#archlinux32 | Logs for 2019-11-16

[00:13:05] <buildmaster> pentium4/cypari2 is broken (says buildknecht2): https://archlinux32.org
[00:29:19] <{levi}> I'm going for a wash now, so I'll be away for a couple of hours
[00:35:53] -!- isacdaavid has quit [Quit: Leaving.]
[01:08:19] <buildmaster> any/python-nosexcover is broken (says rechenknecht): https://archlinux32.org
[01:12:04] <buildmaster> any/python-testflo is broken (says eurobuild3): https://archlinux32.org
[01:15:49] <buildmaster> any/python-icalendar is broken (says rechenknecht): https://archlinux32.org
[02:20:18] -!- MrBIOS has quit [Read error: Connection reset by peer]
[02:20:46] -!- MrBIOS has joined #archlinux32
[02:21:45] <buildmaster> i486/nimble is broken (says eurobuild6-7-i486): https://archlinux32.org
[02:24:04] <buildmaster> i686/hexchat is broken (says eurobuild6-2): https://archlinux32.org
[02:24:06] <buildmaster> pentium4/hexchat is broken (says eurobuild6-1): https://archlinux32.org
[02:26:45] <buildmaster> pentium4/nimble is broken (says eurobuild6-5): https://archlinux32.org
[02:29:18] <buildmaster> i686/nimble is broken (says rechenknecht): https://archlinux32.org
[02:30:18] -!- MrBIOS has quit [Quit: MrBIOS]
[02:33:50] <buildmaster> any/proselint is broken (says eurobuild6-2): https://archlinux32.org
[02:35:20] <buildmaster> any/python-click-threading is broken (says nlopc46): https://archlinux32.org
[02:44:17] <buildmaster> any/python-eventlet is broken (says eurobuild6-1): https://archlinux32.org
[03:02:14] <buildmaster> pentium4/python-pplpy is broken (says eurobuild6-4): https://archlinux32.org
[03:08:22] -!- MrBIOS has joined #archlinux32
[03:32:52] <{levi}> I be back
[04:16:32] -!- Buran has quit [Remote host closed the connection]
[04:20:36] <buildmaster> any/python-btchip is broken (says nlopc46): https://archlinux32.org
[04:34:10] <buildmaster> any/python-daemon is broken (says nlopc46): https://archlinux32.org
[04:46:26] -!- Buran has joined #archlinux32
[04:47:32] -!- urraka has quit [Quit: WeeChat 2.6]
[04:51:36] <buildmaster> i486/python-zipp is broken (says nlopc46-i486bs1): https://archlinux32.org
[04:52:04] <buildmaster> i686/python-zipp is broken (says rechenknecht): https://archlinux32.org
[04:56:41] <buildmaster> i486/python-hunter is broken (says eurobuild6-7-i486): https://archlinux32.org
[05:27:09] <buildmaster> i686/python-pplpy is broken (says eurobuild6-4): https://archlinux32.org
[06:01:42] <buildmaster> i686/389-ds-base is broken (says eurobuild6-3): https://archlinux32.org
[06:05:21] <buildmaster> any/m2r is broken (says eurobuild6-1): https://archlinux32.org
[06:35:02] <{levi}> I'll be around tomorrow if anyone feels like helping
[06:35:04] -!- {levi} has quit [Quit: Hurrah for Elw3's new quit message]
[07:11:41] -!- MrBIOS has quit [Quit: MrBIOS]
[07:12:00] -!- MrBIOS has joined #archlinux32
[07:12:27] -!- MrBIOS has quit [Client Quit]
[07:13:00] -!- MrBIOS has joined #archlinux32
[07:13:13] -!- MrBIOS has quit [Client Quit]
[07:13:37] -!- MrBIOS has joined #archlinux32
[07:13:59] -!- MrBIOS has quit [Client Quit]
[07:14:26] -!- MrBIOS has joined #archlinux32
[07:14:45] -!- MrBIOS has quit [Client Quit]
[07:15:10] -!- MrBIOS has joined #archlinux32
[07:15:31] -!- MrBIOS has quit [Client Quit]
[07:16:45] -!- MrBIOS has joined #archlinux32
[07:17:04] -!- MrBIOS has quit [Client Quit]
[07:17:35] -!- MrBIOS has joined #archlinux32
[07:17:50] -!- MrBIOS has quit [Client Quit]
[07:18:17] -!- MrBIOS has joined #archlinux32
[07:18:36] -!- MrBIOS has quit [Client Quit]
[07:18:55] -!- MrBIOS has joined #archlinux32
[07:19:22] -!- MrBIOS has quit [Client Quit]
[07:30:43] -!- darrylrev0k has quit [Ping timeout: 245 seconds]
[07:34:37] <buildmaster> pentium4/389-ds-base is broken (says eurobuild6-3): https://archlinux32.org
[07:57:03] -!- vboxler has joined #archlinux32
[07:57:30] <vboxler> Hello, the forum is now online again
[07:57:34] <vboxler> You are not logged in.Topics: Active | Unanswered
[07:57:36] <vboxler> The recovered forum pages can be found here.
[07:57:37] <vboxler> Info
[07:57:39] <vboxler> This forum is not accepting new registrations.
[07:57:56] <vboxler> but i don can login
[07:58:08] -!- vboxler has quit [Client Quit]
[09:10:42] -!- deep42thought has joined #archlinux32
[09:10:42] <buildmaster> Hi deep42thought!
[09:10:42] <buildmaster> !rq deep42thought
[09:10:44] <phrik> buildmaster: <deep42thought> because you're a pessimist with no hope - or you're a realist afraid of change ... pick your evil
[09:11:21] <deep42thought> abaumann: I put python-setuptools-bootstrap into build-support with the intention to be the bootstrapping package for setuptools (but it looks, like it has some issues)
[09:11:53] <deep42thought> regarding forums ({levi}, vboxler): what exactly is the issue? registration email not arriving or what?
[09:12:18] <deep42thought> (it worked, when I registered, but it might have changed in the meantime ...)
[09:12:40] <deep42thought> regarding "recovering": we decided to not copy over all the content, but rather link the old content
[09:13:13] <deep42thought> let us know if it makes sense to copy over some content (e.g. to continue threads or so) and we can do so on an individual basis
[09:18:41] -!- abaumann has joined #archlinux32
[09:18:42] <buildmaster> Hi abaumann!
[09:18:42] <buildmaster> !rq abaumann
[09:18:44] <phrik> buildmaster: <abaumann> I made the experience that when the config is simple and looking nicely it usually also works. :-)
[09:18:45] <abaumann> hi all
[09:18:56] <abaumann> registration is currently disabled.
[09:19:08] <abaumann> I still don't know to which mail server mail should be sent on the buildmaster
[09:19:26] <abaumann> I don't feel inclined to relay it to private mail servers and domains
[09:19:58] <deep42thought> ah, ok, that explains it :-)
[09:20:01] <abaumann> hi abaumann
[09:20:05] <deep42thought> btw: good morning!
[09:20:06] <abaumann> lol
[09:20:08] <deep42thought> lol
[09:20:09] <abaumann> hi deep42thought
[09:20:15] <deep42thought> hi deep42thought
[09:20:16] * abaumann jawns
[09:20:20] <deep42thought> (just for good measure)
[09:20:25] <abaumann> too early to think properly :-)
[09:20:40] <abaumann> on the buildmater: relayhost = smtp.andreasbaumann.cc
[09:20:42] <deep42thought> you're in the same time zone as I - aren't you?
[09:20:54] <abaumann> then it redistributes it to the mailing list
[09:21:19] <abaumann> not having access to the real MX of archlinux32.org I cannot add a allowed IP for relaying
[09:21:21] <deep42thought> we could add the buildmaster to MX, so it is allowed to send directly
[09:21:30] <abaumann> yep
[09:21:39] <deep42thought> it's a question of DNS, not access to the MX
[09:21:46] <abaumann> no
[09:21:48] <deep42thought> should I add it to MX?
[09:21:51] <abaumann> hopefully not.
[09:21:55] <abaumann> ah
[09:22:07] <abaumann> you mean, the buildmaster should sent mail itself on behalf of the domain
[09:22:12] <deep42thought> yes
[09:22:33] <deep42thought> I thought, the consensus was to set up a mailserver on the buildmaster some time anyways?
[09:22:46] <abaumann> then yes.
[09:23:02] <abaumann> but I have to apply some strong rules there, so we don't end up in problems.
[09:23:24] <deep42thought> rules regarding *what*?
