| *** tarek has quit IRC | 00:01 | |
| *** mwh2 has joined #zope3-dev | 00:04 | |
| *** BjornT has joined #zope3-dev | 00:07 | |
| *** srichter has quit IRC | 00:18 | |
| *** admp has quit IRC | 00:18 | |
| *** mwh2 has quit IRC | 00:42 | |
| *** niemeyer has quit IRC | 00:49 | |
| *** _projekt01 has quit IRC | 01:02 | |
| *** hazmat has quit IRC | 01:33 | |
| *** bskahan has quit IRC | 01:33 | |
| *** hazmat has joined #zope3-dev | 01:33 | |
| *** d2m has quit IRC | 02:08 | |
| *** srichter has joined #zope3-dev | 02:25 | |
| *** ChanServ sets mode: +o srichter | 02:25 | |
| *** mexiKON has joined #zope3-dev | 02:57 | |
| *** philiKON has quit IRC | 03:04 | |
| *** hazmat has quit IRC | 04:24 | |
| *** RaFromBRC has joined #zope3-dev | 04:25 | |
| *** hazmat has joined #zope3-dev | 04:38 | |
| *** stub has joined #zope3-dev | 04:41 | |
| *** hazmat has quit IRC | 05:03 | |
| *** hazmat has joined #zope3-dev | 05:05 | |
| *** netkrom has joined #zope3-dev | 05:20 | |
| *** hazmat has quit IRC | 05:27 | |
| *** MiUlEr has joined #zope3-dev | 07:31 | |
| *** mexiKON has quit IRC | 07:35 | |
| *** viyyer has joined #zope3-dev | 08:10 | |
| *** mooded has joined #zope3-dev | 09:04 | |
| *** Aiste has quit IRC | 09:10 | |
| *** mooded has quit IRC | 09:18 | |
| *** Aiste has joined #zope3-dev | 09:19 | |
| *** hdima has joined #zope3-dev | 09:21 | |
| *** d2m has joined #zope3-dev | 10:03 | |
| *** sashav has joined #zope3-dev | 10:05 | |
| *** vlado|away is now known as vlado | 10:08 | |
| *** Theuni has joined #zope3-dev | 10:19 | |
| *** sashav has quit IRC | 10:34 | |
| *** alga has joined #zope3-dev | 11:09 | |
| *** sashav has joined #zope3-dev | 11:10 | |
| *** lunatik has joined #zope3-dev | 11:12 | |
| *** VladDrac has quit IRC | 11:12 | |
| *** lunatik has quit IRC | 11:15 | |
| *** lunatik has joined #zope3-dev | 11:16 | |
| *** mooded has joined #zope3-dev | 11:51 | |
| *** tarek_ has joined #zope3-dev | 12:16 | |
| *** __gotcha_ has joined #zope3-dev | 12:24 | |
| tarek_ | hello, does a "regular" critical section is enough to serialize a bit of code on any Zope installation (ZEO, non-ZEO, etc..) ? | 12:31 |
|---|---|---|
| tarek_ | or soemthing else has to be done | 12:31 |
| SteveA | well... | 12:33 |
| SteveA | what resource are you concerned will be in contention? | 12:33 |
| *** efge has joined #zope3-dev | 12:34 | |
| *** netkrom has quit IRC | 12:37 | |
| tarek_ | contention ? | 12:38 |
| tarek_ | you mean inside the critical section ? | 12:39 |
| tarek_ | some code that I do not want several threads to run at the same time | 12:40 |
| *** regebro has joined #zope3-dev | 12:40 | |
| SteveA | right | 12:52 |
| SteveA | so, what code is this? | 12:53 |
| SteveA | is this code that is a method on a persistent object? | 12:53 |
| SteveA | is the code something that accesses external files? | 12:53 |
| SteveA | in the first case, the persistent object is the "resource" that may be in contention | 12:53 |
| SteveA | in the second, then the external files are the "resource" that may be in contention | 12:53 |
| SteveA | so, if you're dealing with persistent objects, then you don't need to worry, as each thread gets its own copy of the persistent objects it is using, separate from other threads that use them | 12:59 |
| tarek_ | yes it's the first case indeed | 13:00 |
| tarek_ | ok ok | 13:00 |
| SteveA | if you're dealing with external files, and you don't have one set of external files per thread, then you need to have a critical section | 13:00 |
