*** kosh has joined #zope | 00:47 | |
sqwishy | Suppose I'm trying to use zope to make serializers to turn object into strings. It seems simple to use adapters to get strings from objects. But I'm not sure how to go the other way. | 00:54 |
---|---|---|
sqwishy | I can obtain an adapter from an object which I want serialized, but I can't obtain the adapter from the serialized string. | 00:54 |
*** menesis has quit IRC | 01:02 | |
*** Arfrever has quit IRC | 01:11 | |
*** frapell has quit IRC | 01:15 | |
*** frapell has joined #zope | 01:15 | |
*** frapell has joined #zope | 01:15 | |
*** polysix has quit IRC | 01:24 | |
J1m | sqwishy, see http://svn.zope.org/zope.fssync/trunk/ which does what you're talking about more or less. | 01:37 |
*** polysix has joined #zope | 02:05 | |
*** J1m has quit IRC | 02:21 | |
*** _mup_ has quit IRC | 03:00 | |
*** _mup_ has joined #zope | 03:00 | |
*** benji has quit IRC | 03:16 | |
*** frapell has quit IRC | 03:30 | |
*** frapell has joined #zope | 03:30 | |
*** frapell has joined #zope | 03:30 | |
*** Spanktar has quit IRC | 03:44 | |
*** RichyB has quit IRC | 04:11 | |
*** tiwula has quit IRC | 04:22 | |
*** benji has joined #zope | 04:26 | |
*** alecm has quit IRC | 04:35 | |
*** alecm has joined #zope | 04:42 | |
*** alecm has joined #zope | 04:42 | |
*** alecm has quit IRC | 05:31 | |
*** tiwula has joined #zope | 05:41 | |
*** dayne has joined #zope | 05:49 | |
*** nande has quit IRC | 05:51 | |
*** alecm has joined #zope | 05:52 | |
*** KageSenshi has joined #zope | 05:56 | |
*** kosh has quit IRC | 06:52 | |
*** rbanffy has joined #zope | 07:16 | |
*** mcdonc has joined #zope | 07:26 | |
*** benji has quit IRC | 07:39 | |
*** benji has joined #zope | 07:40 | |
*** __mac__ has joined #zope | 07:47 | |
*** __mac__ has quit IRC | 08:10 | |
*** tiwula has quit IRC | 08:22 | |
*** mcdonc_ has joined #zope | 09:03 | |
*** mcdonc has quit IRC | 09:04 | |
*** mcdonc_ has quit IRC | 09:10 | |
*** __mac__ has joined #zope | 09:21 | |
*** mcdonc has joined #zope | 09:23 | |
*** agroszer has joined #zope | 09:35 | |
*** mcdonc has quit IRC | 10:01 | |
*** giacomos has joined #zope | 10:01 | |
*** KageSenshi has quit IRC | 10:15 | |
*** J1m has joined #zope | 10:21 | |
*** KageSenshi has joined #zope | 10:30 | |
*** eperez has joined #zope | 10:36 | |
*** J1m has quit IRC | 10:58 | |
*** menesis has joined #zope | 11:23 | |
*** mitchell`off is now known as mitchell` | 11:23 | |
*** fredvd has joined #zope | 11:30 | |
*** zenwryly has quit IRC | 11:34 | |
*** zenwryly has joined #zope | 11:34 | |
*** evilbungle has quit IRC | 11:39 | |
*** zenwryly has quit IRC | 11:47 | |
*** zenwryly has joined #zope | 11:47 | |
*** evilbungle has joined #zope | 12:03 | |
*** eperez has quit IRC | 12:14 | |
*** maurits has joined #zope | 12:17 | |
*** giacomos has quit IRC | 12:23 | |
*** zenwryly has quit IRC | 12:23 | |
*** zenwryly has joined #zope | 12:23 | |
*** giacomos has joined #zope | 12:27 | |
*** teix has joined #zope | 12:32 | |
*** zenwryly has quit IRC | 12:37 | |
*** zenwryly has joined #zope | 12:37 | |
*** eperez has joined #zope | 12:48 | |
*** zenwryly has quit IRC | 12:52 | |
*** zenwryly has joined #zope | 12:52 | |
*** eperez has quit IRC | 12:57 | |
*** eperez has joined #zope | 13:00 | |
*** zenwryly has quit IRC | 13:06 | |
*** zenwryly has joined #zope | 13:06 | |
*** zenwryly has quit IRC | 13:11 | |
*** zenwryly has joined #zope | 13:12 | |
*** zenwryly has quit IRC | 13:20 | |
*** zenwryly has joined #zope | 13:21 | |
*** avoinea has joined #zope | 13:21 | |
*** zenwryly has quit IRC | 13:31 | |
*** zenwryly has joined #zope | 13:31 | |
*** zenwryly has quit IRC | 13:35 | |
*** zenwryly has joined #zope | 13:35 | |
*** zenwryly has quit IRC | 13:49 | |
*** zenwryly has joined #zope | 13:49 | |
*** zenwryly has quit IRC | 13:50 | |
*** zenwryly has joined #zope | 13:50 | |
* mgedmin wants to improve z3c.password exception messages | 13:53 | |
