Text 11997, 92 rader
Skriven 2021-02-15 16:37:50 av andrew clarke (3:633/267)
  Kommentar till text 11961 av Alexey Vissarionov (2:5020/545)
Ärende: Future of ftsc.org web site
===================================
14 Feb 21 07:27, you wrote to me:

 AV> These are primarily the git repositories. I'd be happy to use git for
 AV> all our needs several years ago (when I was a member of the FTSC), but
 AV> some old farts appeared unable to learn ever more simple things than git
 AV> and gpg.

An old English phrase springs to mind: "You can't teach old dogs new tricks."
:-)

 AV> GitHub is distrusted (they are known to wipe whole projects due to
 AV> politically "unreliable" people rarticipating there), so it could serve
 AV> only as a mirror.

"Perfect is the enemy of good" - Voltaire

I have no issue with GitHub and the likelyhood of an FTSC repo being shut down
by GitHub is essentially zero, and even if it's non-zero, the point of any repo
is that its users have local copies, so it can always be restored one way or
another.

 AV> Anyway, to do that we have to start using git, so here's a question:
 AV> out of all candidates, who is familiar with it?

Even if not, Git takes 10 minutes to learn if you're familar with CVS or
Subversion. Obviously longer if you've never used revision control software
before. But I'd like to think all the nominated FTSC members have at least some
knowledge of what revision control is, but who knows around here. :-)

 AV> Current ${subj} is a bit unfriendly to a search engines, but it's very
 AV> friendly to mirroring software like wget. That's not what we could have
 AV> with git, but it allows anyone to keep their own FTSC documents archive.

It's pretty rare anyone needs every document.

In any case "git clone xyz" will download them all, probably quicker than wget
can mirror them, and you get a complete log of ever commit.

 ac>> The Wikipedia entry for FidoNet could also point to both the
 ac>> GitHub repo and archive.org snapshot, since they're fairly
 ac>> relevant.

 AV> Seems unwise. Keeping ftsc.org and adding some mirrors would be mush
 AV> better.

So keep it, but the point I was getting at is that an FTSC repo should
encourage feedback, bug reports etc. GitHub (and the other sites like it) give
you all that for free.

 ac>> Hosting all the FTSC documents on GitHub would be particularly
 ac>> useful since it would allow anyone to write bug reports or file
 ac>> "issues" relating to the various FidoNet standards documents, which
 ac>> may help any future developers. (Or historians...)

 AV> "FidoNet is our primary mode of communication" // (q)

 AV> So all reports should go here, to the FTSC_PUBLIC echoarea. Also, git
 AV> can work over a netmail...

The FTSC should not be in the business of restricting the discussion of its own
documents to be within the bounds of FidoNet.

In any case FTSC_PUBLIC is awful for bug reports in comparison to something
like GitHub.

Among other things, it:

- is not really public, or is essentially invisible to public without
considerable effort
- has no archive
- has no search, so the same problems get asked
- has no way to resolve/close bug reports
- requires readers to skip non-technical posts (just like this one);
particularly every single dreaded message about FTSC
membership/nominations/voting/etc that are all nonsense to anyone only
interested in actual technical work


Not directed at Alexey:

It used to be the technical discussions were in NET_DEV. I assume for some
reason that echo died, not my connection to it. But the same questions got
asked over and over there too, mostly because nobody kept a long-term record of
anything except for the FTSC, who only ever published standards and never any
annotations/addendums/clarifications.

Eg. a common thread was how to implement FTS-9 (MSGID/REPLY) properly since it
was was so open to interpretation (and probably still is). (This is only an
example, not an invitation to mansplain MSGID/REPLY to me.)

--- GoldED+/BSD 1.1.5-b20180707
 * Origin: Blizzard of Ozz, Melbourne, Victoria, Australia (3:633/267)