[09:23:37] <abaumann> for instance, SMTP sending is only allowed authenticated, every tool gets its user and password
[09:23:57] <deep42thought> ah, right
[09:24:01] <abaumann> I had twice an open SMTP relay in my admin live, twice is enough :-)
[09:24:23] <deep42thought> ok :-)
[09:24:52] <deep42thought> otoh: relaying via private mail servers will not circumvent that problem - only shift the blamed one
[09:25:03] <deep42thought> "one" = "mail server"
[09:25:15] <abaumann> yes, but I want to be able to send private email independent of the archlinux32 project ;-)
[09:25:32] <abaumann> if my mailserver gets blocked, I have a problem
[09:25:32] <deep42thought> that's not what I meant
[09:25:43] <deep42thought> ah, ok
[09:25:59] <deep42thought> I'll add an mx entry for the buildmaster
[09:26:00] <abaumann> the other 'one' :-)
[09:26:10] <deep42thought> and when you feel lucky, you can set up a mail server on it
[09:26:28] <abaumann> it's already there, but not configured :-)
[09:26:36] <abaumann> and I'll test the registrations then
[09:28:19] <abaumann> smtp.archlinux32.org. 3600 IN CNAME archlinux32.org.
[09:28:19] <abaumann> archlinux32.org. 3515 IN A
[09:28:25] <abaumann> so there is a mailserver there
[09:30:25] <abaumann> so it points there, but there is no mailserver
[09:33:25] <deep42thought> we have set up smtp subdomain???
[09:33:30] <deep42thought> strange, I cannot remember
[09:34:00] <abaumann> I would leave it and just a new MX record
[09:34:07] <abaumann> maybe we find out later :-)
[09:34:36] <deep42thought> sure, we leave it
[09:34:44] <deep42thought> or we add it as mail3.archlinux32.org
[09:34:55] <deep42thought> there is no smtp
[09:35:02] <buildmaster> any/hy is broken (says eurobuild6-1): https://archlinux32.org
[09:35:05] <deep42thought> it's only a wildcard *.archlinux32.org cname
[09:35:14] <abaumann> oh.
[09:35:31] <deep42thought> I thought, mx may not point to cname, but *must* point to a or aaaa
[09:35:32] <abaumann> yes, that's something you don't see from outside, unless you dig flabla.archlinux32.org
[09:36:05] <abaumann> why?
[09:36:29] <abaumann> archlinux32.org itself must be an A record
[09:36:54] <abaumann> ah.
[09:36:55] <deep42thought> not even that
[09:36:59] <abaumann> RFC 2181 section 10.3
[09:37:01] <buildmaster> any/python-oauth2client is broken (says buildknecht2): https://archlinux32.org
[09:37:11] <deep42thought> I think, dns convention dictates, MX must point to a A or AAAA record but not to CNAME
[09:37:30] <buildmaster> any/python-baron is broken (says nlopc46): https://archlinux32.org
[09:37:36] <abaumann> buildmaster: shut up
[09:37:37] <buildmaster> Sorry, I will do.
[09:37:39] <deep42thought> ok, we now have an (low priority) MX to mail3.archlinux32.org which points (via A/AAAA) to the buildmaster
[09:37:41] <abaumann> looks like
[09:39:14] <deep42thought> let me know, when should I kick out the other MX'es, so you can check receiving of emails, too
[09:39:37] <abaumann> ok. should work fine with all MXs in place
[09:39:53] <deep42thought> sending: yes, receiving: no
[09:40:08] <abaumann> we'll have to add SPF, DMARC etc later.. not necessary in the beginning
[09:40:11] <deep42thought> e.g.: it would be good if you did not receive emails in behalf of archlinux32.org yet
[09:40:27] <deep42thought> spf has "mx" in it, so it should work out of the box
[09:41:09] <abaumann> but on behalf of buildmaster.archlinux32.org? I don't think I should change the name of the buildmaster know to mail3.archlinux32.org
[09:41:25] <deep42thought> no
[09:41:34] <deep42thought> you should (some time) receive email for archlinux32.org
[09:41:38] <deep42thought> and send under that domain
[09:41:40] -!- esotericnonsense has joined #archlinux32
[09:42:09] <abaumann> ok
[09:42:48] -!- vboxler has joined #archlinux32
[09:43:06] <vboxler> deep42thought: hello
[09:43:22] <vboxler> did you have a one men forum?
[09:43:45] <vboxler> deep42thought:
[09:43:47] <vboxler> Info
[09:43:48] <vboxler> This forum is not accepting new registrations.
[09:43:50] <vboxler> Go back
[09:43:56] <abaumann> hi vboxler
[09:44:12] <abaumann> yes. Registration is disabled currently till things are properly set up
[09:44:34] <vboxler> abaumann: ok, thank you for the information, i will wait
[09:44:50] <vboxler> cu
[09:44:50] <abaumann> you will be one of the first test guinea pigs ;-)
[09:45:05] <abaumann> cu :-)
[09:45:10] -!- vboxler has quit [Client Quit]
[09:46:52] <deep42thought> pre-alpha tester?
[09:47:22] <abaumann> :-)
[09:48:48] <deep42thought> what's the standard way to secure an email server against unauthorized sending?
[09:49:04] <deep42thought> I mean: everyone with a shell on that server could send on behalf of it, right?
[09:49:18] <abaumann> yes, that's one of the big traps.
[09:49:37] <deep42thought> firewall rules blocking outgoing connections to port 25 except ... what?
[09:49:48] <abaumann> no,
[09:49:54] <abaumann> the firewall is open on port 25
[09:50:01] <abaumann> maybe
[09:50:17] <deep42thought> I meant: block *outgoing*
[09:50:17] <abaumann> because it's only a sending MX we could close it for receiving
[09:50:21] <abaumann> ah
[09:50:40] <abaumann> unless you want to actually deliver messages
[09:50:47] <deep42thought> and only postfix (or whoever checks authentication) is allowed to connect to port 25 of another server
[09:51:02] <abaumann> yes.
[09:51:09] <deep42thought> but: I'm no sysadmin, I have no idea what the standard way is ...
[09:51:12] <abaumann> it listens to 25 and sends to other 25s
[09:51:18] <deep42thought> yes
[09:51:29] <abaumann> the standard is what works and doesn't get hacked :-)
[09:51:50] <abaumann> for delivering you can use 587 for instance
[09:52:02] <abaumann> so external users can log and authenticate and then send messages
[09:52:27] <abaumann> in my company we have whole subnets of server, which are allowed to send emails and relay them to the main mailserver
[09:52:31] <abaumann> it really all depends
[09:52:53] <abaumann> we should get away with a very strict and simple postfix configuration..
[09:52:57] <abaumann> ..simple for postfix. :-)
[09:53:19] <deep42thought> but what prevents anyone with a shell (e.g. cron) to send arbitrary emails?
[09:53:41] <abaumann> nothing
[09:54:02] <abaumann> the trick is: know what you are installing on the machine and who has an SSH account to it
[09:54:06] <abaumann> limit that.
[09:54:42] <abaumann> cron is a typical example where you would go unauthenticated and just forward it in the aliases to root
[09:54:53] <abaumann> the root you distribute to an admin or list of admins
[09:55:21] <abaumann> the dangerous part is when external unknwon people can trigger sending of an email, as a forum or the bugreporting tool
[09:56:07] <abaumann> first the software has to have some multi-send/IP whatever security measures, so that it cannot use the mailserver as open relay
[09:56:32] <abaumann> and on postfix side we can have some throttling and sanity rules to check for the contents of the email which are sent
[09:56:49] <deep42thought> hmm, ok
[09:57:04] <abaumann> I'm only part-time administrator :-)
[09:57:08] <deep42thought> sounds reasonable, we do not want to give ssh access to the buildmaster away freely, anyways ;-)
[09:57:16] <abaumann> no.
[09:57:39] <abaumann> and if somebody gets root or a wheel user, the game is over anyway :-)
[09:58:02] <deep42thought> yes, but mail-problems already arise when someone gets shell access at all
[09:58:11] <abaumann> true
[09:58:56] <abaumann> if _anybody_ can log in, it's over :-)
[09:59:42] <deep42thought> especially you or me ;-)
[09:59:50] <abaumann> :-)
[10:00:09] <deep42thought> -A input -j drop
[10:00:15] <deep42thought> better be sure
[10:09:32] <deep42thought> abaumann: [build-support] packages can now be added to the git like any other packages *without* an upstream variant, too - it would be a good idea to put all those packages, you created to manually circumvent stuff, into build-support/$pkgname-shim (or similar) for future use
[10:10:10] <deep42thought> (I tried this with python-setuptools-bootstrap, but apparently, I did something wrong)
[10:10:36] <deep42thought> also I would move extra/python-pip-bootstrap into [build-support] as it should not be in [extra] imho
[10:11:01] <abaumann> ah, good. I remembered yesterday I had a python-pip-bootstrap AUR project for that purpose
[10:11:18] <abaumann> I didn't understand what python-setuptools-bootstrap is supposed to do, it looks rather complex.
