libravatar-fans team mailing list archive
-
libravatar-fans team
-
Mailing list archive
-
Message #00083
Re: IRC meeting (2018-10-28)
Hello,
we finished today's bi-weekly IRC meeting.
We discussed the following topics:
* Review of last two weeks
* Tests
* Automatic Deployment
* Design for the wiki
* Backup
* Data migration
The summary of the discussion is in the wiki:
https://wiki.libravatar.org/shutdown-coordination/?updated#index2h2
You find the full log attached.
We postponed a discussion about the "fallback to gravatar" feature to the
next meeting.
The next IRC meeting will happen on the 11th of November at 19:00 UTC in
#libravatar at freenode.
Cheers,
Lars
20:17 <sumpfralle1> What happened in your world since our last meeting?
20:17 <sumpfralle1> (your -> all of you)
20:18 <nipos> Too much to write here since I missed a lot of meetings :(
20:18 <sumpfralle1> hehe :)
20:18 <clime> not much in mine regarding libravatar
20:18 <sumpfralle1> It sounds like you (or someone) sorted out the rename issue?
20:18 <sumpfralle1> ivatar/libravatar?
20:19 <falko> the software will be named ivatar, everything else will stay as it is.
20:20 <falko> i think this was pretty ok for everyone
20:20 <nipos> Yes,that makes sense,good decision
20:20 <sumpfralle1> I was glad to hear this. So now we have a good name for distinguishing between the "old" and the "new" implementation.
20:21 <sumpfralle1> Other things: I did a bit of mail forwarding setup and documentation in the wiki (as written to the mailing list).
20:22 <sumpfralle1> Did anything else happen?
20:22 <falko> software-wise no, i'm afraid, it was a pretty busy week.
20:23 <sumpfralle1> btw.: did you already write a bit of tests for the API? Could I fiddle a bit in order to let them run against the old implementation?
20:24 <nipos> Busy weeks for me,too.I already finished the ivatar domain a long time ago but maybe I should also redesign the wiki,too?Users of my own software are waiting for updates,too but maybe I can find some time for it.
20:25 <tleguern> I think a redesign of the wiki is good idea, so it is more on par with the new style.
20:25 <sumpfralle1> The wiki: I think, it is more oriented to development and internals - thus personally I am fine with the current layout.
20:25 <sumpfralle1> Regarding the website style: is it already part of the current code?
20:25 <sumpfralle1> (aka "merged"?)
20:26 <nipos> Yes,it's already merged
20:26 <nipos> And if I remember correctly it's live on the demo site
20:26 <sumpfralle1> Good - so now you are running out of tasks :)
20:26 <falko> yes, it is merged and yes, it's online.
20:27 <nipos> Yeah that's why I talked about redesigning the wiki.I can't do anything else because I don't understand Python :/
20:28 <tleguern> sumpfralle1: I have a few tests for my implementation of libravatar here https://github.com/Aversiste/libravatar.cgi/blob/master/regress/regress.sh
20:28 <tleguern> It can be iterated upon I guess
20:28 <sumpfralle1> Hm - I am not sure, how "themable" the wiki is (see https://branchable.com) - maybe it is possible?
20:29 <clime> tleguern: looks nice!
20:29 <falko> I'm open for more tests - lots of them.
20:29 <nipos> Oh,I thought it runs as software on the Libravatar VM
20:29 <sumpfralle1> tleguern: interesting approach (not using python, but shell instead) - in this case it is great, since it should "just work" against the current implementation?
20:30 <nipos> On some special Ikiwiki Hoster the possibilities are likely very limited
20:30 <sumpfralle1> I am quite confident, it is a bit themeable. If you enjoy such a mission, I would be glad, if you take a look at that ...
20:31 <nipos> And how should I do that at Branchable?
20:31 <tleguern> sumpfralle1: I will scrub the script a bit so it can be used with other implementations.
20:31 <sumpfralle1> I hope, there is documentation :)
20:31 <sumpfralle1> tleguern: cool!
20:32 <nipos> Every software can be redesigned but without the ability to run a custom source code,that's difficult/maybe impossible
20:32 <clime> I can do some tests this/next week or play with the migration
20:32 <sumpfralle1> tleguern: in case you prefer shell for stuff (sometimes I do) - there is "sharness" - this is a nice shell-based testing thingy - it just landed in Debian
20:33 <sumpfralle1> nipos: I imagine, that a bit of CSS would be a very good start (and probably sufficient).
20:34 <clime> I Will also setup the auto-deploy Hook probly on tue/wed
20:34 <tleguern> I will take a look, I do a lot of things in shell (perhaps a bit too much sometimes :P)
20:34 <clime> (auto-deploy for stg)
20:34 <sumpfralle1> clime: sounds good - thus new code will automatically go there?
20:34 <falko> +1 for auto-deploy
20:34 <sumpfralle1> -> yes - ok
20:34 <clime> Shell Is great
20:35 <sumpfralle1> :)
20:35 <clime> sumpfralle yes fór auto-deploy q
20:35 <sumpfralle1> good
20:36 <sumpfralle1> Are there other topics?
20:36 <clime> We also need to take existing clients and test them a bit
20:36 <sumpfralle1> Good point! How can we tackle this without a running instance with the current data?
20:37 <@fmarier> regarding the wiki discussion, it is possible to change the theme, but also, we can tweak the CSS of the current theme: https://github.com/libravatar/wiki.libravatar.org/blob/master/local.css
20:37 <clime> YouTube can test against libravatar-stg.fedorainfracloud.org
20:37 <sumpfralle1> Should we also use "staging" as a target for the data migration?
