IRC Logs

2008 9
Mo Tu We Th Fr Sa So
1 2 3 4 5 6 7
8 9 10 11 12 13 14
15 16 17 18 19 20 21
22 23 24 25 26 27 28
29 30          

24. 09 2008

[00:00:44] <tux21b> things i would like to see in python is reloadable and revisioned modules, better optimazations and no gil :)
[00:05:56] * luks has quit IRC
[00:05:59] * luks has joined #pocoo
[00:08:11] * jinks1 has joined #pocoo
[00:10:13] * Kaelten has quit IRC
[00:10:38] * highwaychile has quit IRC
[00:17:11] <mitsuhiko> yeah. revisioned modules. that would be it
[00:19:19] * jinks has quit IRC
[00:21:41] <xorAxAx> mitsuhiko: how would that work?
[00:22:08] <mitsuhiko> xorAxAx: what. revisioned modules?
[00:22:16] <mitsuhiko> *versioned modules
[00:22:36] <tux21b> xorAxAx, do you know erlang? :)
[00:22:37] <xorAxAx> yes
[00:22:47] <xorAxAx> erlang has completly different binding semantics
[00:22:55] <mitsuhiko> xorAxAx: from jinja[>=2] import Template
[00:23:07] <mitsuhiko> requires of course that sys.modules goes baba
[00:23:18] <xorAxAx> thats something else than tux21b meant
[00:23:27] <tux21b> ah, that would be a good idea too, but i was speaking of code changes
[00:23:35] <xorAxAx> and its surely funny if you have components A and B both using C but in different versions
[00:23:41] <xorAxAx> so you have a split brain globals issue
[00:23:49] <xorAxAx> but globals are bad anyway
[00:24:13] <mitsuhiko> xorAxAx: yes. but i'm more interested in different versions of the same library in the same process
[00:24:30] <mitsuhiko> reason: sqlalchemy for example replaces the api every month :)
[00:24:34] <xorAxAx> yes, with the aforementioned issue
[00:24:41] <tux21b> i like the idea too, would be rally useful
[00:24:45] <xorAxAx> yeah, sqlalchemy is using a lot of globals
[00:25:14] * mitsuhiko likes the assembly loading system of .net
[00:25:23] <mitsuhiko> unfortunately that won't work in python
[00:25:33] <xorAxAx> imagine an object of an old sqlalchemy version passing from component a to b
[00:25:34] * luks has quit IRC
[00:25:36] * luks has joined #pocoo
[00:26:06] <mitsuhiko> xorAxAx: why would it do that?
[00:26:18] <xorAxAx> well, depends on your app
[00:26:29] <xorAxAx> whether it separates both domains correctly
[00:26:37] <xorAxAx> and whether they are clearly separated
[00:26:47] <mitsuhiko> i have django and sa running in the same process
[00:26:55] <mitsuhiko> i want sa1 and sa2 running there too :)
[00:26:58] <mitsuhiko> separated from each other
[00:27:17] * yashh has joined #pocoo
[00:27:21] <mitsuhiko> because for example lodgeit wants sa 0.4 and my textpress blog wants 0.5 and i want them to be "one website" :)
[00:27:24] <yashh> hello all
[00:27:38] <xorAxAx> mitsuhiko: yeah, if its separate apps ...