| SteveA | with a global lock | 13:00 |
| tarek_ | so i guess my critical section wouldbe enough | 13:00 |
| tarek_ | the use case is : | 13:00 |
| tarek_ | i call an external server to update persistent object | 13:00 |
| SteveA | if it's the first case, you don't need a critical section. just write all your code as if it is single-threaded | 13:00 |
| tarek_ | that call can't be done simultaenously by several threads | 13:01 |
| SteveA | you mean that you'd get a conflict error? | 13:01 |
| SteveA | if you want to avoid that, you can write conflict resolution code, that resolves the conflict caused by multiple concurrent writes. | 13:02 |
| tarek_ | yes, i mean that if one thread enters this function, the state of persistents objects will change, thus making other writes obsolete | 13:02 |
| SteveA | you're using transactions | 13:02 |
| SteveA | so, only one write will "win", unless you write conflict resolution code | 13:02 |
| tarek_ | ok | 13:03 |
| SteveA | that's special code that the zodb calls to resolve a conflict for your objects | 13:03 |
| SteveA | you'll need to look up the special method you need to write for that | 13:03 |
| tarek_ | is there a particular way to hook zodb on this ? (i guess it's time for me to look on this in zodb doc) | 13:03 |
| SteveA | you could also use something like the "QueuedCatalog" system to queue your changes, and then process them in a batch every so often | 13:04 |
| SteveA | collapsing a bunch of changes into a single one | 13:04 |
| tarek_ | oh yeah sounds good, | 13:04 |
| SteveA | you need to read the fine manual. the zodb docs on the wiki are good. | 13:04 |
| tarek_ | ok thanks SteveA, I've got all puzzle pieces now :) | 13:05 |
| SteveA | cool | 13:05 |
| *** apoirier has joined #zope3-dev | 13:20 | |
| *** mkerrin has joined #zope3-dev | 13:41 | |
| *** mgedmin has joined #zope3-dev | 13:52 | |
| *** lunatik has quit IRC | 13:57 | |
| *** lunatik has joined #zope3-dev | 13:58 | |
| *** bskahan has joined #zope3-dev | 14:13 | |
| *** srichter has quit IRC | 14:15 | |
| *** __gotcha__ has joined #zope3-dev | 14:16 | |
| *** __gotcha has quit IRC | 14:31 | |
| *** viyyer has quit IRC | 14:34 | |
| *** SteveA_ has joined #zope3-dev | 14:36 | |
| *** SteveA has quit IRC | 14:37 | |
| mgedmin | does zope 3 handle the http/1.1 Range: header? | 14:38 |
| * mgedmin guesses not | 14:38 | |
| SteveA_ | i think zserver does | 14:40 |
| mgedmin | grep says it doesn't | 14:41 |
| *** faassen has joined #zope3-dev | 14:44 | |
| SteveA_ | oh | 14:44 |
| SteveA_ | it did in zope2 at some point, i'm sure | 14:45 |
| *** lunatik has quit IRC | 14:45 | |
| *** lunatik has joined #zope3-dev | 14:46 | |
| *** ignas has joined #zope3-dev | 14:50 | |
| *** lunatik has quit IRC | 14:53 | |
| *** lunatik has joined #zope3-dev | 14:56 | |
| *** mooded has quit IRC | 15:00 | |
| *** philiKON has joined #zope3-dev | 15:06 | |
| *** srichter has joined #zope3-dev | 15:28 | |
| *** ChanServ sets mode: +o srichter | 15:29 | |
| *** niemeyer has joined #zope3-dev | 15:40 | |
| *** VladDrac has joined #zope3-dev | 15:51 | |
| *** [apoirier] has joined #zope3-dev | 16:00 | |
| *** bradb has quit IRC | 16:00 | |
| *** admp has joined #zope3-dev | 16:07 | |
| *** apoirier has quit IRC | 16:08 | |
| *** admp has quit IRC | 16:21 | |
| *** BjornT has quit IRC | 16:23 | |
| *** SteveA_ has quit IRC | 16:32 | |
| *** admp has joined #zope3-dev | 16:32 | |
| *** sashav has quit IRC | 16:34 | |
| *** MiUlEr has quit IRC | 16:35 | |
