| [00:00:07] |
<alect> |
trailing : |
| [00:01:19] |
<johnp> |
http://pastebin.com/818858 |
| [00:01:37] |
<coderanger> |
what alec said |
| [00:02:07] |
<johnp> |
im just modifying code adding one if statement |
| [00:02:16] |
<johnp> |
for a check for caps and numbers |
| [00:02:28] |
<coderanger> |
if needs to end with a ':' |
| [00:02:30] |
<johnp> |
if re.match('[0-9A-Z]', user) |
| [00:02:30] |
<johnp> |
req.hdf['registration.error'] = 'Username cannot contain numbers or capital letters.<br>Hopefully you are not a spambot. :)' |
| [00:02:30] |
<johnp> |
return |
| [00:02:37] |
<coderanger> |
all block statements in Python do |
| [00:02:43] |
<coderanger> |
if blah(): |
| [00:03:06] |
<johnp> |
ah |
| [00:03:07] |
<johnp> |
doh |
| [00:03:11] |
<johnp> |
im a python noob |
| [00:03:15] |
<johnp> |
;) |
| [00:03:22] |
<coderanger> |
diveintopython.org :) |
| [00:04:41] |
<johnp> |
NameError: global name 're' is not defined |
| [00:04:47] |
<johnp> |
do i need to at the top |
| [00:04:49] |
<johnp> |
import re |
| [00:05:11] |
<coderanger> |
yep |
| [00:12:50] |
<johnp> |
mmm getting weird errors |
| [00:13:52] |
<coderanger> |
such as? |
| [00:16:20] |
* |
groton has joined #trac |
| [00:17:01] |
* |
cmlenz has joined #trac |
| [00:17:18] |
* |
Guardian has quit IRC |
| [00:20:07] |
<pacopablo> |
alect: I'm going to have to try to figure it out another day. |
| [00:20:39] |
<pacopablo> |
alect: it looks like a query may be failing inside the same transaction, and is then nullifying the schema setting |
| [00:20:54] |
<pacopablo> |
alect: problem being is why it would do that on your server and not mine |
| [00:21:08] |
<pacopablo> |
alect: btw, trac-hacks.org is still up and running :) |
| [00:21:11] |
<pacopablo> |
night |
| [00:21:14] |
<coderanger> |
gnite |
| [00:21:51] |
* |
Guardian has joined #trac |
| [00:25:36] |
<alect> |
okay paco, thanks :) |
| [00:28:59] |
<asmodai> |
and back |
| [00:30:59] |
<johnp> |
so what are .pyc and .pyo files used for |
| [00:31:08] |
* |
s0undt3ch_ has joined #trac |
| [00:31:10] |
<alect> |
magic! |
| [00:33:20] |
<coderanger> |
johnp: They are (in various forms) compiled parse trees |
| [00:33:41] |
<coderanger> |
python will save itself some time if it can, and not reparse/recompile a module |
| [00:34:08] |
<johnp> |
so if i update the py file |
| [00:34:19] |
<johnp> |
does that mean it will show old results |
| [00:34:23] |
* |
flox_ has joined #trac |
| [00:34:23] |
<coderanger> |
it checks the mtimes, so its all automated |
| [00:34:29] |
* |
flox_ is now known as flox |
| [00:34:36] |
<johnp> |
if changed it should recompile? |
| [00:34:55] |
<coderanger> |
yes |
| [00:35:32] |
<coderanger> |
its not really "compiling" but thats about the closest thing to describe it |
| [00:35:47] |
<coderanger> |
they are intermediary forms used by the interpreter |
| [00:36:17] |
<coderanger> |
in short, ignore them, they almost never matter |
| [00:43:07] |
* |
flox_ has joined #trac |
| [00:43:13] |
* |
flox has quit IRC |
| [00:43:27] |
* |
flox_ is now known as flox |
| [00:44:22] |
<johnp> |
if re.match('/[0-9A-Z]/', user): |
| [00:44:38] |
<johnp> |
shouldn't that catch all Caps and numbers in the username |
| [00:44:38] |
<alect> |
read the re module docs |
| [00:44:48] |
<coderanger> |
no / |
| [00:45:32] |
<coderanger> |
you mean '^[0-9A-Z]*$' |
| [00:46:03] |
<coderanger> |
the first ^ is implicit in the .match(), so you can leave it out if desired |
| [00:46:34] |
<johnp> |
