| [00:00:12] |
* |
stifal has quit IRC |
| [00:02:17] |
* |
Kaelten has quit IRC |
| [00:26:04] |
* |
mtrichardson has joined #pocoo |
| [00:37:38] |
<s0undt3ch> |
ppl, using py2.5's xml.etree.ElementTree is there any way to prettyprint it? like on elementtree.ElementLib.prettyprint |
| [00:39:02] |
<s0undt3ch> |
sorry, saw someone mentioning XML ;) |
| [00:43:22] |
<s0undt3ch> |
_paneb: have any idea? |
| [00:46:39] |
* |
davidcramer_ has joined #pocoo |
| [00:46:40] |
* |
davidcramer has quit IRC |
| [00:51:04] |
<s0undt3ch> |
got it |
| [01:03:50] |
* |
Hogey-Dawg has joined #pocoo |
| [01:14:50] |
* |
Hogey-Dawg has left #pocoo |
| [01:52:07] |
* |
alowry has quit IRC |
| [01:53:55] |
* |
mtrichardson has quit IRC |
| [02:02:59] |
* |
damjan has quit IRC |
| [02:02:59] |
* |
damjan_ has joined #pocoo |
| [02:36:12] |
* |
Crast has joined #pocoo |
| [02:37:20] |
* |
maix has quit IRC |
| [02:41:06] |
* |
aconbere has quit IRC |
| [02:45:37] |
* |
davidcramer has joined #pocoo |
| [02:45:38] |
* |
davidcramer_ has quit IRC |
| [03:18:43] |
* |
EnTeQuAk_ has joined #pocoo |
| [03:20:47] |
* |
EnTeQuAk has quit IRC |
| [03:59:22] |
* |
Kaelten has joined #pocoo |
| [04:42:24] |
* |
Kaelten has quit IRC |
| [04:42:46] |
* |
Kaelten has joined #pocoo |
| [05:29:01] |
* |
alowry has joined #pocoo |
| [05:45:50] |
* |
alowry has quit IRC |
| [07:15:17] |
* |
sashav has quit IRC |
| [07:46:16] |
* |
Kaelten has quit IRC |
| [07:54:52] |
* |
EnTeQuAk_ is now known as EnTeQuAk |
| [08:28:11] |
* |
jpcw has joined #pocoo |
| [08:59:52] |
* |
kib2 has joined #pocoo |
| [09:04:10] |
<kib2> |
birkenfeld: ping ? |
| [09:39:56] |
* |
kib2 has quit IRC |
| [10:26:07] |
* |
Tik-Tok has joined #pocoo |
| [10:47:34] |
* |
izibi has joined #pocoo |
| [11:08:03] |
* |
stifal has joined #pocoo |
| [11:45:00] |
* |
Baumichel has joined #pocoo |
| [11:53:08] |
* |
maix has joined #pocoo |
| [11:59:18] |
* |
highwaychile has joined #pocoo |
| [12:05:26] |
<asmodai> |
weirdos! |
| [12:16:25] |
<prencher> |
nou |
| [12:23:48] |
<asmodai> |
wat nou? |
| [12:23:54] |
<prencher> |
NO U |
| [12:24:05] |
<asmodai> |
so |
| [12:24:09] |
<asmodai> |
did mitsuhiko stir awake yet? |
| [12:24:34] |
<prencher> |
apparently not for the last 38hrs 51mins 26secs and counting |
| [12:25:14] |
<ronny> |
oO |
| [12:25:51] |
<prencher> |
oO what |
| [12:27:19] |
<ronny> |
oO nothing |
| [12:27:34] |
<asmodai> |
oO something! |
| [12:29:25] |
<asmodai> |
prencher: http://www.danwei.org/electronic_games/gambling_your_life_away_in_zt.php |
| [12:30:35] |
<prencher> |
asmodai - reading the first paragraph.. just sounds like someone made a game that's free but with micropayments for shit? |
| [12:30:40] |
<prencher> |
e.g. what the koreans have been doing for years? |
| [12:30:41] |
<asmodai> |
yea |
| [12:30:57] |
<asmodai> |
but it provides some insight into these games and the psychology |
| [12:31:33] |
<asmodai> |
And that is why I prefer these western MMORPGs. The Asian ones are mostly based on 'pachinko'-style gambling |
| [12:32:25] |
<ronny> |
lol |
| [12:32:30] |
<asmodai> |
I abhor micro payment games |
| [12:32:51] |
<asmodai> |
Since basically the only way to get ahead is by continuously pouring in money. |
| [12:33:07] |
<asmodai> |
Reminds me also of Habbo Hotel, where young children spend cash to decorate their virtual rooms. |
| [12:35:43] |
<prencher> |
asmodai - imo the happy balance is free game with micropayments, but nothing to stop you from gaining items by skill |
| [12:35:55] |
<prencher> |
since that's exactly what people already do, only they buy it from shady companies |
| [12:36:00] |
<asmodai> |
*nod* |
| [12:36:10] |
<asmodai> |
But most games will never do that. |
| [12:36:15] |
<prencher> |
so why not monetize that and keep the game cheap or free |
| [12:36:16] |
<prencher> |
well |
| [12:36:30] |
<prencher> |
koreans have bene doing this exact thing for years |
| [12:36:35] |
<prencher> |
very successfully |
| [12:36:44] |
<prencher> |
and they're starting to expand to the west now |
| [12:37:15] |
<prencher> |
the main issue i see would be if westerners will accept it |
| [12:37:25] |
<asmodai> |
I doubt it. |
| [12:37:28] |
<prencher> |
even if people blatantly already do it in e.g. wow |
| [12:37:47] |
<prencher> |