| *** bradb has joined #zope3-dev | 16:38 | |
| *** BjornT has joined #zope3-dev | 16:51 | |
| *** SteveA_ has joined #zope3-dev | 16:53 | |
| *** admp has quit IRC | 16:58 | |
| *** hdima has quit IRC | 17:00 | |
| *** BjornT has quit IRC | 17:05 | |
| *** SteveA_ has quit IRC | 17:08 | |
| *** BjornT has joined #zope3-dev | 17:09 | |
| *** SteveA_ has joined #zope3-dev | 17:10 | |
| *** mohsen-away is now known as mohsen | 17:15 | |
| *** SteveA_ has quit IRC | 17:20 | |
| *** admp has joined #zope3-dev | 17:25 | |
| *** admp has quit IRC | 17:53 | |
| *** admp has joined #zope3-dev | 18:19 | |
| *** vlado is now known as vlado|away | 18:33 | |
| *** regebro has quit IRC | 18:43 | |
| *** __gotcha_ has quit IRC | 18:58 | |
| *** xena has quit IRC | 19:02 | |
| *** mohsen is now known as mohsen-away | 19:09 | |
| *** bradb is now known as bradb|out | 19:15 | |
| *** ignas_ has joined #zope3-dev | 19:16 | |
| *** ignas has quit IRC | 19:16 | |
| *** hazmat has joined #zope3-dev | 19:19 | |
| *** [apoirier] has quit IRC | 19:21 | |
| *** tvon has joined #zope3-dev | 19:28 | |
| *** lunatik has left #zope3-dev | 19:34 | |
| *** lunatik has joined #zope3-dev | 19:37 | |
| *** deo has quit IRC | 19:59 | |
| *** tarek_ has quit IRC | 20:11 | |
| *** RaFromBRC has quit IRC | 20:43 | |
| srichter | oh wow, the XML Tree finally works in Konqui (KDE 3.4); yipee! | 20:45 |
| srichter | oh, no, nevermind | 20:46 |
| srichter | I was in the StaticTree skin :-) | 20:46 |
| *** tvon has quit IRC | 20:51 | |
| *** tvon has joined #zope3-dev | 20:52 | |
| *** palmTree has joined #zope3-dev | 20:54 | |
| *** projekt01 has joined #zope3-dev | 20:55 | |
| tvon | heh | 20:56 |
| *** efge has left #zope3-dev | 21:01 | |
| *** deo has joined #zope3-dev | 21:06 | |
| *** RaFromBRC has joined #zope3-dev | 21:09 | |
| srichter | philiKON: Are you there? | 21:10 |
| *** admp has joined #zope3-dev | 21:18 | |
| philiKON | srichter, yes | 21:23 |
| srichter | do you have some time? | 21:24 |
| philiKON | a little; why | 21:24 |
| srichter | I am writing a WSGI-compliant wrapper for ZServer's HTTP server | 21:24 |
| philiKON | cool | 21:24 |
| srichter | I am pretty much done, i.e. everything runs | 21:24 |
| srichter | except that security seems to be ignored | 21:25 |
| philiKON | zserver as in zope.server? | 21:25 |
| srichter | so I need someone to brainstorm with :-) | 21:25 |
| srichter | yes | 21:25 |
| philiKON | hmm | 21:25 |
| philiKON | i haven't looked a single bit at wsgi | 21:25 |
| srichter | I should check in my code in a branch first though | 21:25 |
| srichter | don't worry | 21:25 |
| srichter | it's not that important | 21:25 |
| srichter | it is just a standard that describes a certain flow of function calls | 21:26 |
| srichter | which masterfully separate server from application | 21:26 |
| srichter | there is very little new code and I do not think it has much to do with the problem | 21:26 |
| srichter | so first, how do I check in my local changes to a branch? | 21:27 |
| philiKON | first, you need to create a branch | 21:27 |
| philiKON | you do that best with a svn cp call that uses absolute URLs | 21:27 |
| srichter | ok | 21:28 |
| philiKON | svn cp svn+ssh://srichter@svn.zope.org/Zope3/trunk svn+ssh://srichter@svn.zope.org/Zope3/branches/srichter-wsgi-zserver | 21:28 |
| philiKON | something like that | 21:28 |
| srichter | done | 21:29 |
| srichter | next? :-) | 21:29 |
| srichter | I need to switch roots, right? | 21:29 |
| philiKON | right | 21:29 |