[10:11:34] <deep42thought> it is identical to the one, upstream used for bootstrapping
[10:11:55] <abaumann> python-pip-bootstrap simply uses the official python binary blob (which BTW is Python bytecode) and creates the basic tools
[10:11:59] <abaumann> ah.
[10:12:00] <deep42thought> but I might have missed bootstrap variants of some other packages they were using
[10:12:19] <abaumann> python-pyparsing was the package I had to build against python-pip-bootstrap
[10:12:22] <abaumann> and only that.
[10:12:36] <abaumann> then python-setuptools would build normally (minus check())
[10:12:51] <abaumann> I'll remove that at the end of rebuilding puthon
[10:12:58] <deep42thought> this will work automatically if the first package is in [build-support]
[10:13:12] <deep42thought> besides check()
[10:13:21] <abaumann> I still hvae trouble with the syntax of the script used for build-support
[10:13:29] <abaumann> it's not clear to be how to use it
[10:13:45] <abaumann> that's why I started with my own local bootstrapping repos in the first place
[10:13:46] <deep42thought> there are now two ways to put packages into build-support
[10:14:23] <abaumann> you saw build-support/extra?
[10:14:23] <deep42thought> you can put packages into build-support the "normal" way: put a PKGBUILD into the git (build-support/$pkgname/PKGBUILD)
[10:14:26] <deep42thought> yes
[10:14:34] <deep42thought> this is one layer to deep
[10:14:37] <abaumann> ah
[10:14:43] <deep42thought> it needs to be build-support/$pkgname/PKGBUILD
[10:14:54] <deep42thought> and it will *not* be appended to any upstream PKGBUILD
[10:15:02] <deep42thought> because upstream does not have "build-support"
[10:15:06] <abaumann> yep
[10:15:19] <abaumann> and they are built when checked into git and published in build-support?
[10:15:23] <abaumann> how are they removed?
[10:15:41] <deep42thought> they are not
[10:15:56] <deep42thought> or: manually with delete-package
[10:16:00] <abaumann> ah, ok.
[10:16:05] <deep42thought> (it has some option to remove build-support packages)
[10:16:40] <deep42thought> -n
[10:16:45] <deep42thought> no, -b
[10:18:19] <abaumann> I like to have manual control, especially if rust-bin is suddendly silently used for i686 adding SSE2 instructions everywhere. That's BTW how I ended up in the mess, I forgot to remove a rust-bin providing rust
[10:18:20] <deep42thought> ah, I should read more thoroughly
[10:18:54] <deep42thought> the part about building is right, my comment was regarding the deletion
[10:19:14] <deep42thought> you can remove it from git and it should get marked for deletion then, too
[10:19:22] <abaumann> ah. ok
[10:19:35] <deep42thought> or you just change the PKGBUILD to not provide rust anymore
[10:19:39] <deep42thought> but this feels like a hack
[10:19:47] <abaumann> yeah, agreed.
[10:20:09] <abaumann> but I have to keep them somewhere else, so I can resuse them next time.
[10:20:30] <abaumann> and more importantly, somebody else should be able to do so too.
[10:20:37] <abaumann> so, I kept them in the AUR
[10:20:47] <deep42thought> or keep them in git, but "disable" them
[10:20:56] <abaumann> but this then triggered people to request deletion request like "it contains binary blobs" :-)
[10:21:02] <abaumann> better maybe
[10:21:03] <deep42thought> e.g. comment out the provides and put a return into package()
[10:22:45] <deep42thought> I was considering a possibility to easily pull in upstream PKGBUILDs in build-support, but everything I could think of seems pretty cumbersome ...
[10:23:26] <abaumann> manual build-support is fine. As you have to do some other manual stuff anyway likely - like thinking :-)
[10:23:57] <deep42thought> "manual thinking" lol
[10:26:53] <abaumann> uff. the big backup finished, in 2.6 days.
[10:27:02] <abaumann> I could invest in more than 60MBit/s maybe..
[10:32:38] <deep42thought> only the initial backup should take that long, right?
[10:32:49] <abaumann> yep
[10:38:45] <deep42thought> ah, to finish my monologue: there is a second option to put packages into build-support, namely the copy-to-build-support script, which takes a given package, modifies it and puts it into build-support
[10:39:11] <deep42thought> this is done in two steps, because someone needs to sign the "built" package - and the key for that cannot be expected to reside on the buildmaster
[10:39:56] <deep42thought> hmm, no, this seems to be the *third* way
[10:39:59] * deep42thought is confused
[10:40:01] <abaumann> ah, I do something similar, I do a local-build-package, then I sign it and upload it onto my local bootstrap.
[10:40:19] <abaumann> It's the old "I want to build a packages without the buildmaster" approach
[10:40:24] <abaumann> completlely without
[10:40:42] <deep42thought> yeah, but this is not traceable in git
[10:40:50] <abaumann> yes, true
[10:40:51] <deep42thought> and I dislike not having some kind of history
[10:41:07] <deep42thought> though it might be faster / more convenient
[10:41:32] <abaumann> if I test options to bootstrap or build rust, it's essential I can tweak things as I want.
[10:41:38] <deep42thought> copy-to-build-support copies packages 1:1
[10:41:47] <deep42thought> yes, for testing
[10:41:58] <abaumann> But this can also be done outside in a testing environment, when I have the findings, I put them into git
[10:41:58] <deep42thought> but the package which ends up in the repositories should be traceable
[10:42:07] <deep42thought> exactly
[10:42:23] <abaumann> That's why a local bootstrap is handy: those packages are just used to force build something.
[10:42:27] <deep42thought> the problem arises, when you need to build x with special options in order to properly build y which depends on x
[10:42:34] <abaumann> afterwards I rebuild the packages always in the standard way
[10:42:40] <deep42thought> good
[10:42:44] <abaumann> so far I had this only with rust
[10:43:01] <abaumann> going over the rust-bin and binary blob is bad, I know
[10:43:08] <abaumann> but there was simply no other option
[10:43:23] <deep42thought> yes, that's sad
[10:43:33] <abaumann> nowadays mrustc bootstrapping might work, as mrustc now supports newest rust versions (that's the rumour)
[10:43:34] <deep42thought> but I think we/you invested enough time into finding a proper way
[10:43:57] <abaumann> I'm against monocultur in computer languages, there is no way one company/organisation controls a language like Mozilla is controlling rust
[10:44:04] <abaumann> *monoculture
[10:44:44] * abaumann is watching a youtube interview Bjarne Stroustrup, so the monocultur thing might be something I just heard :-)
[10:44:48] <abaumann> *with
[10:45:48] <abaumann> "error: couldn't load codegen backend "/usr/lib/rustlib/i686-unknown-linux-gnu/codegen-backends/librustc_"
[10:46:05] <abaumann> ah cool. my llvm move broke rust bootstrapping (not rust in i686)
[10:46:54] <buildmaster> any/python-pyte is broken (says rechenknecht): https://archlinux32.org
[10:52:12] <buildmaster> any/python-click-repl is broken (says eurobuild3): https://archlinux32.org
[10:58:01] <abaumann> /usr/bin/update-mime-database: error while loading shared libraries: libicuuc.so.64: cannot open shared y
[10:58:04] <abaumann> error: command failed to execute correctly
[10:58:11] <abaumann> aha, an icu problem is ahead :-)
[11:00:16] <abaumann> /usr/lib/rustlib/i686-unknown-linux-gnu/codegen-backends/librustc_codegen_llvm-llvm.so has libicuuc.so.64 => not found, so it's not a LLVM issue
[11:02:09] <deep42thought> so: everything is a icu-issue ;-)
[11:02:19] <deep42thought> except for the python-pyparse issues ;-)
[11:02:20] <abaumann> ah, and half of python is now 3.8 and half 3.7, as many build tools depend on python, we can expect some failures everywhere
[11:02:25] <abaumann> lddtree /usr/lib/rustlib/i686-unknown-linux-gnu/codegen-backends/librustc_codegen_llvm-llvm.so
[11:02:28] <abaumann> ModuleNotFoundError: No module named 'elftools'
[11:02:35] <abaumann> so python-pyelftools is broken in staging
[11:02:51] <abaumann> it's interesting to learn about those build problems
[11:03:11] <abaumann> so, it doesn't make sense to test rust bootstrapping during a major rebuild of the userland
[11:03:20] <deep42thought> should we reschedule all python-* just for good measure?
