IRC log of #zope3-dev for Friday, 2005-06-24

« Thursday, 2005-06-23 Index Saturday, 2005-06-25 »
*** FarcePest has quit IRC00:02
*** hazmat has joined #zope3-dev00:17
*** bskahan is now known as bska|dinner00:18
*** hazmat has quit IRC00:21
*** projekt01 has joined #zope3-dev00:41
*** bska|dinner has quit IRC00:44
*** SureshZ has left #zope3-dev01:06
*** bskahan has joined #zope3-dev01:33
projekt01Does somebody know a method for testing if a there is a missing tearDown somewhere?01:34
philiKONprojekt01, you can use the test runner (test.py) to run a certain test repeatedly (see --help for info)01:35
philiKONusually, if you're missing a tearDown, a repeated execution of a test fails01:35
philiKONof course, there's no guarantee01:35
projekt01One test fails if I run only this test, because the test uses some settings from another test setUp()01:36
projekt01philiKON, Ok, I try this, thanks01:37
*** SureshZ has joined #zope3-dev02:18
*** tvon has joined #zope3-dev02:22
*** bskahan has quit IRC02:29
*** tvon has quit IRC02:34
*** yota has quit IRC02:43
*** tvon has joined #zope3-dev03:12
*** projekt01 has quit IRC03:22
*** kaczordek has joined #zope3-dev04:18
*** bskahan has joined #zope3-dev04:52
*** jbb666 has quit IRC05:04
*** jbb666 has joined #zope3-dev05:08
*** RaFromBRC has quit IRC05:19
*** bskahan has quit IRC05:44
*** stub has joined #zope3-dev05:54
*** SureshZ has left #zope3-dev08:33
*** projekt01 has joined #zope3-dev09:44
*** tvon has quit IRC09:58
*** yota has joined #zope3-dev10:00
*** guido_g has joined #zope3-dev10:39
*** kaczordek has quit IRC10:46
*** d2m_ has quit IRC10:52
*** d2m has quit IRC10:52
*** d2m_ has joined #zope3-dev10:56
*** d2m has joined #zope3-dev11:02
*** lunatik has joined #zope3-dev11:12
*** tarek has joined #zope3-dev11:29
*** philiKON has quit IRC11:52
*** stub has quit IRC11:56
*** anguenot has joined #zope3-dev12:08
*** guido_g has quit IRC12:24
*** lunatik has left #zope3-dev12:38
*** hdima has joined #zope3-dev12:54
*** bcsaller has quit IRC13:07
*** efge has joined #zope3-dev13:08
*** J1m has joined #zope3-dev13:15
*** bskahan has joined #zope3-dev13:23
*** andrewX_ has joined #zope3-dev13:27
*** andrew_m has quit IRC13:28
*** andrewX_ is now known as andrew_m13:29
andrew_mi'm trying to access the underlying content object from a custom widget - is that possible or generally a bad idea?13:30
J1mit is possible and usually a bad idea. :)13:31
J1mself.context.context13:32
andrew_moh, nice, thanks13:32
efgethere's this comment in zope.interface.interface.__call__: An embedded function is used to allow an optional argument to __call__ without resorting to a global marker13:33
efgewhat's wrong with resorting to a global marker ? speed ?13:33
J1mI don't like global markers because they seem to be misinterpreted.13:34
J1mI've seen bugs where people treated them as part of a module's interface, leading to subtle bugs.13:34
efgecan't you just del them afterwards ?13:34
J1mHading them in a nested function makes them truly private.13:35
efgethe contorsions of defining a function in a function makes things not very readable13:35
J1mNo, you can't delete them because the function using them needs them at run time.13:35
efgehm right13:35
efgea decorator would solve this :)13:35
efgeprobably13:36
J1mI *thought* this was also a speed advantage, but it turns out that the global lookup is faster, much to my surprise.13:36
J1mIn general, I don't like the use of an extra argument to change semantics of a call.13:37
J1mThis is why Zope 3 generaly has get and query calls.13:37
J1mI made an exception in this case because I find calling nterfaces so comelling.13:37
J1mI made an exception in this case because I find calling nterfaces so compelling.13:37
J1mYes, a decorator could probably solve this.13:38
J1mRewriting this in C would solve it too.13:38
J1mWhich I do plan to do at some point.13:38
*** bskahan has quit IRC13:50
efgesrichter: you've made the adapter_hook hookable, (in zope/component/__init__.py), is there any current use of this ?13:53