ok i think it works now |
| [00:46:41] |
<johnp> |
http://svn.phpontrax.com/register |
| [00:47:48] |
<johnp> |
but this still passes user = john123 |
| [00:48:00] |
<johnp> |
Bob123 fails |
| [00:48:27] |
<coderanger> |
pastebin that section of code somewhere |
| [00:49:34] |
<johnp> |
ok |
| [00:49:56] |
* |
s0undt3ch has quit IRC |
| [00:49:57] |
* |
s0undt3ch_ is now known as s0undt3ch |
| [00:52:38] |
* |
tenshiKur0 has joined #trac |
| [00:54:02] |
<johnp> |
http://pastebin.com/818879 |
| [00:54:29] |
<coderanger> |
That is not the correct regex |
| [00:54:37] |
<coderanger> |
coderanger: you mean '^[0-9A-Z]*$' |
| [00:55:05] |
<coderanger> |
also you probably want to raise an error, not return |
| [00:55:27] |
<coderanger> |
raise TracError('error message') |
| [00:55:30] |
<johnp> |
thats how the other code in there is |
| [00:55:40] |
<johnp> |
for that plugin |
| [00:55:56] |
<johnp> |
let me try ur regex |
| [00:56:51] |
<johnp> |
that regex doesn't catach anything |
| [00:58:10] |
<coderanger> |
do you want to match something with all caps/numbers or just any? |
| [00:58:29] |
<asmodai> |
http://www.adobe.com/aboutadobe/pressroom/pressreleases/200611/110706Mozilla.html |
| [00:58:40] |
* |
Semhirage has quit IRC |
| [00:59:56] |
<coderanger> |
asmodai: Wow, thats pretty cool |
| [01:00:15] |
<asmodai> |
mmm, not sure whether to be happy or cry |
| [01:00:19] |
<coderanger> |
johnp: If you want to match any you want re.search('[0-9A-Z]') |
| [01:00:31] |
<asmodai> |
It might stiffle acceptance of SVG |
| [01:00:40] |
<johnp> |
k let me try that |
| [01:01:09] |
<coderanger> |
asmodai: Eh, they are for very different things |
| [01:01:30] |
<coderanger> |
asmodai: Its a nice gesture overall |
| [01:02:11] |
<johnp> |
sweet that worked! |
| [01:02:22] |
<johnp> |
ok my anti spambot is in place |
| [01:02:25] |
<johnp> |
:) |
| [01:02:25] |
<johnp> |
thanks |
| [01:03:28] |
<coderanger> |
asmodai: More frightening is the possibility of more my-whole-site-is-a-flash-app places |
| [01:03:44] |
<asmodai> |
*nod* |
| [01:04:03] |
<coderanger> |
asmodai: But those are the kind of idiots who didn't care how much it cost |
| [01:05:16] |
<misc> |
well, they just released a javas^W ecmascript virtual machine, no ? |
| [01:07:35] |
<asmodai> |
<<+TMS>> could be its Adobe's proactive move against Canvas |
| [01:09:47] |
<johnp> |
http://pastebin.com/818886 |
| [01:09:56] |
<johnp> |
when tryin to login in |
| [01:12:30] |
<coderanger> |
Are you using the 0.10 version of the plugin |
| [01:15:38] |
* |
Semhirage has joined #trac |
| [01:18:59] |
* |
hpnadig_ has quit IRC |
| [01:19:01] |
<johnp> |
crap |
| [01:20:32] |
* |
hpnadig_ has joined #trac |
| [01:22:50] |
<johnp> |
if re.search('[0-9A-Z]', user): |
| [01:22:50] |
<johnp> |
raise TracError('Username cannot contain numbers or capital letters. Hopefully you are not a spambot. :)') |
| [01:23:02] |
<johnp> |
yeah ur right about the TracError ;) |
| [01:24:26] |
* |
ekimus has joined #trac |
| [01:25:28] |
<johnp> |
cool everything is working |
| [01:25:35] |
<johnp> |
thanks ranger |
| [01:31:04] |
* |
cmlenz has quit IRC |
| [01:36:44] |
<ekimus> |
hello, which variable in apache controls how trac displays dates? (aside from PythonOption TracLocale, i'd like to do it with SetEnv from apache) |
| [01:37:20] |
<coderanger> |
TRAC_LOCALE |
| [01:37:29] |
<coderanger> |
that won't affect mod_python though |
| [01:37:55] |
<coderanger> |
it takes the system locale by default |
| [01:38:05] |