[11:03:38] <abaumann> I wouldn't
[11:03:38] <abaumann> https://buildmaster-status.archlinux32.org
[11:03:39] <phrik> Title: Buildmaster for Archlinux32 packages (i486, i686, pentium4, any) (at buildmaster-status.archlinux32.org)
[11:03:46] <abaumann> is looking good
[11:04:02] <abaumann> when the redline doesn't approach the base, then yes, maybe
[11:04:07] <abaumann> *red line
[11:04:24] <abaumann> I can force build python packages individually, if they block something now
[11:04:27] <deep42thought> broken python-* in staging sounds worrysome
[11:04:44] <deep42thought> yes, but it would be bad if you missed something
[11:04:50] <abaumann> yeah, but we should also test the self-healing capabilites of the build system :-)
[11:05:01] <deep42thought> seed-build-list -a
[11:05:11] <deep42thought> but it's not executed automatically anymore, I think
[11:05:37] <deep42thought> oh, it *is* - so everything should heal automagically :-)
[11:05:48] <abaumann> exactly :-)
[11:06:49] <deep42thought> ah, no, the timer is there, but it is not automatically started (e.g. it's not enabled)
[11:07:41] * deep42thought feels lucky and asks Dirty Harry to pull the trigger once more
[11:08:10] <abaumann> :-)
[11:08:21] <abaumann> oh. the incremental backup finished.
[11:08:26] <deep42thought> :-)
[11:08:34] <abaumann> so that's really efficient.
[11:08:58] <deep42thought> looks, like *a*lot* of stuff is scheduled by seed-build-list -a :-/
[11:09:16] <deep42thought> ... aaand it failed :-(
[11:09:40] <abaumann> :-(
[11:11:12] <deep42thought> some deleted package was tried to be scheduled again
[11:12:10] <deep42thought> ... next trial :-)
[11:15:19] <buildmaster> any/absl-py is broken (says rechenknecht): https://archlinux32.org
[11:17:59] <deep42thought> ok, it does not schedule *everything* - just *almost* everything ;-)
[11:20:26] <abaumann> well, eurobuild6 has power :-)
[11:21:25] <buildmaster> i686/accounts-qml-module is broken (says eurobuild6-4): https://archlinux32.org
[11:21:26] <deep42thought> if it does not schedule that much every time, then all should be ok
[11:21:57] <buildmaster> i686/alsa-utils are broken (says eurobuild6-2): https://archlinux32.org
[11:21:59] <buildmaster> pentium4/alsa-utils are broken (says eurobuild6-5): https://archlinux32.org
[11:22:10] <deep42thought> my fear is, that we are hosting a race between the (broken) scheduler and the overwhelmed build-slaves here ...
[11:22:26] <buildmaster> pentium4/accounts-qml-module is broken (says eurobuild6-6): https://archlinux32.org
[11:26:49] <deep42thought> on a second run, `seed-build-list -a` schedules some packages *again* - this is wrong
[11:27:03] <abaumann> the same packages?
[11:27:11] <deep42thought> yes
[11:27:13] <abaumann> mmh
[11:27:23] <deep42thought> firefox-developer-edition-i18n for example
[11:27:37] <deep42thought> that's the reason, why I disabled the timer, I think
[11:27:46] <abaumann> which is dangling because firefox-developer-edition is blacklisted?
[11:27:54] <abaumann> I would leave it disabled for now..
[11:28:20] <deep42thought> yes
[11:28:33] <deep42thought> ah, so we should blacklist that one now, too
[11:30:04] <deep42thought> remember: I removed the dependency tracking from the blacklisting logic
[11:30:25] <deep42thought> (which accellerated get-package-updates from ~30min to a few minutes runtime)
[11:30:42] <abaumann> I vaguely remember :-)
[11:31:32] <deep42thought> ah, there are all i18n-$lang variants blacklisted, but not the pkgbase itself :-/
[11:33:03] <deep42thought> interestingly, firefox-developer-edition is *not* blacklisted
[11:34:08] <abaumann> ah. maybe my love for it blacklisted it only in my mind? ;-)
[11:34:27] <deep42thought> ah, no
[11:34:33] <deep42thought> it's blacklisted on pentium4
[11:34:39] <deep42thought> the i18n is blacklisted as any
[11:34:43] <deep42thought> my bad, sry
[11:36:13] <deep42thought> ok, so the only mistake was to blacklist each pkgname instead of the pkgbase of the split package
[11:37:22] <deep42thought> kernel update, I'll be back in a minute
[11:37:25] -!- deep42thought has quit [Quit: leaving]
[11:38:31] <abaumann> so, the convoluted old mailsetup via my host for buildmaster mails has gone, the machine can now send status messgaes as root.
[11:39:32] <abaumann> postfix is on loopback-only, so it can not listen to anything from outside
[11:39:38] <abaumann> seems to be the most effective way
[11:39:47] <abaumann> additionally to the firewall
[11:39:47] -!- deep42thought has joined #archlinux32
[11:39:48] <buildmaster> Hi deep42thought!
[11:39:48] <buildmaster> !rq deep42thought
[11:39:50] <phrik> buildmaster: <deep42thought> I like to move it, move it; I like to move it, move it
[11:40:30] <deep42thought> ok, nice :-)
[11:40:53] <deep42thought> and we can - at some point - set up aliases and mailing lists on the buildmaster itself, right?
[11:41:14] <abaumann> but then it has to listen
[11:41:27] <deep42thought> yes
[11:41:42] <deep42thought> but that's a rather trivial change, isn't it?
[11:42:05] <abaumann> yes
[11:46:55] <abaumann> so, me cooking now..
[11:47:11] <abaumann> I will continue with registration emails in the late afternoon, evening, most likely
[11:47:42] <deep42thought> we were offline for some weeks now, I think a day or two more don't really matter now ;-)
[11:47:45] <deep42thought> thanks!
[11:48:20] <abaumann> np
[11:48:24] <abaumann> true :-)
[11:52:01] <buildmaster> pentium4/avahi is broken (says eurobuild6-6): https://archlinux32.org
[11:52:18] <buildmaster> i686/avahi is broken (says eurobuild6-3): https://archlinux32.org
[11:52:20] <deep42thought> seed-build-list -a ran through once (manually)
[11:52:28] <deep42thought> let's see, what changes it brought :-)
[11:55:19] <abaumann> a peak - in packages :-)
[11:55:25] <deep42thought> that for sure
[11:55:41] <deep42thought> don't burn your meal ;-)
[11:59:05] <buildmaster> i486/gpgme is broken (says nlopc46-i486bs0): https://archlinux32.org
[12:00:13] <abaumann> nono, it's self-cooking :-)
[12:00:33] <deep42thought> !grab abaumann
[12:00:34] <phrik> deep42thought: Tada!