will they accept it being an encouraged thing, even if the game is free, or will they proclaim cheat |
| [12:37:50] |
<asmodai> |
I still see Warhammer, WoW and such a bit differently though. |
| [12:38:04] |
<asmodai> |
since there it is the amount of time you put in as opposed to money |
| [12:38:13] |
<prencher> |
it isn't though |
| [12:38:27] |
<prencher> |
you plop down $100 to a shady dealer and you've just brought 3 months of work |
| [12:38:49] |
<asmodai> |
For the character you mean? Yeah sure, but that's not what Blizzard themselves offer. |
| [12:38:59] |
<prencher> |
no, but that's the thing |
| [12:39:00] |
<asmodai> |
Which is different from this company that owns both game and the items. |
| [12:39:15] |
<prencher> |
imagine if they did offer it, and wow itself was cheap or free per month |
| [12:39:36] |
<prencher> |
you could spend $10 / mo getting a couple nice items instead of paying and grinding your ass off |
| [12:39:43] |
<prencher> |
hell thats the main reason people buy shit anyway.. they got jobs |
| [12:40:03] |
<prencher> |
im not saying i particularly agree with that business model, but thats the way things are, why not build for it |
| [12:40:05] |
<asmodai> |
But that's the problem, you then only focus on the loot. |
| [12:40:13] |
<asmodai> |
So you get the virtual equivalent of showing off. |
| [12:40:14] |
<prencher> |
people already only focus on the loot in wow |
| [12:40:19] |
<asmodai> |
(which already exists of course) |
| [12:40:30] |
<asmodai> |
prencher: Which is why my guild rocks :) |
| [12:40:40] |
<prencher> |
the game is nothing BUT loot, regardless of your view on it |
| [12:41:01] |
<prencher> |
there's no content endgame.. it's all grind for better loot - you do everything hundreds of times to get a shinier piece |
| [12:41:03] |
<prencher> |
or you stop playing |
| [12:41:29] |
<asmodai> |
Which is why I am enjoying my questing and all that. ^^ |
| [12:41:39] |
<prencher> |
that stuff runs out fast |
| [12:42:08] |
<asmodai> |
I wonder how long off persistent changes are |
| [12:42:22] |
<prencher> |
never going to happen |
| [12:42:33] |
<prencher> |
however wotlk is doing some interesting things with phasing |
| [12:42:35] |
<asmodai> |
There's actually a company working on that now. |
| [12:42:47] |
<prencher> |
yeah, there was 10 years ago as well |
| [12:43:06] |
<prencher> |
persistent changes just don't work when you have 5000 people doing shit to your world and all are the heroes |
| [12:43:25] |
<asmodai> |
That's the thing, of course. |
| [12:43:33] |
<asmodai> |
What phasing are you talking about? |
| [12:43:36] |
<prencher> |
phasing might become the happy medium |
| [12:43:40] |
<prencher> |
well, in wotlk |
| [12:43:51] |
<prencher> |
as you complete quests (or other stuff - anything c an trigger it) |
| [12:43:58] |
<prencher> |
thye can now change your characters viewstate on the world |
| [12:44:10] |
<prencher> |
example is the DK starting zone (they do it all over northrend too) |
| [12:44:25] |
<prencher> |
you do a few quests for mr lich king & buddies |
| [12:44:44] |
<asmodai> |
ah, to realize you've been his puppet |
| [12:44:49] |
<prencher> |
as you do, the area you're attacking changes - buildings get burned down, key characters get killed et al |
| [12:44:55] |
<asmodai> |
oh |
| [12:44:56] |
<prencher> |
and stuff stays burned down |
| [12:44:58] |
<asmodai> |
nice |
| [12:46:30] |
<prencher> |
it's not instanced - it's simply a viewstate for your character |
| [12:46:42] |
<asmodai> |
oh, that's nice |
| [12:46:43] |
<prencher> |
it's basically liek when you die.. you can see others but they can't see you |
| [12:46:45] |
<asmodai> |
got some background? |
| [12:46:54] |
<prencher> |
only for the entire world |
| [12:47:03] |
<prencher> |
hm? |
| [12:47:09] |
<asmodai> |
on that phasing |
| [12:47:28] |
<prencher> |
well.. here's perhaps a more tangible example |
| [12:47:36] |
<prencher> |
in northrend, there's an npc you can rescue |
| [12:47:45] |
<prencher> |
so you rescue it |
| [12:47:52] |
<prencher> |
take it back to the town |
| [12:48:00] |
<prencher> |
when you do so.. it stays at the town for you and everybody that's completed it |
| [12:48:05] |
<prencher> |
and that npc will now give you quests |
| [12:48:16] |
<prencher> |
for people that haven't completed it, that npc isn't at the town, it's at the place where you get the rescue quest |