20:37 <clime> *you
20:37 <sumpfralle1> funny auto-completion :)
20:37 <clime> Yep, probly
20:37 <clime> Ye
20:38 <sumpfralle1> falko: do you have access to the current dataset for trying out the data migration?
20:40 <falko> nope, i don't sumpfralle1. i wanted to drop fmarier a mail, if it would be possible to get access to the data somehow.
20:40 <nipos> fmarier: I can try to do it with the css but I'm not sure how much I can do with only css
20:41 <@fmarier> falko: do you want a postgres dump? the images?
20:41 <falko> pretty cool would be if I can download all data as xml exports, as provided by the site. since i can import them already, just not as a mass-import yet.
20:43 <@fmarier> falko: you'd have to write your own mass-export functionality for that though, the current codebase doesn't have that
20:44 <falko> ack fmarier - will dig into that then.
20:44 <sumpfralle1> so with a bit of luck (and free time of falko) we could hope for "staging" running with the current dataset in two weeks?
20:46 <falko> sounds like a reasonable plan. if IBM is not blocking me :-P
20:46 <@fmarier> Ideally, we should do our tests with the prod data but with anoymized email addresses (to ensure we don't accidentally spam people out as we're testing things).
20:46 <@fmarier> I can provide a postgres dump with anonymized email addresses once falko is ready to turn it into the format he needs.
20:46 <sumpfralle1> good point - so there should be just no working outgoing mail, I guess?
20:50 <@fmarier> yeah that's the idea
20:51 <@fmarier> though just in case, I'll just turn all of the email addresses into invalid ones
20:51 <sumpfralle1> ok - better safe than sorry :)
20:51 <@fmarier> like <hash>@libravatar.org or something like that
20:51 <falko> would be perfect, i think
20:51 <sumpfralle1> or even @dummy.libravatar.org
20:52 <sumpfralle1> Do we have other topics?
20:53 <@fmarier> I guess I should just give a quick update that backups have been migrated and also incoming emails.
20:54 <falko> cool, fmarier
20:54 <sumpfralle1> yeah
20:55 <sumpfralle1> for the backups: we will probably want to add another storage place somewhen and also add one or two more gpg keys in the future. (not yet, I guess)
20:55 <@fmarier> I can easily add a second GPG key right now.
20:56 <sumpfralle1> maybe falko would be a good candidate?
20:56 <falko> C5637919EF252943 would be my public key
20:57 <@fmarier> https://pgp.mit.edu/pks/lookup?search=C5637919EF252943&op=index doesn't return anything
20:57 <@fmarier> hm, speaking of incoming emails. they don't seem to be getting through anymore...
20:58 <sumpfralle1> fmarier: do you have a specific example address I could try?
20:58 <@fmarier> oh, nevermind, I just got one
20:58 <sumpfralle1> :)
20:58 <@fmarier> (trying to log into the wiki)
20:59 <sumpfralle1> I do not think, we use greylisting - but I can be wrong.
20:59 <-- falko (~falko@xxxxxxxxxxxxxxxxxxxxxxxxxxxxx) hat den IRC verlassen (Remote host closed the connection)
20:59 <sumpfralle1> anyway: other topics?
20:59 --> falko (~falko@86.59.13.244) hat den Channel #libravatar betreten
21:00 <sumpfralle1> falko: other topics?
21:00 --> falko_ (~falko@xxxxxxxxxxxxxxxxxxxxxxxxxxxxx) hat den Channel #libravatar betreten
21:00 <tleguern> We can discuss about the need to redirect to gravatar, but it can wait unitl next time
21:00 <sumpfralle1> sounds good - I will a note
21:00 <sumpfralle1> add a note
21:01 <@fmarier> On that topic, I will say that it is currently optional but I suspect that most users of Libravatar rely on it very heavily.
21:02 <falko_> # gpg --keyserver pgp.mit.edu --send-keys EF252943
21:02 <falko_> gpg: sending key EF252943 to hkp server pgp.mit.edu
21:02 <falko_> @fmarier it should be there.
21:02 <@fmarier> My guess is that many would stop using Libravatar if it didn't fall back to Gravatar.
21:03 <falko_> regarding gravatar, ok, let's discuss this next time, but yes, that's an important point
21:03 <@fmarier> So that's something to keep in mind.
21:03 <sumpfralle1> Is this fallback currently done by client implementations or by the libravatar server?
21:03 <tleguern> It is made by libravatar
21:03 <@fmarier> I'm not saying it's necessarily a bad thing, but it's how it has worked for the past 8 years.
21:03 <@fmarier> Yes, it's server-side
21:03 <sumpfralle1> So we should probably keep this.
21:04 <falko_> let's discuss this next time please, because i have an opinion here and i want to discuss this... :-)
21:04 <tleguern> Ok :)
21:04 <sumpfralle1> good - I added a note
21:05 -- falko_ ist nun als falko bekannt
21:05 <@fmarier> I won't be able to make the next meeting, but that's probably all I wanted to say on the topic, so feel free to discuss without me :)
21:05 <sumpfralle1> ok - it will not be forgotten
21:05 <falko> ack!
21:05 <sumpfralle1> So we are finished - what do you think?
21:06 <falko> yep. i think so.
21:07 <clime> Take care till next meeting.
21:07 <-- clime (5918392c@gateway/web/freenode/ip.89.24.57.44) hat den IRC verlassen
21:07 <tleguern> See you next time !
21:08 <sumpfralle1> Have a good remaining Sunday evening!
21:08 <nipos> Goodbye
21:08 <@fmarier> bye everyone
21:08 <-- fmarier (~francois@fsf/member/fmarier) hat den IRC verlassen (Quit: leaving)
21:09 <falko> have a nice evening and good start into the coming week!
References