<ekimus> |
so the easiest way would still be to use TracLocale separately? |
| [01:38:32] |
<ekimus> |
by separately I mean not to let apache handle it but trac |
| [01:39:22] |
<coderanger> |
what frontend are you using |
| [01:40:24] |
* |
otaku42_away is now known as otaku42 |
| [01:40:25] |
<ekimus> |
mod_python with apache if that's what you mean |
| [01:41:04] |
<coderanger> |
Why can't you use PythonOption TracLocale? |
| [01:51:54] |
* |
cmlenz has joined #trac |
| [01:52:26] |
<coderanger> |
cmlenz: Make that 2 spams ;-) |
| [01:53:15] |
<cmlenz> |
where? |
| [01:53:18] |
<cmlenz> |
deleted already? |
| [01:53:34] |
<coderanger> |
yeah |
| [01:53:58] |
<cmlenz> |
ah |
| [01:54:05] |
<coderanger> |
Should be to last allowed ticket modification in the log |
| [01:54:36] |
<cmlenz> |
well, that's the problem when bayes thinks something is ham... it gets upmodded quite a bit |
| [01:54:48] |
<cmlenz> |
7 points in this case, outweighing the IP blacklist |
| [01:55:41] |
<coderanger> |
I'd say 2 spams in as many weeks is a manageable amount though :) |
| [02:07:05] |
<cmlenz> |
absolutely |
| [02:16:57] |
* |
hpnadig has joined #trac |
| [02:17:15] |
* |
hpnadig_ has quit IRC |
| [02:17:24] |
* |
Hydra_ has joined #trac |
| [02:28:52] |
* |
omry has quit IRC |
| [02:29:55] |
* |
omry has joined #trac |
| [02:51:05] |
* |
idnar has quit IRC |
| [02:51:08] |
* |
idnar_ has joined #trac |
| [02:53:24] |
* |
coderanger -> sleep |
| [02:59:43] |
* |
rsalveti has joined #trac |
| [03:03:31] |
<alect> |
evening |
| [03:07:53] |
* |
idnar_ is now known as idnar |
| [03:08:55] |
* |
wkornewald has joined #trac |
| [03:09:28] |
<wkornewald> |
hi, is the trac-hacks svn repos inaccessible? although I use --username and --password it still says "authorization required" |
| [03:13:10] |
<alect> |
to check out code? |
| [03:13:13] |
<alect> |
or check in? |
| [03:13:16] |
<wkornewald> |
also, what about a nice authentication mechanism which automaticallycheck in |
| [03:13:19] |
<wkornewald> |
oops |
| [03:13:20] |
<wkornewald> |
check in |
| [03:13:33] |
<alect> |
check-ins are restricted to your own hacks |
| [03:13:56] |
<wkornewald> |
I should be member of dbauthplugin (except if someone removed me) |
| [03:15:26] |
<coderanger> |
wkornewald: What username are you using? |
| [03:15:33] |
<wkornewald> |
oh, my username is wkornew (can I change that somehow?) |
| [03:17:57] |
<alect> |
not really |
| [03:18:19] |
<alect> |
better to keep consistent log messages and trac change histories |
| [03:18:49] |
<wkornewald> |
okay, but why does it not work for me. I'm using svn co --username wkornew --password ... https://trac-hacks.org/svn/dbauthplugin |
| [03:18:54] |
<coderanger> |
also tickets and whatnot |
| [03:19:03] |
<wkornewald> |
and then svn ci --username ....... |
| [03:19:42] |
<alect> |
[Tue Nov 07 21:45:58 2006] [error] [client 89.53.219.245] access to /svn/!svn/ver/1303/dbauthplugin/simple/dbauth/auth.py failed, reason: verification of user id '<null>' not configured |
| [03:20:33] |
<alect> |
i'd suggest something is wrong with your client |
| [03:20:37] |
<wkornewald> |
why user id null? |
| [03:20:51] |
<alect> |
given that there are no other log entries of a similar nature for anybody lese |
| [03:20:52] |
<coderanger> |
wkornewald: You shouldn't need to use --username |
| [03:20:54] |
<wkornewald> |
I tried to upgrade it from 1.3 to 1.4, but it didn't help |
| [03:20:57] |
* |
jooopyme has joined #trac |
| [03:21:00] |
