Oct  5 17:44:34 2018
From: salo at saloits.com (Timothy J. Salo)
Date: Fri, 5 Oct 2018 12:44:34 -0500
Subject: list and weblogs
In-Reply-To: <20181005173201.GA91555@mysterium.local>
References: <20181005140336.GA21213@nobelware.com>
 <20181005173201.GA91555@mysterium.local>
Message-ID: <84753e4e-1069-e09e-8379-e0b806206f22@saloits.com>

Hi,

I suggest that mail from this list be identified by something like
[list-name] in the subject line (specifically at the start of the
subject line).

I get a lot of email.  I won't necessarily recognize email to this list
unless it is somehow visually marked.

-tjs

From samuel.stuewe at gmail.com  Fri Oct  5 18:50:36 2018
From: samuel.stuewe at gmail.com (Sam Stuewe)
Date: Fri, 5 Oct 2018 13:50:36 -0500
Subject: list and weblogs
In-Reply-To: <20181005173201.GA91555@mysterium.local>
References: <20181005140336.GA21213@nobelware.com>
 <20181005173201.GA91555@mysterium.local>
Message-ID: <CABSwF4xRnqmTMNbXfbnwFZQOvWhQnEuAUXvJXji+ETqbYS28ug@mail.gmail.com>

On 10/5/18, Joe Nelson <joe at begriffs.com> wrote:
> There are ten subscribers now. If any of you want to jump in and
> introduce yourselves, feel free. :)
Hello all! My name is Sam (aka halosghost around the web).

>> Also, Dave and Joe regularly maintain weblogs, which I follow. What
>> software,
>> if any, do you gys use to write your weblog? I have a "weblog" to add
>> random
>> content on a website of mine, but it is ad-hoc PHP written by the idiot
>> typing
>> this message.
I host an in-development website at <https://halosgho.st>. It is
written in C and technically not static, so it may not be of-interest,
but I'll happily discuss its architecture if anyone would like. I have
not decided whether or not to publish a blog underneath it.

I am definitely down for murmur meetups, or in-person get-togethers
whenever, though I tend to be on the busier side of life these days,
so a week of advance notice would be much appreciated!

Good to meet you all!

All the best,

-Sam

From joe at begriffs.com  Fri Oct  5 20:32:11 2018
From: joe at begriffs.com (Joe Nelson)
Date: Fri, 5 Oct 2018 15:32:11 -0500
Subject: list and weblogs
In-Reply-To: <84753e4e-1069-e09e-8379-e0b806206f22@saloits.com>
References: <20181005140336.GA21213@nobelware.com>
 <20181005173201.GA91555@mysterium.local>
 <84753e4e-1069-e09e-8379-e0b806206f22@saloits.com>
Message-ID: <20181005203211.GA1190@mysterium.local>

Timothy J. Salo wrote:
> I suggest that mail from this list be identified by something like
> [list-name] in the subject line (specifically at the start of the
> subject line).

I can't modify the subject line when relaying from the list or it will
break some DKIM signatures and hence send mail to spam for certain
domains.

However, the list does identify its messages with a "List-Id" header and
a "Sender" header. You can use either of these headers as the basis for
a rule that will move messages to a custom folder; see section three in
my howto: https://talk.begriffs.com/howto/#configure

> I get a lot of email.  I won't necessarily recognize email to this list
> unless it is somehow visually marked.

Yeah, I know what you mean. I subscribe to a number of high-volume lists
like postgresql-hackers and openbsd-ports. I have them each go to
separate imap folders so they don't mix in my mail client.

From salo at saloits.com  Fri Oct  5 22:26:28 2018
From: salo at saloits.com (Timothy J. Salo)
Date: Fri, 5 Oct 2018 17:26:28 -0500
Subject: list and weblogs
In-Reply-To: <20181005203211.GA1190@mysterium.local>
References: <20181005140336.GA21213@nobelware.com>
 <20181005173201.GA91555@mysterium.local>
 <84753e4e-1069-e09e-8379-e0b806206f22@saloits.com>
 <20181005203211.GA1190@mysterium.local>
Message-ID: <878a4a19-f769-6632-58a3-5273c0cb8641@saloits.com>

On 10/5/2018 3:32 PM, Joe Nelson wrote:
> Timothy J. Salo wrote:
> [...]
> However, the list does identify its messages with a "List-Id" header and
> a "Sender" header. You can use either of these headers as the basis for
> a rule that will move messages to a custom folder; see section three in
> my howto: https://talk.begriffs.com/howto/#configure
> 
>> I get a lot of email.  I won't necessarily recognize email to this list
>> unless it is somehow visually marked.

Yes, I _could_ filter on "List-Id".  But, for me, that treatment is
reserved for high-volume lists that I don't want to read in real-time.
I'm not quite sure what this list is or will be, but it probably won't
warrant this special processing (for me).

> Yeah, I know what you mean. I subscribe to a number of high-volume lists
> like postgresql-hackers and openbsd-ports. I have them each go to
> separate imap folders so they don't mix in my mail client.

Most high-volume lists I subscribe to mark the subject line with
"[...]".  It seems to be prevailing practice, as well as very handy.

-tjs


From joe at begriffs.com  Fri Oct  5 23:54:30 2018
From: joe at begriffs.com (Joe Nelson)
Date: Fri, 5 Oct 2018 18:54:30 -0500
Subject: list and weblogs
In-Reply-To: <878a4a19-f769-6632-58a3-5273c0cb8641@saloits.com>
References: <20181005140336.GA21213@nobelware.com>
 <20181005173201.GA91555@mysterium.local>
 <84753e4e-1069-e09e-8379-e0b806206f22@saloits.com>
 <20181005203211.GA1190@mysterium.local>
 <878a4a19-f769-6632-58a3-5273c0cb8641@saloits.com>
Message-ID: <20181005235430.GB1190@mysterium.local>

Timothy J. Salo wrote:
> Yes, I _could_ filter on "List-Id".  But, for me, that treatment is
> reserved for high-volume lists that I don't want to read in real-time.

What's the downside of filtering on List-Id? Your user agent header says
you're using Thunderbird -- aren't folders listed along the side in that
client along with counts of unread messages in each? Participation is
just a click away.

> Most high-volume lists I subscribe to mark the subject line with
> "[...]".  It seems to be prevailing practice, as well as very handy.

It's a practice from before the major mail providers honored DMARC. The
only way I can have the list modify the subject line is if I also spoof
the From to make the message appear to come from the list rather than
the true author. This creates a degraded experience for mail clients'
address book (since the authors addresses are not mentioned properly).

Here's an article with my research on the
topic, let me know if I'm overlooking anything.
https://begriffs.com/posts/2018-09-18-dmarc-mailing-list.html

Also the Postgres, OpenBSD, and GnuPG lists I'm on don't modify the
subject header. I see that happening only in Google Groups lists,
because Google doesn't respect the standards. See discussions online for
instance about the way Gmail fudges IMAP.

From pstelzig at gmx.com  Sat Oct  6 01:30:15 2018
From: pstelzig at gmx.com (pstelzig)
Date: Fri, 5 Oct 2018 20:30:15 -0500
Subject: list and weblogs
In-Reply-To: <20181005235430.GB1190@mysterium.local>
References: <20181005140336.GA21213@nobelware.com>
 <20181005173201.GA91555@mysterium.local>
 <84753e4e-1069-e09e-8379-e0b806206f22@saloits.com>
 <20181005203211.GA1190@mysterium.local>
 <878a4a19-f769-6632-58a3-5273c0cb8641@saloits.com>
 <20181005235430.GB1190@mysterium.local>
Message-ID: <f80b6241-ff1c-5093-8fdd-5254988c6570@gmx.com>



On 10/05/2018 06:54 PM, Joe Nelson wrote:

> Also the Postgres, OpenBSD, and GnuPG lists I'm on don't modify the
> subject header. I see that happening only in Google Groups lists,
> because Google doesn't respect the standards. See discussions online for
> instance about the way Gmail fudges IMAP.
> 

The linux kernel lists I'm on do alter the header, I'm used to filtering 
be that, but filtering be list ID will also work for me.

From salo at saloits.com  Sat Oct  6 05:57:44 2018
From: salo at saloits.com (Timothy J. Salo)
Date: Sat, 6 Oct 2018 00:57:44 -0500
Subject: list and weblogs
In-Reply-To: <20181005235430.GB1190@mysterium.local>
References: <20181005140336.GA21213@nobelware.com>
 <20181005173201.GA91555@mysterium.local>
 <84753e4e-1069-e09e-8379-e0b806206f22@saloits.com>
 <20181005203211.GA1190@mysterium.local>
 <878a4a19-f769-6632-58a3-5273c0cb8641@saloits.com>
 <20181005235430.GB1190@mysterium.local>
Message-ID: <0ad7a0f0-f8b3-9309-8945-ce13fb609255@saloits.com>

On 10/5/2018 6:54 PM, Joe Nelson wrote:
> [...]
> Also the Postgres, OpenBSD, and GnuPG lists I'm on don't modify the
> subject header. I see that happening only in Google Groups lists,
> because Google doesn't respect the standards. See discussions online for
> instance about the way Gmail fudges IMAP.

_All_ of the IETF (Internet Engineering Task Force) mail lists that I
subscribe to (which is a fair number) modify the Subject: header.  It's
not just a Google thing.

(For what it is worth, the IETF is an international standards organization.)

It's not my list, so you can certainly do what you want.  But, I think
that you really are out of step with prevailing practice.

(Google group lists provide inserting [...] as a configuration option.)

-tjs

From dave.bucklin at gmail.com  Sat Oct  6 14:03:06 2018
From: dave.bucklin at gmail.com (Dave Bucklin)
Date: Sat, 6 Oct 2018 09:03:06 -0500
Subject: list and weblogs
In-Reply-To: <20181005173201.GA91555@mysterium.local>
References: <20181005140336.GA21213@nobelware.com>
 <20181005173201.GA91555@mysterium.local>
Message-ID: <20181006140306.n2zfg55wryoz65fr@19a6.tech>

On Fri, Oct 05, 2018 at 12:32:01PM -0500, Joe Nelson wrote:
> There are ten subscribers now. If any of you want to jump in and
> introduce yourselves, feel free. :)

Greetings, I'm Dave. I blog over at davebucklin.com. I also maintain a
gopher hole at SDF (gopher.club). Most of the content is duplicated.
>From 1998 until 2011 I was a developer on a very old system called Pick.
It was a contemporary of Unix that felt more like a TOPS-20 system. I'm
still fascinated by tech of that vintage. I'm also interested in
mechanical keyboards, and there's actually a meetup today from 1:30 to
4:30 at the Shoreview Library, 4560 Victoria St. North, Shoreview. This
happens maybe once a year, so if you are at all interested, I encourage
you to swing by. Tell 'em Large Marge sent ya!

Hack the planet.

From nompelis at nobelware.com  Sat Oct  6 16:43:41 2018
From: nompelis at nobelware.com (Ioannis Nompelis)
Date: Sat, 6 Oct 2018 16:43:41 +0000
Subject: list and weblogs
In-Reply-To: <20181005235430.GB1190@mysterium.local>
References: <20181005140336.GA21213@nobelware.com>
 <20181005173201.GA91555@mysterium.local>
 <84753e4e-1069-e09e-8379-e0b806206f22@saloits.com>
 <20181005203211.GA1190@mysterium.local>
 <878a4a19-f769-6632-58a3-5273c0cb8641@saloits.com>
 <20181005235430.GB1190@mysterium.local>
Message-ID: <20181006164341.GD14821@nobelware.com>

The local TC-LUG (Twin Cities Linux Users Group) modifies the subject line.

I have no preference, but modifying the subject sounds good. I also like a
reply to go to the list instead of the sender, but I do not mind doing it
with a group reply.

------

I will introduce myself for verbosely. I am Yanni, and you will see me signing
as "IN" in most places. I am a research scientist and sometimes teaching staff
at the UofM (Minneapolis) in an engineering department. In real life I look
and act like this: http://umn.edu/~nompelis

I got my first computer in 1984, and I wanted one since 1983 (I remember like
it was yesterday). It was an Amstrad CPC 8128 (A UK company from the 80s). I
jumped to the Amiga in 1990 (took a while to gather the funds), and have been
in love with it ever since but rarely touch it now. On Linux since kernel 1.1
from 1993 and a fanatic, but I can deal with most Unix-like environments like
I am at home. I like learning (mostly by reading and experimentation) things
on many topics; I am a learning junkie. I encrypt. I like to create stuff,
mostly software and art. I speak a couple of languages other than English. I
love chocolate, music, sports and cars.

I am a hacker... in the true, old sense of the word, and I like to find unusual
uses for things. I can be easily pursuaded to work on something cool, but make
sure you keep my interest high so that nobody else pursuades me to work on
something else...
IN


From joe at begriffs.com  Sat Oct  6 21:16:24 2018
From: joe at begriffs.com (Joe Nelson)
Date: Sat, 6 Oct 2018 16:16:24 -0500
Subject: list and weblogs
In-Reply-To: <0ad7a0f0-f8b3-9309-8945-ce13fb609255@saloits.com>
References: <20181005140336.GA21213@nobelware.com>
 <20181005173201.GA91555@mysterium.local>
 <84753e4e-1069-e09e-8379-e0b806206f22@saloits.com>
 <20181005203211.GA1190@mysterium.local>
 <878a4a19-f769-6632-58a3-5273c0cb8641@saloits.com>
 <20181005235430.GB1190@mysterium.local>
 <0ad7a0f0-f8b3-9309-8945-ce13fb609255@saloits.com>
Message-ID: <20181006211624.GA55716@mysterium.local>

I hope I'm not becoming tedious to everyone by pursuing this topic, but
figuring out the answer seems to be in the spirit of our group.  It's an
opportunity to learn about email, a fundamental internet technology.

Timothy J. Salo wrote:
> _All_ of the IETF (Internet Engineering Task Force) mail lists that I
> subscribe to (which is a fair number) modify the Subject: header.  It's
> not just a Google thing.

First, I feel like it was intellectually lazy of me to exhibit the
behavior of various lists. We ought to judge the options by their merits
rather than by whichever organizations rightly or wrongly employ them.

Second, senders whose domains have a p=reject DMARC policy have indeed
complained of the IETF list behavior:
https://www.ietf.org/mail-archive/web/dmarc/current/msg03229.html

> It's not my list, so you can certainly do what you want.  But, I think
> that you really are out of step with prevailing practice.

I'm happy to change the list behavior if we determine that it's not
working correctly. I think we can come to a conclusion we all agree
with and are happy about if we consider the arguments.

So far here's my paraphrase of what you're saying:
* List emails without a subject prefix get lost in general mail
  on your mail client.
* Filtering by List-Id is undesirable (can you clarify this?)

To attempt to clarify what I was saying
* The List-Id and subject line prefix are equally filterable
  with mail client rules. Prefixing the subject line with a
  list identifier seems to be a hack since a list identifier
  is not the "subject" of discussion. RFC2919 proposed List-Id
  as a standard location for list identification, to "make it easier for
  mail clients to provide automated tools for users to perform list
  functions. https://tools.ietf.org/html/rfc2919
* Some mail clients have built-in support for displaying the List-Id
  header, and can distinguish mailing list messages from personal mail
  without needing to create separate folders. In Mutt (with the right
  configuration) you can see the List-Id as visibly as a subject prefix:
  http://www.mutt.org/doc/manual/#using-lists
* If we do want to go the route of adding a subject prefix, we
  will be required to change the From address to be that of the
  list rather than the original author. (Prefixing a subject
  while retaining the original From often runs afoul of DMARC,
  which makes messages go to Spam or returns them to sender.)
* For lists that transmit patches for git, a rewritten From address
  breaks the "git am" command, because the From is used in the commit.
  That's not a problem for this list per se, but indicates the way that
  conflating From and Sender in SMTP breaks tools that are written in
  conformance with RFC5322.
* If we change the From address we will want to add a Reply-To with the
  original author's address. Otherwise there's no way for list members to
  reply off-list unless they know each other's address. Reply-to addresses
  are usually not displayed in mail clients' message listing, nor are they
  added to a client address book. (Although mutt can display it with %A in
  the index_format config parameter.)