[00:27:49] <xorAxAx> mitsuhiko: but then you usually do http proxying
[00:27:57] <tux21b> or we would like to rewrite the django parts to sa 0.5 while the forum still uses 0.4
[00:28:06] <mitsuhiko> sure. but it's a stupid limitation of python because sys.modules is a singleton
[00:28:28] <mitsuhiko> you can somewhat work around that issue in ruby
[00:28:46] <mitsuhiko> but it's not solved there either
[00:29:05] <mitsuhiko> because... ruby loves monkey business :)
[00:29:07] <mitsuhiko> hi yashh
[00:29:15] <yashh> hi yashh
[00:29:18] <yashh> oops
[00:29:21] <yashh> hi mitsuhiko
[00:29:50] <xorAxAx> mitsuhiko: one could imagine some kind of import context
[00:29:58] <xorAxAx> mitsuhiko: that replaces sys.modules with different sets
[00:30:24] <mitsuhiko> xorAxAx: that's basically what i do in zine for plugins loaded from different application instances
[00:30:36] <mitsuhiko> but that only works because the code limits plugins in what they are doing
[00:30:48] <mitsuhiko> eg, they are not allowed to monkey patch or access other versions of itself
[00:30:55] <mitsuhiko> or interact with other applications at all
[00:31:11] * maze_ has joined #pocoo
[00:31:12] * maze has quit IRC
[00:38:30] <yashh> hey
[00:38:41] <yashh> whats the adv of SQLAlchemy
[00:38:54] <yashh> basically I am a Django guy
[00:40:14] <ronny> yashh: sqlalchemy actually is usefull when it comes to databases, while djangos orm basicaly uses them just for object persistence
[00:40:31] <yashh> yes
[00:40:42] <yashh> i heard its more good than Django ORM
[00:41:03] <ronny> well, djangos orm is like a nasty toy compared to sqlalchemy
[00:41:20] <ronny> also there is the sql expression layer
[00:42:06] <ronny> its basically the most powerfull python tool for database interaction
[00:42:39] <yashh> cool
[00:42:54] * maze_ is now known as maze
[00:43:01] <yashh> there was a small discussion on #django regarding Sql Alchemy
[00:43:12] <yashh> but to use it with django it needs a lot of customization
[00:45:45] * yashh has left #pocoo
[02:20:16] * heikkit has quit IRC
[02:22:10] * davidcramer has joined #pocoo
[02:32:09] * mtrichardson has quit IRC
[02:33:21] * alowry has quit IRC
[02:52:51] * kleinweby_ has joined #pocoo
[02:59:45] * alowry has joined #pocoo
[03:01:49] * kleinweby has quit IRC
[03:08:09] * saptah has quit IRC
[03:15:55] * maze_ has joined #pocoo
[03:20:43] * maze has quit IRC
[03:43:52] * Baumichel has quit IRC
[05:13:14] * saptah has joined #pocoo
[05:31:04] * rcscomp2 has quit IRC
[05:31:06] * aconbere has joined #pocoo
[05:36:37] * rcscomp has joined #pocoo
[05:39:33] <rcscomp> has anyone had problems with the werkzeug http server being really slow when being used form another machine? When I test a website built on my desktop with a browser on my laptop, the page takes about 12 seconds to load. When testing from my desktop its intantaneous. I have a gigabit connection between the two and don't have any other connection issues between the two computers.
[05:55:03] <prencher> rcscomp - i had a very odd stalling issue when my ip had no reverse dns.. maybe you're seeing similar
[05:55:56] <prencher> rcscomp - try to go to your hosts file, add something to the effect of <your ip> localhost, on the computer with the werkzeug instance running on
[05:56:08] <prencher> it's a stretch but it worked for me back when my isp didnt assign rdns to my ip
[05:56:33] * davidcramer has quit IRC
[05:56:45] <rcscomp> prencher: hmmmm...that might be the culprit since I do my own internal DNS. Do you know if I can turn off resolution on the HTTP server output?
[05:57:09] * davidcramer has joined #pocoo
[06:01:20] <rcscomp> prencher: bingo!!!! I did a test entry in my hosts file and that solved the delay. Thank you soooooooooooo much!
[06:04:52] <prencher> hehe, you're welcome
[06:06:46] <prencher> rcscomp - as far as turning off the rdns.. im going to write a patch for it soon.. mostly because it just pisses me off to have rdns as long as the shell in the dev logs
[06:08:32] <hads> I've run across the same issue actually, meant to look into patching it too.
[06:12:49] <prencher> remind me tomorrow
[06:12:56] <prencher> for now, bed time
[06:14:01] <rcscomp> prencher: thanks for your help. Good night.
[06:15:21] * aconbere has quit IRC
[06:27:03] * aconbere has joined #pocoo
[06:33:58] * KirkMcDonald has quit IRC
[06:34:02] * KirkMcDonald has joined #pocoo
[07:04:40] * rcscomp has quit IRC
[07:05:02] * rcscomp has joined #pocoo
[07:07:43] <rcscomp> anyone still around? I have a patch for the --no-dnslookups option but can't create a ticket on the Werkzeug Trac
[07:11:23] <rcscomp> details and link to patch: http://groups.google.com/group/pocoo-libs/browse_thread/thread/54846669425914f1
[07:11:40] * rcscomp has left #pocoo
[07:15:52] * rcscomp has joined #pocoo
[07:16:03] <hads> rcscomp: Good stuff
[07:19:11] <rcscomp> hads: thanks. Not sure if the name works, seems kind of long. Maybe --no-rdns would be better. But it works for now.