<jooopyme> |
hi |
| [03:21:14] |
<coderanger> |
wkornewald: subversion will prompt you for a username/password |
| [03:21:24] |
<wkornewald> |
coderanger: I also tried without --username |
| [03:21:28] |
<wkornewald> |
it doesn't work |
| [03:22:06] |
<wkornewald> |
I'm using the Windows build which might be a problem (there was some incompatibility with Apache 2.2, for example)... |
| [03:22:13] |
<wkornewald> |
but it's from the official website |
| [03:22:18] |
<jooopyme> |
I want to create alias wiki in apache2 in order to just enter the server name without /trac then trac will load. How do I do it? Someone in the mailing list told me to use DocumentRoot, but I don't know how |
| [03:24:49] |
<alect> |
wkornewald: me too |
| [03:25:08] |
<coderanger> |
jooopyme: What frontend |
| [03:25:13] |
<wkornewald> |
alect: what does svn info show for you? |
| [03:26:29] |
<alect> |
it shows a lot. what specifically? |
| [03:26:46] |
<wkornewald> |
the checkout URL |
| [03:26:56] |
<alect> |
https://trac-hacks.org/svn |
| [03:27:33] |
<wkornewald> |
alect: which client are you using? 1.4? |
| [03:28:37] |
<alect> |
yep |
| [03:28:41] |
<alect> |
under linux |
| [03:28:47] |
<wkornewald> |
ah |
| [03:28:51] |
<wkornewald> |
I thought it was Windows |
| [03:30:05] |
<alect> |
very strange |
| [03:30:07] |
* |
alect hmmms |
| [03:30:43] |
<coderanger> |
I know the 1.4 build of Tortise was having issues at some point |
| [03:30:54] |
<alect> |
see if you can commit something coderanger |
| [03:31:28] |
<wkornewald> |
alect: maybe you need to upgrade the server? at least, it didn't work with 1.3, either, so it can't just be my client |
| [03:31:43] |
<alect> |
server is on 1.4.0 |
| [03:31:57] |
<coderanger> |
alect: Auth failure ... |
| [03:32:02] |
<coderanger> |
It was working a few hours ago |
| [03:32:10] |
<alect> |
ffs |
| [03:32:23] |
<alect> |
last commit was a couple of days ago no? |
| [03:32:34] |
<wkornewald> |
ah, good, we're coming closer to the problem :) |
| [03:32:49] |
<alect> |
[1486] |
| [03:33:08] |
<alect> |
3 days ago |
| [03:33:08] |
<coderanger> |
I thought I checked in stuff on tracmarks |
| [03:33:22] |
<coderanger> |
ahh no, that was to a local repo |
| [03:33:32] |
<alect> |
i'll do a system upgrde |
| [03:33:35] |
<alect> |
and a bounce |
| [03:33:38] |
<alect> |
maybe it will magically go away |
| [03:33:39] |
<wkornewald> |
okay |
| [03:33:41] |
* |
alect wishes |
| [03:34:11] |
<alect> |
th was going so well for a while |
| [03:34:17] |
<alect> |
zero maintenance problems |
| [03:34:23] |
<alect> |
last few weeks have been...trouble |
| [03:34:50] |
<coderanger> |
alect: At least the median echoping times seems to be stabalzing |
| [03:34:56] |
<alect> |
ah good |
| [03:35:01] |
<alect> |
any more alerts? |
| [03:35:27] |
<coderanger> |
Yeah, the upper bound is still pretty high |
| [03:35:31] |
<wkornewald> |
another problem we're experiencing with Trac: we're getting those "32, broken pipe" errors. I don't know if it's MySQL closing the connection too early or a fastcgi problem. I've contacted my host for an upgrade, but it seems that others are experiencing it, too. it's pretty annoying because Trac often doesn't show any page, at all |
| [03:35:31] |
<alect> |
yes |
| [03:35:45] |
<wkornewald> |
#3957 |
| [03:35:53] |
<wkornewald> |
http://trac.edgewall.org/ticket/3957 |
| [03:36:07] |
<wkornewald> |
we need a bot which automatically converts tickets into links :) |
| [03:36:18] |
<alect> |
ooh good idea |