From nompelis at nobelware.com  Sun Oct  7 16:33:28 2018
From: nompelis at nobelware.com (Ioannis Nompelis)
Date: Sun, 7 Oct 2018 16:33:28 +0000
Subject: list and weblogs
In-Reply-To: <20181006211624.GA55716@mysterium.local>
References: <20181005140336.GA21213@nobelware.com>
 <20181005173201.GA91555@mysterium.local>
 <84753e4e-1069-e09e-8379-e0b806206f22@saloits.com>
 <20181005203211.GA1190@mysterium.local>
 <878a4a19-f769-6632-58a3-5273c0cb8641@saloits.com>
 <20181005235430.GB1190@mysterium.local>
 <0ad7a0f0-f8b3-9309-8945-ce13fb609255@saloits.com>
 <20181006211624.GA55716@mysterium.local>
Message-ID: <20181007163328.GB20747@nobelware.com>

I will correct you here Joe; in Mutt, without configuration you see all the
headers and start to feel bad for the NSA having to store all those 2-3kb of
headers for messages with body: "Ok. See you there tonight."

I propose we try it the way it is now, at least for a while. We have TJS set
up his thunderbird to filter things from this list into a separate folder and
tell us how to do it. Then, we decide if we want to try a different format
on purpose, just to spice up life.

I adapt to circumstances, because I was told that having this ability is a
sign of mental capacity (little do they know...)


From nompelis at nobelware.com  Sun Oct  7 16:38:41 2018
From: nompelis at nobelware.com (Ioannis Nompelis)
Date: Sun, 7 Oct 2018 16:38:41 +0000
Subject: list and weblogs
In-Reply-To: <20181007163328.GB20747@nobelware.com>
References: <20181005140336.GA21213@nobelware.com>
 <20181005173201.GA91555@mysterium.local>
 <84753e4e-1069-e09e-8379-e0b806206f22@saloits.com>
 <20181005203211.GA1190@mysterium.local>
 <878a4a19-f769-6632-58a3-5273c0cb8641@saloits.com>
 <20181005235430.GB1190@mysterium.local>
 <0ad7a0f0-f8b3-9309-8945-ce13fb609255@saloits.com>
 <20181006211624.GA55716@mysterium.local>
 <20181007163328.GB20747@nobelware.com>
Message-ID: <20181007163841.GC20747@nobelware.com>

Oh. Let me ask one more thing on the subject of email. I have my mailserver
configured such that it can use SMTP with authentication. This is great because
when the IP of the system where the mail client was used is NOT the same as the
mailserver doing SMTP, the mail can be rejected by other servers receiving it.
But when I am on my server in a terminal, the Mutt I am using is not doing
authenticated SMTP (it is the good old internal mailng system). And mail gets
delivered just fine.

So far things have been working great.

Is there any obvious problem with this practice? Should I switch the local
emling from the terminal on the systme to do authenticated SMTP?

From joe at begriffs.com  Sun Oct  7 17:07:33 2018
From: joe at begriffs.com (Joe Nelson)
Date: Sun, 7 Oct 2018 12:07:33 -0500
Subject: list and weblogs
In-Reply-To: <20181007163328.GB20747@nobelware.com>
References: <20181005140336.GA21213@nobelware.com>
 <20181005173201.GA91555@mysterium.local>
 <84753e4e-1069-e09e-8379-e0b806206f22@saloits.com>
 <20181005203211.GA1190@mysterium.local>
 <878a4a19-f769-6632-58a3-5273c0cb8641@saloits.com>
 <20181005235430.GB1190@mysterium.local>
 <0ad7a0f0-f8b3-9309-8945-ce13fb609255@saloits.com>
 <20181006211624.GA55716@mysterium.local>
 <20181007163328.GB20747@nobelware.com>
Message-ID: <20181007170733.GA24699@mysterium.local>

Ioannis Nompelis wrote:
> I will correct you here Joe; in Mutt, without configuration you see all the
> headers

Are you talking about the index view or the message view?  For me,
without customizing index_format, the index view shows lines like this:

  10     Oct 03 pstelzig        (0.1K) Working now

Just a message number, flags(none for this message), date, author,
size and subject. To show list information in the index, there's a %L
specifier for index_format that will replace the From display with "To
list at foo.com" if that address has been declared with the subscribe
command in the muttrc. Also NeoMutt has a %K specifier that is kind of
cool:
https://kdecherf.com/blog/2017/04/11/neomutt-gmail-and-the-mailing-lists/

> and start to feel bad for the NSA having to store all those 2-3kb of
> headers for messages with body: "Ok. See you there tonight."

Ha! :)

From joe at begriffs.com  Sun Oct  7 17:22:40 2018
From: joe at begriffs.com (Joe Nelson)
Date: Sun, 7 Oct 2018 12:22:40 -0500
Subject: list and weblogs
In-Reply-To: <20181007163841.GC20747@nobelware.com>
References: <20181005140336.GA21213@nobelware.com>
 <20181005173201.GA91555@mysterium.local>
 <84753e4e-1069-e09e-8379-e0b806206f22@saloits.com>
 <20181005203211.GA1190@mysterium.local>
 <878a4a19-f769-6632-58a3-5273c0cb8641@saloits.com>
 <20181005235430.GB1190@mysterium.local>
 <0ad7a0f0-f8b3-9309-8945-ce13fb609255@saloits.com>
 <20181006211624.GA55716@mysterium.local>
 <20181007163328.GB20747@nobelware.com>
 <20181007163841.GC20747@nobelware.com>
Message-ID: <20181007172240.GB24699@mysterium.local>

Ioannis Nompelis wrote:
> Should I switch the local emling from the terminal on the systme to do
> authenticated SMTP?

Personally I use msmtp [0] to send mail from the terminal. In the config
file you create "accounts" [1] where you can specify to use TLS for a
given server.

Then tell mutt to send via msmtp:

  set sendmail="/usr/local/bin/msmtp"


0: https://marlam.de/msmtp/
1: https://github.com/begriffs/obsd/blob/master/home/.msmtprc

From joe at begriffs.com  Sun Oct  7 17:53:55 2018
From: joe at begriffs.com (Joe Nelson)
Date: Sun, 7 Oct 2018 12:53:55 -0500
Subject: list and weblogs
In-Reply-To: <CABSwF4xRnqmTMNbXfbnwFZQOvWhQnEuAUXvJXji+ETqbYS28ug@mail.gmail.com>
References: <20181005140336.GA21213@nobelware.com>
 <20181005173201.GA91555@mysterium.local>
 <CABSwF4xRnqmTMNbXfbnwFZQOvWhQnEuAUXvJXji+ETqbYS28ug@mail.gmail.com>
Message-ID: <20181007175355.GA28139@mysterium.local>

Sam Stuewe wrote:
> I host an in-development website at <https://hal0osgho.st>. It is
> written in C and technically not static, so it may not be of-interest,
> but I'll happily discuss its architecture if anyone would like. I have
> not decided whether or not to publish a blog underneath it.

This sounds really interesting, I'd love to hear more!

The server is sending back some cool looking security headers:

$ curl -I https://halosgho.st/
HTTP/1.1 200 OK
Content-Length: 2493
Content-Type: text/html; charset=UTF-8
Connection: keep-alive
content-security-policy: default-src 'self'
x-frame-options: SAMEORIGIN
x-xss-protection: 1; mode=block
x-content-type-options: nosniff
referrer-policy: no-referrer
strict-transport-security: max-age=31536000; includeSubDomains
Date: Sun, 07 Oct 2018 17:37:19 GMT
Expires: Sun, 14 Oct 2018 17:37:19 GMT
Server: lwan

The readme for your site's repo has a link to an auditing tool
https://securityheaders.com for header security... Oh crap, when I run
my own site through it I get a D grade.

> I am definitely down for murmur meetups, or in-person get-togethers

Great, can you send me a public SSH key? I'll add it to our shared
server so you can log in for pair programming or updating our web site.
http://frostbyte.cc/wiki.txt

From samuel.stuewe at gmail.com  Sun Oct  7 19:08:14 2018
From: samuel.stuewe at gmail.com (Sam Stuewe)
Date: Sun, 7 Oct 2018 14:08:14 -0500
Subject: list and weblogs
In-Reply-To: <20181007175355.GA28139@mysterium.local>
References: <20181005140336.GA21213@nobelware.com>
 <20181005173201.GA91555@mysterium.local>
 <CABSwF4xRnqmTMNbXfbnwFZQOvWhQnEuAUXvJXji+ETqbYS28ug@mail.gmail.com>
 <20181007175355.GA28139@mysterium.local>
Message-ID: <CABSwF4x5jQBa-3CDgWJwvjv539jLSEk75t_nXCfbe=UeyzZwmg@mail.gmail.com>

On 10/7/18, Joe Nelson <joe at begriffs.com> wrote:
> Sam Stuewe wrote:
>> I host an in-development website at <https://halosgho.st>. It is
>> written in C and technically not static, so it may not be of-interest,
>> but I'll happily discuss its architecture if anyone would like. I have
>> not decided whether or not to publish a blog underneath it.
>
> This sounds really interesting, I'd love to hear more!
Sure! I use lwan (<https://lwan.ws>) as a dynamically-linked library,
so the site and the server are one and the same. Lwan offers mustache
templating (<https://mustache.github.io/mustache.5.html>). At the
moment, I only leverage the templating engine to update the copyright
year in the footer dynamically (super wasteful because it'll check the
time on every page-load, but the site loads between 0.5 and 3s
depending on where you are in the world, so I don't feel like that's a
real bottleneck for performance at the moment), but the plan is to
eventually have project pages setup so you could go look up some
information about some of my utilities before being kicked over to
github (planning to migrate to my own hosted scm eventually). The CSS
and HTML are written by-hand (not that there's much to it) and are
minified through a very basic shell script. :P

Performance, usability, portability, and security (in no particular
order) are large concerns of mine, so I have tried to leverage a lot
of things to make my site secure for end-users (and for me). For
instance, Lwan offers a built-in system to drop privileges and lock
itself in a chroot jail (I am still working on fully implementing
this, but it's next-up on my to-do list).

Technically, I have two servers running: one that serves on 8080, and
either serves files purely to meet Let's Encrypt ACME challenges
(which are fulfilled by acme-client
(<https://kristaps.bsd.lv/acme-client/>)) or forcibly 307s (/not/ 301)
to https://halosgho.st. The second is the actual site running on 8443.
nftables then redirects 80 -> 8080. And hitch
(<https://hitch-tls.org/>), which terminates my TLS endpoint runs on
443 and uses my site directly as the backend.

> The server is sending back some cool looking security headers:
>
> $ curl -I https://halosgho.st/
> HTTP/1.1 200 OK
> Content-Length: 2493
> Content-Type: text/html; charset=UTF-8
> Connection: keep-alive
> content-security-policy: default-src 'self'
> x-frame-options: SAMEORIGIN
> x-xss-protection: 1; mode=block
> x-content-type-options: nosniff
> referrer-policy: no-referrer
> strict-transport-security: max-age=31536000; includeSubDomains
> Date: Sun, 07 Oct 2018 17:37:19 GMT
> Expires: Sun, 14 Oct 2018 17:37:19 GMT
> Server: lwan
Yeah, I benchmark myself a bit on some various security metrics. I get
an A+ on SSLLabs and on securityheaders.com (the difference between an
A and an A+ on both is the HSTS header). These headers should
drastically reduce the easy of MITM, XSS, drive-by, embedding and
other attacks.

> The readme for your site's repo has a link to an auditing tool
> https://securityheaders.com for header security... Oh crap, when I run
> my own site through it I get a D grade.
What can I say, those benchmarks tickle me a bit. They are really easy
to implement (though some of them, when set as high as I have (read:
content-security-policy), can make development a bit more painful), so
if you wanted to do better on them, it's not hard.

>> I am definitely down for murmur meetups, or in-person get-togethers
>
> Great, can you send me a public SSH key? I'll add it to our shared
> server so you can log in for pair programming or updating our web site.
> http://frostbyte.cc/wiki.txt
I will generate one here in a bit and send it along!

From nompelis at nobelware.com  Sun Oct  7 19:41:52 2018
From: nompelis at nobelware.com (Ioannis Nompelis)
Date: Sun, 7 Oct 2018 14:41:52 -0500
Subject: list and weblogs
In-Reply-To: <20181007170733.GA24699@mysterium.local>
References: <20181005140336.GA21213@nobelware.com>
 <20181005173201.GA91555@mysterium.local>
 <84753e4e-1069-e09e-8379-e0b806206f22@saloits.com>
 <20181005203211.GA1190@mysterium.local>
 <878a4a19-f769-6632-58a3-5273c0cb8641@saloits.com>
 <20181005235430.GB1190@mysterium.local>
 <0ad7a0f0-f8b3-9309-8945-ce13fb609255@saloits.com>
 <20181006211624.GA55716@mysterium.local>
 <20181007163328.GB20747@nobelware.com>
 <20181007170733.GA24699@mysterium.local>
Message-ID: <20181007194152.GA3054@nobelware.com>

> 
> Are you talking about the index view or the message view?  For me,
> without customizing index_format, the index view shows lines like this:
> 
>   10     Oct 03 pstelzig        (0.1K) Working now
>

My index view is similar, and I know about the %L formatting.

I was talking about "message view" where you filter headers individually.

I did not know about memtp, but I will skip the ad-hoc pieaces for now. I am
a "sendmail" person, probably from legacy, and even dis-like postfix to some
extent. (Do not hate me for that; I also use it on my VPS.)


From nompelis at nobelware.com  Sun Oct  7 19:48:55 2018
From: nompelis at nobelware.com (Ioannis Nompelis)
Date: Sun, 7 Oct 2018 14:48:55 -0500
Subject: list and weblogs
In-Reply-To: <CABSwF4x5jQBa-3CDgWJwvjv539jLSEk75t_nXCfbe=UeyzZwmg@mail.gmail.com>
References: <20181005140336.GA21213@nobelware.com>
 <20181005173201.GA91555@mysterium.local>
 <CABSwF4xRnqmTMNbXfbnwFZQOvWhQnEuAUXvJXji+ETqbYS28ug@mail.gmail.com>
 <20181007175355.GA28139@mysterium.local>
 <CABSwF4x5jQBa-3CDgWJwvjv539jLSEk75t_nXCfbe=UeyzZwmg@mail.gmail.com>
Message-ID: <20181007194855.GB3054@nobelware.com>

> 
> Technically, I have two servers running: one that serves on 8080, and
> either serves files purely to meet Let's Encrypt ACME challenges
> (which are fulfilled by acme-client
> (<https://kristaps.bsd.lv/acme-client/>)) or forcibly 307s (/not/ 301)
> to https://halosgho.st. The second is the actual site running on 8443.
> nftables then redirects 80 -> 8080. And hitch
> (<https://hitch-tls.org/>), which terminates my TLS endpoint runs on
> 443 and uses my site directly as the backend.
>

I will admit here that his is 50% outside of my understanding.

So, I will say this first, I do not use automatic certificate rewnewal clinets,
and I will ask a question: why didn't multi-domain Let's Encrypt certificate
ever work for me?


From salo at saloits.com  Mon Oct  8 01:04:20 2018
From: salo at saloits.com (Timothy J. Salo)
Date: Sun, 7 Oct 2018 20:04:20 -0500
Subject: list and weblogs
In-Reply-To: <20181006211624.GA55716@mysterium.local>
References: <20181005140336.GA21213@nobelware.com>
 <20181005173201.GA91555@mysterium.local>
 <84753e4e-1069-e09e-8379-e0b806206f22@saloits.com>
 <20181005203211.GA1190@mysterium.local>
 <878a4a19-f769-6632-58a3-5273c0cb8641@saloits.com>
 <20181005235430.GB1190@mysterium.local>
 <0ad7a0f0-f8b3-9309-8945-ce13fb609255@saloits.com>
 <20181006211624.GA55716@mysterium.local>
Message-ID: <e11a9d46-bb11-757c-4285-b323d0e3efaa@saloits.com>

On 10/6/2018 4:16 PM, Joe Nelson wrote:
> Timothy J. Salo wrote:
>> _All_ of the IETF (Internet Engineering Task Force) mail lists that I
>> subscribe to (which is a fair number) modify the Subject: header.  It's
>> not just a Google thing.
> 
> First, I feel like it was intellectually lazy of me to exhibit the
> behavior of various lists. We ought to judge the options by their merits
> rather than by whichever organizations rightly or wrongly employ them.

No, widespread prevailing practice often _is_ relevant.

It is not uncommon for a standard to conflict with prevailing practice.
In some cases, prevailing practice changes to conform to the standard.
In other cases, the standard is revised to conform with the prevailing
practice that the standard tried to change (intentionally or
inadvertently).  I suppose that in some cases, a standard may be simply
ignored, and never gain traction.

Moreover, the fact that IETF, the organization responsible for the DMARC
standard, has decided that its mail lists will not conform to the
DMARC standard is a strong hint that the DMARC standard may be flawed.