| philiKON | now in your working copy | 21:29 |
| philiKON | you do svn switch to the new location (branch) | 21:30 |
| * philiKON just read zope.app.wsgi/README.txt, interfaces and __init__ | 21:31 | |
| philiKON | seems plausible | 21:31 |
| philiKON | so, zope.app is wsgi-ready | 21:31 |
| srichter | yes | 21:31 |
| philiKON | and you want to make zope.server wsgi-ready, too? | 21:31 |
| philiKON | so, basically, coming from the other end, eh? | 21:31 |
| srichter | yes, and its done :-) | 21:31 |
| srichter | right | 21:31 |
| philiKON | so that in the future, zope.server and zope.app are glued thru wsgi? | 21:31 |
| srichter | ok, checked in | 21:32 |
| *** alga has quit IRC | 21:32 | |
| srichter | well, zope.server is actually going away this weekend | 21:32 |
| philiKON | twisted? | 21:32 |
| srichter | I am meeting with Itamar and maybe foom to integrate Twisted | 21:33 |
| philiKON | but on a branch, right? | 21:33 |
| srichter | yes | 21:33 |
| philiKON | ("YAY!" btw) :)) | 21:33 |
| srichter | but I would like to get it into 3.1; I am tired of supporting the zserver with its async issues | 21:33 |
| srichter | anyway, the code is in the branch | 21:34 |
| philiKON | well, mmh, i would like to see 3.1 out there | 21:34 |
| srichter | me too! | 21:34 |
| srichter | but if noone fixes the critical bugs, it's not going to happen | 21:34 |
| srichter | soon | 21:34 |
| FarcePest | so twisted is now using zope.interface, and zope will be using a twisted component? | 21:34 |
| philiKON | yup | 21:35 |
| srichter | FarcePest: we will use twisted.web2 and twisted.protocol | 21:35 |
| philiKON | twisted.web2 is wsgi compliant? | 21:35 |
| srichter | twisted.web2 is WSGI-compliant | 21:35 |
| srichter | yes | 21:35 |
| philiKON | very cool | 21:36 |
| srichter | it should be very very easy to integrate, once we change to their async main loop | 21:36 |
| FarcePest | zope and twisted sleeping together, MASS HYSTERIA | 21:36 |
| philiKON | lol | 21:36 |
| srichter | philiKON: ok, so if you change the servertype from HTTP to WSIG-HTTP, the new server component will be used | 21:36 |
| philiKON | srichter, we'll see how it works out | 21:37 |
| * philiKON still readiing code | 21:37 | |
| srichter | you will notice that you can do anything on the site without needing to login as a user | 21:37 |
| srichter | the simplest test is to: telnet localhost 8080 | 21:37 |
| srichter | GET /manage | 21:37 |
| * philiKON suspects it's a problem with WSGIPublisherApplication, not the server type | 21:38 | |
| srichter | the old server throws you "Unauthorized" while the new server just gives you a redirect | 21:38 |
| srichter | yep, but if you follow the steps carefully, you will notice that we do exactely the same thing as the other publisher server | 21:38 |
| srichter | maybe the problem is with the custom IHeaderOutput object we are using, but I would not know why | 21:39 |
| philiKON | man, you caught me in the middle of hacking a Chirikov chaotic system in python :) | 21:39 |
| * philiKON still checking out | 21:40 | |
| srichter | ok | 21:40 |
| philiKON | srichter, ever used scipy or Numeric? | 21:40 |
| srichter | I used Numeric once, but just to try it out | 21:41 |
| srichter | I have not looked at the optimizations too much | 21:41 |
| philiKON | i'm taking an advanced class now (that i'm actually not supposed to take) that gives an introduction to computational physics | 21:41 |
| philiKON | they use python here in dresden | 21:42 |
| philiKON | quite cool | 21:42 |