| [03:36:23] |
<coderanger> |
@ticket 3957 |
| [03:36:23] |
<evil_twin> |
coderanger: http://trac.edgewall.org/ticket/3957 |
| [03:36:23] |
<alect> |
paco: do it |
| [03:36:31] |
<wkornewald> |
ah |
| [03:36:32] |
<coderanger> |
like that? ;-) |
| [03:36:34] |
<wkornewald> |
but why not #...? |
| [03:36:36] |
<alect> |
that's not automatic! |
| [03:36:53] |
<coderanger> |
Because # means channel in IRC generally |
| [03:36:53] |
<alect> |
@wiki SandBox |
| [03:36:53] |
<evil_twin> |
alect: http://trac.edgewall.org/wiki/SandBox |
| [03:37:05] |
<alect> |
#\d+ |
| [03:37:18] |
<wkornewald> |
but we don't normally talk about channels with numbers, only |
| [03:37:24] |
<coderanger> |
heh |
| [03:37:34] |
<alect> |
mmm at least the smoke ping trend is downwards, not up |
| [03:38:06] |
<alect> |
we'll see how pg holds up |
| [03:38:20] |
<alect> |
then if it's still fucked, it'll be time for a dedicated server |
| [03:38:38] |
<wkornewald> |
does trac.org use fastcgi? |
| [03:38:46] |
<alect> |
th does |
| [03:38:50] |
<alect> |
edgewall does not, afaik |
| [03:38:55] |
<cmlenz> |
it does |
| [03:38:57] |
<cmlenz> |
with light |
| [03:38:59] |
<cmlenz> |
y |
| [03:39:11] |
<alect> |
aha |
| [03:39:33] |
<wkornewald> |
and why is it always fast? we're running with Apache+fastcgi and often it takes ages to load the first request, but after that it's fast enough |
| [03:40:03] |
<wkornewald> |
also, can nobody reproduce the broken pipe error? |
| [03:40:41] |
<cmlenz> |
people can... there have been reports, but AFAIK no patches |
| [03:40:50] |
<cmlenz> |
we're not seeing broken pipes on edgewall.org |
| [03:40:56] |
<cmlenz> |
(AFAICT) |
| [03:41:15] |
<cmlenz> |
maybe it's the fcgi apache module that's screwing up? |
| [03:42:08] |
<alect> |
i read about some bugs in mod_fastcgi too |
| [03:42:10] |
* |
rsalveti has left #trac |
| [03:42:12] |
<wkornewald> |
maybe, but others had the same error with Rails and TG even with the most recent fastcgi version. the real fix seemed to be to catch the exception and reestablish all SQL connections |
| [03:42:49] |
<alect> |
anecdotally, mod_fcgid is better |
| [03:43:26] |
<cmlenz> |
is this really DB related?? |
| [03:43:32] |
<wkornewald> |
unfortunately, we can only use what our host has pre-installed (and maybe ask for upgrades sometimes) |
| [03:44:13] |
<wkornewald> |
cmlenz: I google'd and many people wrote that it was MySQL closing the connection |
| [03:44:23] |
<wkornewald> |
BTW, I'm using MySQL |
| [03:44:45] |
<wkornewald> |
maybe you can reproduce it with MySQL, too? |
| [03:45:08] |
<cmlenz> |
I don't have factcgi setup locally, and we're using postgres on edgewall.org |
| [03:45:21] |
<wkornewald> |
cmlenz: at least, Rails fixed the bug by catching the error and reopening the DB connection |
| [03:45:41] |
<wkornewald> |
(that's what I saw in a changelog entry) |
| [03:45:44] |
<cmlenz> |
I can't really imagine that's the same problem |
| [03:46:02] |
<cmlenz> |
we're catching all kinds of exceptions in the DB code |
| [03:46:07] |
<wkornewald> |
it's the same situation: fastcgi, Apache, MySQL, broken pipe (32) |
| [03:46:13] |
<wkornewald> |
even with TurboGears |
| [03:46:20] |
<wkornewald> |
though, I don't know if TG fixed it |
| [03:49:28] |
<wkornewald> |
some people also reported that they fixed it by downgrading fastcgi and some by upgrading fastcgi...strange and very annoying problem (our developers are pretty unhappy with Trac and I'd like to fix it or we have to switch to flyspray or something like that which I don't really want) |