> Second, senders whose domains have a p=reject DMARC policy have indeed
> complained of the IETF list behavior:
> https://www.ietf.org/mail-archive/web/dmarc/current/msg03229.html

Yes, but look at the responses to this email, and perhaps more
importantly, who wrote those responses:

o Theodore Ts'o: a long-time contributor to the Linux Kernel, a past
   member of the IETF Security Area Directorate, and a past chair of the
   IETF IPsec working group.  He refers to: "the DMARC brain-damage".

o Brian Carpenter: a past chair of the IETF.

o Dave Crocker: who has been active with Internet engineering and the
   IETF pretty much since the beginning.  He was involved with RFC 822,
   "Standard for the Format of ARPA Internet Text Messages", one of the
   early email standards.  And, he is a Senior Advisor of the Messaging
   Anti-Abuse Working Group.

When people highly respected in the Internet community think that DMARC
has some flaws, we should probably listen.

> * If we do want to go the route of adding a subject prefix, we
>    will be required to change the From address to be that of the
>    list rather than the original author. (Prefixing a subject
>    while retaining the original From often runs afoul of DMARC,
>    which makes messages go to Spam or returns them to sender.)

No, this isn't required.  The IETF mail lists don't do this.  And,
read the responses to the IETF mail you referenced.

> * If we change the From address we will want to add a Reply-To with the
>    original author's address. Otherwise there's no way for list members to
>    reply off-list unless they know each other's address. Reply-to addresses
>    are usually not displayed in mail clients' message listing, nor are they
>    added to a client address book. (Although mutt can display it with %A in
>    the index_format config parameter.)

Right.  Don't change the From: address.

I am not a mail expert.  And, I don't want to become.  But, I believe
that the IETF is working on a revision to the DMARC specification that
will permit more reasonable mail list behavior:

<https://datatracker.ietf.org/doc/draft-ietf-dmarc-arc-protocol/>

It is worth noting that this Internet Draft is a working group document,
not an individual submission.  As such, the general idea has been
exposed to a certain amount of scrutiny, before it was accepted as a
working group document.  And, this Internet Draft is on revision -18,
so it has experienced considerable review and discussion within the
working group.

If you are interested in the technical details of this proposal, you
should look at the IETF DMARC Working Group work on Authenticate
Received Chain (ARC).

-tjs


From samuel.stuewe at gmail.com  Mon Oct  8 01:43:35 2018
From: samuel.stuewe at gmail.com (Sam Stuewe)
Date: Sun, 7 Oct 2018 20:43:35 -0500
Subject: list and weblogs
In-Reply-To: <20181007194855.GB3054@nobelware.com>
References: <20181005140336.GA21213@nobelware.com>
 <20181005173201.GA91555@mysterium.local>
 <CABSwF4xRnqmTMNbXfbnwFZQOvWhQnEuAUXvJXji+ETqbYS28ug@mail.gmail.com>
 <20181007175355.GA28139@mysterium.local>
 <CABSwF4x5jQBa-3CDgWJwvjv539jLSEk75t_nXCfbe=UeyzZwmg@mail.gmail.com>
 <20181007194855.GB3054@nobelware.com>
Message-ID: <CABSwF4zVEz3dJLrmr24EJPMRNc8D-uixz7V4AARCDBZ5BviN0Q@mail.gmail.com>

On 10/7/18, Ioannis Nompelis <nompelis at nobelware.com> wrote:
> I will admit here that his is 50% outside of my understanding.
Never a better opportunity to learn! :D

> So, I will say this first, I do not use automatic certificate rewnewal
> clinets,
> and I will ask a question: why didn't multi-domain Let's Encrypt
> certificate
> ever work for me?
I am unsure, but I may have been unclear. When I said two servers, I
meant software-wise, not hardware-wise. Both of those lwan instances
are running on the same machine. Additionally, since I explicitly
didn't want to use certbot (the official LE client), it took me a
while to get it working, though it works reasonably well now. What
went wrong for you? I'd happily lend my assistance if I can.

Tangentially related: I don't have a personal blog (I have in the
past, though never managed to keep it updated), but I have been
thinking about it. For those of you that do host a blog, is there a
particular motivation that keeps you involved in writing posts or is
it just for your personal musings?

All the best,

-Sam

From nompelis at nobelware.com  Mon Oct  8 02:41:52 2018
From: nompelis at nobelware.com (Ioannis Nompelis)
Date: Sun, 7 Oct 2018 21:41:52 -0500
Subject: list and weblogs
In-Reply-To: <CABSwF4zVEz3dJLrmr24EJPMRNc8D-uixz7V4AARCDBZ5BviN0Q@mail.gmail.com>
References: <20181005140336.GA21213@nobelware.com>
 <20181005173201.GA91555@mysterium.local>
 <CABSwF4xRnqmTMNbXfbnwFZQOvWhQnEuAUXvJXji+ETqbYS28ug@mail.gmail.com>
 <20181007175355.GA28139@mysterium.local>
 <CABSwF4x5jQBa-3CDgWJwvjv539jLSEk75t_nXCfbe=UeyzZwmg@mail.gmail.com>
 <20181007194855.GB3054@nobelware.com>
 <CABSwF4zVEz3dJLrmr24EJPMRNc8D-uixz7V4AARCDBZ5BviN0Q@mail.gmail.com>
Message-ID: <20181008024152.GC3054@nobelware.com>

> I am unsure, but I may have been unclear. When I said two servers, I
> meant software-wise, not hardware-wise. Both of those lwan instances
> are running on the same machine. Additionally, since I explicitly
> didn't want to use certbot (the official LE client), it took me a
> while to get it working, though it works reasonably well now. What
> went wrong for you? I'd happily lend my assistance if I can.
>

Much clearer now. My issue was with putting multiple domains that are
virtual hosts (same IP, and Apache makes the distinction using the domain
name from the URL). I found no way of getting LE to certify multiple
domains. That is where I need some help. We should probably open a new
topic on the list and not pollute this thread; your call.
 
> Tangentially related: I don't have a personal blog (I have in the
> past, though never managed to keep it updated), but I have been
> thinking about it. For those of you that do host a blog, is there a
> particular motivation that keeps you involved in writing posts or is
> it just for your personal musings?
>

I, generally, keep a lot of notes, which are not made public. Sometimes
making things public has the added benefit of forcing one to structure
their ideas and make a story coherent for all, rather than essoteric.
(I had a brief discussion about this with Dave Bucklin.) The main motivation
for me is to "give back," as in, I have been that poor soul that wishes
they knew and would gladly stumble on a solution that is meant for them
to read. One motto I have addopted is "give more than you taken" but I
also like the "better than you found it" one, which is related.

But do not believe that I am that nice given these ideas.


From salo at saloits.com  Mon Oct  8 04:01:15 2018
From: salo at saloits.com (Timothy J. Salo)
Date: Sun, 7 Oct 2018 23:01:15 -0500
Subject: list and weblogs
In-Reply-To: <20181008024152.GC3054@nobelware.com>
References: <20181005140336.GA21213@nobelware.com>
 <20181005173201.GA91555@mysterium.local>
 <CABSwF4xRnqmTMNbXfbnwFZQOvWhQnEuAUXvJXji+ETqbYS28ug@mail.gmail.com>
 <20181007175355.GA28139@mysterium.local>
 <CABSwF4x5jQBa-3CDgWJwvjv539jLSEk75t_nXCfbe=UeyzZwmg@mail.gmail.com>
 <20181007194855.GB3054@nobelware.com>
 <CABSwF4zVEz3dJLrmr24EJPMRNc8D-uixz7V4AARCDBZ5BviN0Q@mail.gmail.com>
 <20181008024152.GC3054@nobelware.com>
Message-ID: <b900f176-00b7-c13a-0679-824f75cd5e0d@saloits.com>

On 10/7/2018 9:41 PM, Ioannis Nompelis wrote:

> Much clearer now. My issue was with putting multiple domains that are
> virtual hosts (same IP, and Apache makes the distinction using the domain
> name from the URL). I found no way of getting LE to certify multiple
> domains. That is where I need some help. We should probably open a new
> topic on the list and not pollute this thread; your call.

Have you configured SNI?

See, for example, for a description of SNI and why you need it to use
multiple SSL certs with multiple virtual servers:

   <https://stuff-things.net/2016/11/16/apache-sni/>

   This [figuring out which virtual host to use, based on the URL]
   doesn?t work for HTTPS, the server send the SSL Certificate long
   before the browser sends any data about the request. SNI was added
   to the (TLS) negotiation so that the browser could request a specific
   certificate when making the initial connection, allowing the server
   to handle many certificates on a single IP.

Googling [apache sni] should yield lots of good info.

Once I got SNI working, the Lets Encrypt certbot-auto script gets or
renews certs for all of my virtual hosts.  The script even asks which
virtual hosts you want to create certs for.

Is this what you were looking for?

-tjs

From dave.bucklin at gmail.com  Mon Oct  8 12:06:47 2018
From: dave.bucklin at gmail.com (Dave Bucklin)
Date: Mon, 8 Oct 2018 07:06:47 -0500
Subject: list and weblogs
In-Reply-To: <CABSwF4zVEz3dJLrmr24EJPMRNc8D-uixz7V4AARCDBZ5BviN0Q@mail.gmail.com>
References: <20181005140336.GA21213@nobelware.com>
 <20181005173201.GA91555@mysterium.local>
 <CABSwF4xRnqmTMNbXfbnwFZQOvWhQnEuAUXvJXji+ETqbYS28ug@mail.gmail.com>
 <20181007175355.GA28139@mysterium.local>
 <CABSwF4x5jQBa-3CDgWJwvjv539jLSEk75t_nXCfbe=UeyzZwmg@mail.gmail.com>
 <20181007194855.GB3054@nobelware.com>
 <CABSwF4zVEz3dJLrmr24EJPMRNc8D-uixz7V4AARCDBZ5BviN0Q@mail.gmail.com>
Message-ID: <20181008120647.2war4owy566ljsua@19a6.tech>

On Sun, Oct 07, 2018 at 08:43:35PM -0500, Sam Stuewe wrote:
> For those of you that do host a blog, is there a
> particular motivation that keeps you involved in writing posts or is
> it just for your personal musings?

For me, it's two things. First, it's a creative outlet. I get to write
about this weird stuff that almost no one else I know is interested in.
Second, it gives potential employers something to look at beyond my
resume and, I hope, demonstrates my ability to work through technical
problems and explain them to a non-technical audience.

From joe at begriffs.com  Wed Oct 10 02:11:56 2018
From: joe at begriffs.com (Joe Nelson)
Date: Tue, 9 Oct 2018 21:11:56 -0500
Subject: list and weblogs
In-Reply-To: <e11a9d46-bb11-757c-4285-b323d0e3efaa@saloits.com>
References: <20181005140336.GA21213@nobelware.com>
 <20181005173201.GA91555@mysterium.local>
 <84753e4e-1069-e09e-8379-e0b806206f22@saloits.com>
 <20181005203211.GA1190@mysterium.local>
 <878a4a19-f769-6632-58a3-5273c0cb8641@saloits.com>
 <20181005235430.GB1190@mysterium.local>
 <0ad7a0f0-f8b3-9309-8945-ce13fb609255@saloits.com>
 <20181006211624.GA55716@mysterium.local>
 <e11a9d46-bb11-757c-4285-b323d0e3efaa@saloits.com>
Message-ID: <20181010021156.GA59713@mysterium.local>

Timothy J. Salo wrote:
> It is not uncommon for a standard to conflict with prevailing practice.
> In some cases, prevailing practice changes to conform to the standard.
> In other cases, the standard is revised to conform with the prevailing
> practice that the standard tried to change (intentionally or
> inadvertently).  I suppose that in some cases, a standard may be simply
> ignored, and never gain traction.

I see, so the prevailing practice in this case is for lists to prepend
their identifier to the subject, and add a footer to the body with
administrative links. DMARC admittedly conflicts with this practice:
the SPF check will fail due to the unapproved sender (list server) and
the DKIM signature will be invalidated by the message modification in
subject or body.

> Moreover, the fact that IETF, the organization responsible for the DMARC
> standard, has decided that its mail lists will not conform to the
> DMARC standard is a strong hint that the DMARC standard may be flawed.

As you mentioned, the IETF is developing Authenticated Received Chain
(ARC), a standard which complements DMARC and allows the recipient
to determine that no party other than a certain intermediate (for
us, a list server) has modified the sender's message. This allows
the recipient, trusting the intermediary, to accept a message which
otherwise fails DMARC.

The IETF email thread seems to object to DMARC for two main reasons:
a) the standard was decided by big mail companies and thrust on the
   world de-facto
b) it lacks the notion of trusted intermediaries. Any changes to the
   DKIM-signed fields means instant failure, no matter who modified them
This makes sense, I can see how DMARC is shortsighted there.

However, here's what confuses me about prevailing practice. Why are
people adamant about using the subject and body to perform list id and
administration? Some list admins would rather wait for ARC to get rolled
out (while some of their messages to go spam) rather than switch over
to using List-* headers instead. List-* headers work fine with DMARC
because the sender doesn't sign them with DKIM.

My best guess is that the admins don't want to switch over because
a) it would mess with the filtering rules of their current subscribers
b) using those specialized headers is not compatible with simplistic
   mail clients. The less sophisticated clients hide headers and are
   unable to perform filter actions on them
c) admins don't like google/yahoo/etc bossing them around, and would
   rather some of their mail bounces than change anything to comply
   while ARC is being developed

I can definitely acknowledge the argument that the subject and body are
a good place to display list information because they are guaranteed to
be visible to everyone. It's a bit of a kludge and a deficiency of MUAs,
but that's kind of where we are.

I value deliverability in the friends@ list more than this UI issue, so
propose to keep the current behavior. See below.

> > * If we do want to go the route of adding a subject prefix, we
> >    will be required to change the From address to be that of the
> >    list rather than the original author. (Prefixing a subject
> >    while retaining the original From often runs afoul of DMARC,
> >    which makes messages go to Spam or returns them to sender.)
> 
> No, this isn't required.  The IETF mail lists don't do this.  And,
> read the responses to the IETF mail you referenced.

Mailman does not yet support ARC, although it's targeted for the 3.3
release https://gitlab.com/mailman/mailman/merge_requests/204

Until we are able to enable ARC, I don't see what you mean. A message
from a p=reject sending domain *will* be rejected by recipients if our
list modifies it, right? At least for recipients that check DMARC.

When ARC is available for us, then we could consider putting the list id
in the subject line.

From salo at saloits.com  Wed Oct 10 08:40:38 2018
From: salo at saloits.com (Timothy J. Salo)
Date: Wed, 10 Oct 2018 03:40:38 -0500
Subject: list and weblogs
In-Reply-To: <20181010021156.GA59713@mysterium.local>
References: <20181005140336.GA21213@nobelware.com>
 <20181005173201.GA91555@mysterium.local>
 <84753e4e-1069-e09e-8379-e0b806206f22@saloits.com>
 <20181005203211.GA1190@mysterium.local>
 <878a4a19-f769-6632-58a3-5273c0cb8641@saloits.com>
 <20181005235430.GB1190@mysterium.local>
 <0ad7a0f0-f8b3-9309-8945-ce13fb609255@saloits.com>
 <20181006211624.GA55716@mysterium.local>
 <e11a9d46-bb11-757c-4285-b323d0e3efaa@saloits.com>
 <20181010021156.GA59713@mysterium.local>
Message-ID: <f7525631-71cf-3609-969b-fbda8da388d8@saloits.com>

On 10/9/2018 9:11 PM, Joe Nelson wrote:
> The IETF email thread seems to object to DMARC for two main reasons:
> a) the standard was decided by big mail companies and thrust on the
>     world de-facto
> b) it lacks the notion of trusted intermediaries. Any changes to the
>     DKIM-signed fields means instant failure, no matter who modified them
> This makes sense, I can see how DMARC is shortsighted there.

I didn't read the IETF DMARC email.  But, standardization efforts such
as DMARC are often under a lot of pressure to finalize _something_,
even though it isn't perfect.  I suspect that the DMARC working
group felt this pressure.  While the IETF (as a standards organization)
generally tries to not break stuff, tries to not conflict with existing
practice or implementations, it doesn't always succeed.  I suspect that
this is another case where "moving fast" overcame the IETF's reluctance
to "breaking things".  But, I don't keep up on the DMARC working
group, so this is just my speculation based on how standards groups
typically work.

By the way, this reluctance to conflict with existing implementations
is why we have URLs of the form "http://", rather than something
more concise like "http:".  I don't know what the existing
implementation was that used "http://", but I suspect that it was
something we haven't heard of for decades.  If I had all those wasted
two-bytes from every URL send over the Internet...