| [12:48:53] |
<prencher> |
in the DK starting zone you change phase at least 5 times that i can think of |
| [12:49:18] |
<prencher> |
all the while you can only see players in the same phase you are in, and the world in that phase |
| [12:49:22] |
<asmodai> |
That is indeed a nice change slowly working to a semi-persistent world |
| [12:49:43] |
<prencher> |
it just means they can do persistent-ish changes for you |
| [12:49:57] |
<prencher> |
within reason - they can't really remove buildings for example |
| [12:50:19] |
<asmodai> |
No of course, but it means you could at least work more on story based questing |
| [12:50:20] |
<prencher> |
well they could in areas where the phase is the entire world |
| [12:50:28] |
<asmodai> |
instead of having everything stay very static |
| [12:50:32] |
<prencher> |
like the DK area - where a keep will be burned down eventually |
| [12:50:57] |
<prencher> |
but take the npc for example.. such a viewstate change wouldn't work with a building getting burned down if you could access the upper floor before |
| [12:51:01] |
<prencher> |
you'd see people running in midair |
| [12:51:05] |
<prencher> |
or running into it and dissappearing |
| [12:51:18] |
<prencher> |
and yes.. story wise |
| [12:51:22] |
<prencher> |
well.. you'll see :) |
| [12:51:34] |
<prencher> |
they are taking shit to whole new levels in that regard with lots of phasing |
| [12:51:44] |
<prencher> |
it's cool, but it doesn't change the endgame issue |
| [12:52:17] |
<asmodai> |
true, but then again, I wonder if you could ever solve that |
| [12:52:45] |
<asmodai> |
I mean, it remains a game where you have to set a challenge that people need to overcome |
| [12:53:03] |
<asmodai> |
of course, common/uncommon/rare/epic/legandary reminds me of M:TG |
| [12:53:09] |
<asmodai> |
legendary |
| [12:54:01] |
<prencher> |
i really liked guild wars' approach .. since you go from instance to instance, they changed the stuff lots over the course of the storyline |
| [12:54:11] |
<prencher> |
you still ended up with the endgame of course, but |
| [12:54:57] |
<prencher> |
but the key example in gw is the very beginning.. the first 5ish levels you do pre-charring.. in this pristine world |
| [12:55:26] |
<prencher> |
then at one point you get to the charring, and once that's over you're now in the present day, and those other areas are not available anymore |
| [12:55:29] |
<prencher> |
LOTRO did similar |
| [12:55:41] |
<prencher> |
but the areas you were in are now destroyed and burned down et al |
| [12:56:09] |
<asmodai> |
*nods* |
| [12:56:22] |
<asmodai> |
But is there a solution to end game raiding at all |
| [12:58:11] |
<prencher> |
imo you have to take the guild wars or daoc approach |
| [12:58:26] |
<prencher> |
gw did it to a lesser extent, but the daoc model worked well |
| [12:58:44] |
<prencher> |
which is a perceptual struggle of "back and forth" |
| [12:58:46] |
<asmodai> |
not sure about daoc |
| [12:58:55] |
<asmodai> |
only played it for a short while |
| [12:58:56] |
<prencher> |
well, there were 3 realms |
| [12:59:12] |
<prencher> |
each realm had a frontier - rvr zone; in there there were 11 keeps to capture and hold |
| [12:59:17] |
<prencher> |
also there were relic keeps |
| [12:59:28] |
<prencher> |
relics conveyed realm-wide bonuses |
| [12:59:40] |
<prencher> |
so you have these huge raids on the relic keeps to steal the relics for your own realm |
| [13:00:08] |
<asmodai> |
rofl: If we raised mana just because players asked for it in the forums, everyone would be in nine digits by now. |
| [13:00:13] |
<prencher> |
also, whoever had the most keeps had access to darkness falls - an endgame epic-loot dungeon that was rvr enabled |
| [13:00:21] |
<asmodai> |
mmm |
| [13:00:33] |
<asmodai> |
but doesn't that depend on the factions being 33/33/33 split? |
| [13:00:44] |
<prencher> |
and had bigass raid bosses |
| [13:01:15] |
<prencher> |
also, since it was rvr enabled, you had lots of fun down there when you took control - your realm would gain control and a ton of people would just rush down there before another raealm got control (within a couple hours usually - sometimes iwthin 20 mins) |
| [13:01:38] |
<prencher> |