[07:35:33] * alowry has quit IRC
[07:45:30] * marchon has joined #pocoo
[07:45:52] <marchon> mitsuhiko - are you here?
[07:49:07] * marchon has quit IRC
[08:02:58] * rcscomp has left #pocoo
[08:10:58] * sashav has joined #pocoo
[08:13:07] <birkenfeld> moin!
[08:24:16] * davidcramer has quit IRC
[09:19:10] * davidcramer has joined #pocoo
[09:31:42] <asmodai> birkenfeld: dude!
[09:32:48] <asmodai> http://www.openplans.org/projects/topp-engineering/blog/2008/09/24/pyinstall-a-new-hope/
[09:36:09] <birkenfeld> let's hope the next chapter isn't "The Eby Strikes Back"
[09:37:51] <asmodai> but did you see Ian's mail
[09:39:14] <birkenfeld> where to?
[09:40:46] <birkenfeld> mitsuhiko: ping
[09:40:53] <asmodai> the TIP list
[09:41:18] <asmodai> part of said email:
[09:41:20] <asmodai> Single-file executable for you to try:
[09:41:20] <asmodai> https://svn.openplans.org/svn/pyinstall/trunk/pyinstall.py
[09:41:20] <asmodai>
[09:41:20] <asmodai> I think it's good.
[09:49:09] <mitsuhiko> birkenfeld: pong
[09:49:43] <birkenfeld> mitsuhiko: keyword search doesn't seem to work
[09:50:54] <mitsuhiko> birkenfeld: hmm. i tried with "sphinx.ext.refcount" yesterday and some other modules, and i got output
[09:50:59] <mitsuhiko> which keyword doesn't work?
[09:51:15] <birkenfeld> e.g. "os.path" in the python docs
[09:51:50] <mitsuhiko> birkenfeld: can you do "Search._index.keywords['os.path']" in firebug?
[09:51:55] <mitsuhiko> does that return an item?
[09:52:16] <birkenfeld> yes
[09:53:45] <mitsuhiko> hmm
[09:53:51] <birkenfeld> also there's something wrong with incremental index building, but I can't exactly say if it's new with that commit
[09:54:01] <mitsuhiko> that could be
[09:54:06] <mitsuhiko> i changed the indexer format
[09:54:08] <mitsuhiko> looking into that
[09:54:20] * luks has quit IRC
[09:54:47] * luks has joined #pocoo
[09:56:22] <mitsuhiko> birkenfeld: hwere is the search index loaded?
[09:56:45] <birkenfeld> load_indexer
[09:57:12] <mitsuhiko> hmm
[09:57:27] <mitsuhiko> ah. got it
[09:59:46] * aconbere has quit IRC
[10:01:44] <mitsuhiko> birkenfeld: if i raise the environment version, will it still try to load the index?
[10:02:22] <birkenfeld> it seems so, yes
[10:02:42] <birkenfeld> but that can be changed
[10:05:03] <asmodai> birkenfeld: finished BH btw
[10:05:08] <birkenfeld> asmodai: nice
[10:11:11] <mitsuhiko> birkenfeld: sphinx eats 100% of the cpu here and doesn't progress
[10:11:15] <mitsuhiko> (when building the python docs)
[10:11:20] <mitsuhiko> reference/datamodel
[10:12:02] <birkenfeld> mitsuhiko: didn't happen here just now
[10:12:12] <mitsuhiko> hmm
[10:12:21] * mitsuhiko reruns and straces
[10:13:22] <mitsuhiko> birkenfeld: heisenbug
[10:13:29] <mitsuhiko> but it's weird
[10:13:58] * davidcramer has quit IRC
[10:18:30] <birkenfeld> mitsuhiko: how do I do "ab" in "abc" in JS?
[10:19:01] <mitsuhiko> "abc".indexOf("ab") > -1 afair
[10:22:48] * jpcw has joined #pocoo
[10:23:25] <mitsuhiko> birkenfeld: does the search in the python docs work at all for you?
[10:23:29] <mitsuhiko> it looks up my browser
[10:23:37] <asmodai> hehehe
[10:23:42] * asmodai points and laughs
[10:24:00] <birkenfeld> mitsuhiko: yes
[10:25:13] <mitsuhiko> what the hack? looks like it's built against a terrible old version of sphinx
[10:25:28] <birkenfeld> docs.python.org?