> However, here's what confuses me about prevailing practice. Why are
> people adamant about using the subject and body to perform list id and
> administration? Some list admins would rather wait for ARC to get rolled
> out (while some of their messages to go spam) rather than switch over
> to using List-* headers instead. List-* headers work fine with DMARC
> because the sender doesn't sign them with DKIM.
> 
> My best guess is that the admins don't want to switch over because
> a) it would mess with the filtering rules of their current subscribers
> b) using those specialized headers is not compatible with simplistic
>     mail clients. The less sophisticated clients hide headers and are
>     unable to perform filter actions on them
> c) admins don't like google/yahoo/etc bossing them around, and would
>     rather some of their mail bounces than change anything to comply
>     while ARC is being developed

I think that it is probably users who don't like the change.

Stuff I actually read goes to my inbox.  I filter stuff in my inbox
(based on To:, From:, Cc: and List-Id:) and tag it with color.  I
currently use only the "important" (red) tag.  Red seems to be the
only color that I am assured of seeing under all light conditions.

I visually scan the unread email in my inbox for "[...] in the subject
line (which is a lot of email) and for email lines in red (which is
still quite a bit of email).  Those are things that I am assured of
at looking at before I mark them as junk and delete them.

I filter a lot of email based on List-Id: and move it to folders.  But,
I don't particularly read the mail that gets filtered to  folders.  I
have a _lot_ of folders, and many of them are nested.  I only look at
email that has been filtered into folders if I am searching for email
(often long after it was sent) on some topic or have some reason to
actually look in these folders.  Much of the email in these folders is
never read.  I have dozens of folders, perhaps hundreds.  Again, I use
folders for archiving email, not for reading email.

> I can definitely acknowledge the argument that the subject and body are
> a good place to display list information because they are guaranteed to
> be visible to everyone. It's a bit of a kludge and a deficiency of MUAs,
> but that's kind of where we are.

I don't see this as a deficiency of MUAs.  It's simply how I (and
apparently a lot of other people) use email.

> I value deliverability in the friends@ list more than this UI issue, so
> propose to keep the current behavior. See below.
> 
>>> * If we do want to go the route of adding a subject prefix, we
>>>     will be required to change the From address to be that of the
>>>     list rather than the original author. (Prefixing a subject
>>>     while retaining the original From often runs afoul of DMARC,
>>>     which makes messages go to Spam or returns them to sender.)
>>
>> No, this isn't required.  The IETF mail lists don't do this.  And,
>> read the responses to the IETF mail you referenced.
> 
> Mailman does not yet support ARC, although it's targeted for the 3.3
> release https://gitlab.com/mailman/mailman/merge_requests/204
> 
> Until we are able to enable ARC, I don't see what you mean. A message
> from a p=reject sending domain *will* be rejected by recipients if our
> list modifies it, right? At least for recipients that check DMARC.

Isn't a white-list function available?

> When ARC is available for us, then we could consider putting the list id
> in the subject line.

It is important to keep the "[...]" string short, so you might want to
use something much shorter.  Maybe, "[mn-maker]" rather than
"[Let's learn and code together <friends.talk.begriffs.com>]".  Long
"[...]" strings pretty much don't work.

By the way, "[...]" tags in the Subject: line work for direct replies
to a list message (for example, direct email from a list recipient
to the author, which isn't copied to the list), but List-Id: isn't
present in this direct email.  The receiver either has to filter on
something other than List-Id: or has to remember all of the active
topics for this group (since there is no visual marker).

If email doesn't have a "[...]" in Subject: line, if I don't recognize
the sender, and if the Subject: doesn't catch my eye for some reason,
email usually gets manually marked as junk and deleted.

-tjs

From nompelis at nobelware.com  Wed Oct 10 19:16:27 2018
From: nompelis at nobelware.com (Ioannis Nompelis)
Date: Wed, 10 Oct 2018 19:16:27 +0000
Subject: list and weblogs
In-Reply-To: <f7525631-71cf-3609-969b-fbda8da388d8@saloits.com>
References: <20181005173201.GA91555@mysterium.local>
 <84753e4e-1069-e09e-8379-e0b806206f22@saloits.com>
 <20181005203211.GA1190@mysterium.local>
 <878a4a19-f769-6632-58a3-5273c0cb8641@saloits.com>
 <20181005235430.GB1190@mysterium.local>
 <0ad7a0f0-f8b3-9309-8945-ce13fb609255@saloits.com>
 <20181006211624.GA55716@mysterium.local>
 <e11a9d46-bb11-757c-4285-b323d0e3efaa@saloits.com>
 <20181010021156.GA59713@mysterium.local>
 <f7525631-71cf-3609-969b-fbda8da388d8@saloits.com>
Message-ID: <20181010191627.GA23413@nobelware.com>

Tim, my recommendation is to "simplicate" life. Getting email that you never
read and keeping it would bother me knowing it is happening! Oh man, how
different people are surprises me every time.

People generally try to get out of mass mailings, not archive them, unless
they are a three-letter agency.

Alright, thank you guys for teaching me a bunch about email! It was worth the
read so far.

Also, I am looking into the virtual hosts and HTTPS certificate validation.
I have just been too busy to report back... writing code for work, testing it,
and being in meetings. Plus, I had to get my car ready for a weekend at BIR
starting this Friday. So far, I am, miraculously, meeting all deadlines.


From salo at saloits.com  Wed Oct 10 23:24:11 2018
From: salo at saloits.com (Timothy J. Salo)
Date: Wed, 10 Oct 2018 18:24:11 -0500
Subject: list and weblogs
In-Reply-To: <20181010191627.GA23413@nobelware.com>
References: <20181005173201.GA91555@mysterium.local>
 <84753e4e-1069-e09e-8379-e0b806206f22@saloits.com>
 <20181005203211.GA1190@mysterium.local>
 <878a4a19-f769-6632-58a3-5273c0cb8641@saloits.com>
 <20181005235430.GB1190@mysterium.local>
 <0ad7a0f0-f8b3-9309-8945-ce13fb609255@saloits.com>
 <20181006211624.GA55716@mysterium.local>
 <e11a9d46-bb11-757c-4285-b323d0e3efaa@saloits.com>
 <20181010021156.GA59713@mysterium.local>
 <f7525631-71cf-3609-969b-fbda8da388d8@saloits.com>
 <20181010191627.GA23413@nobelware.com>
Message-ID: <798f3d67-8357-fb5c-d34d-d0a869ab1a58@saloits.com>

On 10/10/2018 2:16 PM, Ioannis Nompelis wrote:
> Tim, my recommendation is to "simplicate" life. Getting email that you never
> read and keeping it would bother me knowing it is happening! Oh man, how
> different people are surprises me every time.

Filtering certain email to folders does greatly simplify my life.  My
inbox is for mail that I want to be aware of in near-real-time.

Mail filtered to folders is mail that I don't need to look at in
near-real-time.  These are mostly mail lists that I want to track,
although not necessarily interact with.  It includes a bunch of
email lists related to amateur radio (mostly digital communications),
amateur satellites, the IETF mail lists that don't land in my
inbox, and some public policy discussions.

Some of these archived folders I will read in batches, when I get
around to it.  Other folders, I look at when I am searching for
email on a particular topic, or when I want to see discussions about
some current event.  I prefer to archive these emails on my machine,
rather than rely on internet-accessible archives, for several
reasons:

1) It's faster to read local email.  Reading a lot of messages in
    an Internet archive can be really slow.

2) It's easier to search.  Google search has its place, but sometimes
    grep is really nice.  Plus, some email archives refuse to let
    search engines index them.

3) Some Internet archives are incomplete.  They probably don't (and
    probably shouldn't) include sender email addresses; sometimes I
    want that.  Sometimes, old email is deleted, or sometimes individual
    messages are deleted; sometimes I want those messages.

I'm not recommending this practice, but it is what works for me.

> People generally try to get out of mass mailings, not archive them, unless
> they are a three-letter agency.

My folders contain filtered mail, generally email from lists, so they
are generally pretty clean.

> Alright, thank you guys for teaching me a bunch about email! It was worth the
> read so far.

I hope that someone found what I write to be useful or educational.

> Also, I am looking into the virtual hosts and HTTPS certificate validation.

I found it a bit tricky to get configured, but "Google is your friend".
Start by Googling [https sni].

You can find some good in info at: <https://letsencrypt.org/> .

There are sites that will analyze your SSL configuration.  I used:
<https://www.ssllabs.com/ssltest/> .

-tjs

From nompelis at nobelware.com  Wed Oct 10 23:31:36 2018
From: nompelis at nobelware.com (Ioannis Nompelis)
Date: Wed, 10 Oct 2018 23:31:36 +0000
Subject: list and weblogs
In-Reply-To: <798f3d67-8357-fb5c-d34d-d0a869ab1a58@saloits.com>
References: <20181005203211.GA1190@mysterium.local>
 <878a4a19-f769-6632-58a3-5273c0cb8641@saloits.com>
 <20181005235430.GB1190@mysterium.local>
 <0ad7a0f0-f8b3-9309-8945-ce13fb609255@saloits.com>
 <20181006211624.GA55716@mysterium.local>
 <e11a9d46-bb11-757c-4285-b323d0e3efaa@saloits.com>
 <20181010021156.GA59713@mysterium.local>
 <f7525631-71cf-3609-969b-fbda8da388d8@saloits.com>
 <20181010191627.GA23413@nobelware.com>
 <798f3d67-8357-fb5c-d34d-d0a869ab1a58@saloits.com>
Message-ID: <20181010233136.GA481@nobelware.com>

Tim sounds like my kind of guy based on topics of interest!

I _love_ Mutt for the search functionality, which is an evolution of the
functionality of Elm, the evolution of which is Mutt itself. Can't beat the
functionality, but I admit of using grep tactically on occasion.

I have been with Let's Encrypt for a long while, and my issue was SNI, which I
will be studying next week. Joe's weblog has a lot of info on random topics,
and one of them is exactly that or related.


From nompelis at nobelware.com  Thu Oct 11 04:47:10 2018
From: nompelis at nobelware.com (Ioannis Nompelis)
Date: Wed, 10 Oct 2018 23:47:10 -0500
Subject: my PGP public key
Message-ID: <20181011044710.GE2160@nobelware.com>

Never a bad time to remind and encourage people to encrypt. Here is my public
key, whihc you can also find on key servers.


-----BEGIN PGP PUBLIC KEY BLOCK-----
Version: GnuPG v2.0.18 (GNU/Linux)
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=dB/f
-----END PGP PUBLIC KEY BLOCK-----

From joe at begriffs.com  Thu Oct 11 06:41:33 2018
From: joe at begriffs.com (Joe Nelson)
Date: Thu, 11 Oct 2018 01:41:33 -0500
Subject: list and weblogs
In-Reply-To: <20181010021156.GA59713@mysterium.local>
References: <20181005140336.GA21213@nobelware.com>
 <20181005173201.GA91555@mysterium.local>
 <84753e4e-1069-e09e-8379-e0b806206f22@saloits.com>
 <20181005203211.GA1190@mysterium.local>
 <878a4a19-f769-6632-58a3-5273c0cb8641@saloits.com>
 <20181005235430.GB1190@mysterium.local>
 <0ad7a0f0-f8b3-9309-8945-ce13fb609255@saloits.com>
 <20181006211624.GA55716@mysterium.local>
 <e11a9d46-bb11-757c-4285-b323d0e3efaa@saloits.com>
 <20181010021156.GA59713@mysterium.local>
Message-ID: <20181011064133.GB21311@mysterium.local>

I previously wrote:
> I can definitely acknowledge the argument that the subject and body are
> a good place to display list information because they are guaranteed to
> be visible to everyone. ...

Actually here's another argument in favor of using a subject prefix:
it's preserved by even direct replies.  I've noticed on this list
that if someone replies to me and CC's the list, their direct message
doesn't include a List-Id header. Lacking that header, it doesn't match
my filter and lands in my inbox. It doesn't appear as part of the rest
of its thread inside my filter folder.  Filtering instead on e.g.
"[friends]" in the subject would catch direct replies too.

The list configuration could be changed to send a duplicate message to
people who are already in the To and CC lines, but I think that would
be annoying in its own way.

This is making me wonder what would happen for the current list
subscribers if we did try enabling subject line prefixing. Do any of
us actually have a strict enough DMARC policy to mess up the delivery?

gmail.com
  v=DMARC1; p=none; sp=quarantine; rua=mailto:mailauth-reports at google.com
christianwood.net
  (no dmarc)
riseup.net
  v=DMARC1;p=none;pct=100
nobelware.com
  (no dmarc)
gmx.com
  (no dmarc)
saloits.com
  (no dmarc)
begriffs.com
  v=DMARC1; p=reject; rua=mailto:cred+rua at begriffs.com; ruf=mailto:cred+ruf at begriffs.com; adkim=r; aspf=r; pct=100; rf=afrf; ri=86400; sp=reject

Those domains with p=none are instructing the receiver to not take any
special action if DMARC fails. So in reality it looks like only my
own domain has the settings turned up to 11. I could try switching to
p=none as well. (I'll still get daily reports emailed to me about anyone
spoofing me, so I'm comfortable trying this change.)

What do you all think, would you prefer me to enable a subject prefix on
the list? I believe we've found a few real advantages of the practice,
and determined it won't hurt the current subscribers. (If a yahoo.com
person signs up, well then their messages would have problems...)

From joe at begriffs.com  Thu Oct 11 17:38:18 2018
From: joe at begriffs.com (Joe Nelson)
Date: Thu, 11 Oct 2018 12:38:18 -0500
Subject: list and weblogs
In-Reply-To: <20181011064133.GB21311@mysterium.local>
References: <20181005173201.GA91555@mysterium.local>
 <84753e4e-1069-e09e-8379-e0b806206f22@saloits.com>
 <20181005203211.GA1190@mysterium.local>
 <878a4a19-f769-6632-58a3-5273c0cb8641@saloits.com>
 <20181005235430.GB1190@mysterium.local>
 <0ad7a0f0-f8b3-9309-8945-ce13fb609255@saloits.com>
 <20181006211624.GA55716@mysterium.local>
 <e11a9d46-bb11-757c-4285-b323d0e3efaa@saloits.com>
 <20181010021156.GA59713@mysterium.local>
 <20181011064133.GB21311@mysterium.local>
Message-ID: <20181011173818.GA53625@mysterium.local>

Joe Nelson wrote:
> Lacking that header, it doesn't match my filter and lands in my
> inbox. It doesn't appear as part of the rest of its thread inside my
> filter folder.  Filtering instead on e.g. "[friends]" in the subject
> would catch direct replies too.

Actually there's another solution to this. Just make the filter rule
compound:

  List-Id contains friends.talk.begriffs.com
  OR   CC contains friends at talk.begriffs.com

This works nicely because if someone replies off-list then it will not
match this rule, but if the list sends the message or is included in the
reply then the rule will match. On-list=filter, off-list=inbox.

As a side note, I think it's possible to do filtering even if the list
hadn't set List-Id, because it preserves "To: Friends at talk.begriffs.com"
from the original sender, so the rule could match on just To and CC.

With that taken care of I don't feel a compelling reason to switch to
prefixing. What do you think?

From nompelis at nobelware.com  Thu Oct 11 22:07:54 2018
From: nompelis at nobelware.com (Ioannis Nompelis)
Date: Thu, 11 Oct 2018 22:07:54 +0000
Subject: list and weblogs
In-Reply-To: <20181011173818.GA53625@mysterium.local>
References: <84753e4e-1069-e09e-8379-e0b806206f22@saloits.com>
 <20181005203211.GA1190@mysterium.local>
 <878a4a19-f769-6632-58a3-5273c0cb8641@saloits.com>
 <20181005235430.GB1190@mysterium.local>
 <0ad7a0f0-f8b3-9309-8945-ce13fb609255@saloits.com>
 <20181006211624.GA55716@mysterium.local>
 <e11a9d46-bb11-757c-4285-b323d0e3efaa@saloits.com>
 <20181010021156.GA59713@mysterium.local>
 <20181011064133.GB21311@mysterium.local>
 <20181011173818.GA53625@mysterium.local>
Message-ID: <20181011220754.GA18938@nobelware.com>

> 
>   List-Id contains friends.talk.begriffs.com
>   OR   CC contains friends at talk.begriffs.com
> 
> This works nicely because if someone replies off-list then it will not
> match this rule, but if the list sends the message or is included in the
> reply then the rule will match. On-list=filter, off-list=inbox.
> 

You had me convinced with your previous email that one (recipient) would want
the personal responses to go into the same folder (be filtered in the same way)
as the general discussion on the list. This, assuming that all relevant talk
gets put in the same place. Now, see below.