and then you'd hunt down the people from the oither realms down there and kill them - once they die they can't come back, since htey no longer had control |
| [13:01:55] |
<prencher> |
awesome fun |
| [13:02:23] |
<prencher> |
it did sorta depend on an somewhat even realm split, but i was on the underpopulated realm on our server and we did fine |
| [13:02:54] |
* |
sashav has joined #pocoo |
| [13:02:55] |
<asmodai> |
ok, 'cause say EU-Thunderhorn has a 2,7:1 Alliance:Horde ratio |
| [13:02:59] |
<asmodai> |
effective 1,7:1 |
| [13:03:39] |
<prencher> |
asmodai - well that's why i thought gw had the edge there - it just didn't have the incentive to do the tug of war, since there was nothing global to be had |
| [13:04:03] |
<prencher> |
but since their system was based on guild vs guild, you didn't get the issue of "80% picked horde" |
| [13:08:16] |
* |
davidcramer has quit IRC |
| [13:12:44] |
<asmodai> |
mmm |
| [13:12:53] |
<asmodai> |
but doesn't that mean bigger guilds are stronger? |
| [13:13:06] |
<prencher> |
in theory.. in practice it just doesn't work out that wy |
| [13:13:48] |
<asmodai> |
prencher: got flash? |
| [13:13:53] |
<prencher> |
hm? |
| [13:13:57] |
<asmodai> |
http://FantasticContraption.com/?designId=2468622 |
| [13:14:16] |
<prencher> |
thinking about it though.. EVE-Online nails every aspect of persistent changes and endgame, if you like the gameplay itself |
| [13:14:29] |
<asmodai> |
I found it boring as hell |
| [13:15:13] |
<prencher> |
sure, i find it hard to stay interested in it too, but |
| [13:15:37] |
<prencher> |
it's the corps that run that game outside of empire space, and there's a constant battle of politics and zone control, not to mention all out wars that's purely between players |
| [13:15:48] |
<prencher> |
they really nailed it, as far as that aspect is concerned |
| [13:16:02] |
<asmodai> |
But that's if you're into this PvP stuff |
| [13:16:25] |
<prencher> |
it isn't though |
| [13:16:36] |
<prencher> |
there's tons and tons of things to do that never ever involves pvp |
| [13:16:46] |
<prencher> |
just because you're working against other corps doesn't mean you're doing pvp |
| [13:16:57] |
<prencher> |
you don't have ot be working against other corps either for that matter |
| [13:17:07] |
<asmodai> |
But in the 0,0 - 0,3 zones it is though |
| [13:17:13] |
<prencher> |
there are plenty of contractors building stuff and playing politics and the markets and stuff |
| [13:17:19] |
* |
sashav has quit IRC |
| [13:17:20] |
<prencher> |
you are mistaken |
| [13:17:33] |
<asmodai> |
I often am, but I learn. :P |
| [13:17:49] |
<prencher> |
first off, you don't have to go into those zones.. and there are plenty of areas you can get allowed into by the corps controlling them |
| [13:18:49] |
<prencher> |
anyways.. FC is kinda cool but meh |
| [13:18:55] |
<asmodai> |
FC? |
| [13:18:57] |
<prencher> |
not my kind of minigame |
| [13:19:03] |
<asmodai> |
ah |
| [13:19:30] |
<zepolen> |
is there a doc for the werkzeug libraries or just read the source? |
| [13:20:25] |
<prencher> |
yes.. the "Documentation" part of the website |
| [13:21:07] |
<zepolen> |
but it doesnt give a comprehensive module overview like sqlalchemy's docs have, its more like a bunch of tutorials that touch upon various parts of werkzeug |
| [13:28:15] |
* |
sashav has joined #pocoo |
| [13:49:06] |
* |
grumpy has joined #pocoo |
| [13:49:44] |
<grumpy> |
prencher: are you using Python 3.0 yet. Python 2.6 is so yesterdays news. ;-) |
| [13:51:01] |
<_paneb> |
if my service gets a PUT with an attached XML document (and the client send the right Content-Type header), i will have to use the request's header dict? |
| [13:51:26] |
<grumpy> |
How come mitsuhiko isn't around. Need to grill him about threading.local() use in Werkzeug? |
| [13:52:07] |
<asmodai> |
prencher: I had totally missed this: http://www.crummy.com/software/BeautifulSoup/documentation.html#Searching by CSS class |
| [13:53:37] |
* |
highwaychile has quit IRC |
| [13:54:08] |
* |
sashav_ has joined #pocoo |
| [14:12:15] |
* |
sashav has quit IRC |
| [14:13:12] |
<aa_> |
grumpy: womething wrong with threading.local ? |
| [14:14:25] |
* |
sashav_ has quit IRC |
| [14:19:10] |
<prencher> |
grumpy - i'd happily use 3.0 if it was out and all libs worked on it |
| [14:21:02] |
<prencher> |
need to get 2.6 deployed along with mod_wsgi |
| [14:21:52] |
* |