| srichter | yeah, great | 21:42 |
| * philiKON now compiling | 21:42 | |
| philiKON | btw, i just bought the commemorative edition of the feynman lectures | 21:43 |
| philiKON | very nice set of books | 21:43 |
| srichter | yes | 21:43 |
| srichter | I got them too | 21:43 |
| philiKON | srichter, maybe this is the problem: | 21:46 |
| philiKON | def setAuthUserName(self, name): | 21:47 |
| philiKON | """See zope.publisher.interfaces.http.IHeaderOutput""" | 21:47 |
| philiKON | pass | 21:47 |
| srichter | but in the IHeaderOutput it says that this is for logging purposes only | 21:47 |
| philiKON | ok | 21:47 |
| srichter | which concures with what I know; the user is set in the interaction.participations[0].principal | 21:48 |
| srichter | where a participation is simply a Request instance | 21:48 |
| philiKON | right | 21:49 |
| philiKON | but still | 21:49 |
| philiKON | if authentication wouldn't work, you'd never be able to authenticate | 21:49 |
| philiKON | we have a different problem | 21:49 |
| philiKON | we don't even need to authenticate | 21:50 |
| srichter | right | 21:50 |
| srichter | I am so lost | 21:51 |
| srichter | I know that | 21:51 |
| philiKON | i get: | 21:51 |
| srichter | - the application object (root folder) is proxied | 21:51 |
| philiKON | ZopeXMLConfigurationError: File "/Users/philipp/dev/Zope3-wsgi/src/zope/app/server/configure.zcml", line 19.2-23.8 | 21:51 |
| philiKON | ConfigurationError: ('Invalid value for', 'component', "Couldn't import zope.app.server.wsgi, No module named wsgihttpserver") | 21:51 |
| srichter | oops, forgot to check in a file | 21:52 |
| srichter | checked in | 21:52 |
| srichter | I know that | 21:52 |
| srichter | - the application object (root folder) is proxied | 21:52 |
| srichter | - the interaction is correctly set | 21:53 |
| srichter | - request.principal == UnauthenticatedPrincipal | 21:53 |
| *** tvon has quit IRC | 21:53 | |
| srichter | - the checkers are setup correctly, because doing "object.setitem" on the application obejct from a debugger prompt gives me a Forbidden error | 21:54 |
| srichter | <over> | 21:54 |
| *** palmTree has quit IRC | 21:56 | |
| *** tvon has joined #zope3-dev | 21:57 | |
| philiKON | srichter, i'm lost too | 21:57 |
| srichter | I just do not understand why the security is not kicking in | 21:58 |
| *** RaFromBRC is now known as RaFromBRC|afk | 21:59 | |
| srichter | "manage" is registered for all objects, right? | 21:59 |
| philiKON | right | 21:59 |
| philiKON | are views that are looked up security proxied? | 21:59 |
| srichter | maybe pdb'ing somewhere in the traversal process will yield some clues | 21:59 |
| srichter | they should be | 22:00 |
| philiKON | srichter, dude, i'm asked to log in | 22:04 |
| srichter | noooooooo! | 22:04 |
| srichter | did you ever reproduce my problem before? | 22:04 |
| philiKON | no, i was still mangled with some other things and only now managed to start z3 up and log in | 22:04 |
| *** mgedmin has quit IRC | 22:05 | |
| srichter | did you change zope.conf? | 22:05 |
| srichter | to use WSGI-HTTP instead of HTTP? | 22:05 |
| philiKON | yes | 22:05 |
| philiKON | 2005-04-12T20:59:45 INFO WSGIHTTPServer zope.server.http (WSGI-HTTP) started. | 22:05 |
| philiKON | Hostname: bender.local | 22:05 |
| philiKON | Port: 8080 | 22:05 |
| srichter | ahhhhh! | 22:06 |
| philiKON | 127.0.0.1 - anonymous [12/Apr/2005:21:04:50 +0200] "GET /@@contents.html HTTP/1.1" 401 4333 "" "Mozilla/5.0 (Macintosh; U; PPC Mac OS X Mach-O; en-US; rv:1.7.6) Gecko/20050317 Firefox/1.0.2" | 22:06 |