> As a side note, I think it's possible to do filtering even if the list
> hadn't set List-Id, because it preserves "To: Friends at talk.begriffs.com"
> from the original sender, so the rule could match on just To and CC.
>

Sure, filtering is possible by the recipient. But now what I thought was a good
notion to bear in mind with regard to keeping list-talk to the list-related
folder, etc, including personal replies, goes out the door.

Maybe we need to have an answer to this semantic before deciding on the
proposition to infect the subject line -- of which I am in strong favour.


From joe at begriffs.com  Fri Oct 12 07:56:53 2018
From: joe at begriffs.com (Joe Nelson)
Date: Fri, 12 Oct 2018 00:56:53 -0700
Subject: list and weblogs
In-Reply-To: <f7525631-71cf-3609-969b-fbda8da388d8@saloits.com>
References: <20181005173201.GA91555@mysterium.local>
 <84753e4e-1069-e09e-8379-e0b806206f22@saloits.com>
 <20181005203211.GA1190@mysterium.local>
 <878a4a19-f769-6632-58a3-5273c0cb8641@saloits.com>
 <20181005235430.GB1190@mysterium.local>
 <0ad7a0f0-f8b3-9309-8945-ce13fb609255@saloits.com>
 <20181006211624.GA55716@mysterium.local>
 <e11a9d46-bb11-757c-4285-b323d0e3efaa@saloits.com>
 <20181010021156.GA59713@mysterium.local>
 <f7525631-71cf-3609-969b-fbda8da388d8@saloits.com>
Message-ID: <20181012075653.GB5539@mysterium.local>

My apologies, didn't see this message until just now because it went to
my spam folder. Here are the factors that made fastmail think it was
spam:

X-Spam-score: 5.2
X-Spam-hits: ME_CMSPAM_100 4, RDNS_NONE 1.274, SPF_HELO_PASS -0.001, SPF_PASS -0.001,
  LANGUAGES en, BAYES_USED none, SA_VERSION 3.4.0

I don't know what ME_CMSPAM_100 means, but RDNS_NONE seems like it might
be a saloits reverse-DNS misconfiguration
https://wiki.apache.org/spamassassin/Rules/RDNS_NONE

If you could solve the rdns part I think future messages won't cross the
spam threshold.

Timothy J. Salo wrote:
> By the way, this reluctance to conflict with existing implementations
> is why we have URLs of the form "http://", rather than something more
> concise like "http:".

Interesting, I never stopped to think about how pointless
the slashes are. After some searching, found this:
https://www.w3.org/People/Berners-Lee/FAQ.html#etc
Guess it comes from Apollo Unix paths.

> I have a _lot_ of folders, and many of them are nested. ... I have
> dozens of folders, perhaps hundreds.

Dang, you're an email power user! Perhaps my few folders are destined to
multiply over time.

> > Until we are able to enable ARC, I don't see what you mean. A message
> > from a p=reject sending domain *will* be rejected by recipients if our
> > list modifies it, right? At least for recipients that check DMARC.
> 
> Isn't a white-list function available?

Well what do you know, I can whitelist a whole domain in fastmail by
adding "*@<domain>.<tld>" to the address book on there. I had assumed
that I would have to run my own mailserver to do whitelisting, but guess
not. Glad you pointed that out.

> It is important to keep the "[...]" string short, so you might want to
> use something much shorter.  Maybe, "[mn-maker]" rather than
> "[Let's learn and code together <friends.talk.begriffs.com>]".  Long
> "[...]" strings pretty much don't work.

For sure, don't know why mailman made such a long list-id. It pulled
from a description property that I added to populate part of the web
GUI.

> By the way, "[...]" tags in the Subject: line work for direct replies
> to a list message (for example, direct email from a list recipient
> to the author, which isn't copied to the list), but List-Id: isn't
> present in this direct email.  The receiver either has to filter on
> something other than List-Id: or has to remember all of the active
> topics for this group (since there is no visual marker).

Wish I would have seen this message earlier! I arrived at the same
conclusion but only after some time. Although also realized you can add
a search for the list name in the CC field as well as the list-id which
will catch those direct replies that also CC the list. (And maybe direct
replies that don't CC the list belong in the inbox.)

> If email doesn't have a "[...]" in Subject: line, if I don't recognize
> the sender, and if the Subject: doesn't catch my eye for some reason,
> email usually gets manually marked as junk and deleted.

Could you consider treating the folder for this list's messages as less
of an archive like your other folders, and more as another inbox-like
place? Given previous research in another message, I saw that none of
our domains except mine is going to cause DMARC problems, so I *could*
enable subject prefixes. I kind of like the purity of the unadorned
subjects we have now, but want to gauge how much of a pain it is for you
to adapt to that way. Maybe given all your existing folders this one
wouldn't stand out enough, unless you can reorder the folders.

From nompelis at nobelware.com  Fri Oct 12 17:13:16 2018
From: nompelis at nobelware.com (Ioannis Nompelis)
Date: Fri, 12 Oct 2018 17:13:16 +0000
Subject: Books: The Laws of Cryptography with Java Code - Neal R. Wagner
Message-ID: <20181012171316.GA28732@nobelware.com>

I figured we can throw in an interesting link fo online material worth
archiving for collective or individual use. I have many links, but let's
start with one.

I found this link a while ago and it is still active.
http://www.cs.utsa.edu/~wagner/lawsbookcolor/laws.pdf

I am always interested in getting good books in PDF. This one is exactly what
one would expect. If you want more depth, look elsewhere. But if you wnat to
code something quickly, look here.


From nompelis at nobelware.com  Fri Oct 12 18:59:37 2018
From: nompelis at nobelware.com (Ioannis Nompelis)
Date: Fri, 12 Oct 2018 18:59:37 +0000
Subject: Books: Project Oberon; The Design of an Operating System and Compiler
Message-ID: <20181012185937.GB28732@nobelware.com>

This is another interesting book in PDf. Written by Niklaus Wirth (legendary
and pioneering computer scientist -- still alive) and Jurg Gutknecht.

Books: Project Oberon; The Design of an Operating System and Compiler

Wirth's "Algorithms + Data Structures = Programs", which I have on hard-cover,
is a classic. Mine has the elusive chapter 5 on parsing languages.


From joe at begriffs.com  Sat Oct 20 15:59:11 2018
From: joe at begriffs.com (Joe Nelson)
Date: Sat, 20 Oct 2018 10:59:11 -0500
Subject: Usenet over ham radio?
Message-ID: <20181020155911.GA5778@mysterium.local>

Anyone know if it's feasible to execute UUCP over ham radio to run a
radio-based usenet? The list we have now is certainly more convenient,
but the challenge is intriguing. It's enticing to think a local group
can carry on communication without ISPs or cell phone companies.

Also we have our resident radio networking guru Tim on the list who
probably has some ideas. :)

From dave.bucklin at gmail.com  Sun Oct 21 03:02:51 2018
From: dave.bucklin at gmail.com (Dave Bucklin)
Date: Sat, 20 Oct 2018 22:02:51 -0500
Subject: OpenBSD Upgrade GtG?
Message-ID: <B6EAA746-449F-41C6-B1BA-69A72EE42784@gmail.com>

So, the VPS could be upgraded to OpenBSD 6.4. Seems like a good excuse to get together and get it done.

Pick dates that work for you:
https://doodle.com/poll/uati2i8dcddpzbqr

OpenBSD upgrade process:
https://www.romanzolotarev.com/openbsd/vultr-upgrade.html


From kurtis at riseup.net  Sun Oct 21 06:19:00 2018
From: kurtis at riseup.net (kurtis)
Date: Sun, 21 Oct 2018 06:19:00 +0000
Subject: Usenet over ham radio?
In-Reply-To: <20181020155911.GA5778@mysterium.local>
References: <20181020155911.GA5778@mysterium.local>
Message-ID: <e907e2f3-954c-693d-2ab3-3608c0aaa178@riseup.net>

Hi Joe,

Joe Nelson:
> Anyone know if it's feasible to execute UUCP over ham radio to run a
> radio-based usenet? The list we have now is certainly more convenient,
> but the challenge is intriguing. It's enticing to think a local group
> can carry on communication without ISPs or cell phone companies.

This sounds similar to the WOM project:
https://wom.community/

Here's an interview he did about it:
https://www.youtube.com/watch?v=g1-NZvZPdds

Here's a presentation he did with slides:
https://livestream.com/internetsociety/radnets17/videos/164804778
https://ossguy.com/talks/20171022_radical_networks/

In Solidarity,
kurtis

From joe at begriffs.com  Sun Oct 21 06:43:43 2018
From: joe at begriffs.com (Joe Nelson)
Date: Sun, 21 Oct 2018 01:43:43 -0500
Subject: OpenBSD Upgrade GtG?
In-Reply-To: <B6EAA746-449F-41C6-B1BA-69A72EE42784@gmail.com>
References: <B6EAA746-449F-41C6-B1BA-69A72EE42784@gmail.com>
Message-ID: <20181021064343.GB46826@begriffs.com>

> So, the VPS could be upgraded to OpenBSD 6.4. Seems like a good excuse
> to get together and get it done.

Yeah, sounds like fun. Could get an IRC bouncer and voice chat on the
machine too while we're at it.

> Pick dates that work for you:
> https://doodle.com/poll/uati2i8dcddpzbqr

As I commented on the doodle form, Wednesdays are the open-house day at
the hack factory, so we can get in without anyone needing to pay for
membership. This Weds works for me, but not next (Halloween). The one
after that is fine too.

Maybe we could get some cheap food at China Wok (2800 27th Ave S) that
night too, go out for a group dinner.

> OpenBSD upgrade process:
> https://www.romanzolotarev.com/openbsd/vultr-upgrade.html

Nice, looks pretty straightforward.

From dave.bucklin at gmail.com  Sun Oct 21 15:01:54 2018
From: dave.bucklin at gmail.com (Dave Bucklin)
Date: Sun, 21 Oct 2018 10:01:54 -0500
Subject: Usenet over ham radio?
In-Reply-To: <20181020155911.GA5778@mysterium.local>
References: <20181020155911.GA5778@mysterium.local>
Message-ID: <20181021150154.t4ycconlzw5x7z2v@19a6.tech>

On Sat, Oct 20, 2018 at 10:59:11AM -0500, Joe Nelson wrote:
> Anyone know if it's feasible to execute UUCP over ham radio to run a
> radio-based usenet? The list we have now is certainly more convenient,
> but the challenge is intriguing. It's enticing to think a local group
> can carry on communication without ISPs or cell phone companies.
> 
> Also we have our resident radio networking guru Tim on the list who
> probably has some ideas. :)

I've heard of this being done. It wasn't UUCP per se; I think it might
have been an XMODEM download or something. UUCP has been noted as being
particularly flexible in terms of the transport.

From nompelis at nobelware.com  Mon Oct 22 15:54:12 2018
From: nompelis at nobelware.com (Ioannis Nompelis)
Date: Mon, 22 Oct 2018 15:54:12 +0000
Subject: OpenBSD Upgrade GtG?
In-Reply-To: <20181021064343.GB46826@begriffs.com>
References: <B6EAA746-449F-41C6-B1BA-69A72EE42784@gmail.com>
 <20181021064343.GB46826@begriffs.com>
Message-ID: <20181022155412.GB4164@nobelware.com>

I think I can make the Wednesday work. I also have a $50 set aside for the
VPS/domain, on which I do not want to grow mold!

We can run the chat server on my VPS; I tested it and it executes.


From joe at begriffs.com  Wed Oct 24 00:07:13 2018
From: joe at begriffs.com (Joe Nelson)
Date: Tue, 23 Oct 2018 19:07:13 -0500
Subject: OpenBSD Upgrade GtG?
In-Reply-To: <20181022155412.GB4164@nobelware.com>
References: <B6EAA746-449F-41C6-B1BA-69A72EE42784@gmail.com>
 <20181021064343.GB46826@begriffs.com>
 <20181022155412.GB4164@nobelware.com>
Message-ID: <20181024000713.GA69549@begriffs.com>

Ioannis Nompelis wrote:
> I think I can make the Wednesday work. I also have a $50 set aside for
> the VPS/domain, on which I do not want to grow mold!

Should we meet tomorrow the 24th at 7 for the open house?
http://www.tcmaker.org/blog/hack-factory/

I can bring some money to contribute toward server costs too.

> We can run the chat server on my VPS; I tested it and it executes.

Cool, I'm curious to try it out.

From dave.bucklin at gmail.com  Wed Oct 24 02:16:24 2018
From: dave.bucklin at gmail.com (Dave Bucklin)
Date: Tue, 23 Oct 2018 21:16:24 -0500
Subject: OpenBSD Upgrade GtG?
In-Reply-To: <20181024000713.GA69549@begriffs.com>
References: <B6EAA746-449F-41C6-B1BA-69A72EE42784@gmail.com>
 <20181021064343.GB46826@begriffs.com>
 <20181022155412.GB4164@nobelware.com>
 <20181024000713.GA69549@begriffs.com>
Message-ID: <20181024021624.zxfekr5tedizjb52@19a6.tech>

On Tue, Oct 23, 2018 at 07:07:13PM -0500, Joe Nelson wrote:
> Ioannis Nompelis wrote:
> > I think I can make the Wednesday work. I also have a $50 set aside for
> > the VPS/domain, on which I do not want to grow mold!
> 
> Should we meet tomorrow the 24th at 7 for the open house?
> http://www.tcmaker.org/blog/hack-factory/
> 
> I can bring some money to contribute toward server costs too.

I could, in theory, but next week (Halloween!) or the week after works
better.

From joe at begriffs.com  Wed Oct 24 02:40:42 2018
From: joe at begriffs.com (Joe Nelson)
Date: Tue, 23 Oct 2018 21:40:42 -0500
Subject: OpenBSD Upgrade GtG?
In-Reply-To: <20181024021624.zxfekr5tedizjb52@19a6.tech>
References: <B6EAA746-449F-41C6-B1BA-69A72EE42784@gmail.com>
 <20181021064343.GB46826@begriffs.com>
 <20181022155412.GB4164@nobelware.com>
 <20181024000713.GA69549@begriffs.com>
 <20181024021624.zxfekr5tedizjb52@19a6.tech>
Message-ID: <20181024024042.GA79416@begriffs.com>

> > Should we meet tomorrow the 24th at 7 for the open house?

> I could, in theory, but next week (Halloween!) or the week after works
> better.

K, let's do it later then. I'm busy on Halloween though.

From nompelis at nobelware.com  Wed Oct 24 03:22:52 2018
From: nompelis at nobelware.com (Ioannis Nompelis)
Date: Tue, 23 Oct 2018 22:22:52 -0500
Subject: OpenBSD Upgrade GtG?
In-Reply-To: <20181024024042.GA79416@begriffs.com>
References: <B6EAA746-449F-41C6-B1BA-69A72EE42784@gmail.com>
 <20181021064343.GB46826@begriffs.com>
 <20181022155412.GB4164@nobelware.com>
 <20181024000713.GA69549@begriffs.com>
 <20181024021624.zxfekr5tedizjb52@19a6.tech>
 <20181024024042.GA79416@begriffs.com>
Message-ID: <20181024032252.GA1815@nobelware.com>

> 
> > > Should we meet tomorrow the 24th at 7 for the open house?
> 
> > I could, in theory, but next week (Halloween!) or the week after works
> > better.
> 
> K, let's do it later then. I'm busy on Halloween though.

Joe, should I still meet you at the Hack Factory for the open house? I can
make tomorrow work anyway, we might as well do the open house. Your call.


From samuel.stuewe at gmail.com  Wed Oct 24 14:35:13 2018
From: samuel.stuewe at gmail.com (Sam Stuewe)
Date: Wed, 24 Oct 2018 09:35:13 -0500
Subject: OpenBSD Upgrade GtG?
In-Reply-To: <20181024032252.GA1815@nobelware.com>
References: <B6EAA746-449F-41C6-B1BA-69A72EE42784@gmail.com>
 <20181021064343.GB46826@begriffs.com> <20181022155412.GB4164@nobelware.com>
 <20181024000713.GA69549@begriffs.com>
 <20181024021624.zxfekr5tedizjb52@19a6.tech>
 <20181024024042.GA79416@begriffs.com> <20181024032252.GA1815@nobelware.com>
Message-ID: <CABSwF4y7_OxoDOGBRG727PCwvVXOUhxOhDaOaHxw48CAXSPSMw@mail.gmail.com>

Let's try that again.