*** RichyB has joined #zope | 14:02 | |
*** zenwryly has quit IRC | 14:04 | |
*** zenwryly has joined #zope | 14:04 | |
* mgedmin wants to kill himself | 14:08 | |
*** zenwryly has quit IRC | 14:10 | |
*** zenwryly has joined #zope | 14:11 | |
*** J1m has joined #zope | 14:14 | |
*** zenwryly has quit IRC | 14:18 | |
*** zenwryly has joined #zope | 14:18 | |
*** zenwryly has quit IRC | 14:33 | |
*** zenwryly has joined #zope | 14:34 | |
* mgedmin gets first signs of success! | 14:39 | |
*** sunew has joined #zope | 14:42 | |
*** zenwryly has quit IRC | 14:47 | |
*** zenwryly has joined #zope | 14:47 | |
*** zenwryly has quit IRC | 15:01 | |
*** zenwryly has joined #zope | 15:01 | |
*** mcdonc has joined #zope | 15:06 | |
*** nande has joined #zope | 15:11 | |
*** zenwryly has quit IRC | 15:16 | |
*** zenwryly has joined #zope | 15:16 | |
mgedmin | anybody cares about z3c.password and would like to take a look at http://zope3.pov.lt/trac/changeset/127450 ? | 15:23 |
*** zenwryly has quit IRC | 15:30 | |
*** zenwryly has joined #zope | 15:31 | |
*** J1m has quit IRC | 15:34 | |
mgedmin | what was that tool that used SETUP.cfg files? | 15:44 |
mgedmin | is it dead dead dead? | 15:44 |
*** zenwryly has quit IRC | 15:45 | |
*** zenwryly has joined #zope | 15:45 | |
mgedmin | zpkg, I think? | 15:45 |
mgedmin | oh, gmane... "14 Nov 18:25" | 15:46 |
mgedmin | WHAT YEAR? | 15:46 |
mgedmin | turns out, 2005 | 15:46 |
mgedmin | but nooo, time of day is way more important to see in a discussion than the fricking year | 15:46 |
* mgedmin realizes the year and month are visible in the sidebar | 15:47 | |
nande | xd i have no idea what z3c.password is | 15:47 |
betabug | mgedmin: gmane's date thing is really idiotic, I think you have to dig for the year in the headers | 15:51 |
mgedmin | I dug it out of svn logs given revision numbers (the zpkg mailing list is quiet -- spam and svn log messages) | 15:51 |
*** eperez has quit IRC | 15:53 | |
agroszer | mgedmin, go ahead and cut a release | 15:54 |
mgedmin | of z3c.password? I can do that... | 15:54 |
agroszer | yup | 15:54 |
*** zenwryly has quit IRC | 15:59 | |
*** zenwryly has joined #zope | 15:59 | |
*** zenwryly has quit IRC | 16:13 | |
*** zenwryly has joined #zope | 16:15 | |
*** kosh has joined #zope | 16:16 | |
*** mcdonc_ has joined #zope | 16:20 | |
*** mcdonc__ has joined #zope | 16:21 | |
*** mcdonc_ has quit IRC | 16:25 | |
*** mcdonc has quit IRC | 16:25 | |
*** zenwryly has quit IRC | 16:28 | |
*** zenwryly has joined #zope | 16:28 | |
*** mcdonc has joined #zope | 16:29 | |
*** mcdonc__ has quit IRC | 16:34 | |
*** J1m has joined #zope | 16:34 | |
*** __mac__ has quit IRC | 16:39 | |
*** zenwryly has quit IRC | 16:42 | |
*** zenwryly has joined #zope | 16:43 | |
*** frapell has quit IRC | 16:45 | |
*** frapell has joined #zope | 16:46 | |
*** frapell has joined #zope | 16:46 | |
*** goschtl has joined #zope | 16:46 | |
*** frapell has joined #zope | 16:46 | |
*** frapell has joined #zope | 16:46 | |
*** __mac__ has joined #zope | 16:51 | |
*** menesis