| philiKON | see the 401 (Unauthorized) | 22:06 |
| srichter | aaaaarrrrrrrggggggggggghhhhhhhhhhhhh | 22:06 |
| srichter | can you try: | 22:06 |
| srichter | telnet localhost 8080 | 22:06 |
| srichter | GET /manage | 22:06 |
| philiKON | Connected to localhost. | 22:07 |
| philiKON | Escape character is '^]'. | 22:07 |
| philiKON | GET /manage HTTP/1.1 | 22:07 |
| philiKON | HTTP/1.1 401 Unauthorized | 22:07 |
| srichter | argh | 22:07 |
| srichter | here is what I get from the traverser looking up manage: | 22:08 |
| srichter | > /opt/zope/Zope3/Zope3-Fresh/src/zope/app/container/traversal.py(72)publishTraverse() | 22:08 |
| srichter | -> import pdb; pdb.set_trace() | 22:08 |
| srichter | (Pdb) self.context | 22:08 |
| srichter | <zope.app.folder.folder.Folder object at 0x41057aec> | 22:08 |
| srichter | (Pdb) type(self.context) | 22:08 |
| srichter | <type 'zope.security._proxy._Proxy'> | 22:08 |
| srichter | (Pdb) request | 22:08 |
| srichter | <zope.publisher.browser.BrowserRequest instance URL=http://localhost:8080/manage> | 22:08 |
| srichter | (Pdb) request.principal | 22:08 |
| srichter | <zope.app.security.principalregistry.UnauthenticatedPrincipal object at 0x419e388c> | 22:08 |
| srichter | (Pdb) zapi.queryMultiAdapter((self.context, request), name='manage') | 22:08 |
| srichter | <zope.app.publisher.browser.viewmeta.ManagementViewSelector object at 0x44f2f74c> | 22:08 |
| srichter | (Pdb) | 22:08 |
| srichter | this is strange | 22:09 |
| philiKON | i guess so | 22:10 |
| srichter | can you try to put a pdb call somewhere in the WSGI code just to make sure that this code path is used? | 22:10 |
| philiKON | ok | 22:10 |
| philiKON | srichter, i get the pdb trace | 22:13 |
| srichter | where did you place it? | 22:13 |
| philiKON | WSGIPublisherApplication.__call__, just before the clal to publish() | 22:14 |
| srichter | ok, that's perfect | 22:14 |
| srichter | ok, so it must be me | 22:14 |
| srichter | I am going to build Zope 3 from SVN and from scratch | 22:14 |
| srichter | and see what happens | 22:14 |
| philiKON | do that | 22:15 |
| philiKON | and toss your zodb | 22:15 |
| srichter | yes | 22:15 |
| srichter | OT: does pyxml ship with vanilla Python? | 22:16 |
| srichter | because it's suppose to have an XPath parser | 22:16 |
| philiKON | really?!? | 22:17 |
| srichter | are using Python 2.4? | 22:17 |
| srichter | that's what Itamar said today | 22:17 |
| * philiKON is using python 2.3 | 22:18 | |
| srichter | ok, me too | 22:18 |
| srichter | with XPath in place, we could implement the disable tag | 22:18 |
| srichter | that would be very powerful | 22:18 |
| philiKON | yes | 22:18 |
| srichter | ok, with a fresh checkout I get the login request as well | 22:20 |
| srichter | I am relieved now :-) | 22:20 |
| philiKON | :) | 22:20 |
| philiKON | maybe you check svn stat again on your not-working sandbox | 22:21 |
| srichter | first I am tossing my ZODB | 22:21 |
| srichter | wow, so if I get this done today, I could start working on Twisted, maybe | 22:22 |
| srichter | he he, deleting the Data.fs did not help | 22:23 |
| philiKON | wsig will make it easier for me to integrate modzope | 22:23 |
| srichter | I hope so | 22:24 |
| srichter | the code in zope.app.wsgi is very simple | 22:24 |
| *** mohsen-away has quit IRC | 22:25 | |
| srichter | if I switch my checkout to the trunk again, I can just make the checkin, right? | 22:25 |