| [03:53:04] |
<wkornewald> |
guys, I really need your help, could you please take a look at this bug? I don't know Trac well enough to fix it the way I suggested and I don't have enough time to dive in |
| [03:53:35] |
* |
tenshiKur0 has quit IRC |
| [03:55:17] |
<cmlenz> |
I don't see how we can fix that in Trac except for ignoring the error |
| [03:56:06] |
<cmlenz> |
who is "our developers" btw? |
| [03:56:18] |
<cmlenz> |
is that haiku? |
| [03:59:46] |
<wkornewald> |
cmlenz: yes |
| [04:00:04] |
<wkornewald> |
cmlenz: yes, Haiku, I mean |
| [04:00:33] |
* |
hpnadig has quit IRC |
| [04:01:29] |
<wkornewald> |
cmlenz: does Trac keep DB connections alive for a long time or does it create a new DB connection for every request? maybe it should be changed to reconnect with every request? AFAIK, PHP does it that way, too, because of the too low MySQL timeout |
| [04:01:32] |
* |
hpnadig has joined #trac |
| [04:01:49] |
<wkornewald> |
at leat, I did it that way, but maybe things have changed |
| [04:02:11] |
<alect> |
that would be very inefficient |
| [04:02:16] |
<cmlenz> |
PHP does it that way because that's just the only way you can work in PHP ;-) |
| [04:02:29] |
<LarstiQ> |
wkornewald: can't you switch away from mysql? |
| [04:03:36] |
<cmlenz> |
or use mod_python, or lighty |
| [04:03:44] |
<wkornewald> |
LarstiQ: at least, our user DB is stored in MySQL (we're using Drupal's DB), so we can't completely switch, but I think it would still be pretty important for Trac to have proper MySQL support ;) |
| [04:03:53] |
<cmlenz> |
(although people have had the same problem in lighty, apparently) |
| [04:03:57] |
<wkornewald> |
cmlenz: we're on a shared host and cannot choose anything |
| [04:04:44] |
<wkornewald> |
cmlenz: even PHP has the same problem with fastcgi, sometimes... |
| [04:04:47] |
<cmlenz> |
wkornewald: a shared server for a project of that size/scope is kinda sad ;-) |
| [04:05:25] |
<cmlenz> |
you should at least go with a hosting provider that makes sure things actually work |
| [04:06:02] |
<alect> |
AHA!!! |
| [04:06:03] |
<wkornewald> |
cmlenz: site5 is pretty good, actually. we didn't have any problems before and they reacted very quickly when we wanted something. the problems started with Trac :) |
| [04:06:15] |
* |
alect is thankful for sweet, sweet backups |
| [04:06:35] |
<wkornewald> |
cmlenz: real hosting is pretty expensive. how much do you pay? |
| [04:06:39] |
<wkornewald> |
which provider do you use? |
| [04:06:40] |
<alect> |
wkornewald: commits should work |
| [04:06:47] |
<cmlenz> |
wkornewald: so now you have to come running to us because you're using |
| [04:06:54] |
<alect> |
th is running on a dedicated server with 1tb of data for $60/month |
| [04:06:56] |
<cmlenz> |
an experimental DB :-P |
| [04:07:04] |
<alect> |
hosting is cheap as hell |
| [04:07:05] |
<cmlenz> |
(experimental in Trac, that is) |
| [04:07:15] |
<cmlenz> |
I don't know what hosting edgewall.org is using |
| [04:07:20] |
<alect> |
managed hosting, not so much |
| [04:07:22] |
<cmlenz> |
but I know it's completely paid for by adsense |
| [04:07:24] |
<wkornewald> |
alect: thanks! it worked!!! :)) |
| [04:07:38] |
<alect> |
yeah, adsense pays for th too :) |
| [04:07:41] |
* |
alect pats google |
| [04:08:23] |
<alect> |
strangely, "AuthType Basi" did not cause apache to error out at all |
| [04:08:34] |
<cmlenz> |
hehe |
| [04:08:41] |
<wkornewald> |