KirkMcDonald has quit IRC |
| [14:26:27] |
* |
KirkMcDonald has joined #pocoo |
| [14:28:29] |
* |
Baumichel has quit IRC |
| [14:42:06] |
<grumpy> |
aa_: sorry, had to drive mother in law home. |
| [14:42:50] |
<grumpy> |
aa_: Been noticed that in mod_wsgi that because C level thread state object created per request, that threading.local() does't sruvive between requests even for same Apache thread. |
| [14:43:57] |
<grumpy> |
aa_: in general I believe that threading.local() use only for life of request and then is cleared of attributes for that request. |
| [14:44:01] |
<prencher> |
grumpy - pretty sure werkzeug doesn't use thread locals at all itself, it just exposes the local module, which is generally used for request-duration "globals" |
| [14:44:21] |
<grumpy> |
aa_; thus not sure it is a big deal, but want to query mitsuhiko about uses cases he has seen for threading.local() |
| [14:44:25] |
<prencher> |
common idiom is to assign request to a thread local so as to have it available everywhere, rather than passing it around |
| [14:44:50] |
<grumpy> |
aa_: one bad example is someone using threading.local() to hold database handle as cheat way to avoid doing locking themselves on common data structure. |
| [14:45:20] |
<grumpy> |
baby crying, time to go feed her, back later .... |
| [14:49:25] |
<tux21b> |
grumpy, we use it to pass the request object and the application context around. and my last changeset in werkzeug was about to cache the data from memcache too in the context of request to avoid querying >50 times the same key |
| [14:49:58] |
<tux21b> |
but something there is behaving odd, so I need Armin too :D |
| [14:52:02] |
<tux21b> |
grumpy, and sqlalchemy does the same for connections and the object store, but they are having their own thread.local |
| [14:53:51] |
<grumpy> |
It is the sqlalchemy like case I am worried about. If it is doing it only for request context then fine. |
| [14:54:17] |
<grumpy> |
But if it is using it to stash a database handle per thread and expects to persist across requests handled by that thread, then it is a problem. |
| [14:54:43] |
<tux21b> |
grumpy, you should call session.remove() after each requests, which deletes the object store and commits/rollback open transactions |
| [14:55:08] |
<prencher> |
tux21b - that isn't the issue, the issue would be the connection pooling sqlalchemy can do |
| [14:55:29] |
<grumpy> |
Am not talking in this case about per request. Ie., talking about case where threading.local() created a global scope for use by thread for whole life of process. |
| [14:55:30] |
<prencher> |
if you can't persist the pool of connections between requests you've just completely nullified it |
| [14:56:06] |
<prencher> |
i forget how exactly sqlalch does it, but look at their connection pool documentation grumpy |
| [14:56:09] |
<grumpy> |
For a pool of connections, threading.local() wouldn't be used as would use global variable and proper mutex locking. |
| [14:56:47] |
<grumpy> |
This is because in that case you don't want to bind one connection per thread as may want less connections than threads and share limited number. |
| [14:56:47] |
<prencher> |
not for the global pool, it might be using thread locals as well for it |
| [14:57:15] |
<tux21b> |
ok, then they are probably just binding connections with open transactions and the object store of course. Otherwise it wouldn't make sense ;) |
| [14:57:58] |
<grumpy> |
Problem is naive people who think it is easier to bind one connection direct to thread. |
| [14:58:12] |
<grumpy> |
In mod_wsgi such a connection right now would get killed off at end of request. |
| [14:58:36] |
<tux21b> |
Have you looked at the sources? Does sa really handle it this way? |
| [14:58:43] |
<grumpy> |
In CherryPy it can also reclaim threads when it feels like it, so connections in that hosting environment could also get cleaned up before you expect. |
| [14:58:50] |
<prencher> |
http://www.sqlalchemy.org/docs/04/pooling.html here is the relevant documentation grumpy |
| [14:59:02] |
<prencher> |
it *can* use threadlocals - but it wouldn't be an issue with the way mod_wsgi does things |
| [14:59:11] |
<grumpy> |
Am not saying that anything does it using threading.local(). Would expect they do it properly. |
| [14:59:26] |
<prencher> |