has quit IRC | 16:56 | |
*** zenwryly has quit IRC | 16:57 | |
*** zenwryly has joined #zope | 16:57 | |
*** sunew has quit IRC | 17:00 | |
*** __mac__ has quit IRC | 17:01 | |
*** dayne has quit IRC | 17:05 | |
*** zenwryly has quit IRC | 17:11 | |
*** zenwryly has joined #zope | 17:11 | |
*** menesis has joined #zope | 17:23 | |
*** zenwryly has quit IRC | 17:25 | |
*** zenwryly has joined #zope | 17:26 | |
*** zenwryly has quit IRC | 17:40 | |
*** zenwryly has joined #zope | 17:41 | |
*** Fs0L has joined #zope | 17:43 | |
*** zenwryly has quit IRC | 17:54 | |
*** zenwryly has joined #zope | 17:54 | |
*** maurits has quit IRC | 18:07 | |
*** zenwryly has quit IRC | 18:09 | |
*** zenwryly has joined #zope | 18:09 | |
*** daMaestro has joined #zope | 18:10 | |
*** zenwryly has quit IRC | 18:23 | |
*** zenwryly has joined #zope | 18:24 | |
*** benji has quit IRC | 18:26 | |
*** benji has joined #zope | 18:27 | |
*** zenwryly has quit IRC | 18:38 | |
*** zenwryly has joined #zope | 18:38 | |
*** benji has quit IRC | 18:40 | |
*** benji has joined #zope | 18:41 | |
*** agroszer has quit IRC | 18:49 | |
*** goschtl has quit IRC | 18:49 | |
*** zenwryly has quit IRC | 18:52 | |
*** zenwryly has joined #zope | 18:53 | |
*** tiwula has joined #zope | 19:02 | |
*** avoinea has quit IRC | 19:06 | |
*** zenwryly has quit IRC | 19:06 | |
*** zenwryly has joined #zope | 19:06 | |
*** kosh has quit IRC | 19:08 | |
*** giacomos has quit IRC | 19:11 | |
*** zenwryly has quit IRC | 19:21 | |
*** zenwryly has joined #zope | 19:21 | |
*** Spanktar has joined #zope | 19:31 | |
*** mcdonc_ has joined #zope | 19:31 | |
*** mcdonc has quit IRC | 19:35 | |
*** zenwryly has quit IRC | 19:35 | |
*** zenwryly has joined #zope | 19:35 | |
*** mcdonc_ has quit IRC | 19:37 | |
*** zenwryly has quit IRC | 19:50 | |
*** zenwryly has joined #zope | 19:50 | |
*** RichardB has joined #zope | 20:00 | |
*** RichyB has quit IRC | 20:00 | |
*** RichardB has quit IRC | 20:04 | |
*** zenwryly has quit IRC | 20:04 | |
*** zenwryly has joined #zope | 20:04 | |
*** RichyB has joined #zope | 20:05 | |
*** zenwryly has quit IRC | 20:18 | |
*** zenwryly has joined #zope | 20:20 | |
*** evilbungle has quit IRC | 20:27 | |
*** yvl has quit IRC | 20:28 | |
*** frapell has quit IRC | 20:30 | |
*** frapell has joined #zope | 20:30 | |
*** frapell has joined #zope | 20:30 | |
*** frapell has quit IRC | 20:30 | |
*** Guest95083 has joined #zope | 20:30 | |
*** zenwryly has quit IRC | 20:33 | |
*** zenwryly has joined #zope | 20:33 | |
*** Guest95083 has quit IRC | 20:34 | |
*** maker has joined #zope | 20:36 | |
maker | Hello. | 20:36 |
maker | I have an interface and a mixin class, with a few abstractmethods. I was wondering if there was something to achieve the result of "raise NotImplementedError" if the specific method was not overrided withoud breaking verify.verifyObject. | 20:38 |
*** zenwryly has quit IRC | 20:47 | |
*** zenwryly has joined #zope | 20:47 | |
*** zenwryly has quit IRC | 21:02 | |
*** zenwryly has joined #zope | 21:03 | |
*** frapell has joined #zope | 21:08 | |
*** frapell has joined #zope | 21:08 | |
*** TresEquis has joined #zope | 21:13 | |
*** menesis has quit IRC | 21:15 | |
*** zenwryly has quit IRC | 21:16 | |
*** zenwryly has joined #zope | 21:16 | |
*** frapell has quit IRC | 21:26 | |
*** frapell has joined #zope | 21:26 | |
*** frapell has joined #zope | 21:27 | |
*** zenwryly has quit IRC | 21:30 | |
*** zenwryly has joined #zope | 21:31 | |
*** zenwryly has