[12:03:41] <buildmaster> i486/samba is broken (says nlopc46-i486bs1) - already flagged out-of day upstream on 2019-11-01: https://archlinux32.org
[12:04:22] <buildmaster> i686/libproxy is broken (says eurobuild6-1): https://archlinux32.org
[12:04:48] <buildmaster> pentium4/libproxy is broken (says nlopc46): https://archlinux32.org
[12:05:01] <buildmaster> pentium4/qt5-base is broken (says eurobuild6-6): https://archlinux32.org
[12:06:11] <buildmaster> i686/qt5-base is broken (says eurobuild6-3): https://archlinux32.org
[12:06:13] <buildmaster> pentium4/krb5 is broken (says eurobuild6-5): https://archlinux32.org
[12:07:50] <buildmaster> i686/gpgme is broken (says eurobuild6-4): https://archlinux32.org
[12:08:21] <buildmaster> i686/krb5 is broken (says rechenknecht): https://archlinux32.org
[12:09:59] -!- deep42thought has quit [Quit: leaving]
[12:10:56] <buildmaster> i486/libpsl is broken (says nlopc46-i486bs1): https://archlinux32.org
[12:12:28] <buildmaster> i686/libpsl is broken (says eurobuild6-2): https://archlinux32.org
[12:12:46] -!- abaumann has quit [Quit: leaving]
[12:16:38] <buildmaster> i486/krb5 is broken (says nlopc46-i486bs0): https://archlinux32.org
[12:18:47] <buildmaster> pentium4/gpgme is broken (says eurobuild3): https://archlinux32.org
[12:22:44] <buildmaster> i486/pacman is broken (says nlopc46-i486bs1): https://archlinux32.org
[12:26:51] <buildmaster> i486/shadow is broken (says nlopc46-i486bs0): https://archlinux32.org
[12:26:56] <buildmaster> pentium4/systemd is broken (says rechenknecht): https://archlinux32.org
[12:32:10] <buildmaster> i486/systemd is broken (says nlopc46-i486bs1): https://archlinux32.org
[12:37:35] <buildmaster> i686/samba is broken (says eurobuild3) - already flagged out-of day upstream on 2019-11-01: https://archlinux32.org
[12:38:17] <buildmaster> pentium4/libaccounts-glib is broken (says nlopc46): https://archlinux32.org
[12:39:34] <buildmaster> i686/pacman is broken (says eurobuild6-4): https://archlinux32.org
[12:41:35] -!- buildmaster has quit [Remote host closed the connection]
[12:41:48] -!- buildmaster has joined #archlinux32
[12:41:48] <buildmaster> !rq buildmaster
[12:41:49] <phrik> buildmaster: <buildmaster> I might be insane, but never confused ... ;-)
[12:43:04] <buildmaster> i486/xfsprogs are broken (says nlopc46-i486bs0): https://archlinux32.org
[12:45:14] <buildmaster> pentium4/libgit2-glib is broken (says buildknecht2): https://archlinux32.org
[12:46:35] <buildmaster> pentium4/gom is broken (says eurobuild6-6): https://archlinux32.org
[12:47:48] <buildmaster> i686/systemd is broken (says eurobuild6-2): https://archlinux32.org
[12:48:52] <buildmaster> pentium4/libplist is broken (says nlopc46): https://archlinux32.org
[12:50:53] -!- buildmaster has quit [Remote host closed the connection]
[12:51:05] -!- buildmaster has joined #archlinux32
[12:51:05] <buildmaster> !rq buildmaster
[12:51:06] <phrik> buildmaster: <buildmaster> I might be insane, but never confused ... ;-)
[12:59:53] <buildmaster> pentium4/mpc is broken (says rechenknecht): https://archlinux32.org
[13:00:53] <buildmaster> pentium4/libimobiledevice is broken (says eurobuild6-4): https://archlinux32.org
[13:01:28] -!- samantaz_ has joined #archlinux32
[13:09:31] <buildmaster> pentium4/libsignon-glib is broken (says eurobuild6-6): https://archlinux32.org
[13:10:13] <buildmaster> pentium4/xapps are broken (says nlopc46) - already flagged out-of day upstream on 2019-11-14: https://archlinux32.org
[13:12:12] -!- urraka has joined #archlinux32
[13:20:38] <buildmaster> any/dbeaver-plugin-apache-poi is broken (says nlopc46): https://archlinux32.org
[13:23:00] <buildmaster> any/python-wsgi-intercept is broken (says rechenknecht): https://archlinux32.org
[13:24:14] <buildmaster> any/dbeaver-plugin-batik is broken (says eurobuild6-2): https://archlinux32.org
[13:24:27] -!- MrBIOS has joined #archlinux32
[13:39:00] <buildmaster> pentium4/qgo is broken (says buildknecht2): https://archlinux32.org
[13:42:08] -!- vboxler has joined #archlinux32
[13:42:49] -!- vboxler has quit [Client Quit]
[13:42:59] <buildmaster> pentium4/graphviz is broken (says eurobuild6-1): https://archlinux32.org
[13:46:31] <buildmaster> any/python-zope.testing is broken (says nlopc46): https://archlinux32.org
[13:47:01] <buildmaster> pentium4/speedcrunch is broken (says rechenknecht): https://archlinux32.org
[13:55:26] * buildmaster goes insane.
[14:07:48] -!- Alina-malina has quit [Ping timeout: 245 seconds]
[14:17:17] -!- Alina-malina has joined #archlinux32
[14:20:25] * buildmaster resumes sanity.
[14:20:31] <buildmaster> pentium4/kphotoalbum is broken (says eurobuild6-6): https://archlinux32.org
[14:20:50] <buildmaster> pentium4/pantheon-videos are broken (says eurobuild6-2): https://archlinux32.org
[14:20:51] <buildmaster> pentium4/pantheon-screenshot is broken (says eurobuild6-1): https://archlinux32.org
[14:23:52] <buildmaster> pentium4/pantheon-shortcut-overlay is broken (says eurobuild6-4): https://archlinux32.org
[14:28:37] <buildmaster> any/dbeaver-plugin-bouncycastle is broken (says nlopc46): https://archlinux32.org
[14:32:56] -!- deep42thought has joined #archlinux32
[14:32:56] <buildmaster> Hi deep42thought!
[14:32:56] <buildmaster> !rq deep42thought
[14:32:58] <phrik> buildmaster: <deep42thought> I've read, that the new cairo may add huuuugeeee spacing t o e a c h l e t t e r
[14:35:35] <deep42thought> abaumann: we might also consider adding a DMARC record to improve ratings?
[14:37:01] <buildmaster> pentium4/pantheon-camera is broken (says buildknecht2): https://archlinux32.org
[14:41:09] <buildmaster> i686/llvm is broken (says eurobuild6-5): https://archlinux32.org
[14:41:13] <buildmaster> pentium4/llvm is broken (says eurobuild6-3): https://archlinux32.org
[14:42:07] <buildmaster> pentium4/yubico-pam is broken (says eurobuild6-2): https://archlinux32.org
[14:42:38] <buildmaster> pentium4/pantheon-terminal is broken (says eurobuild6-1): https://archlinux32.org
[14:42:57] <buildmaster> pentium4/libmirage is broken (says eurobuild6-4): https://archlinux32.org
[14:47:38] <buildmaster> pentium4/repose is broken (says eurobuild6-4): https://archlinux32.org
[14:51:29] <buildmaster> pentium4/switchboard is broken (says eurobuild6-2): https://archlinux32.org
[14:51:52] -!- samantaz_ has quit [Ping timeout: 252 seconds]
[14:58:37] <buildmaster> any/riot is broken (says eurobuild6-6): https://archlinux32.org
[15:03:54] <deep42thought> pyparsing is still missing/broken, hmmm
[15:04:03] <buildmaster> pentium4/tipp10 is broken (says eurobuild6-3): https://archlinux32.org
[15:04:20] <buildmaster> pentium4/fcitx is broken (says buildknecht2): https://archlinux32.org
[15:04:53] <deep42thought> buildmaster: shut up
[15:04:54] <buildmaster> Sorry, I will do.
[15:24:42] -!- deep42thought has quit [Quit: leaving]
[15:56:32] -!- djmoch has quit [Quit: ZNC 1.7.5 - https://znc.in]
[15:59:24] -!- djmoch has joined #archlinux32
[16:05:05] -!- AquilaIrreale has joined #archlinux32
[16:17:33] <buildmaster> any/python-flask-babelex is broken (says eurobuild6-6): https://archlinux32.org
[16:18:15] <buildmaster> any/python-flask-gravatar is broken (says eurobuild6-2): https://archlinux32.org
[16:21:01] <buildmaster> any/python-flask-mail is broken (says eurobuild6-3): https://archlinux32.org
[16:22:17] <buildmaster> any/python-flask-paranoid is broken (says eurobuild6-1): https://archlinux32.org
[16:22:56] <buildmaster> any/python-flask-principal is broken (says eurobuild6-4): https://archlinux32.org
[16:26:10] -!- vboxler has joined #archlinux32
[16:26:43] <vboxler> titzus
[16:27:12] -!- vboxler has parted #archlinux32
[16:30:20] <buildmaster> any/python-sshtunnel is broken (says eurobuild6-1): https://archlinux32.org
[16:34:31] <buildmaster> any/python-urwidtrees are broken (says eurobuild6-1): https://archlinux32.org
[16:36:37] -!- AquilaIrreale has quit [Remote host closed the connection]
[16:53:34] -!- urraka has quit [Quit: WeeChat 2.6]
[17:04:08] <buildmaster> any/python-flask-sqlalchemy is broken (says buildknecht2): https://archlinux32.org
[17:14:43] <buildmaster> any/python-sqlalchemy-continuum is broken (says eurobuild6-6): https://archlinux32.org
[17:32:20] <nit-picker> key 2FF1E976D6EB2E954A87DC14443904EC9EC51A8A (from Roelf Wichertjes (archlinux32 master key) <contact@roelf.org>) in package archlinux32-keyring-transition-20191103-1-any.pkg.tar.xz expires on 2019-11-13 (in -3 < 100 days).