[10:25:47] <mitsuhiko> no, locally
[10:25:52] <mitsuhiko> i have a searchindex.json
[10:26:01] <mitsuhiko> but i just built it :)
[10:26:06] <birkenfeld> oh
[10:26:07] <birkenfeld> :)
[10:36:56] <mitsuhiko> birkenfeld: got a searchindex.json again
[10:37:02] <mitsuhiko> do the python docs have a special builder?
[10:37:15] <birkenfeld> nope
[10:37:22] <mitsuhiko> argh. tools/sphinx
[10:37:24] <birkenfeld> is your tools/sphinx
[10:37:28] <birkenfeld> yeah :)
[10:37:32] <birkenfeld> it's a symlink here
[10:37:48] <birkenfeld> mitsuhiko: how much did you change in searchtools.js?
[10:38:10] <mitsuhiko> birkenfeld: that explains soooo much now
[10:38:24] <mitsuhiko> oh boy. that explains a lot
[10:38:54] <mitsuhiko> birkenfeld: locally or on the server?
[10:38:56] <mitsuhiko> in general not that much
[10:38:58] <birkenfeld> locally
[10:39:02] <mitsuhiko> a bit. why?
[10:39:07] <birkenfeld> ok, because I'm editing it right now
[10:39:10] <mitsuhiko> ic an check it in
[10:39:10] <mitsuhiko> sec
[10:39:34] <mitsuhiko> checked in
[10:40:36] <CIA-49> python-doctools: armin.ronacher * r66573 /doctools/trunk/sphinx/ (static/searchtools.js search.py builder.py): Some tiny search fixes.
[10:40:44] <birkenfeld> ok
[10:40:56] <birkenfeld> mitsuhiko: I'm encoding the "module" etc. now
[10:41:12] <mitsuhiko> encoding -> giving it a number?
[10:41:16] <birkenfeld> yes
[10:41:21] <mitsuhiko> actually, that information is unused at the moment i think
[10:41:26] <birkenfeld> no more
[10:41:35] <mitsuhiko> ah
[10:46:52] * davidcramer has joined #pocoo
[10:47:22] * davidcramer has joined #pocoo
[10:47:30] <CIA-49> python-doctools: georg.brandl * r66574 /doctools/trunk/sphinx/ (static/searchtools.js search.py):
[10:47:31] <CIA-49> python-doctools: * Search for partial keyword matches and be case insensitive.
[10:47:31] <CIA-49> python-doctools: * Show keyword results before regular ones.
[10:47:31] <CIA-49> python-doctools: * Show full name, type of keyword and title of containing doc.
[10:47:39] <davidcramer> how do you all typically handle large siege attempts?
[10:47:52] <davidcramer> i have awesome response times, im just not that familiar with using siege to pusht he limits
[10:48:27] <davidcramer> e.g. can only pull 25-30 req/s but the response time is 0.01
[10:49:07] <davidcramer> (i am sieging from the same machine as well)
[10:54:23] <CIA-49> python-doctools: georg.brandl * r66575 /doctools/trunk/sphinx/ (static/searchtools.js search.py): Compress the index more, by not storing the label name for keywords.
[11:07:23] <CIA-49> python-doctools: georg.brandl * r66576 /doctools/trunk/sphinx/search.py: Add stopword list.
[11:24:13] * davidcramer has quit IRC
[11:32:45] <CIA-49> python-doctools: georg.brandl * r66578 /doctools/trunk/sphinx/ (static/searchtools.js search.py): More compression: group objects by prefix.
[11:33:25] * delroth has quit IRC
[11:33:46] * delroth has joined #pocoo
[11:34:23] <CIA-49> python-doctools: georg.brandl * r66579 /doctools/trunk/CHANGES: Add changelog entry about search change.
[11:44:06] <birkenfeld> mitsuhiko: ping
[11:44:12] <mitsuhiko> birkenfeld: pong
[11:44:30] <birkenfeld> mitsuhiko: isn't it somehow possible in JS to not quote strings?
[11:44:35] <birkenfeld> when used as dict keys
[11:48:55] <mitsuhiko> {foo: "bar"}
[11:49:00] <mitsuhiko> unless foo is a keyword or contains spaces
[11:49:06] <mitsuhiko> however. that's not valid json, just valid javascript
[11:49:17] <birkenfeld> it only seems to work for assignments?
[11:51:33] <mitsuhiko> birkenfeld: assignments?
[11:51:41] <mitsuhiko> what do you mean?