quit IRC | 21:45 | |
*** zenwryly has joined #zope | 21:45 | |
*** __mac__ has joined #zope | 21:46 | |
*** zenwryly has quit IRC | 21:59 | |
*** zenwryly has joined #zope | 21:59 | |
*** zenwryly has quit IRC | 22:14 | |
*** zenwryly has joined #zope | 22:14 | |
*** Fs0L has quit IRC | 22:21 | |
*** RichyB has quit IRC | 22:25 | |
*** zenwryly has quit IRC | 22:28 | |
*** zenwryly has joined #zope | 22:28 | |
*** zenwryly has quit IRC | 22:43 | |
*** zenwryly has joined #zope | 22:44 | |
*** TresEquis has quit IRC | 22:45 | |
*** TresEquis has joined #zope | 22:45 | |
benbangert | J1m: around? | 22:46 |
benbangert | got a weird heisenbug in logging thread madness with gevent :) | 22:46 |
*** zenwryly has quit IRC | 22:57 | |
*** zenwryly has joined #zope | 22:57 | |
benbangert | J1m: your logging trick isn't good enough. https://github.com/python-zk/kazoo/issues/16 | 22:58 |
*** teix has quit IRC | 22:58 | |
benbangert | J1m: because the logger has a lock, and the monkey patching mucks up the lock with a gevent lock.... | 22:58 |
J1m | Have I mentioned my opinion of gevent monkey patching? | 23:05 |
benbangert | J1m: found a solution.... grabbing the logging Handler instance itself, and setting lock = None on it | 23:07 |
benbangert | will prevent it from attempting to lock at all during writing output | 23:07 |
benbangert | yea, I know, it sucks | 23:08 |
benbangert | :) | 23:08 |
J1m | benbangert, so this is really an issue w gevent and logging. | 23:08 |
benbangert | since this is the only user of the ZooKeeper logging handler, it shouldn't need to lock | 23:08 |
benbangert | yea, cause logging handlers make a threading.RLock instance..... and when monkey-patched..... boom | 23:08 |
J1m | That's a stretch imo. | 23:08 |
benbangert | I could replace the lock on the handler instance with a proper threading lock | 23:09 |
J1m | anyone can use any logger they want. | 23:09 |
benbangert | though I have to dig up the real un-monkey-patched one :) | 23:09 |
benbangert | its really nasty to get an unadulterated one though :( | 23:09 |
J1m | It's nasty to run around monkey patching things in such a profound way. | 23:10 |
J1m | Monkey patching in gevent is optional. | 23:10 |
J1m | I use gevent wo monkey patching. | 23:10 |
*** zenwryly has quit IRC | 23:11 | |
benbangert | sure, but not everyone does | 23:12 |
J1m | This is a hard problem to solve. You're basically saying that gevent is going to break a basic feature of the standard library and that kazoo has to somehow work around that. | 23:12 |
benbangert | thats correct | 23:12 |
benbangert | gevent monkey patching breaks logging | 23:12 |
J1m | k, whatever :) | 23:12 |
benbangert | if you want to log from a separate thread | 23:13 |
*** zenwryly has joined #zope | 23:13 | |
J1m | I wouldn't want to have maintain that. | 23:13 |
J1m | when zc.zk is reimimplemented on top of kazoo, I won't have to worry about it. | 23:13 |
benbangert | not supporting it rules out gevent entirely for web use | 23:13 |
J1m | It does not. | 23:13 |
benbangert | since the webapp server monkey patch up front | 23:13 |
J1m | No way. | 23:13 |
benbangert | gunicorn + gevent will monkey patch as it loads | 23:13 |
J1m | OK, thanks for the warning. :) | 23:14 |
benbangert | once it does, the logging done the way kazoo/zc.zk will break at some point | 23:14 |
J1m | so basically, you're saying gunicorn is broken. | 23:14 |
benbangert | if you ask it to use gevent, it monkey-patches with gevent | 23:14 |