[17:42:21] <nit-picker> key 2FF1E976D6EB2E954A87DC14443904EC9EC51A8A (from Roelf Wichertjes (archlinux32 master key) <contact@roelf.org>) in package archlinux32-keyring-20191103-1.0-any.pkg.tar.xz expires on 2019-11-13 (in -3 < 100 days).
[17:50:09] -!- abaumann has joined #archlinux32
[17:50:10] <buildmaster> Hi abaumann!
[17:50:10] <buildmaster> !rq abaumann
[17:50:11] <phrik> buildmaster: <abaumann> they do all kind of weird stuff, nobody else does, because they are known not to work. :-)
[17:50:45] <abaumann> DMARC, yeah, maybe later. It should all not matter, Yahoo is just not accepting those emails..
[17:51:17] <abaumann> as the buildmaster is not in SPF, I'm relaying currently over my mail server
[17:52:53] <abaumann> levi: as you where moderator before, you get the privilege to register first again - after admins, that is..:-)
[17:52:59] <abaumann> *were
[17:53:27] <abaumann> I also have now a registration for bugs.archlinux32.org: this manual registration before keeps people away.
[17:53:40] <abaumann> this also means, you have to solve now Google puzzles :->
[18:05:56] -!- samantaz_ has joined #archlinux32
[18:28:20] <buildmaster> any/cairosvg is broken (says eurobuild6-6): https://archlinux32.org
[18:30:34] -!- {levi} has joined #archlinux32
[18:31:25] <abaumann> levi: check if you have moderators right now.
[18:32:09] <abaumann> https://bbs.archlinux32.org
[18:32:10] <phrik> Title: Installation of pentium4 optimized packages / Installation / Archlinux32 Forum (at bbs.archlinux32.org)
[18:32:21] <abaumann> we should try not to simply copy-paste stuff from the old static pages.
[18:32:32] <abaumann> We should clean up and correct wrong information
[18:32:46] <abaumann> and we should really stick more to the forum structure as in the past.
[18:33:13] <abaumann> also, we should invite people to use the bug reporter for bugs, as this make managing the bugs so much easier than search through the forum.
[18:33:28] <{levi}> Yes, I seem to be able to see at least parts of the moderator system
[18:33:45] <abaumann> I might have to fine tune some permissions..
[18:34:22] <{levi}> Okay, I can search for users and get details, but I couldn't see the move post/delete options on the post I just checked
[18:35:14] <{levi}> But there are limited posts to check now, and they're all made by administrators which might be why
[18:35:28] <abaumann> yes. I think so.
[18:35:39] <abaumann> "The recovered forum pages can be found here."
[18:35:50] <abaumann> there are the archived pages. I put a note on the top.
[18:36:35] <abaumann> It will take a while to get the forum back as it was, I'm afraid. :-(
[18:36:49] <abaumann> On the positive side: we don't look dead anymore :-)
[18:37:44] <{levi}> Yes, I had a dig into those pages yesterday. Thanks for finding an archiving them all
[18:37:53] <buildmaster> any/pifpaf is broken (says eurobuild6-6): https://archlinux32.org
[18:38:03] <abaumann> np
[18:38:26] <{levi}> I do tend to agree we should have fewer categories, but I'm not sure that going through all of the old posts one by one and recategorising them will get us very far
[18:38:32] <{levi}> And certainly not very fast
[18:38:49] <buildmaster> pentium4/opendht is broken (says eurobuild6-4): https://archlinux32.org
[18:39:03] <{levi}> But if we devising many->one mappings from old categories to new ones turns out to be easy, we could probably do that
[18:39:22] <buildmaster> any/jupyter-notebook is broken (says eurobuild6-3) - already flagged out-of day upstream on 2019-11-08: https://archlinux32.org
[18:39:40] <{levi}> s/devising/could devise/
[18:39:44] <abaumann> In flyspray I know exactly what is missing and could recover most things.
[18:39:49] <buildmaster> pentium4/openvdb is broken (says buildknecht2): https://archlinux32.org
[18:39:51] <abaumann> If unkown I made a placeholder bug.
[18:39:58] <abaumann> For the forum it's very different..
[18:40:09] <abaumann> ..I have no clue how much I could rescue
[18:40:11] <{levi}> Indeed
[18:40:47] <abaumann> I would say: prio 1 is installation and internal documentation (how to set up slaves, mirrors, etc.)
[18:41:05] <abaumann> second is maybe hints on how to build things, but those should IMHO go to bugs.archlinux32.org
[18:41:42] <abaumann> hardware setups and guides for specific problems (like xorg.conf setting for specific hardware) comes maybe next.
[18:42:10] <{levi}> Can bugs-dot capture things that aren't bugged like unusual but valid build processes?
[18:42:57] <abaumann> mmh. dunno.
[18:42:58] <{levi}> I'm thinking we need categories that are easy for users to select from when posting.
[18:43:24] <buildmaster> any/pylama is broken (says nlopc46): https://archlinux32.org
[18:43:28] <abaumann> yeah. the fact we got everything everywhere is an indiacation that the forum structure was not obvious
[18:43:32] <abaumann> buildmaster: shut up
[18:43:32] <buildmaster> Sorry, I will do.
[18:43:38] <{levi}> But most posts were bug reports, and if we want to discourage them then maybe we want to eliminate all places they could reasonabally post
[18:44:05] <{levi}> And have a notice saying don't post bug reports here somewhere obvious
[18:44:16] <abaumann> Image a new user:
[18:44:31] <abaumann> He/she doesn't know whether it's a bug, has to open accounts in two systems
[18:44:36] <abaumann> this is hurdle.
[18:45:00] <abaumann> so maybe a process could also be: let them post first in the forum, then we can make a bug out of it and link it.
[18:45:19] <{levi}> Yes. Should we encorage them to only register for the bug tracker, and make their experiments there? Still seems something of a stretch with many users.
[18:45:21] <abaumann> also the reported problem way have a n:m relation to bugs
[18:45:50] <abaumann> main rule: we should just restore what is really important information for users
[18:46:02] <{levi}> Yes, I don't have an existential problem with bug reports starting in the forum, but they should migrate to bug reports sooner rather than later
[18:47:03] <abaumann> yes. that's why I'm also allowing automatic registration in flyspray (bugs.archlinux32.org) now
[18:47:17] <abaumann> this manual process was a little bit cumbersome
[18:47:33] <abaumann> let's see, how much spam I get on my mail server :-)
[18:47:58] -!- deep42thought has joined #archlinux32
[18:48:06] <deep42thought> abaumann: the buildmaster _is_ in spf
[18:48:30] <deep42thought> it has a mx record and that appears in spf
[18:48:42] <abaumann> deep42thought: maybe I have a stale DNS cache entry?
[18:48:44] <abaumann> mmh
[18:48:52] <deep42thought> for which?
[18:48:59] <deep42thought> spf had the mx since a long time
[18:49:10] <abaumann> archlinux32.org. 600 IN TXT "v=spf1 mx include:andreasbaumann.cc include:tyzoid.com include:aperture-laboratories.science -all"
[18:49:23] <deep42thought> there is the "mx"
[18:49:43] <abaumann> ah.
[18:49:47] <abaumann> didn't read that one
[18:49:48] <deep42thought> :-)
[18:49:48] <abaumann> sorry
[18:49:51] <deep42thought> np
[18:50:21] <abaumann> so far I see no difference going over my mailserver or mail3/buildmaster
[18:50:47] <abaumann> so I could switch now actually
[18:52:39] <abaumann> "Nov 16 18:52:23 buildmaster.archlinux32.org postfix/smtp[1314901]: C77A950E1359: host gmail-smtp-in.l.google.com[] said: 421-4.7.0 [ 15] Our system has detected an unusual rate of 421-4.7.0 unsolicited mail originating from your IP address. To protect our 421-4.7.0 users from spam, mail sent from your IP address has been temporarily 421-4.7.0 rate limited. Please
[18:52:45] <abaumann> visit 421-4.7.0 https://support.google.com to 421 4.7.0 review our Bulk Email Senders Guidelines. v15si4872158wrc.423 - gsmtp (in reply to end of DATA command)
[18:52:46] <phrik> Title: Sender Guidelines - Gmail Help (at support.google.com)
[18:52:49] <abaumann> *gggr*
[18:53:04] <abaumann> rate limit, wtf
[18:53:08] <deep42thought> this is the buildmaster's ip?