[11:52:00] <birkenfeld> mitsuhiko: at least in firebug, when I eval {a:1} it doesn't work, but a={a:1} works
[11:52:11] <mitsuhiko> ({a: 1}) works
[11:52:15] <mitsuhiko> {} alone is not a valid statement
[11:52:29] <birkenfeld> ah
[11:52:31] <birkenfeld> interesting :)
[11:52:40] <birkenfeld> and somehow understandable
[11:52:48] <mitsuhiko> one of the better design decisions of javascript
[11:52:51] <birkenfeld> in a language with braces...
[11:59:23] <mitsuhiko> birkenfeld: i'm thinking about adding jinja2 support to sphinx. should it prefer jinja2 over jinja or provide a separate template bridge to set in the config?
[11:59:36] * davidcramer has joined #pocoo
[12:00:12] <birkenfeld> mitsuhiko: depends on the template compatibility
[12:00:15] <birkenfeld> mitsuhiko: urgs
[12:00:29] <mitsuhiko> urgs?
[12:00:30] <birkenfeld> ({"true": 1, true: 2}) is {true: 2}
[12:00:40] <mitsuhiko> you can't use variables in object literals
[12:00:47] <birkenfeld> isn't true a keyword?
[12:01:04] <mitsuhiko> birkenfeld: welcome to the wonderful world of "microsoft does it wrong, let's make our code compatible"
[12:01:19] <mitsuhiko> birkenfeld: class is a keyword in javascript too, but you can use it as identifier
[12:01:20] <birkenfeld> grr
[12:01:48] <mitsuhiko> is especially great if you load your code into rhino or some standard compliant interpreter that gives a syntax error
[12:02:03] <birkenfeld> ok, I'm not surprised anymore that "1" and 1 are also treated the same
[12:02:12] <mitsuhiko> birkenfeld: http://www.javascripter.net/faq/reserved.htm
[12:02:28] <mitsuhiko> birkenfeld: sure. that's however intentional because there are no non string keys allowed
[12:02:48] <mitsuhiko> that's actually how arrays are implemented
[12:02:50] <mitsuhiko> very, very crazy
[12:03:14] <birkenfeld> reminds me of PHP
[12:03:18] <mitsuhiko> no, it's worse
[12:03:23] <birkenfeld> haha
[12:03:41] <mitsuhiko> birkenfeld: the worst are older jscript versions that implemented objects as linked lists
[12:03:58] <mitsuhiko> nice for arrays, but god. that sucks for everything else :)
[12:04:53] <birkenfeld> ahh, now I know why incremental building doesn't work for docs.python.org
[12:04:58] <birkenfeld> the probably don't have simplejson
[12:08:41] * maze_ is now known as maze
[12:15:42] <davidcramer> http://code.google.com/p/django-command-extensions/wiki/RunServerPlus
[12:15:49] <davidcramer> ok the werkzeug utils are shiny :)
[12:15:50] <davidcramer> props on those
[12:19:17] * grumpy_ has joined #pocoo
[12:46:16] * rodfersou[away] is now known as rodfersou
[12:51:25] <asmodai> nice to see jquery.com dropping the drawing from the homepage
[13:06:21] * grumpy_ has quit IRC
[13:11:57] <davidcramer> jquery site looks nice
[13:12:16] <davidcramer> i still prefer mootools :)
[13:12:58] <davidcramer> mmm even has nice plugin hosting now
[13:13:34] * grumpy_ has joined #pocoo
[13:16:05] <birkenfeld> mitsuhiko: okay. simplejson is soon history
[13:16:56] <asmodai> no!
[13:17:04] <asmodai> It will get speed ups in 2.6.1!
[13:17:57] <CIA-49> python-doctools: georg.brandl * r66581 /doctools/trunk/sphinx/static/jquery.js: Update jQuery.
[13:21:16] <birkenfeld> hah, another 50k saved
[13:22:20] <mitsuhiko> heh
[13:23:05] <birkenfeld> mitsuhiko: okay, one more thing
[13:23:10] <mitsuhiko> yep?
[13:23:55] <birkenfeld> if the list of filenumbers contains only one, it needn't be a list
[13:24:33] <birkenfeld> but after that, I can't imagine any more ways to make it smaller
[13:24:47] <mitsuhiko> birkenfeld: implement gzip for javascript :_D
[13:24:58] <birkenfeld> hehe
[13:31:25] <CIA-49> python-doctools: georg.brandl * r66582 /doctools/trunk/sphinx/ (util/_json.py builder.py util/json.py):
[13:31:26] <CIA-49> python-doctools: * Always use our own JS dumper instead of simplejson.