benbangert | so that your webapp works 'properly' with gevent | 23:14 |
J1m | pywsgi doesn't require monkey patching. | 23:15 |
J1m | This is wildly more invasive that zope ever was. | 23:15 |
J1m | well, whatever. | 23:15 |
J1m | If you feel like maintaining this in kazoo, more power to you. I wouldn't. | 23:16 |
benbangert | pywsgi must not be spawning greenlets to handle requests then | 23:17 |
J1m | yes | 23:17 |
J1m | no | 23:17 |
J1m | It does spawn greenlets. | 23:17 |
benbangert | ah, so it relies on you actually making sure to use gevent in your app | 23:17 |
J1m | You don't need monkey patching to use pywsgi. | 23:17 |
benbangert | gunicorn takes the 'magic' approach by using monkey patching so your app "just works" | 23:17 |
J1m | Uh huh. | 23:18 |
benbangert | cause odds are most of your libraries somewhere, are not using gevent in a large app | 23:18 |
benbangert | so without the monkey-patching, it'll have crappy performance | 23:18 |
J1m | imo, anyone who has a clue and uses an async network api also uses thread pools, unless all they're doing is moving bits between sockets. | 23:19 |
benbangert | agreed | 23:19 |
J1m | and if you do that, you don't need monkey patching. | 23:19 |
*** __mac__ has quit IRC | 23:20 | |
J1m | These people in the gevent (and twisted) community who think you can do everything async and that everyone else has to conform are drinking too much of their own cool-ade. | 23:21 |
J1m | That's my (increasingly grumpy :) opinion. :) | 23:21 |
benbangert | ya know, I agree entirely, this code is vomit-worthy | 23:23 |
benbangert | :) | 23:23 |
*** zenwryly has quit IRC | 23:26 | |
*** zenwryly has joined #zope | 23:26 | |
J1m | benbangert, ftr zc.resumelb is a wsgi server built on gevent (and zc.zk) that doesn't require apps to be gevent aware and doesn't monkey patch. | 23:32 |
benbangert | J1m: yea, but I doubt apps with threads would 'just work' | 23:32 |
benbangert | cause the second you make a lock object and use it across threads and a gevent thread, gevent will vomit | 23:32 |
J1m | No, because there's no monkey patching in zc.resumelb. In fact, zc.resumelb provides thread pools, which is how we use it. | 23:33 |
J1m | My point is that you can have a gevent based wsgi server without monkey patching and without apps having to be gevent aware. | 23:34 |
J1m | zc.resumelb doesn't use thread pools by default, though maybe it should. | 23:35 |
benbangert | how does it handle the socket without monkey patching it? | 23:36 |
benbangert | oh, you use the gevent one | 23:36 |
J1m | Yes. | 23:36 |
benbangert | so most libraries use a socket to talk to other services. if they all use blocking sockets, then your 100 greenlets are gonna go one at a time on the socket | 23:36 |
J1m | Yes, unless you use thread pools. | 23:37 |
J1m | which is what we do. | 23:37 |
benbangert | sure, your app will 'work' without the monkey patching, but it won't get a very big performance boost | 23:37 |
J1m | sure it will, cuz the the HTTP IO is still asyncronous. | 23:37 |
TresEquis | join #breckprog | 23:38 |
J1m | You aren't creating a thread for each request. | 23:38 |
J1m | TresEquis, you talkin to us? | 23:38 |
TresEquis | sorry, missed the '/' | 23:38 |
J1m | heh | 23:39 |
*** zenwryly has quit IRC | 23:40 | |
*** zenwryly has joined #zope | 23:40 | |
*** zenwryly has quit IRC | 23:55 | |
*** zenwryly has joined #zope | 23:55 | |
*** TresEquis has quit IRC | 23:57 |
Generated by irclog2html.py 2.15.1 by Marius Gedminas - find it at mg.pov.lt!