hmm, adsense...not very nice for us, I think |
| [04:09:16] |
<wkornewald> |
we're paying merely $8 per month and switching to $60 would be a *lot* of money for us |
| [04:09:28] |
<cmlenz> |
check out http://www.webfaction.com/ |
| [04:09:36] |
<wkornewald> |
I mean, with that money we can do a lot more interesting things (e.g.: we recently bought a projector) |
| [04:10:30] |
<jooopyme> |
coderanger: about the apache configuration. You have asked "what frontend?" What do you meen? |
| [04:10:58] |
<jooopyme> |
coderanger: I run apache2 |
| [04:11:04] |
<alect> |
cmlenz: i've made dns_blacklist create a table which "training" data can be inserted into |
| [04:11:29] |
<alect> |
it uses that table in preference to a real db lookup |
| [04:11:40] |
<alect> |
so it can be used for whitelisting as well |
| [04:15:08] |
<cmlenz> |
ah |
| [04:15:32] |
<cmlenz> |
have you had problems with incorrect blacklisting? |
| [04:15:36] |
<alect> |
happy with those semantics? |
| [04:15:46] |
<alect> |
no i haven't actually |
| [04:16:17] |
<alect> |
but i figured it couldn't hurt, particularly as there are both "spam" and "ham" training commands |
| [04:16:27] |
<cmlenz> |
I'm not sure, for local blacklisting doesn't it make more sense to block requests before they even reach the web server? |
| [04:16:58] |
<alect> |
how do you train that with trac though? |
| [04:17:10] |
<alect> |
ie. i have my list of spam that's gotten through |
| [04:17:29] |
<alect> |
and there are a *lot* of duplicate attempts from the same addresses on th |
| [04:18:21] |
<cmlenz> |
wkornewald: do you know which exact apache version and fastcgi module/version your using? |
| [04:18:22] |
<alect> |
the data could then be extracted and inserted into a real dns bl for dissemination |
| [04:18:25] |
<wkornewald> |
cmlenz: what does webfaction offer that is so special? dedicated hosting is extremely expensive (we could pay a developer with that) |
| [04:18:37] |
<wkornewald> |
cmlenz: unfortuantely, I don't know. how can I check it on my shared host? |
| [04:21:21] |
<cmlenz> |
apparently apache 1.3.37 and mod_fastcgi 2.4.2 |
| [04:21:27] |
<cmlenz> |
(Server header) |
| [04:21:34] |
<wkornewald> |
cmlenz: ah, thanks :) |
| [04:21:46] |
<cmlenz> |
the webfaction shared package for 7.50 includes Trac/Subversion/PHP/etc |
| [04:23:00] |
<cmlenz> |
plus they offer free trac/svn hosting for open-source projects |
| [04:23:49] |
<wkornewald> |
cmlenz: I once had Trac running on dreamhost (which has comparable price) and it was dog slow. I think we need something faster (and we probably can pay around $30/month) |
| [04:24:10] |
<wkornewald> |
but $30 is probably the limit |
| [04:24:31] |
<wkornewald> |
is webfaction fast enough for a hosting a medium-big site? |
| [04:24:37] |
<cmlenz> |
well, I'm not sure you can compare dreamhost with webfaction |
| [04:24:42] |
<wkornewald> |
well, they have that $30 package |
| [04:24:59] |
<wkornewald> |
me neither. site5 is a lot faster than dreamhost for us |
| [04:25:49] |
<cmlenz> |
well, but it doesn't work ;-) |
| [04:26:05] |
<wkornewald> |
cmlenz: right :) |
| [04:26:15] |
<wkornewald> |
how much transfer do you need for SVN hosting? |
| [04:26:16] |
<cmlenz> |
see http://blog.webfaction.com/static/atchs/screencasts/trac-svn/trac-svn-screencast.html |
| [04:27:35] |
<ekimus> |
wkornewald: why not rent a vserver you can get 3 for that budget.... |
| [04:28:36] |
<ekimus> |