all it's threadlocals support does is make sure the same thread gets the same connection object if one is already allocated to it from the pool |
| [14:59:39] |
<prencher> |
so, nothing to worry about from sqlalch's side |
| [14:59:45] |
<grumpy> |
But only in context of same request. Correct? Yes/No. |
| [15:00:00] |
<prencher> |
sqlalch doesn't deal with that |
| [15:00:22] |
<prencher> |
it's pool is not bound to threads, what it can do is assign the same connection to the same thread always, that's all |
| [15:00:33] |
<prencher> |
so that mod_wsgi clears the thread with request won't be an issue |
| [15:00:56] |
<grumpy> |
Only in as much as next time that thread used for a request, it will get a new one from the pool. |
| [15:01:11] |
<prencher> |
yup |
| [15:01:11] |
<grumpy> |
Issues is whether this may mean that mod_wsgi is making things less efficient than it could be. |
| [15:01:29] |
<prencher> |
oh i doubt it really |
| [15:01:59] |
<grumpy> |
One never knows though. In case of SA, if it is grabbing from persistent pool then okay. |
| [15:02:08] |
<prencher> |
it is, thats exactly what the pooling is |
| [15:02:33] |
<grumpy> |
But for other systems, if they create new connection if can't find one in thread.local(), ie., not taking them from pool, that would be inefficient. |
| [15:02:38] |
<prencher> |
it's just N connections it can keep open and then assign to threads ad hoc - what use_threadlocals does is ensure that the same thread always gets the same connection instead of several if it requests a connection multiple times |
| [15:02:49] |
<prencher> |
indeed, you should document that |
| [15:03:23] |
<grumpy> |
Or I can change mod_wsgi code so it keeps table of thread state objects for Apache foreign threads and reuses them. |
| [15:03:59] |
<grumpy> |
Just one to research possibility that changing behaviour of what it does now will trap anyone. |
| [15:04:09] |
<grumpy> |
FWIW, mod_python behaves same as mod_wsgi. |
| [15:04:53] |
<grumpy> |
Something like Paste serve or CherryPy WSGI server, or FASTCGI solutions, because threads are Python threads and not foreign threads, the threading.local() would persist. |
| [15:05:07] |
<grumpy> |
Thus mod_wsgi would then just work like command line Python would. |
| [15:05:25] |
<grumpy> |
If someone was dependent on mod_wsgi behaviour, it would break on command line Python. |
| [15:05:38] |
<grumpy> |
Thus don't really see an issue in changing how mod_wsgi works. |
| [15:05:41] |
<prencher> |
being closer to "pure python" behaviour is always good |
| [15:06:30] |
<prencher> |
grumpy - the idiom in werkzeug is always to clear out your thread locals at the end of the request, similar with recycling the sqlalch session at the end of the request |
| [15:06:43] |
<grumpy> |
Hmmm, Colloquy is stuffing up on me. Shows posts in other channels but when I go there they vanish. :-( |
| [15:07:17] |
<prencher> |
or assign the thread locals at each request invocation - for example that 'request' sticks around doesn't matter, since you always do locals.request = Request(env) or similar at the start of each request |
| [15:07:41] |
<prencher> |
zine does similar with it's application context - it's assigned on each request invocation |
| [15:07:51] |
<prencher> |
and i believe inyoka does that too, right tux21b? |
| [15:08:06] |
<zepolen> |
what about thread locals which you want to persist between requests |
| [15:08:15] |
<grumpy> |
The threading.local() is stored in C structure of threadstate. Thus not accessible in Python directly. |
| [15:08:25] |
<prencher> |
that's what you can't do presently zepolen |
| [15:08:46] |
<zepolen> |
unless you make your own LocalManager correct? |
| [15:09:07] |
<grumpy> |
zepolen: works with Python command line, but not that way with mod_wsgi and mod_python. Thus what we are talking about, ie., making mod_wsgi match Python command line behaviour. |
| [15:09:10] |
<prencher> |
werkzeugs LocalManager just wraps threading.local |
| [15:09:15] |
<prencher> |
so no |
| [15:09:28] |
<grumpy> |
Hmmm, this time I think I have to feed baby, not going back to sleep. |
| [15:26:04] |
* |
Kaelten has joined #pocoo |
| [15:27:06] |
<asmodai> |
Heya Kaelten |
| [15:28:05] |
<Kaelten> |
hiya |
| [15:28:38] |
<grumpy> |
Time for me to go. I'll be away from home for the next week. Not sure I will be allowed on the net that much, so talk to you guys in a week if I can't get on before then. |