[18:53:15] <abaumann> yes
[18:53:32] <abaumann> yeah. that's Google being Google
[18:54:03] <abaumann> bad postmaster tools, no email contact, intransparent rules, I got blocked 2 months after a boo-boo from my side :-)
[18:54:21] <abaumann> ok, relaying it is for now.
[18:54:59] <{levi}> Oh, do I have a log in to fly spray automatically now?
[18:55:11] <{levi}> I assume my old login is nixed
[18:55:33] <abaumann> you assume correctly :-)
[18:56:08] <{levi}> I don't have a login to flyspray though
[18:56:16] <deep42thought> I see no advantage of starting maybe-bugs in the forum instead of the bug tracker
[18:56:32] <abaumann> levi: you can register for a new account now
[18:57:00] <{levi}> Okay, am doing
[18:57:18] <abaumann> I tested before with a Gmail address, so I hope it works
[18:57:43] <deep42thought> now, that the registration on the bugtracker is easier, we could enforce opening issues there instead in the forums
[18:57:56] <abaumann> deep42thought: yeah, but the fact is, people rather use a forum to complain than a bug report system which actually asks them to investigate before posting. :-)
[18:58:08] <abaumann> that was exactly my thinking
[18:58:57] <deep42thought> well, we have too few hands to hold each user's hand ;-)
[18:59:20] <abaumann> !grab deep42thought
[18:59:21] <phrik> abaumann: Tada!
[18:59:22] <deep42thought> brb
[19:00:09] <{levi}> Few, finally got the google recaptcha thing to work
[19:00:14] <{levi}> s/Few/Phew/
[19:00:26] <abaumann> :->
[19:00:55] <abaumann> You officially qualified as non-robot now :-)
[19:01:27] <{levi}> Yeah, I always have to click on reams of images due to my security settings and the fact that google doesn't know me any more
[19:01:45] <{levi}> No email yet though
[19:02:12] <{levi}> I spoke too soon, and maybe I was checking the wrong account
[19:02:17] <abaumann> uff.
[19:02:36] <abaumann> I'm setting up a mailserver roughly every 5 years, so I might be a little bit rusty. :-)
[19:04:32] <{levi}> All seems good now. actually
[19:04:58] <abaumann> Ok, let me also raise your permissions there a little bit.
[19:05:35] <{levi}> Ta, it's time to eat here now, so I'll test later
[19:05:45] <abaumann> ok, cu later
[19:20:17] <abaumann> 32.arlm.tyzoid.com/ is down, http://arch32.mirrors.simplysam.us shows Access forbidden
[19:33:10] <deep42thought> abaumann: that's hosted on tyzoid's server, too, I think
[19:47:10] <abaumann> deep42thought: yeah, guessed so.
[19:55:09] <{levi}> Okay, cool, I have extra buttons now
[19:55:14] <{levi}> On flyspray that is
[19:55:22] <deep42thought> :-D
[19:55:40] * deep42thought imagines a shirt with more buttons than holes
[19:59:15] <{levi}> I have a couple of those. Spare buttons at the bottom below the waist band :)
[19:59:26] <deep42thought> ah, right, they exist :-)
[20:00:13] <{levi}> :)
[20:01:20] <deep42thought> abaumann: I would remove your i486 doc-patches from shadow - it looks, like it misses files in package() otherwise
[20:03:12] <{levi}> BTW, on the adminstration menu of the bbs, it tells me that install.php still exists
[20:03:25] <deep42thought> I saw that, too
[20:03:31] <deep42thought> we do not need it anymore, right?
[20:03:38] <abaumann> levi: try to execute it :-)
[20:04:02] <abaumann> I have removed access from the outside in Apache, I want to check in first into git.
[20:04:50] <{levi}> Okay, it reckons it's been deleted
[20:05:00] <{levi}> I assume that's what you meant
[20:05:06] <abaumann> deep42thought: if you remove the 486 doc patches from shadow will no longer build
[20:05:17] <abaumann> I doubt gnome-doc-utils is now available on i486
[20:05:49] <abaumann> oh. I don't have a git yet for that.
[20:05:57] <abaumann> ok, let me move install.php out of the way for now
[20:06:01] <deep42thought> I'm building it, currently
[20:06:16] <deep42thought> shadow^
[20:06:23] <abaumann> ah yeah.
[20:06:36] <abaumann> oh. xml2po is available now?
[20:06:43] <abaumann> then yes. it can be dropped.
[20:06:54] <abaumann> was just there during bootstrapping IIRC
[20:07:27] <abaumann> the missing files were the man-pages of shadow :-)
[20:09:29] <deep42thought> looks, good - I update the git
[20:09:37] <abaumann> :-)
[20:10:25] <deep42thought> gettext has icu-problems on i486 :-(
[20:12:24] <abaumann> i486 has i486 problems :-)
[20:12:36] <abaumann> the are many things broken there in core I noticed.
[20:13:10] <deep42thought> shouldn't we fix it, then?
[20:13:12] <deep42thought> :-D
[20:13:29] <deep42thought> otherwise I might be offline soon ;-)
[20:13:30] <abaumann> yep.
[20:13:32] <abaumann> " Query failed: Deadlock found when trying to get lock; try restarting transaction"
[20:13:35] <abaumann> oups.
[20:13:36] <abaumann> https://buildmaster-status.archlinux32.org
[20:14:21] <abaumann> yeah. updating tables which printing a long error list might result in that
[20:14:40] <abaumann> pacman fails on i686
[20:14:48] <abaumann> systemd fails on pentium4
[20:14:57] <deep42thought> otoh: truncating that list still needs us to have the full list in the database
[20:14:57] <abaumann> llvm on pentium4
[20:19:16] <deep42thought> should I simply add some "LIMIT" to that mysql query?
[20:20:16] <abaumann> maybe yes.
[20:20:51] <abaumann> we should aim at less broken packages anyway ;-)
[20:20:56] <deep42thought> hmm, this query employs a lot of temporary tables, I'm not certain *where* I should limit :-/
[20:21:02] <deep42thought> !grab abaumann
[20:21:07] <phrik> deep42thought: Tada!
[20:21:07] <deep42thought> lol
[20:21:12] <abaumann> I'm not even sure where this happens
[20:21:20] <abaumann> it was the very first time I saw this
[20:22:03] <deep42thought> that's the problem with running the web frontend on the buildmaster: any query on the buildmaster can block the web frontend and vice versa
[20:22:35] <abaumann> Yes, I'm especially worried about the web frontend blocking the buildmaster and causing data havock there.
[20:22:38] <deep42thought> otherwise, the webfrontend will simply display stale information ("stale" meaning "a few seconds/minutes old")
[20:22:40] <abaumann> the other way round is not dramatic.
[20:22:56] <abaumann> why did I do that in the first place?
[20:22:58] <abaumann> ah.
[20:23:06] <deep42thought> with the web frontend on archlinux32.org, the buildmaster cannot block the webfrontend
[20:23:07] <abaumann> buildmaster-status was only a backup
[20:23:14] <deep42thought> right
[20:23:22] <deep42thought> you should probably block public access
[20:23:28] <abaumann> so, let's close that one down.
[20:23:29] <abaumann> exactly
[20:23:32] <deep42thought> :-)
[20:24:51] <abaumann> so, let's hop nagios munin to buildmaster.archlinux32.org/nagios resp. munin and close down buildmaster-status.archlinux32.org
[20:25:20] <deep42thought> nagios and munin can be hosted on a different machine?
[20:25:45] <abaumann> in principle
[20:25:55] <abaumann> if I move it, use munin-node and nrpe to communicate
[20:26:01] <deep42thought> you could also keep them on the buildmaster ...
[20:26:06] <abaumann> yeah.
[20:26:13] <deep42thought> seems more simple and more robust
[20:26:14] <abaumann> i just want to hov the virtual host
[20:26:20] <abaumann> *hop
[20:26:28] <deep42thought> hop?
[20:26:29] <abaumann> https://buildmaster.archlinux32.org and https://buildmaster.archlinux32.org
[20:26:35] <abaumann> instead of buildmaster-status
[20:26:58] <abaumann> hop == confiugre :-)
[20:27:53] <deep42thought> hmm, why not keep buildmaster-status for those two subdirs?