Ioannis, pardon the off-list reply and the top-post.

Message was as follows:

The week following Halloween is much more possible for me. Life is
somewhat hectic (house renovations) lately.

All the best,

-Sam

From joe at begriffs.com  Wed Oct 24 15:06:47 2018
From: joe at begriffs.com (Joe Nelson)
Date: Wed, 24 Oct 2018 10:06:47 -0500
Subject: OpenBSD Upgrade GtG?
In-Reply-To: <20181024032252.GA1815@nobelware.com>
References: <B6EAA746-449F-41C6-B1BA-69A72EE42784@gmail.com>
 <20181021064343.GB46826@begriffs.com>
 <20181022155412.GB4164@nobelware.com>
 <20181024000713.GA69549@begriffs.com>
 <20181024021624.zxfekr5tedizjb52@19a6.tech>
 <20181024024042.GA79416@begriffs.com>
 <20181024032252.GA1815@nobelware.com>
Message-ID: <20181024150647.GA11056@begriffs.com>

Ioannis Nompelis wrote:
> Joe, should I still meet you at the Hack Factory for the open
> house? I can make tomorrow work anyway, we might as well do the open
> house. Your call.

Actually how about meeting at Fueled Collective tonight on the 4th floor
of the Grain Exchange Building, 400 4th St S, Minneapolis? (Pretty easy
to remember the address: 4x4x4.)

I'll be there today and can stay for an evening meetup. Maybe it's
closer to you anyways?

To get in, take the elevators to the 4th floor, and then to your right
will be an entrance to what looks like a huge ballroom, and that's the
place.

Anyone who can make it is welcome. Should we say 6:00?

From nompelis at nobelware.com  Wed Oct 24 16:09:47 2018
From: nompelis at nobelware.com (Ioannis Nompelis)
Date: Wed, 24 Oct 2018 16:09:47 +0000
Subject: OpenBSD Upgrade GtG?
In-Reply-To: <20181024150647.GA11056@begriffs.com>
References: <B6EAA746-449F-41C6-B1BA-69A72EE42784@gmail.com>
 <20181021064343.GB46826@begriffs.com> <20181022155412.GB4164@nobelware.com>
 <20181024000713.GA69549@begriffs.com>
 <20181024021624.zxfekr5tedizjb52@19a6.tech>
 <20181024024042.GA79416@begriffs.com> <20181024032252.GA1815@nobelware.com>
 <20181024150647.GA11056@begriffs.com>
Message-ID: <20181024160947.GA12968@nobelware.com>

> Actually how about meeting at Fueled Collective tonight on the 4th floor
> of the Grain Exchange Building, 400 4th St S, Minneapolis? (Pretty easy
> to remember the address: 4x4x4.)

> 
> Anyone who can make it is welcome. Should we say 6:00?

OK. I will meet you there.


Also, sam just ran into an artifact of the listserv not over-writing the To:
field.


From samuel.stuewe at gmail.com  Wed Oct 24 16:12:19 2018
From: samuel.stuewe at gmail.com (Sam Stuewe)
Date: Wed, 24 Oct 2018 11:12:19 -0500
Subject: OpenBSD Upgrade GtG?
In-Reply-To: <20181024160947.GA12968@nobelware.com>
References: <B6EAA746-449F-41C6-B1BA-69A72EE42784@gmail.com>
 <20181021064343.GB46826@begriffs.com> <20181022155412.GB4164@nobelware.com>
 <20181024000713.GA69549@begriffs.com>
 <20181024021624.zxfekr5tedizjb52@19a6.tech>
 <20181024024042.GA79416@begriffs.com> <20181024032252.GA1815@nobelware.com>
 <20181024150647.GA11056@begriffs.com> <20181024160947.GA12968@nobelware.com>
Message-ID: <CABSwF4xsbm=oKnp6+3OQOXZVF8wzd1me1cLdeEL_e4q+ejc1=A@mail.gmail.com>

On 10/24/18, Ioannis Nompelis <nompelis at nobelware.com> wrote:
> Also, sam just ran into an artifact of the listserv not over-writing the
> To:
> field.

I suppose. I'm unfortunately, still bound to the Gmail web-client. I
have tried so hard to switch so many times and have mostly given up
for now. I accidentally hit "Reply" rather than "Reply-All".

All the best,

-Sam

From joe at begriffs.com  Wed Oct 24 16:25:49 2018
From: joe at begriffs.com (Joe Nelson)
Date: Wed, 24 Oct 2018 11:25:49 -0500
Subject: OpenBSD Upgrade GtG?
In-Reply-To: <CABSwF4xsbm=oKnp6+3OQOXZVF8wzd1me1cLdeEL_e4q+ejc1=A@mail.gmail.com>
References: <B6EAA746-449F-41C6-B1BA-69A72EE42784@gmail.com>
 <20181021064343.GB46826@begriffs.com>
 <20181022155412.GB4164@nobelware.com>
 <20181024000713.GA69549@begriffs.com>
 <20181024021624.zxfekr5tedizjb52@19a6.tech>
 <20181024024042.GA79416@begriffs.com>
 <20181024032252.GA1815@nobelware.com>
 <20181024150647.GA11056@begriffs.com>
 <20181024160947.GA12968@nobelware.com>
 <CABSwF4xsbm=oKnp6+3OQOXZVF8wzd1me1cLdeEL_e4q+ejc1=A@mail.gmail.com>
Message-ID: <20181024162549.GA15643@begriffs.com>

Sam Stuewe wrote:
> I'm unfortunately, still bound to the Gmail web-client. I have tried
> so hard to switch so many times and have mostly given up for now.

I can help you get set up with Mutt if you like. You can hook it up to
your existing Gmail account and then choose to use mutt or the regular
web client depending on the task. They can coexist. It's a nice way to
ease into it.

From samuel.stuewe at gmail.com  Wed Oct 24 16:31:05 2018
From: samuel.stuewe at gmail.com (Sam Stuewe)
Date: Wed, 24 Oct 2018 11:31:05 -0500
Subject: OpenBSD Upgrade GtG?
In-Reply-To: <20181024162549.GA15643@begriffs.com>
References: <B6EAA746-449F-41C6-B1BA-69A72EE42784@gmail.com>
 <20181021064343.GB46826@begriffs.com> <20181022155412.GB4164@nobelware.com>
 <20181024000713.GA69549@begriffs.com>
 <20181024021624.zxfekr5tedizjb52@19a6.tech>
 <20181024024042.GA79416@begriffs.com> <20181024032252.GA1815@nobelware.com>
 <20181024150647.GA11056@begriffs.com> <20181024160947.GA12968@nobelware.com>
 <CABSwF4xsbm=oKnp6+3OQOXZVF8wzd1me1cLdeEL_e4q+ejc1=A@mail.gmail.com>
 <20181024162549.GA15643@begriffs.com>
Message-ID: <CABSwF4w9MiKCHGebyFnQWkw8=E-27TJF1Z1urScRfwseY-1LRw@mail.gmail.com>

On 10/24/18, Joe Nelson <joe at begriffs.com> wrote:
> I can help you get set up with Mutt if you like. You can hook it up to
> your existing Gmail account and then choose to use mutt or the regular
> web client depending on the task. They can coexist. It's a nice way to
> ease into it.

I have successfully switched to mutt for gmail in the past, and I love
the idea of going back to it. It's mostly inertia that stops me.
Though there are enough other things that I am working on at the
moment, taking the time for that switch isn't high on the list. When
we manage to meet up, I may take you up on that offer though! :)

All the best,

-Sam

From nompelis at nobelware.com  Wed Oct 24 16:31:33 2018
From: nompelis at nobelware.com (Ioannis Nompelis)
Date: Wed, 24 Oct 2018 16:31:33 +0000
Subject: OpenBSD Upgrade GtG?
In-Reply-To: <20181024162549.GA15643@begriffs.com>
References: <20181021064343.GB46826@begriffs.com>
 <20181022155412.GB4164@nobelware.com> <20181024000713.GA69549@begriffs.com>
 <20181024021624.zxfekr5tedizjb52@19a6.tech>
 <20181024024042.GA79416@begriffs.com> <20181024032252.GA1815@nobelware.com>
 <20181024150647.GA11056@begriffs.com> <20181024160947.GA12968@nobelware.com>
 <CABSwF4xsbm=oKnp6+3OQOXZVF8wzd1me1cLdeEL_e4q+ejc1=A@mail.gmail.com>
 <20181024162549.GA15643@begriffs.com>
Message-ID: <20181024163133.GA13786@nobelware.com>

Mutt. I am writing this from a terminal with Mutt.

But I will warn you against using both. I do not think they coexist well. Mutt
will just use IMAP, which is great, and which Gmail just "fakes" in general.
I found, through my email @umn.edu, that IMAP flags set by Mutt and those set
from Gmail itself do not play well together.


From dave.bucklin at gmail.com  Wed Oct 24 22:36:40 2018
From: dave.bucklin at gmail.com (Dave Bucklin)
Date: Wed, 24 Oct 2018 17:36:40 -0500
Subject: OpenBSD Upgrade GtG?
In-Reply-To: <CABSwF4y7_OxoDOGBRG727PCwvVXOUhxOhDaOaHxw48CAXSPSMw@mail.gmail.com>
References: <B6EAA746-449F-41C6-B1BA-69A72EE42784@gmail.com>
 <20181021064343.GB46826@begriffs.com>
 <20181022155412.GB4164@nobelware.com>
 <20181024000713.GA69549@begriffs.com>
 <20181024021624.zxfekr5tedizjb52@19a6.tech>
 <20181024024042.GA79416@begriffs.com>
 <20181024032252.GA1815@nobelware.com>
 <CABSwF4y7_OxoDOGBRG727PCwvVXOUhxOhDaOaHxw48CAXSPSMw@mail.gmail.com>
Message-ID: <20181024223640.q6vhh27dkyltka74@19a6.tech>

On Wed, Oct 24, 2018 at 09:35:13AM -0500, Sam Stuewe wrote:
> The week following Halloween is much more possible for me. Life is
> somewhat hectic (house renovations) lately.

Shall we get together at the Hack Factory open house at 7pm on the 7th
of November, then?

From nompelis at nobelware.com  Wed Oct 24 23:39:38 2018
From: nompelis at nobelware.com (Ioannis Nompelis)
Date: Wed, 24 Oct 2018 23:39:38 +0000
Subject: OpenBSD Upgrade GtG?
In-Reply-To: <20181024223640.q6vhh27dkyltka74@19a6.tech>
References: <B6EAA746-449F-41C6-B1BA-69A72EE42784@gmail.com>
 <20181021064343.GB46826@begriffs.com> <20181022155412.GB4164@nobelware.com>
 <20181024000713.GA69549@begriffs.com>
 <20181024021624.zxfekr5tedizjb52@19a6.tech>
 <20181024024042.GA79416@begriffs.com> <20181024032252.GA1815@nobelware.com>
 <CABSwF4y7_OxoDOGBRG727PCwvVXOUhxOhDaOaHxw48CAXSPSMw@mail.gmail.com>
 <20181024223640.q6vhh27dkyltka74@19a6.tech>
Message-ID: <20181024233938.GA29517@nobelware.com>

> Shall we get together at the Hack Factory open house at 7pm on the 7th
> of November, then?


Joe, who is sitting next to me, says "7/7 sounds lucky." I guess we are
tentatively ON for this gathering.

Look for my next message.


From nompelis at nobelware.com  Wed Oct 24 23:42:03 2018
From: nompelis at nobelware.com (Ioannis Nompelis)
Date: Wed, 24 Oct 2018 23:42:03 +0000
Subject: Murmur server is temporarily up
Message-ID: <20181024234203.GB29517@nobelware.com>

Joe and I just got to test the Murmu server. To make things simple, I took
the pre-compiled binary (do not hate me...) and executed it on my VPS. It
is using my SSL certificate, and we have a password for the server. Try it
out:

host: nobelware.com
port: 64738
password: the domain name that we have without the ".cc" part

It works. I also got an app for my iphone, which worked fine, although
it does throw some ads.

This was easier than I thought.
IN


From joe at begriffs.com  Thu Oct 25 03:42:19 2018
From: joe at begriffs.com (Joe Nelson)
Date: Wed, 24 Oct 2018 22:42:19 -0500
Subject: Experimental mailman change
Message-ID: <20181025034219.GA47830@begriffs.com>

I disabled this setting:
  "Filter out duplicate messages to list members (if possible)"

Mailman was removing people from the CC in its re-broadcasted messages
when those people are already subscribed to the list. It does this
to prevent duplicate messages. The modified CC was making some of my
messages fail their DKIM check.

I'm wondering whether mail clients will be smart enough to see a
matching Message-Id and not the messages twice. If the configuration
change ends up sucking I can put it back the way it was. Let me know if
you start seeing dups.

From joe at begriffs.com  Thu Oct 25 05:38:13 2018
From: joe at begriffs.com (Joe Nelson)
Date: Thu, 25 Oct 2018 00:38:13 -0500
Subject: Forming an IRC network
Message-ID: <20181025053813.GA48953@begriffs.com>

Hey want to try creating a multi-server IRC network? We've got an IRC
server running on frostbyte.cc, on nobelware.com, and now I'm running
one too.

host: talk.begriffs.com
port: 6697
SSL : required
pass: hchn

It provides user IP cloaking for a little privacy. The server password
is there to keep out random casual internet connections that notice an
open port. The server is also using a chroot and dedicated daemon user
for security.

The frostbyte server doesn't yet support SSL because it doesn't have a
certificate. The nobelware server's SSL setup is in progress. Ioannis
and I were experimenting with it tonight.

Anyway, I thought it would be fun to tie these servers
together into a single network, just to see how that works.
https://tools.ietf.org/html/rfc2810#section-3

From nompelis at nobelware.com  Thu Oct 25 12:48:23 2018
From: nompelis at nobelware.com (Ioannis Nompelis)
Date: Thu, 25 Oct 2018 07:48:23 -0500
Subject: Forming an IRC network
In-Reply-To: <20181025053813.GA48953@begriffs.com>
References: <20181025053813.GA48953@begriffs.com>
Message-ID: <20181025124823.GA1529@nobelware.com>

> 
> Hey want to try creating a multi-server IRC network? We've got an IRC
> server running on frostbyte.cc, on nobelware.com, and now I'm running
> one too.
> 
> host: talk.begriffs.com
> port: 6697
> SSL : required
> pass: hchn
> 

OK. I am getting it ready. You will need to know the following:

        # Own password for the connection. This password has to be configured
        # as "PeerPassword" on the other server.
        ;MyPassword = MySecret
        MyPassword = frostbyte

        # Foreign password for this connection. This password has to be
        # configured as "MyPassword" on the other server.
        ;PeerPassword = PeerSecret
        PeerPassword = hchn


> It provides user IP cloaking for a little privacy. The server password
> is there to keep out random casual internet connections that notice an
> open port. The server is also using a chroot and dedicated daemon user
> for security.
> 

I did not do cloaking; it is not needed among us. But the password is the
same as the Murmur server's "frostbyte".

Now we are plain-texting on the list archive our passwords... Oh well.

> The frostbyte server doesn't yet support SSL because it doesn't have a
> certificate. The nobelware server's SSL setup is in progress. Ioannis
> and I were experimenting with it tonight.

Nobelware has a certificate and I do not know why SSL connections by users
were being dropped.

I will -HUP the nobelware server and see if it connects to yours.


From nompelis at nobelware.com  Thu Oct 25 12:57:04 2018
From: nompelis at nobelware.com (Ioannis Nompelis)
Date: Thu, 25 Oct 2018 07:57:04 -0500
Subject: Forming an IRC network
In-Reply-To: <20181025124823.GA1529@nobelware.com>
References: <20181025053813.GA48953@begriffs.com>
 <20181025124823.GA1529@nobelware.com>
Message-ID: <20181025125704.GB1529@nobelware.com>

Joe, I tried connecting my ngIRC server to yours and there is an SSL error.

[2911:6   75] Establishing connection for "talk.begriffs.com" to "talk.begriffs.com:6697" (45.76.26.49), socket 7 ...
[2911:3   75] SSL protocol error: SSL_connect (error:140740B5:SSL routines:SSL23_CLIENT_HELLO:no ciphers available)
[2911:3   75] SSL connection on socket 7 failed!

Tell me which cipher you are using. I may have to upgrade my 10 year old SSL
libraries...