[13:31:26] <CIA-49> python-doctools: * Compress JS further by omitting quotes for dict keys where possible.
[13:31:43] <birkenfeld> mitsuhiko: how do I test if a value is int or list?
[13:41:45] * grumpy_ has quit IRC
[13:42:00] <mitsuhiko> birkenfeld: if (typeof obj === 'number')
[13:42:25] <birkenfeld> mitsuhiko: is if (obj.length == undefined) bad? ;)
[13:45:07] * davidcramer_ has joined #pocoo
[13:45:30] * davidcramer has quit IRC
[13:45:34] * davidcramer_ has left #pocoo
[13:46:42] <birkenfeld> 'nother 20k
[13:48:05] <asmodai> hahaha
[13:48:09] <asmodai> silly
[13:48:36] <mitsuhiko> birkenfeld: implemented for the json dumper only or all searchindex formats?
[13:49:12] <birkenfeld> mitsuhiko: in search.py, so it will be in all formats
[13:49:20] <birkenfeld> makes no sense otherwise
[13:50:22] <CIA-49> python-doctools: georg.brandl * r66583 /doctools/trunk/ (sphinx/templates/layout.html CHANGES doc/templating.rst): Rename rellinks to linktags.
[13:51:10] <mitsuhiko> birkenfeld: can't say that i like the idea :) but currently the search only works for js anyways
[13:52:07] <CIA-49> python-doctools: georg.brandl * r66584 /doctools/trunk/sphinx/ (static/searchtools.js search.py util/json.py): A fix in the JS dumper and further compression by omitting redundant braces.
[13:53:34] * rafael has joined #pocoo
[14:03:01] * davidcramer has joined #pocoo
[14:03:42] <CIA-49> python-doctools: georg.brandl * r66585 /doctools/trunk/sphinx/ (util/jsdump.py search.py util/json.py builder.py):
[14:03:43] <CIA-49> python-doctools: Rename util.json to util.jsdump because it doesn't generate valid JSON anymore.
[14:03:43] <CIA-49> python-doctools: The JSON html builder still needs simplejson to work.
[14:03:45] <birkenfeld> oh yes, another optimization: convert [120,129,193,283] to "120129193283" :D
[14:04:45] <mitsuhiko> birkenfeld: short integers :D
[14:05:07] <birkenfeld> mitsuhiko: as a string of course :)
[14:05:12] <mitsuhiko> right. not escaping :)
[14:10:32] <rafael> hi
[14:10:50] <rafael> mitsuhiko: how to make default values in forms?
[14:16:01] <mitsuhiko> rafael: a) hg pull
[14:16:08] <mitsuhiko> i pushed the zeml stuff
[14:16:44] <CIA-49> Zine: mitsuhiko default * 499:44f12da69ccb /zine/ (30 files in 11 dirs):
[14:16:45] <CIA-49> Zine: Monster commit: fragment goes, zeml comes. database changes and totally
[14:16:45] <CIA-49> Zine: incompatible with current zine installations. If one runs a blog currently,
[14:16:45] <CIA-49> Zine: change the schema with "alter table" yourself and refresh the parser_data with a
[14:16:45] <CIA-49> Zine: little script.
[14:16:47] <CIA-49> Zine: mitsuhiko default * 498:bc438ac05cea /zine/utils/forms.py: Improved field descriptor for forms.
[14:17:31] <mitsuhiko> b) what do you mean by default
[14:17:34] <mitsuhiko> what kind of default
[14:17:46] <rafael> nice
[14:17:56] <rafael> b) i mean i saved an option eg for pages options
[14:18:11] <rafael> but then the fields should have the config value by default
[14:18:25] <mitsuhiko> that's what the initial parameter in the form constructor is for
[14:18:32] <mitsuhiko> f = Form({'your_field': 'your_value'})
[14:18:59] <rafael> ah, alright :)
[14:19:51] <mitsuhiko> rafael: make sure to merge, i changed a few things
[14:21:30] <rafael> mitsuhiko: and how to label correctly? currently i use <label> in the template
[14:21:45] <mitsuhiko> rafael: check out the new changepasswordform
[14:21:49] <mitsuhiko> it's not part of the form system
[14:22:10] <rafael> k
[14:23:00] <CIA-49> python-doctools: georg.brandl * r66586 /doctools/trunk/ (CHANGES sphinx/util/jsdump.py): Group changelog entries and fix docstring for jsdump.