[20:28:02] <deep42thought> I really liked the previous way
[20:28:44] <deep42thought> ah, no, we never redirected from buildmaster-status
[20:28:56] <abaumann> ah. ok
[20:29:05] <abaumann> slightly more complicated configuration then :-)
[20:29:21] <deep42thought> oh, now I see, what you meant
[20:29:27] <deep42thought> ok, your proposal is fine
[20:29:32] <abaumann> :-)
[20:29:51] <deep42thought> buidlmaster.archlinux32.org redirects to archlinux32.org/buildmaster - except for nagios/ and munin/ which is hosted like before
[20:29:59] <abaumann> exactly
[20:30:16] <deep42thought> (somehow I was confused and thought, buildmaster.archlinux32.org was hosted on archlinux32.org, too)
[20:30:25] -!- waldner has quit [Quit: leaving]
[20:30:52] <buildmaster> pentium4/vtk is broken (says eurobuild6-2): https://archlinux32.org
[20:46:41] <buildmaster> i486/gettext is broken (says nlopc46-i486bs0): https://archlinux32.org
[20:47:16] <buildmaster> pentium4/gettext is broken (says eurobuild6-3): https://archlinux32.org
[20:48:00] <buildmaster> i686/gettext is broken (says eurobuild6-6): https://archlinux32.org
[20:48:08] <deep42thought> how can this be? I just built gettext successfully on nlopc46-i486bs0 !!!
[20:49:02] <abaumann> mmh.
[20:49:08] <abaumann> automagically fixed?
[20:49:14] <deep42thought> hmm, I built the version from trunk
[20:49:33] <abaumann> ah.
[20:49:36] <deep42thought> asp32 -u export core/shadow
[20:49:37] <deep42thought> error: package 'shadow' not found in repo 'core-i486'
[20:49:45] <deep42thought> this is not optimal :-/
[20:49:49] <abaumann> yeah. something broke there.
[20:50:02] <abaumann> but I'm actually not even sure: is asp still supported upstream?
[20:50:10] <deep42thought> why not?
[20:50:22] <deep42thought> everything is in git, so why should support cease?
[20:50:38] <abaumann> ah, mistake: ABS is not asp
[20:50:50] <deep42thought> :-)
[20:54:59] <abaumann> https://git.archlinux32.org
[20:55:29] <deep42thought> nothing hosted yet?
[20:55:42] <abaumann> yes, it is
[20:55:54] <abaumann> just customization and the repos are missing
[20:55:55] <deep42thought> "No repositories found"
[20:56:00] <deep42thought> ah, ok :-)
[20:56:11] <deep42thought> you allow me to be nit-picky?
[20:56:18] <abaumann> sure.
[20:56:28] <deep42thought> drop the /cgit/
[20:56:32] <deep42thought> :-)
[20:56:34] <abaumann> you wrote the nit-picker :-)
[20:56:50] <abaumann> yeah. that's rather ugly. I also noticed
[20:57:04] <nit-picker> he sure did write me
[20:57:12] <abaumann> lol
[20:57:43] <deep42thought> I'm wondering, why I actually log in to chat here - I could send via buildmaster and nit-picker and read in the log from titus_livius ...
[20:58:07] <abaumann> could get confusing :-)
[20:59:02] <deep42thought> btw: Is it possible to order the nagios services in a hierarchy?
[20:59:14] <abaumann> "chat-robot-induced-scizophrenia" ;-)
[20:59:17] <abaumann> I'm sure it is.
[20:59:21] <abaumann> somehow
[20:59:26] <deep42thought> e.g. if some server fails completely, there is not a notification for each of its services
[20:59:40] <abaumann> ah. yes. this can be done in a tree of dependencies
[21:01:05] <abaumann> I just have to remember how
[21:01:28] <deep42thought> Hmm, I don't get it - shadow from core builds fine ...
[21:09:32] <{levi}> Regarding the bbs categories, we can eliminate Press Reviews, combine Pacman/Upgrades with Creating Packages, merge Kernel & Hardware and Servers and Networking and System Administration into a new bug reports forum and probably merge the remaining 'contributing' categories.
[21:10:48] <deep42thought> why have bug report category in the forums?
[21:11:31] <{levi}> Becuase users often need a little chat before they raise a bug in flyspray
[21:12:18] <{levi}> But it should have a description linking to flyspray if possible, and a sticky post saying raising a bug will have better visibility if you are able to.
[21:12:21] <deep42thought> hmm, you're the master of the forums ;-)
[21:12:58] <{levi}> Erk
[21:13:05] <abaumann> yep :-)
[21:14:43] <abaumann> Actually 'contributing' is for people wanting to help (for writting), the 'Technical Issues' if for end user having problems in specific areas.
[21:15:27] <deep42thought> yeah, I think we should make the distinction between "people wanting to hack archlinux32" and "people wanting to use arhclinux32" more clear
[21:15:44] <abaumann> the subsections in there may vary
[21:16:27] <{levi}> Not too much I'd hope. That was the old problem ;)
[21:17:27] <abaumann> I wonder anyway how many people will register?
[21:17:51] <deep42thought> I know of 3 :-D
[21:17:52] <abaumann> deep42thought: shadow fails big times for me in gettext-stuff
[21:18:00] <{levi}> We already have 3
[21:18:10] <deep42thought> yes, those are the ones I meant
[21:18:20] <{levi}> :)
[21:18:46] <abaumann> checking for msgfmt... no
[21:18:47] <abaumann> checking for gmsgfmt... :
[21:18:47] <abaumann> checking for xgettext... no
[21:18:47] <abaumann> checking for msgmerge... no
[21:18:53] <abaumann> this looks like a missing gettext to me
[21:19:03] <deep42thought> it works with packages from testing
[21:19:38] <deep42thought> that's the only difference I can spot
[21:19:44] <abaumann> let's see what happens, when I add a 'gettext' to makedepends
[21:19:46] <deep42thought> but it might be a sufficient difference ;-)
[21:19:50] <abaumann> xmlcatalog: error while loading shared libraries: libicuuc.so.64: cannot open shared object file: No such file or directory
[21:19:53] <abaumann> oups
[21:19:56] <deep42thought> gettext has linking issues
[21:19:59] <deep42thought> icu ones
[21:20:02] <{levi}> Oh that sounds like the old fun
[21:20:06] <abaumann> exactly
[21:20:38] <abaumann> maybe priorize docbook-xsl (xmlcatalog)
[21:20:49] <deep42thought> or gettext
[21:21:00] <abaumann> you see why I want a docdepends=()? in upstream pKGBULDS :-)
[21:21:05] <abaumann> or both
[21:21:34] <abaumann> docdepends draw in tons of unnecessary dependencies
[21:21:52] <abaumann> for man pages and documentation nobody reads usually anyway.. and heads directly to Google nowadays. :-)
[21:21:56] <deep42thought> I'll force gettext onto my build slave
[21:22:03] <abaumann> good idea
[21:22:15] <deep42thought> man.google.com?
[21:23:16] <abaumann> man has to be honoured, as it is basically the reason why Unix exists in the first place.
[21:26:06] <{levi}> Unix exists because Dennis has to be honoured
[21:26:39] <abaumann> I was referring to how they managed to program Unix by telling their bosses they where writing a new documentation tool :-)
[21:27:00] <{levi}> Ah, k
[21:30:07] <abaumann> *grr* this /cgit thingy is stubborn in Apache. I'll try to solve it tomorrow
[21:30:13] <abaumann> enough for today.. :-)
[21:30:18] <deep42thought> good night!
[21:30:25] <abaumann> cu
[21:30:26] -!- abaumann has quit [Quit: leaving]
[21:39:38] <buildmaster> i486/shadow is broken (says nlopc46-i486bs1): https://archlinux32.org
[22:37:35] -!- samantaz_ has quit [Ping timeout: 265 seconds]
[22:38:58] -!- deep42thought has quit [Quit: leaving]
[23:00:41] -!- thePiGrepper has joined #archlinux32
[23:11:39] -!- thePiGrepper has quit [Ping timeout: 240 seconds]
[23:13:11] -!- thePiGrepper has joined #archlinux32
[23:18:12] <buildmaster> pentium4/ispc is broken (says eurobuild6-3): https://archlinux32.org
[23:24:32] <buildmaster> any/mitmproxy is broken (says eurobuild6-4): https://archlinux32.org
[23:26:24] -!- thePiGrepper has quit [Ping timeout: 265 seconds]