From dave.bucklin at gmail.com  Thu Oct 25 13:43:24 2018
From: dave.bucklin at gmail.com (Dave Bucklin)
Date: Thu, 25 Oct 2018 08:43:24 -0500
Subject: Forming an IRC network
In-Reply-To: <20181025125704.GB1529@nobelware.com>
References: <20181025053813.GA48953@begriffs.com>
 <20181025124823.GA1529@nobelware.com>
 <20181025125704.GB1529@nobelware.com>
Message-ID: <20181025134324.2rfenrjkp3zgb2yn@19a6.tech>

On Thu, Oct 25, 2018 at 07:57:04AM -0500, Ioannis Nompelis wrote:
> Joe, I tried connecting my ngIRC server to yours and there is an SSL error.
> 
> [2911:6   75] Establishing connection for "talk.begriffs.com" to "talk.begriffs.com:6697" (45.76.26.49), socket 7 ...
> [2911:3   75] SSL protocol error: SSL_connect (error:140740B5:SSL routines:SSL23_CLIENT_HELLO:no ciphers available)
> [2911:3   75] SSL connection on socket 7 failed!
> 
> Tell me which cipher you are using. I may have to upgrade my 10 year old SSL
> libraries...
> 
In the past, when I've had this problem, I had to manually set the key
size to 1024. Maybe that will work for you in this case.

From nompelis at nobelware.com  Thu Oct 25 14:04:24 2018
From: nompelis at nobelware.com (Ioannis Nompelis)
Date: Thu, 25 Oct 2018 14:04:24 +0000
Subject: Forming an IRC network
In-Reply-To: <20181025134324.2rfenrjkp3zgb2yn@19a6.tech>
References: <20181025053813.GA48953@begriffs.com>
 <20181025124823.GA1529@nobelware.com> <20181025125704.GB1529@nobelware.com>
 <20181025134324.2rfenrjkp3zgb2yn@19a6.tech>
Message-ID: <20181025140424.GA5282@nobelware.com>

Dave, when you get the chance send info to Joe and I. We will make it work.

This is just a learning activity for me, having been an IRC user in the 90s
but never having worked on the admin side of it. I learned a lot from an hour
of work last night.

I think the IRC is of little value compared to the Murmur thing. I will make
the case that sitting on the IRC channel and watching what everybody is talking
about is a full-attention activity. For group or paired programming we are much
better off with Murmur. I can be typing on my terminal, or a common shared X
session via VNC, etc, and be listening and talking on Murmur.

I encourage you guys to connect to it and test it out.

Like Joe said, we had less latency going from Joe's Macbook to my VPS in Dallas
Texas and back to the same room to my phone over the WiFi than we had on a
regular cellular network phonecall!


From joe at begriffs.com  Thu Oct 25 14:52:24 2018
From: joe at begriffs.com (Joe Nelson)
Date: Thu, 25 Oct 2018 09:52:24 -0500
Subject: Experimental mailman change
In-Reply-To: <9C959320E4DE136C.5b3116d5-f6ab-4fcf-ab0a-7d5edfe80a1c@mail.outlook.com>
References: <20181025034219.GA47830@begriffs.com>
 <9C959320E4DE136C.5b3116d5-f6ab-4fcf-ab0a-7d5edfe80a1c@mail.outlook.com>
Message-ID: <20181025145224.GA81732@begriffs.com>

Robbie Herb wrote:
> If you keep getting dupes, could you set it to just not send messages
> to people already on the CC list?

That would be ideal. I had assumed that this is what the "Filter out
duplicate messages to list members" setting did, which is why I had it
enabled initially. We'll see how it behaves. Let me know if you get this
message twice for instance...

From joe at begriffs.com  Thu Oct 25 16:02:49 2018
From: joe at begriffs.com (Joe Nelson)
Date: Thu, 25 Oct 2018 11:02:49 -0500
Subject: Forming an IRC network
In-Reply-To: <20181025140424.GA5282@nobelware.com>
References: <20181025053813.GA48953@begriffs.com>
 <20181025124823.GA1529@nobelware.com>
 <20181025125704.GB1529@nobelware.com>
 <20181025134324.2rfenrjkp3zgb2yn@19a6.tech>
 <20181025140424.GA5282@nobelware.com>
Message-ID: <20181025160249.GC81732@begriffs.com>

Ioannis Nompelis wrote:
> I think the IRC is of little value compared to the Murmur thing. [...]
> For group or paired programming we are much better off with Murmur.

You've got a point there. Maybe we don't even need IRC? Maybe email
is enough to coordinate either in-person meetings or Mumble+SSH
collaboration.

From joe at begriffs.com  Fri Oct 26 01:22:10 2018
From: joe at begriffs.com (Joe Nelson)
Date: Thu, 25 Oct 2018 20:22:10 -0500
Subject: Forming an IRC network
In-Reply-To: <20181025125704.GB1529@nobelware.com>
References: <20181025053813.GA48953@begriffs.com>
 <20181025124823.GA1529@nobelware.com>
 <20181025125704.GB1529@nobelware.com>
Message-ID: <20181026012210.GA21211@begriffs.com>

Ioannis Nompelis wrote:
> Joe, I tried connecting my ngIRC server to yours and there is an SSL
> error.
> ...
> Tell me which cipher you are using. I may have to upgrade my 10 year
> old SSL libraries...

When my local IRC client (LimeChat 2.42) connects to my IRC server, I
see a message in the main channel saying:

Connected using Cipher ECDHE-RSA-AES256-GCM-SHA384 TLSv1.2
                       Kx=ECDH Au=RSA Enc=AESGCM(256) Mac=AEAD

I used acme-client [0] to create my SSL key and letsencrypt [1] to sign
it, if that gives you any other clues. I know very little about how
these programs negotiate crypto during connections or the properties of
the SSL key.

Also here's what Chrome says about the TLS connection to my site (which
uses the same key):

Chrome:
> The connection to this site is encrypted and authenticated using
> TLS 1.2 (a strong protocol), ECDHE_RSA with X25519 (a strong key
> exchange), and AES_256_GCM (a strong cipher).

[0]: https://man.openbsd.org/acme-client
[1]: https://letsencrypt.org/

From joe at begriffs.com  Fri Oct 26 01:38:50 2018
From: joe at begriffs.com (Joe Nelson)
Date: Thu, 25 Oct 2018 20:38:50 -0500
Subject: Forming an IRC network
In-Reply-To: <20181026012210.GA21211@begriffs.com>
References: <20181025053813.GA48953@begriffs.com>
 <20181025124823.GA1529@nobelware.com>
 <20181025125704.GB1529@nobelware.com>
 <20181026012210.GA21211@begriffs.com>
Message-ID: <20181026013850.GA22431@begriffs.com>

Joe Nelson wrote:
> Chrome:
> > The connection to this site is encrypted and authenticated using
> > TLS 1.2 (a strong protocol), ECDHE_RSA with X25519 (a strong key
> > exchange), and AES_256_GCM (a strong cipher).

Just asked Chrome about the nobelware server and it says:

> The connection to this site uses TLS 1.0 (an obsolete protocol),
> RSA (an obsolete key exchange), and AES_128_CBC with HMAC-SHA1 (an
> obsolete cipher).

How does crypto even change so quickly in just a few years?  You
could try creating a fresh key and seeing if our servers are able to
communicate.

From nompelis at nobelware.com  Fri Oct 26 13:07:27 2018
From: nompelis at nobelware.com (Ioannis Nompelis)
Date: Fri, 26 Oct 2018 13:07:27 +0000
Subject: Forming an IRC network
In-Reply-To: <20181026013850.GA22431@begriffs.com>
References: <20181025053813.GA48953@begriffs.com>
 <20181025124823.GA1529@nobelware.com> <20181025125704.GB1529@nobelware.com>
 <20181026012210.GA21211@begriffs.com> <20181026013850.GA22431@begriffs.com>
Message-ID: <20181026130727.GB23373@nobelware.com>

I need to upgrade Appache's SSL module and my OpenSSL.


From salo at saloits.com  Sat Oct 27 18:05:31 2018
From: salo at saloits.com (Timothy J. Salo)
Date: Sat, 27 Oct 2018 13:05:31 -0500
Subject: Usenet over ham radio?
In-Reply-To: <20181020155911.GA5778@mysterium.local>
References: <20181020155911.GA5778@mysterium.local>
Message-ID: <c724d438-e78d-b550-8e8d-a6226296a9ea@saloits.com>

On 10/20/2018 10:59 AM, Joe Nelson wrote:
> Anyone know if it's feasible to execute UUCP over ham radio to run a
> radio-based usenet? The list we have now is certainly more convenient,
> but the challenge is intriguing. It's enticing to think a local group
> can carry on communication without ISPs or cell phone companies.

First, if you are interested in this you should look at Packet BBS.
This is a BBS that runs over the amateur radio packet radio
protocol (AX.25).  A lot of TNCs (Terminal Node Controllers,
roughly a commercial AX.25 modem) support Packet BBS.  However,
commercial TNCs are largely a thing of the past: lots of
software is available to decode AX.25 signals.

Packet BBS probably hasn't changed for decades.  There is a little
bit of packet activity in the area.  I don't know if there are any BBS
attached to the state packet network. Doug Reed, if he is still
around, would know.

Every discussion of connecting Packet BBS and Usenet leads to a
discussion about FCC appropriate use regulations for amateur
radio, specifically how appropriate use can be enforced when
Usenet is gatewayed to amateur radio (and who is responsible
for inappropriate use).  The result is that I don't think there
is much interconnection between Usenet and amateur radio packet.

Depending on what spectrum you use, amateur radio packet radio
can have pretty limited bandwidth.  On 2 meters (144 - 148 Mz),
it is easy to run 1,200 bps.  You could run 9,600 bps on 2 meters,
but you need to tap into your radio on the other side of the
audio section.  Some radios provide this connection; on other
radios you have to go make that connection yourself.

As you move up in frequency, you could run 56K bps.  I think
that you can run up to 56K bps on both 220 MHz and 440 MHz.
But, equipment starts to get expensive, and there aren't a
lot of other stations you can communicate with.  I think some
of the new software-defined radio (SDR) transceivers can
support these frequencies and data rates.  But, they are expensive.

I think that there are a lot of good reasons that the Internet
has (apparently) pretty much killed packet radio.

Having said all that, I think that people interested in data
over amateur radio really ought to look at more modern protocols.
I would start with something like Thread (sort of Nest's variant
of ZigBee) and some more modern physical- and link-layer protocols.
A long time ago, I wrote a paper on this topic, but never got
around to implementing anything:

<https://www.tapr.org/pdf/DCC2013-NIPARnets-AB0DO.pdf>
<https://www.saloits.com/papers/AMSAT2013.pdf>

-tjs

From ericshane at eradman.com  Mon Oct 29 17:11:53 2018
From: ericshane at eradman.com (Eric Radman)
Date: Mon, 29 Oct 2018 12:11:53 -0500
Subject: Comparing oh to other Unix shells
Message-ID: <20181029171153.GB21577@vm2.eradman.com>

Hello!

This year at BSDCAN Michael MacInnis gave an excellent talk on a new
shell he developed called "Oh"

https://www.bsdcan.org/2018/schedule/events/931.en.html

Probably most people would concur that ksh/bash/csh/tcsh are poor
programming languages (word-splitting! *sigh*), but personally I have
not spent the time to learn one of the newer shells.  Instead I spend a
good deal of time eash week carefully crafting scripts with #!/bin/bash

Why not?  The obvious explanation is that we tend to use BASH because
sh/ksh variants are implemented and available everywhere.  Is there
another reason?  What would it take for a shell like Fish or Oh become
more widely used?

-- 
Eric Radman  |  http://eradman.com

From joe at begriffs.com  Tue Oct 30 00:29:34 2018
From: joe at begriffs.com (Joe Nelson)
Date: Mon, 29 Oct 2018 19:29:34 -0500
Subject: Comparing oh to other Unix shells
In-Reply-To: <20181029171153.GB21577@vm2.eradman.com>
References: <20181029171153.GB21577@vm2.eradman.com>
Message-ID: <20181030002934.GA31915@begriffs.com>

Eric Radman wrote:
> This year at BSDCAN Michael MacInnis gave an excellent talk on a new
> shell he developed called "Oh"

Nice video. I like how the author takes the time to survey existing
shells and determine what he believes is lacking, rather than simply
wanting to rewrite things in a language du jour. I also like how he
points out that shells written to exactly imitate certain existing
programming languages (like scheme) are often just cumbersome when used
interactively.

> Why not?  The obvious explanation is that we tend to use BASH because
> sh/ksh variants are implemented and available everywhere.  Is there
> another reason?  What would it take for a shell like Fish or Oh become
> more widely used?

My biggest reaction to the video is that maybe it's unwise to try
to merge the interactive use case with a full-featured programming
language. We might be better served by using an established language
like perl or python for big tasks, and use ordinary shells for more
interactive or simpler things.

Oh has a lot of features and constructs -- starting at about the
"Environments" slide -- that it feels like it wants to wade into full
programming language territory. Once treated as a full language, is
there perhaps less to recommend it over better known languages? The
author calls the UNIX shell "strange" but honestly Oh is a little
strange too.

I'm curious to learn more about the portable subset of shell
scripting. I'd like to carefully study what POSIX guarantees that "sh"
can do across practically any system:
http://pubs.opengroup.org/onlinepubs/9699919799/utilities/V3_chap02.html

My shell scripting has hitherto been a mess of things I've copied and
pasted over time, and is probably full of bashisms. I want to learn to
write polished and portable scripts instead. I wonder whether there is a
tool to lint for POSIX compatibility. Maybe shellcheck can do it.

As a final digression, does anyone know where I can buy a printed copy
of the Open Group Base Specifications? Their online store doesn't have
it. The closest thing I can find is a PDF for what they call "C181:"
https://publications.opengroup.org/c181
It's unclear how much it costs to become an individual member to
download the PDF, or whether individuals can even become members. Also,
assuming I'm able to get the PDF, can anyone recommend a good service to
turn PDFs into softcover books?

From joe at begriffs.com  Tue Oct 30 03:56:52 2018
From: joe at begriffs.com (Joe Nelson)
Date: Mon, 29 Oct 2018 22:56:52 -0500
Subject: Usenet over ham radio?
In-Reply-To: <c724d438-e78d-b550-8e8d-a6226296a9ea@saloits.com>
References: <20181020155911.GA5778@mysterium.local>
 <c724d438-e78d-b550-8e8d-a6226296a9ea@saloits.com>
Message-ID: <20181030035652.GB31915@begriffs.com>

Timothy J. Salo wrote:
> On 10/20/2018 10:59 AM, Joe Nelson wrote:
> > Anyone know if it's feasible to execute UUCP over ham radio to run a
> > radio-based usenet?
>
> First, if you are interested in this you should look at Packet BBS.

That's a good lead, thanks!

What if the "BBS" in our case were actually telnet access to our
frostbyte.cc server? That would simplify the radio traffic by requiring
only a link layer protocol while allowing users to still interact with
the internet via unix commands.

I'm imagining using callsigns for usernames on the server. Then we put
a little gateway in the Hack Factory, and have it SSH into frostbyte
as the callsign/user identified in the radio transmission. The gateway
would send characters from the radio through to SSH, and send the
plaintext response out over the air.

This seems like it's actually somewhat secure because:

a) packet radio is probably pretty obscure, so there would probably be
few people eavesdropping. (Maybe?)

b) it's illegal to misrepresent a callsign, so the callsign itself
functions as both a username and password all at once.  Behind the
scenes frostbyte and the gateway would be using a shared SSH key that
each of the frostbyte radio unix users would allow for login

c) we can decide what level of access to give those radio unix users,
so the damage from another person getting access to the system would be
minimized

> This is a BBS that runs over the amateur radio packet radio
> protocol (AX.25).  A lot of TNCs (Terminal Node Controllers,
> roughly a commercial AX.25 modem) support Packet BBS.  However,
> commercial TNCs are largely a thing of the past: lots of
> software is available to decode AX.25 signals.

Can you recommend a cheap dedicated hardware TNC, or a well-regarded
software TNC? We would probably want to amplify the transmission at the
gateway too -- what's a good amplifier that is compatible with the TNC?

Finally we'd want to put a UHF/VHF antenna on the roof at the Hack
Factory. They already have a RF antenna up there, so they should
be OK adding another. What do you think about the Comet GP-3?
http://www.cometantenna.com/amateur-radio/base-antennas/ba-dual-band/

> Every discussion of connecting Packet BBS and Usenet leads to a
> discussion about FCC appropriate use regulations for amateur
> radio, specifically how appropriate use can be enforced when
> Usenet is gatewayed to amateur radio

I see what you mean. I guess that's another argument in favor of just
doing simple telnet over AX.25. The server environment is better
controlled than random messages coming in from the internet. It helps
prevent accidental transmission of ads, swearing or other prohibited
amateur radio content.

