#foswiki 2012-09-30,Sun

↑back Search ←Prev date Next date→ Show only urls(Click on time to select a line by its url)

WhoWhatWhen
gac410Is the Logger API part of our protected API. Reviewing http://foswiki.org/Development/PluginsApiPolicies it looks like a major API change for Logger->log() would be perfectly acceptable.
I'm looking at gyrations to preserve backwards compat for log() calls, but tbh, after a quick scan of svn, it would probably be fine to revamp the logger.
The only impact would be to core, to users who "break" the Func:: barrier, and to anyone who has written their own logger
[00:13]
Based on that, if the redesigned logger API is accepted, older user written loggers would need to be updated for 1.2. [00:22]
........................... (idle for 2h12mn)
pharveysounds good
I don't think the logger API is in the same category as Func :) Did we even have a logger API prior to 1.1.0?
[02:34]
..... (idle for 20mn)
gac410hm. good point pharvey - [02:54]
SvenDowideitgac410 i think cdot and i are the only people to use the logger api out of core [02:56]
gac410Yup. Foswiki 1.0.9 had a Logger::PlainFile [02:57]
SvenDowideitand you should never worry about compatibility for my stuff (at least while i'm still around)
as its at least60% trying to work out what is possible
[02:57]
gac410I've decided to scrap backwards compat. It's only a few lines of code in PlainFile and Compat to convert the hash back to an array so the old code runs. [02:58]
SvenDowideitsweet
actually, i lie a tiny amount
[02:58]
gac410The Loggers themselfs will all accept the old style call though. [02:58]
SvenDowideitimo having a legacy logger that outputs the old format is needed [02:58]
gac410Oh... they all do that. [02:59]
SvenDowideitbut i recon defaulting to a new more useable format would also be good
as the old form sux
and comes from a day (er, 1998) when the log file was a topic
[02:59]
gac410Even LogDispatch will *output* the legacy format. I'm going to change LogDispatch to be Template based, so you can define the format you want per file. Right now its a prototype line with tokens | $timestamp $level | $action | ... But I'm thinking about it .. maybe some way to avoid regex substitutions. [03:01]
SvenDowideit:) [03:02]
gac410I've got a logger branch on github, but I have not pushed to it yet. I wanted to get the changes and commit messages settled so they would be commitable to svn if the change is approved.
Once I push I can't be rebasing, and amending them.
[03:03]
SvenDowideitwell, you could
if you don't commit from that branch
and instead cherrypick and er, make alot of work for yourself :)
[03:04]
gac410I've added a Logger::setCommonFields() called as a SUPER:: method in the loggers, to pick up user, agent, remoteIP That way *any* event, info, debug, warn, etc. can have a common set of fields.
Rather than depending upon calling the logger via the Foswiki::writeEvent method, that filled in that info.
[03:05]
SvenDowideitexcellent :) [03:11]
.......... (idle for 45mn)
gac410Sven, pharvey. I've pushed my core and logger changes to my github logger branch - https://github.com/gac410/foswiki/commits/logger
I assigned a task, but none of this is in svn. I won't commit to trunk until the 14 days runs out on the proposal. (and assuming the 1.2 release manager agrees ) :P
That is assuming I don't fat finger a dcommit in the wrong branch
[03:56]
.............................................. (idle for 3h46mn)
pharveykewl
is the cool way of spelling meh
[07:44]
Babarwhat do you mean a lot of work for yourself?
I simply never ever work on any branch that's published.
it's just a way to work. When I start doing something, before I issue a commit, I check the branch, and if it's a published one, and I don't have a task, I just create a temporary one
then I can rebase all I want, and once I'm done, I can still dcommit from it
and even if sometimes I realise this too late, it's still easy to fix: you create a new branch to save your work, then you go back to the published branch, you rebase --hard the revision before your commits, et voilĂ .
[07:48]
...................... (idle for 1h49mn)
JeanBaptisteMayeHi
I found a small bug in 1.1.5, I have a one line patch to fix it, what is the best way to contribute it?
[09:39]
............ (idle for 56mn)
CDotJeanBaptisteMaye: raise a task in http://foswiki.org/Tasks
mark it as a release blocker, and it'll get attention
[10:36]
Babaror you can even request SVN access and commit it back yourself :) [10:43]
....................... (idle for 1h54mn)
***JeanBaptisteMaye has quit IRC (Ping timeout: 245 seconds) [12:37]
.... (idle for 15mn)
Babar sets mode: +v [LOGGER_FOSWIKI] [12:52]
.......... (idle for 45mn)
CDot has quit IRC (Quit: Leaving.) [13:37]
....... (idle for 33mn)
gac410hm. Is irclogs.foswiki.org down? It resolves but I don't get any response [14:10]
......... (idle for 40mn)
***kip3f has quit IRC (Ping timeout: 250 seconds) [14:50]
........................................... (idle for 3h31mn)
Babargac410: irclogs.foswiki.org is Colas's machine, afaik. So it could very well be down. [18:21]
..... (idle for 22mn)
***Demosthenes has quit IRC (Ping timeout: 246 seconds) [18:43]
.... (idle for 18mn)
JulianLevensI cannot get thru to irclogs.foswiki.org and/or the logs, is there a problem? [19:01]
ColasHomearg yes, the machine has changed its IP
I need to have irclogs.foswiki.org updated
[19:01]
in the meantime, I have set up another name that should work in ~ 1h at http://irclogs.foswiki.nahaboo.net
I mailed to SvenDowideit the new IP
[19:09]
.... (idle for 18mn)
BabarI'm sure Sven will update the DNS as soon as he gets up, so that should be OK [19:31]
..... (idle for 22mn)
***wdenk1 has quit IRC (Ping timeout: 245 seconds) [19:53]
.... (idle for 18mn)
BabarJeanBaptisteMaye: sorry, I should have thanked you in the commit. Anyway, I've fixed that bug at least. [20:11]
JeanBaptisteMayeNo - I need to thank you for committing on my behalf [20:12]
Babaroh, and gac410 could kill me as I ran the unit tests only on trunk :)
running them on the release branch now
[20:13]
JeanBaptisteMayeAbout this $_[1] thing - how come it can be faster? [20:16]
GithubBot[foswiki] FoswikiBot pushed 1 new commit to master: http://git.io/DhwN1A
[foswiki/master] Item12105: Do not use print directly. All output shall go through the Engine's write() - OlivierRaginel
[20:18]
***GithubBot has left [20:18]
FoswikiBothttp://foswiki.org/Tasks/Item12105 [ Item12105: FastCGIEngineContrib may issue body before headers ] [20:18]
harlanSvenDowideit: ping... [20:18]
.................................... (idle for 2h58mn)
***SvenDowideit has quit IRC (Changing host)
card.freenode.net sets mode: +o SvenDowideit
ChanServ sets mode: -o SvenDowideit
[23:16]
gac410howdy all ... anyone know the status of irclogs.foswiki.org? I'd check the logs .. but ;)
And back to my incessant git questions. So I've pushed my logger branch onto github with several commits that are not in svn. Does that mean now that I shouldn't "git svn rebase" and push svn updates to github?
I suppose the other option is to cherry-pick the commits from my master branch into logger, to avoid the rebase.
[23:25]

↑back Search ←Prev date Next date→ Show only urls(Click on time to select a line by its url)