| srichter | ok, even after rebuilding the old instance does not want to allow me to log in | 22:26 |
| srichter | oh well, I'll nix it :-) | 22:26 |
| philiKON | well, it's a regular branch now | 22:27 |
| philiKON | you need to merge it to the trunk | 22:27 |
| srichter | how do I merge again? :-) | 22:27 |
| srichter | svn merge branch trunk? | 22:27 |
| philiKON | phone | 22:27 |
| srichter | ok | 22:27 |
| philiKON | yup, svn merge -r branch_start:HEAD branch_url . | 22:28 |
| srichter | I just noticed a flaw | 22:30 |
| srichter | in my new checkout I had not changed to the WSGI version of the HTTP server | 22:31 |
| srichter | :-( | 22:31 |
| *** faassen has quit IRC | 22:31 | |
| srichter | but it still works :-) | 22:31 |
| * FarcePest thinks he has found some bad BTreeContainer behavior; can you review something before I put in a bug? | 22:32 | |
| srichter | FarcePest: shoot | 22:33 |
| FarcePest | # c is a BTreeContainer | 22:33 |
| FarcePest | # this fails to iterate over all contained objects | 22:33 |
| FarcePest | for k in c.keys(): # c.keys() is a OOBTreeItems | 22:33 |
| FarcePest | if condition(k): | 22:33 |
| FarcePest | del c[k] | 22:33 |
| FarcePest | # this is at odds with standard dict behavior | 22:33 |
| FarcePest | # workaround: | 22:33 |
| FarcePest | for k in list(c.keys()): | 22:33 |
| FarcePest | if condition(k): | 22:33 |
| FarcePest | del c[k] | 22:33 |
| srichter | what error do you get? | 22:34 |
| srichter | this should work | 22:34 |
| FarcePest | it should work, yes, but some of the objects in c are skipped | 22:34 |
| FarcePest | (the second case works fine) | 22:34 |
| srichter | shrug | 22:34 |
| srichter | send a message to Tim | 22:34 |
| srichter | this is really important, indeed | 22:35 |
| FarcePest | mail, or put in a bug? | 22:35 |
| srichter | I think a mail directly to Tim is best, though a report would be nice too | 22:36 |
| *** admp has joined #zope3-dev | 22:43 | |
| *** hazmat has quit IRC | 22:43 | |
| *** hazmat has joined #zope3-dev | 22:43 | |
| *** Aiste has quit IRC | 22:47 | |
| *** bradb|out is now known as bradb | 22:52 | |
| *** mkerrin has quit IRC | 22:57 | |
| *** ignas_ has quit IRC | 22:57 | |
| *** tarek_ has joined #zope3-dev | 23:00 | |
| *** Damascene has quit IRC | 23:08 | |
| srichter | philiKON: okay, it's all checked in now | 23:34 |
| srichter | philiKON: do you know how to setup SVN externals? | 23:35 |
| *** RaFromBRC|afk is now known as RaFromBRC | 23:35 | |
| *** Damascene has joined #zope3-dev | 23:39 | |
| *** admp has joined #zope3-dev | 23:41 | |
| *** tvon has quit IRC | 23:44 | |
| *** tvon|x31 has joined #zope3-dev | 23:44 | |
| *** tvon|x31 has quit IRC | 23:48 | |
| *** alga has joined #zope3-dev | 23:53 | |
| philiKON | srichter, yes | 23:56 |
| srichter | philiKON: ok, I figured it out | 23:57 |
| srichter | pretty easy actually | 23:57 |
| philiKON | philipp@bender:~/dev/Zope3/src$ svn propget svn:externals . | 23:57 |
| philiKON | BTrees svn://svn.zope.org/repos/main/ZODB/tags/3.4.0a2/src/BTrees | 23:57 |
| philiKON | persistent svn://svn.zope.org/repos/main/ZODB/tags/3.4.0a2/src/persistent | 23:57 |
| philiKON | ThreadedAsync svn://svn.zope.org/repos/main/ZODB/tags/3.4.0a2/src/ThreadedAsy | 23:57 |
| philiKON | ... | 23:57 |
| philiKON | yup | 23:57 |
| *** bskahan has quit IRC | 23:58 | |
| FarcePest | srichter, tim peters says "not a bug" <shrug> | 23:59 |
| srichter | LOL | 23:59 |
| FarcePest | (not an exact quote) | 23:59 |
Generated by irclog2html.py 2.15.1 by Marius Gedminas - find it at mg.pov.lt!