| [15:30:31] |
<asmodai> |
urllib2 has no caching concept, does it? |
| [15:33:23] |
<ronny> |
asmodai: httplib2 has |
| [15:33:27] |
<asmodai> |
yeah |
| [15:33:30] |
<asmodai> |
was looking at it |
| [15:33:44] |
<asmodai> |
was using urllib2, but for some of my screenscraping using httplib2 would be nicer |
| [15:35:08] |
* |
grumpy has quit IRC |
| [15:38:48] |
* |
zepolen has left #pocoo |
| [15:39:13] |
* |
zepolen has joined #pocoo |
| [16:05:17] |
* |
Baumichel has joined #pocoo |
| [16:17:57] |
<_paneb> |
if my service gets a PUT with an attached XML document (and the client send the right Content-Type header), i will have to use the request's header dict? |
| [16:29:01] |
* |
ph has joined #pocoo |
| [16:35:06] |
* |
ph is now known as peterh |
| [16:35:24] |
* |
peterh has left #pocoo |
| [16:45:52] |
* |
alowry has joined #pocoo |
| [17:00:17] |
* |
stifal has quit IRC |
| [17:03:20] |
* |
stifal has joined #pocoo |
| [17:15:04] |
* |
alowry has quit IRC |
| [17:40:28] |
* |
alowry has joined #pocoo |
| [17:46:27] |
* |
damjan_ is now known as damjan |
| [18:59:21] |
* |
EnTeQuAk has quit IRC |
| [19:02:45] |
* |
izibi has quit IRC |
| [19:03:57] |
* |
izibi has joined #pocoo |
| [19:04:49] |
* |
aconbere has joined #pocoo |
| [19:06:27] |
* |
sashav has joined #pocoo |
| [19:44:32] |
* |
davidcramer has joined #pocoo |
| [20:00:58] |
* |
jinks2 is now known as jinks |
| [20:09:39] |
* |
davidcramer has quit IRC |
| [20:10:18] |
* |
davidcramer has joined #pocoo |
| [20:19:04] |
* |
_paneb has quit IRC |
| [20:25:55] |
<birkenfeld> |
a life sign from mitsuhiko! |
| [20:27:52] |
<mitsuhiko> |
heh. where? ^^ |
| [20:31:16] |
<xorAxAx> |
if any electrical wiring issues come up, ask electrical engineer for advice |
| [20:31:24] |
<xorAxAx> |
+mitsuhiko |
| [20:32:20] |
<xorAxAx> |
mitsuhiko: how is pre-student life? :) |
| [20:32:29] |
<mitsuhiko> |
why pre? :) |
| [20:32:43] |
<mitsuhiko> |
very busy already |
| [20:33:35] |
<xorAxAx> |
oh, when did the semester start? |
| [20:35:11] |
<xorAxAx> |
mitsuhiko: |
| [20:35:26] |
<mitsuhiko> |
wednesday :) |
| [20:35:37] |
<mitsuhiko> |
(of tuesday if you count the welcome stuff) |
| [20:35:37] |
<tux21b> |
yeah, finally. hi armin, did you get my last pn? :) |
| [20:35:43] |
<xorAxAx> |
mine starts in 10 days again |
| [20:36:25] |
* |
tux21b has always thought that students don't know what they should do in all their free-time... |
| [20:36:26] |
<xorAxAx> |
mitsuhiko: and, will you use i or j to denote the imaginary unit? |
| [20:36:40] |
<xorAxAx> |
tux21b: to drink or not to drink, thats the question |
| [20:36:48] |
<mitsuhiko> |
tux21b: my network connection is crap :-/ |
| [20:36:54] |
<mitsuhiko> |
haven't had the time yet to check all my stuff |
| [20:37:11] |
<mitsuhiko> |
xorAxAx: j :) |
| [20:37:26] |
<xorAxAx> |
blasphemy |
| [20:37:32] |
<mitsuhiko> |
heh |
| [20:37:35] |
<tux21b> |
mitsuhiko, ah. ok. in short, local_manager seems not to be cleaning up on cleanup() at ubuntuusers |
| [20:40:00] |
* |
sashav has quit IRC |
| [20:43:36] |
<mitsuhiko> |
tux21b: impossible? |
| [20:43:50] |
* |
EnTeQuAk has joined #pocoo |
| [20:43:50] |
<mitsuhiko> |
testcase, then fix :) |
| [20:45:00] |
<tux21b> |
mitsuhiko, my main problem is currently to debug the whole thing. my test-cases (in the cli) where all working, but on the server, with multiple processes it looks different |
| [20:45:30] |
<mitsuhiko> |
tux21b: are you sure cleanup is called? |
| [20:46:11] |
<tux21b> |
i added locally some print's, but it's working fine locally (with 1 process) anyway |
| [20:47:14] |
<tux21b> |
my only changeset was to cache things from the memcache once again in the memory (as part of the local_manager) and now, we are having blink-effects on every request |
| [20:57:35] |
<prencher> |
so where you been mitsuhiko? birkenfeld thought (hoped?) you got hit a bus or something |
| [20:57:41] |
* |
Kaelten has quit IRC |
| [20:57:59] |
* |
Kaelten has joined #pocoo |
| [21:11:36] |
<mitsuhiko> |
prencher: still haven't got my flat |
| [21:11:38] |
<mitsuhiko> |
well, i got it. but it's pretty empty and has no connection yet |
| [21:13:37] |
* |
jpcw has quit IRC |
| [21:16:32] |
* |
sashav has joined #pocoo |
| [21:21:14] |
* |