[14:24:14] <rafael> mitsuhiko: http://paste.pocoo.org/show/86093/ after installation
[14:26:34] * highwaychile has joined #pocoo
[14:32:16] <asmodai> mmm, I am starting to appreciate itsalltext more
[14:42:48] * \3TATUK has joined #pocoo
[14:43:09] * \3TATUK has left #pocoo
[15:12:47] * Kaelten has joined #pocoo
[15:16:57] <asmodai> Kaelten: dude!
[15:17:36] <Kaelten> asmodai: dude!
[15:17:42] <Kaelten> like, where's the car?
[15:19:55] <asmodai> Kaelten: :)
[15:19:59] <asmodai> Kaelten: Nod32 fixed
[15:20:05] <Kaelten> Nod32?
[15:20:11] <asmodai> Kaelten: curse-client.exe starts normally again and can update
[15:20:17] <Kaelten> ah, cool
[15:20:18] <asmodai> the anti-virus
[15:20:29] <Kaelten> ya I think it was some kinda issue with him trying to compress the binary
[15:20:35] <asmodai> possibly!
[15:21:13] * pete__c has quit IRC
[15:21:13] * delroth has quit IRC
[15:21:13] * [diecast] has quit IRC
[15:21:13] * dev has quit IRC
[15:22:45] * delroth has joined #pocoo
[15:22:46] * [diecast] has joined #pocoo
[15:22:46] * pete__c has joined #pocoo
[15:22:46] * dev has joined #pocoo
[15:26:08] * rodfersou is now known as rodfersou[away]
[15:28:44] * davidcramer_ has joined #pocoo
[15:30:03] * davidcramer has quit IRC
[15:32:06] * davidcramer_ has quit IRC
[15:33:04] * davidcramer has joined #pocoo
[16:04:45] * rafael__ has joined #pocoo
[16:06:30] * rafael has quit IRC
[16:14:23] * maix has joined #pocoo
[16:20:17] * sashav has quit IRC
[16:38:12] * tux21b has quit IRC
[16:39:06] * tux21b has joined #pocoo
[16:40:41] <asmodai> well ain't that bizarre
[16:40:46] <asmodai> US army training for US domestic missions
[16:40:57] <asmodai> I wonder what happened to the National Guard.
[16:41:00] * _paneb has quit IRC
[16:50:47] * Kaelten has quit IRC
[17:00:31] <birkenfeld> asmodai: you mean the National Guards?
[17:00:49] <birkenfeld> afaik each state has its own
[17:05:14] * Kaelten has joined #pocoo
[17:09:22] <CIA-49> python-doctools: georg.brandl * r66587 /doctools/trunk/sphinx/locale/pl/LC_MESSAGES/ (sphinx.mo sphinx.js sphinx.po): Update to Polish locale.
[17:24:52] * |[diecast]| has joined #pocoo
[17:25:57] * [diecast] has quit IRC
[17:26:19] * alowry has joined #pocoo
[17:27:35] * |[diecast]| has quit IRC
[17:32:45] <CIA-49> python-doctools: georg.brandl * r66588 /doctools/branches/0.4.x/sphinx/texinputs/sphinx.sty: Fix footer marking.
[17:32:46] <CIA-49> python-doctools: georg.brandl * r66589 /doctools/trunk/sphinx/texinputs/ (Makefile sphinx.sty): #17: fix Makefile syntax.
[17:32:46] <CIA-49> python-doctools: georg.brandl * r66590 /doctools/converter/converter/latexparser.py: Fix url parsing type.
[17:36:55] * mtrichardson has joined #pocoo
[17:38:08] <CIA-49> python-doctools: georg.brandl * r66591 /doctools/trunk/ (CHANGES doc/markup/inline.rst sphinx/environment.py): #16: allow referring to figures without explicit text.
[17:44:09] * alowry has quit IRC
[17:49:24] <CIA-49> python-doctools: georg.brandl * r66592 /doctools/trunk/ (CHANGES sphinx/quickstart.py): Don't overwrite existing sphinx projects from quickstart.
[17:56:00] <CIA-49> python-doctools: georg.brandl * r66593 /doctools/trunk/doc/intro.rst: Converter is now fairly usable for non-Python core projects.