> On 2 meters (144 - 148 Mz), it is easy to run 1,200 bps.  You could
> run 9,600 bps on 2 meters, but you need to tap into your radio on the
> other side of the audio section. ... I think that you can run up to
> 56K bps on both 220 MHz and 440 MHz.

Hm, yeah if we could get 56K on 70cm that would be more than enough
for remote editing in vim etc. And that frequency is accessible to
the tech license level, which would be a lower barrier to entry for
interested folks in the area. Can a TNC accept connections from multiple
clients, or would that require multiple TNCs attached to the gateway
each listening on different frequencies?

> But, equipment starts to get expensive, and there aren't a lot
> of other stations you can communicate with.  I think some of the
> new software-defined radio (SDR) transceivers can support these
> frequencies and data rates.  But, they are expensive.

Maybe SDR is our best bet. Also I wonder how bad the 9600 baud option
would be. Back in my dialup days I never experienced worse than 14.4 but
it's hard to remember how bad that was.

From joe at begriffs.com  Tue Oct 30 04:14:01 2018
From: joe at begriffs.com (Joe Nelson)
Date: Mon, 29 Oct 2018 23:14:01 -0500
Subject: OpenBSD Upgrade GtG?
In-Reply-To: <20181024223640.q6vhh27dkyltka74@19a6.tech>
References: <B6EAA746-449F-41C6-B1BA-69A72EE42784@gmail.com>
 <20181021064343.GB46826@begriffs.com>
 <20181022155412.GB4164@nobelware.com>
 <20181024000713.GA69549@begriffs.com>
 <20181024021624.zxfekr5tedizjb52@19a6.tech>
 <20181024024042.GA79416@begriffs.com>
 <20181024032252.GA1815@nobelware.com>
 <CABSwF4y7_OxoDOGBRG727PCwvVXOUhxOhDaOaHxw48CAXSPSMw@mail.gmail.com>
 <20181024223640.q6vhh27dkyltka74@19a6.tech>
Message-ID: <20181030041401.GC31915@begriffs.com>

Dave Bucklin wrote:
> Shall we get together at the Hack Factory open house at 7pm on the 7th
> of November, then?

This works for me. I could get us pizza/beer like we used to do it if
that sounds good to everyone. Might save time to have food at the event.

Anyone else planning to attend?

From samuel.stuewe at gmail.com  Tue Oct 30 04:30:06 2018
From: samuel.stuewe at gmail.com (Sam Stuewe)
Date: Mon, 29 Oct 2018 23:30:06 -0500
Subject: OpenBSD Upgrade GtG?
In-Reply-To: <20181030041401.GC31915@begriffs.com>
References: <B6EAA746-449F-41C6-B1BA-69A72EE42784@gmail.com>
 <20181021064343.GB46826@begriffs.com> <20181022155412.GB4164@nobelware.com>
 <20181024000713.GA69549@begriffs.com>
 <20181024021624.zxfekr5tedizjb52@19a6.tech>
 <20181024024042.GA79416@begriffs.com> <20181024032252.GA1815@nobelware.com>
 <CABSwF4y7_OxoDOGBRG727PCwvVXOUhxOhDaOaHxw48CAXSPSMw@mail.gmail.com>
 <20181024223640.q6vhh27dkyltka74@19a6.tech>
 <20181030041401.GC31915@begriffs.com>
Message-ID: <CABSwF4xssV+ZEH0oSRcQtO8_1iCoumt95J5goObcMEeNRe6dRw@mail.gmail.com>

On 10/29/18, Joe Nelson <joe at begriffs.com> wrote:
> Anyone else planning to attend?

I am!

All the best,

-Sam

From dave.bucklin at gmail.com  Tue Oct 30 12:25:06 2018
From: dave.bucklin at gmail.com (Dave Bucklin)
Date: Tue, 30 Oct 2018 07:25:06 -0500
Subject: OpenBSD Upgrade GtG?
In-Reply-To: <20181030041401.GC31915@begriffs.com>
References: <B6EAA746-449F-41C6-B1BA-69A72EE42784@gmail.com>
 <20181021064343.GB46826@begriffs.com>
 <20181022155412.GB4164@nobelware.com>
 <20181024000713.GA69549@begriffs.com>
 <20181024021624.zxfekr5tedizjb52@19a6.tech>
 <20181024024042.GA79416@begriffs.com>
 <20181024032252.GA1815@nobelware.com>
 <CABSwF4y7_OxoDOGBRG727PCwvVXOUhxOhDaOaHxw48CAXSPSMw@mail.gmail.com>
 <20181024223640.q6vhh27dkyltka74@19a6.tech>
 <20181030041401.GC31915@begriffs.com>
Message-ID: <20181030122506.op73zyphjhigor5g@19a6.tech>

On Mon, Oct 29, 2018 at 11:14:01PM -0500, Joe Nelson wrote:
> Dave Bucklin wrote:
> > Shall we get together at the Hack Factory open house at 7pm on the 7th
> > of November, then?
> 
> This works for me. I could get us pizza/beer like we used to do it if
> that sounds good to everyone. Might save time to have food at the event.
> 
> Anyone else planning to attend?

Pizza/beer sounds marvelous. See y'all there.

From nompelis at nobelware.com  Tue Oct 30 15:10:41 2018
From: nompelis at nobelware.com (Ioannis Nompelis)
Date: Tue, 30 Oct 2018 15:10:41 +0000
Subject: OpenBSD Upgrade GtG?
In-Reply-To: <20181030122506.op73zyphjhigor5g@19a6.tech>
References: <20181021064343.GB46826@begriffs.com>
 <20181022155412.GB4164@nobelware.com> <20181024000713.GA69549@begriffs.com>
 <20181024021624.zxfekr5tedizjb52@19a6.tech>
 <20181024024042.GA79416@begriffs.com> <20181024032252.GA1815@nobelware.com>
 <CABSwF4y7_OxoDOGBRG727PCwvVXOUhxOhDaOaHxw48CAXSPSMw@mail.gmail.com>
 <20181024223640.q6vhh27dkyltka74@19a6.tech>
 <20181030041401.GC31915@begriffs.com>
 <20181030122506.op73zyphjhigor5g@19a6.tech>
Message-ID: <20181030151041.GA30205@nobelware.com>

I am in as well.

We may want to entertain Chinese food at some point. It is a geek's favourite,
or at least it was in the 90s.


From nompelis at nobelware.com  Tue Oct 30 15:21:11 2018
From: nompelis at nobelware.com (Ioannis Nompelis)
Date: Tue, 30 Oct 2018 15:21:11 +0000
Subject: Comparing oh to other Unix shells
In-Reply-To: <20181030002934.GA31915@begriffs.com>
References: <20181029171153.GB21577@vm2.eradman.com>
 <20181030002934.GA31915@begriffs.com>
Message-ID: <20181030152111.GA30381@nobelware.com>

I think the UofM's printshop will turn PDFs into hard- or soft-cover books.
I will inquire when I am right near there. But expect high prices, more than
buying the book online if it exists, obviously.

I have communicated with the OpenGroup on a couple of topics and they were
prompt in getting back to me. Send an email and inquire about the publications
and lack thereof on their store. I, also, like books more than the PDF version,
but I seldom read whatever technical info I need from a book nowadays.

My "shell scripting" consist of writing a C code to call a bunch of "system()"
functions, and I really want it to remain that way for now. I will tell you
what a good idea would be, to take a particular language format and have a
filter turn it into shell code. This may well exist. Or, take one shell's
code and turn it into another. Being a TCSH fan myself, I would only be
inclined to learn to program a C-shell rather than BASH or anything else.
(Can you tell I am a C person overall...)


From nompelis at nobelware.com  Tue Oct 30 15:31:01 2018
From: nompelis at nobelware.com (Ioannis Nompelis)
Date: Tue, 30 Oct 2018 15:31:01 +0000
Subject: Usenet over ham radio?
In-Reply-To: <20181030035652.GB31915@begriffs.com>
References: <20181020155911.GA5778@mysterium.local>
 <c724d438-e78d-b550-8e8d-a6226296a9ea@saloits.com>
 <20181030035652.GB31915@begriffs.com>
Message-ID: <20181030153101.GA30693@nobelware.com>

I have some thoughts on this.

1. I think it is a pretty bad idea to go down the path where content and
transmission content needs to be "policed" by the gateway owner. If I
understand this correctly, we cannot just relay USENET data to USENET servers
and acting as a gateway like "HAM -> USENET" without monitoring content. Am I
correct?

2. At the very least I would "jail" telnet access on our VPS for any purpose.
I would probably jail activity at a pretty high level, as in, run a whole VM
inside the VPS with NAT, which can be a royal pain to look after.

3. I, sure as Hell, would like to know more about HAM radio because I did not
have the chance to do so in the 90s. And I also think it can be a great service
to the communiity. We give people the means, they can be creative (pushing IoT
stuff over HAM, for example, etc).

I do not have a callsign, but I could be pursuaded to get a license.

From joe at begriffs.com  Wed Oct 31 01:45:01 2018
From: joe at begriffs.com (Joe Nelson)
Date: Tue, 30 Oct 2018 20:45:01 -0500
Subject: Usenet over ham radio?
In-Reply-To: <20181030153101.GA30693@nobelware.com>
References: <20181020155911.GA5778@mysterium.local>
 <c724d438-e78d-b550-8e8d-a6226296a9ea@saloits.com>
 <20181030035652.GB31915@begriffs.com>
 <20181030153101.GA30693@nobelware.com>
Message-ID: <20181031014501.GA1924@begriffs.com>

Ioannis Nompelis wrote:
> 1. I think it is a pretty bad idea to go down the path where content
> and transmission content needs to be "policed" by the gateway owner.

I wasn't really thinking that the gateway would police anything,
just that it's relaying a session on the server rather than random
content from the internet. Amateur radio has a bunch of rules about
what can be put on the air: no music, no profanity, no encryption, no
advertisements. Our server isn't likely to have any of that, so should
be entirely safe to telnet in over the air.

> understand this correctly, we cannot just relay USENET data to
> USENET servers and acting as a gateway like "HAM -> USENET" without
> monitoring content. Am I correct?

If it were just our own usenet messages we'd probably be fine, but
pulling in messages from random sources online could cause trouble.
Whoever sets up the radio gateway is its "control operator" and is
personally on the hook with the FCC for whatever the gateway transmits.

> 2. At the very least I would "jail" telnet access on our VPS for any
> purpose.

Sure, you could help show us how to lock it down. Although it's not as
bad as general telnet access on the internet, it would only be through
the radio interface.

> I do not have a callsign, but I could be pursuaded to get a license.

It's not so bad these days, they don't test you on Morse code. Some
people just practice and memorize the answers to the exam with a site
like https://hamexam.org , but I think it's more interesting to take the
opportunity to learn the material. You can probably get these books for
free from the library:

- https://www.amazon.com/ARRL-Ham-Radio-License-Manual/dp/1625950136
- https://www.amazon.com/Track-Technician-Class-Radio-License/dp/1503077373

Put a little pressure on yourself -- find an exam
session nearby happening in about a month's time with
http://www.arrl.org/find-an-amateur-radio-license-exam-session
Then email the volunteer who is organizing it to thank them and tell
them you'll be there. Just get in the mindset that it's happening!

The technician license is the easiest to get and gives access to VHF and
UHF, which is enough for local packet radio.

From joe at begriffs.com  Wed Oct 31 01:57:25 2018
From: joe at begriffs.com (Joe Nelson)
Date: Tue, 30 Oct 2018 20:57:25 -0500
Subject: Usenet over ham radio?
In-Reply-To: <20181030035652.GB31915@begriffs.com>
References: <20181020155911.GA5778@mysterium.local>
 <c724d438-e78d-b550-8e8d-a6226296a9ea@saloits.com>
 <20181030035652.GB31915@begriffs.com>
Message-ID: <20181031015725.GB1924@begriffs.com>

> Can you recommend a cheap dedicated hardware TNC, or a
> well-regarded software TNC?

Mcsteve on the #tcmaker freenode channel recommended this to
me: https://tnc-x.com/ I can attach it to my Yaesu FT-60 with
https://ww.mfjenterprises.com/Product.php?productid=MFJ-5020

From joe at begriffs.com  Wed Oct 31 04:27:03 2018
From: joe at begriffs.com (Joe Nelson)
Date: Tue, 30 Oct 2018 23:27:03 -0500
Subject: OpenBSD Upgrade GtG?
In-Reply-To: <20181030122506.op73zyphjhigor5g@19a6.tech>
References: <20181021064343.GB46826@begriffs.com>
 <20181022155412.GB4164@nobelware.com>
 <20181024000713.GA69549@begriffs.com>
 <20181024021624.zxfekr5tedizjb52@19a6.tech>
 <20181024024042.GA79416@begriffs.com>
 <20181024032252.GA1815@nobelware.com>
 <CABSwF4y7_OxoDOGBRG727PCwvVXOUhxOhDaOaHxw48CAXSPSMw@mail.gmail.com>
 <20181024223640.q6vhh27dkyltka74@19a6.tech>
 <20181030041401.GC31915@begriffs.com>
 <20181030122506.op73zyphjhigor5g@19a6.tech>
Message-ID: <20181031042703.GB9674@begriffs.com>

Dave Bucklin wrote:
> Pizza/beer sounds marvelous.

Hmm, turns out my wife will be using the car that night, so it'll be
difficult for me to handle the pizza situation, unless we call to have
it delivered which will cost a little more. Maybe one of you guys can
handle providing the food this time?

From dave.bucklin at gmail.com  Wed Oct 31 10:03:43 2018
From: dave.bucklin at gmail.com (Dave Bucklin)
Date: Wed, 31 Oct 2018 05:03:43 -0500
Subject: OpenBSD Upgrade GtG?
In-Reply-To: <20181031042703.GB9674@begriffs.com>
References: <20181022155412.GB4164@nobelware.com>
 <20181024000713.GA69549@begriffs.com>
 <20181024021624.zxfekr5tedizjb52@19a6.tech>
 <20181024024042.GA79416@begriffs.com>
 <20181024032252.GA1815@nobelware.com>
 <CABSwF4y7_OxoDOGBRG727PCwvVXOUhxOhDaOaHxw48CAXSPSMw@mail.gmail.com>
 <20181024223640.q6vhh27dkyltka74@19a6.tech>
 <20181030041401.GC31915@begriffs.com>
 <20181030122506.op73zyphjhigor5g@19a6.tech>
 <20181031042703.GB9674@begriffs.com>
Message-ID: <20181031100343.cknnbg2lrqwda6dl@19a6.tech>

On Tue, Oct 30, 2018 at 11:27:03PM -0500, Joe Nelson wrote:
> Dave Bucklin wrote:
> > Pizza/beer sounds marvelous.
> 
> Hmm, turns out my wife will be using the car that night, so it'll be
> difficult for me to handle the pizza situation, unless we call to have
> it delivered which will cost a little more. Maybe one of you guys can
> handle providing the food this time?

I can make that happen. Do you have any recommendations on where to get
pizza and how much to get?

From nompelis at nobelware.com  Wed Oct 31 15:18:01 2018
From: nompelis at nobelware.com (Ioannis Nompelis)
Date: Wed, 31 Oct 2018 15:18:01 +0000
Subject: Usenet over ham radio?
In-Reply-To: <20181031015725.GB1924@begriffs.com>
References: <20181020155911.GA5778@mysterium.local>
 <c724d438-e78d-b550-8e8d-a6226296a9ea@saloits.com>
 <20181030035652.GB31915@begriffs.com> <20181031015725.GB1924@begriffs.com>
Message-ID: <20181031151801.GB19967@nobelware.com>

Joe, maybe add a "resources" page on our wiki, and make HAM radio a section
of it, and add these links there. Others on this list will like to have the
quick-links, I am sure.


From nompelis at nobelware.com  Wed Oct 31 15:21:25 2018
From: nompelis at nobelware.com (Ioannis Nompelis)
Date: Wed, 31 Oct 2018 15:21:25 +0000
Subject: Software licensing schemes
Message-ID: <20181031152125.GC19967@nobelware.com>

I had told Joe how I was using HDF5, encryption and some obfuscation methods
to write my own software licensing framework. I found an M.S. thesis from
Sweden (May 2014) that discusses a lot of the aspects of both software
licensing and the design and implementation of a licensing scheme. I thought
I'd give my hacker friends the link in case anyone is interested.

https://pdfs.semanticscholar.org/609a/08b9feab4dc1d26120f702527769a9a98135.pdf