sashav has joined #pocoo |
| [21:21:49] |
<apollo13> |
:þ |
| [21:22:59] |
<apollo13> |
mitsuhiko and no internet, must be hard :) |
| [21:23:25] |
<apollo13> |
mitsuhiko: http://doodle.ch/participation.html?pollId=iyts34rhuhsqzhgr |
| [21:23:28] |
<mitsuhiko> |
apollo13: i'm compensating it with getting drunk :P |
| [21:24:21] |
<apollo13> |
k, looking at the empty "sturm" bottles over there *hmm* |
| [21:24:33] |
<apollo13> |
sry for pinging empty :) |
| [21:25:04] |
<apollo13> |
mitsuhiko: you could compensate with studieng |
| [21:25:16] |
<apollo13> |
hmm I guess I misspelled that last word |
| [21:25:20] |
<apollo13> |
studyieng |
| [21:25:26] |
<apollo13> |
not better :/ |
| [21:25:27] |
<mitsuhiko> |
studying? |
| [21:25:32] |
<apollo13> |
ah yeah :) |
| [21:25:47] |
<mitsuhiko> |
i just finished doing that retarded homework for "internet and new media" |
| [21:25:48] |
<apollo13> |
my last english leason is far behind :p |
| [21:26:31] |
<apollo13> |
k, I tried to attend some edv stuff and technical meth, but I can't cause the overlap with my courses |
| [21:27:47] |
<mitsuhiko> |
structured programming is crazy. they enforce a brain dead code style |
| [21:28:08] |
<apollo13> |
Oh, then I am glad I don't have time for it :þ |
| [21:28:20] |
<apollo13> |
which language? c/c++? |
| [21:28:44] |
<mitsuhiko> |
c |
| [21:29:01] |
<mitsuhiko> |
with a styleguide that looks like the bastard child of c++/java and object pascal styleguides |
| [21:29:04] |
<apollo13> |
hmm, maybe I should attend it, my c fu doesn't exist^^ |
| [21:29:50] |
<mitsuhiko> |
there are way too many people |
| [21:30:04] |
<mitsuhiko> |
we were in hs p1 and some people had to stand next to the wall |
| [21:30:12] |
<apollo13> |
wtf |
| [21:30:29] |
<apollo13> |
we are max 70 here in technical physics :) |
| [21:30:59] |
<apollo13> |
and after the first semester about 50% are going to leave |
| [21:31:08] |
<mitsuhiko> |
apollo13: go to the tc, i think they uploaded videos of the lectures there |
| [21:31:24] |
<apollo13> |
tc == what? |
| [21:31:43] |
<mitsuhiko> |
try curry.tugraz.at |
| [21:31:47] |
<mitsuhiko> |
apollo13: tc == teach center |
| [21:31:58] |
<apollo13> |
aspx outch |
| [21:32:09] |
<apollo13> |
oh, that's cool |
| [21:32:38] |
<EnTeQuAk> |
lol |
| [21:33:02] |
<EnTeQuAk> |
this change of mood ;) |
| [21:33:13] |
<apollo13> |
now the videos are cool, not aspx |
| [21:33:23] |
<apollo13> |
no* |
| [21:33:37] |
<mitsuhiko> |
apollo13: what's wrong with aspx? |
| [21:33:57] |
<apollo13> |
I don't like it :þ |
| [21:34:05] |
<apollo13> |
or was it asp *hmm* |
| [21:34:16] |
<mitsuhiko> |
probably asp |
| [21:34:17] |
<EnTeQuAk> |
jsp? |
| [21:35:07] |
<EnTeQuAk> |
well, asp is cool :D |
| [21:35:21] |
<apollo13> |
eeks, http://code.djangoproject.com/ticket/7210 |
| [21:35:31] |
<apollo13> |
I actually would prefer the sa style |
| [21:37:21] |
<apollo13> |
I guess I'll turn off my pc and take a look at linear algebra, it's kinda interesting |
| [21:37:36] |
<EnTeQuAk> |
apollo13, have fun |
| [21:37:58] |
<EnTeQuAk> |
maybe you can develop some cool complex captcha for inyoka? :D |
| [21:38:32] |
<EnTeQuAk> |
would be cool if no dumb guy can register anymore :) |
| [21:40:14] |
<apollo13> |
Oh no, mitsuhiko hates math, would be impossible for him to register *runnnnnnnnnnnnnns* |
| [21:40:34] |
<apollo13> |
cy |
| [21:41:16] |
<EnTeQuAk> |
*g* |
| [21:42:19] |
<apollo13> |
mitsuhiko: btw you are out of luck, it only took me 5 minutes to get my gvb card |
| [21:42:50] |
<mitsuhiko> |
apollo13: but i got it three days earlier :P |
| [21:43:08] |
<apollo13> |
Oh :) |
| [21:43:09] |
* |
mitsuhiko is now configuring vim to this ridiculous coding standard |
| [21:43:15] |
<apollo13> |
lol |
| [21:43:18] |
<apollo13> |
cy tomorrow |
| [21:43:25] |
<mitsuhiko> |
apollo13: why? |
| [21:43:29] |
<apollo13> |
why not? |
| [21:43:30] |
<mitsuhiko> |
or, where :) |
| [21:43:39] |
<apollo13> |
depends |
| [21:43:42] |
<mitsuhiko> |
apollo13: because i'm buying furniture at ikea |
| [21:43:47] |
<apollo13> |
ah |
| [21:43:53] |
<apollo13> |
*hmm* |
| [21:44:00] |
<apollo13> |
won't take the whole day ;) |
| [21:44:20] |
<apollo13> |
I might phone you if I am near jakomini |
| [21:44:37] |
<apollo13> |
btw I am going to by myself a new calculator |
| [21:44:52] |
<apollo13 |