[18:00:22] * [diecast] has joined #pocoo
[18:07:29] * stifal has joined #pocoo
[18:12:09] * alowry has joined #pocoo
[18:14:34] <CIA-49> python-doctools: georg.brandl * r66594 /doctools/trunk/ (4 files in 3 dirs): #18: put footnotes at the correct location in the LaTeX writer.
[18:15:16] <CIA-49> Pygments: thatch trunk * 681:38e669d11284 / (CHANGES pygments/lexers/agile.py): Support simple prompt from irb (Closes #363)
[18:15:18] <CIA-49> Pygments: thatch trunk * 682:6f317b427b60 /pygments/lexers/agile.py: Further fixes for irb lexer
[18:15:18] <CIA-49> Pygments: thatch trunk * 683:c3d87fc41a23 /pygments/lexers/agile.py: Fix RubyLexer heredoc issue as described in #336
[18:15:19] <CIA-49> Pygments: thatch trunk * 684:72573bca3fd0 /tests/examplefiles/irb_heredoc: Regression test for Ruby heredoc fix in #366
[18:15:21] <CIA-49> Pygments: gbrandl trunk * 685:b376b453559d / (3 files in 3 dirs): Merge with Tim.
[18:18:42] * [diecast] has quit IRC
[18:18:54] * [diecast] has joined #pocoo
[18:20:14] <CIA-49> python-doctools: georg.brandl * r66595 /doctools/trunk/doc/ (concepts.rst conf.py): Use palatino for the Sphinx docs.
[18:39:12] * rodfersou[away] is now known as rodfersou
[18:55:28] * rcscomp2 has joined #pocoo
[18:56:54] <rcscomp2> what happens with a WSGI application when a string rather than a list is returned? I.e. r = Response()
[18:56:55] <rcscomp2> r.response = 'hw'
[18:56:55] <rcscomp2> return r(env, start_response)
[18:57:48] <mitsuhiko> rcscomp2: it becomes terrible slow
[18:58:02] <mitsuhiko> because the server will write one char, flush the stream, write one char, flush the stream etc.
[19:03:39] <CIA-49> python-doctools: georg.brandl * r66602 /doctools/branches/0.4.x/sphinx/util/texescape.py: Escape backtick in texescape.
[19:04:10] <rcscomp2> mitsuhiko: interesting. I have been doing this by accident. Would it make sense to make BaseResponse.response a property and if its set as a string, turn it into a one-element list?
[19:04:34] <mitsuhiko> rcscomp2: Response("foo") is perfectly fine
[19:04:50] <mitsuhiko> werkzeug makes sure strings are yielded as one
[19:05:50] <rcscomp2> mitsuhiko: correct. But currently I am needing to create the response object before I know what the actual response output is going to be. So the response object gets set up, I do some processing, and then when I am done with the processing I use response.response to set the actual content.
[19:06:28] <rcscomp2> I can obviously make sure I use a list, but I thought it might be helpful and avoid assumption problems in the future for newbs like myself.
[19:08:40] <mitsuhiko> rcscomp2: set response.data
[19:08:50] <mitsuhiko> if it's a string
[19:09:31] <rcscomp2> mitsuhiko: oh, I see. I was using the wrong property. Thanks.
[19:10:40] <CIA-49> python-doctools: georg.brandl * r66604 /doctools/ (trunk/sphinx/util/texescape.py trunk): (log message trimmed)
[19:10:40] <CIA-49> python-doctools: Merged revisions 66588,66602 via svnmerge from
[19:10:40] <CIA-49> python-doctools: svn+ssh://pythondev@svn.python.org/doctools/branches/0.4.x
[19:10:40] <CIA-49> python-doctools: ........
[19:10:40] <CIA-49> python-doctools: r66588 | georg.brandl | 2008-09-24 17:25:43 +0200 (Wed, 24 Sep 2008) | 2 lines
[19:10:43] <CIA-49> python-doctools: Fix footer marking.
[19:10:45] <CIA-49> python-doctools: ........
[19:10:47] <CIA-49> python-doctools: georg.brandl * r66606 /doctools/trunk/ (CHANGES sphinx/environment.py sphinx/roles.py): Add cmember role (#3875).
[19:11:44] <asmodai> birkenfeld: yeah, but the army is not for domestic deployment
[19:11:55] <birkenfeld> yep
[19:13:06] * leche has joined #pocoo
[19:24:12] * leche has quit IRC
[19:29:14] * leche has joined #pocoo
[19:35:54] <rafael__> mitsuhiko: what do you say to the pasted error?
[19:36:09]<