wkornewald: have one with a database one with apache+trac and the 3rd with a mailserver that should run quite fast. although I think you are well within most limits if you do it on a single server |
| [04:28:39] |
<wkornewald> |
ekimus: what are the limits of vservers? we will have to maintain them ourselves, I guess? |
| [04:29:50] |
<ekimus> |
wkornewald: exactly, but a debian should run quite well, just have a few automated security packages on it and it'll do. what could be a problem is the mailserver. just make sure not to configure it as an open relay |
| [04:31:00] |
<wkornewald> |
ekimus: I really don't want to risk being hacked. we don't have any web admins and if something goes wrong it could take too long to fix the issue |
| [04:31:07] |
<wkornewald> |
cmlenz: do you use webfaction? |
| [04:31:18] |
<cmlenz> |
no |
| [04:31:29] |
<wkornewald> |
cmlenz: and why are you promoting them? :) |
| [04:31:47] |
<cmlenz> |
they seem to be doing a nice job ;-) |
| [04:33:08] |
<wkornewald> |
cmlenz: I like their admin interface. it's a lot easier than doing Trac installs via ssh :) |
| [04:33:21] |
<wkornewald> |
but upgrades are probably more difficult... |
| [04:34:11] |
<wkornewald> |
we would like to host our svn repos, but we don't know how much transfer it generates. does anyone know how much transfer projects like GNOME, KDE, and maybe the Linux kernel generate? |
| [04:36:29] |
<cmlenz> |
no idea |
| [04:39:19] |
<jooopyme> |
Hi, I want to setup apache2 like this: www.mydomain.com will load trac. How do I setup this? |
| [04:39:56] |
<jooopyme> |
I have seen explenations on how to do www.mydomain.com/trac and I use it |
| [04:42:57] |
<alect> |
so, ipbl cmlenz. focus! |
| [04:43:20] |
<cmlenz> |
heh |
| [04:43:40] |
<cmlenz> |
I'm not really sure it's worth the extra complexity (DB upgrade etc) |
| [04:44:03] |
<cmlenz> |
can we submit training to the DNSBL server? |
| [04:44:06] |
<cmlenz> |
servers, even |
| [04:44:25] |
<alect> |
i don't think there's any standard api for it |
| [04:44:34] |
<cmlenz> |
yeah |
| [04:45:26] |
<alect> |
that would be ideal however |
| [04:46:05] |
<cmlenz> |
if we do local blacklisting that means IPs removed from blacklist servers would stay in our table, and would need to be removed manually or sth... maybe just having a whitelist of IPs in trac.ini would be simpler/better? |
| [04:46:34] |
<cmlenz> |
but to be honest, I'm not even sure there's a problem here that needs to be fixed :-P |
| [04:47:13] |
<alect> |
heh |
| [04:47:43] |
<alect> |
fair enough |
| [04:47:56] |
<alect> |
there was something else i was going to do too... |
| [04:48:01] |
<alect> |
aha reverend |
| [04:48:11] |
<cmlenz> |
ah yeah... was just going to ask ;-) |
| [04:48:43] |
<alect> |
that will undoubtedly not be backwards compatible |
| [04:49:10] |
* |
ekimus has quit IRC |
| [04:50:33] |
<cmlenz> |
hm? |
| [04:50:45] |
<alect> |
i'll shutup until i've read the code :) |
| [04:51:08] |
<cmlenz> |
heh |
| [04:52:51] |
<otaku42> |
cmlenz, alect: some dnsbl's allow to submit stuff via a web form... |
| [04:55:56] |
<alect> |
yeah. we could create custom trainers for the different dnsbl's |
| [04:55:58] |
<alect> |
effort |
| [04:57:25] |
<otaku42> |
alect: i'm about to write a plugin for surbl/uribl, by the way. however, as one of the surbl-guys just stated on the mailing list, their main focus is e-mail spamvertisement, not website spam. |
| [04:57:58] |
<otaku42> |
alect: maybe it would be worth to bring up a surbl-like service, focused on trac spam |
| [04:58:22] |
< |