J1myes13:55
J1mactually... not sure13:58
efgeI'm not finding any, but I think it's exactly the mechanism I want, rather than adding another one in the adapter_hooks list14:00
*** stub has joined #zope3-dev14:02
J1mIt is used14:02
J1min zope.app.component.hooks14:03
J1mat the end14:03
J1msee setHooks14:03
*** projekt01 has quit IRC14:06
efgeHm and you cannot hook something that's already hooked, no stacking...14:07
*** regebro has joined #zope3-dev14:08
J1msure, you can get the existing hook and regsiter a new hook that calls the original14:09
efgethx I'll try that14:10
*** J1m has quit IRC14:12
*** philiKON has joined #zope3-dev14:25
*** jhauser has joined #zope3-dev14:29
*** AJC has joined #zope3-dev14:36
srichterefge: right, zope.app.component uses those hooks for its local lookup14:37
efgesrichter: I finally saw that14:37
efgeok, my sethook on top of the existing sethook works, but I'm concerned that zope.app.component.hooks.resetHooks just does a reset() so assumes it's the only one having hooked the thing14:47
*** efge has quit IRC15:04
*** mooded has joined #zope3-dev15:15
srichterI think you can get one back at a time, iirc15:18
srichterif not, we should add this to the hook API15:18
*** bskahan has joined #zope3-dev15:19
*** mkerrin has joined #zope3-dev15:21
*** regebro has quit IRC15:41
*** lunatik has joined #zope3-dev15:49
*** philiKON has quit IRC15:55
*** stub has quit IRC16:07
*** efge has joined #zope3-dev16:13
*** lunatik has left #zope3-dev16:15
*** ignas_ has joined #zope3-dev16:20
*** lunati1 has joined #zope3-dev16:26
*** lunati1 has left #zope3-dev16:28
*** __gotcha has joined #zope3-dev16:40
*** hdima has quit IRC16:45
*** J1m has joined #zope3-dev16:55
*** SureshZ has joined #zope3-dev16:57
efgeJ1m: FWIW I'm abandoning the ideas of proxy-based verisioning for now, too much hard-to-control magic.17:35
efgeI'll be basing my devs on zope.app.versioncontrol, with some changes17:35
J1mI'm glad you are passing on the magic.17:36
J1mAt some point, I'd like to discuss this issue with you in more detail.17:36
efgeSure17:36
J1mI'll note that I'd prefer a simpler model for version control.17:37
J1mWhere versions are just object copies/17:37
efgeRe proxies: I still feel it's possible, but I don't want to tackle it now17:37
efgeWhich z.a.versioncontrol is, right ?17:37
J1mThis is one of the things I percieve and like about CPS's repository model.17:37
J1mNo, version control involves mummification. :)17:38
J1mNo, z.a.versioncontrol involves mummification. :)17:38
J1mWhen you create a version, it is an internal resource in the repository.17:38
efgeI see17:38
J1mI'd prefer that versions be live objects.17:39
efgeIs it implied by the z.a.versioncontrol interfaces, or just the current implementation ?17:39
J1mThat they can be used just like other objects with the exception that they have some state that is fixed and can't change.17:39
J1mis mummification implied?17:40
efgeYes17:40
J1myes17:40
J1mNote that my vision for a simpler version model is incompatible with some use cases.17:40
J1mNamely, use of an external version repository.17:40
J1mOTOH, these use cases have never been realized afaict.17:41
J1mI'm not an expert on content repository, but my feeling is that content repositories should be more like homeless shelters than like inode tables.17:42
J1mCPS's repository model and some of our past models treat respositories like inode tables.17:43
J1mThey then need objects like hard links to tie objects into content space.17:43
*** Aiste has joined #zope3-dev17:44
J1mI'd prefer a simpler modules where repositories are just public folders that contain objects (e.g. old versions) that have nowhere else to live.17:44
J1mI'd prefer a simpler model where repositories are just public folders that contain objects (e.g. old versions) that have nowhere else to live.17:44
VladDracanyone in goteborg already?17:46
efgeVladDrac: lots17:46
efgeJ1m: I agree with that, actually I want the links between the repository content and the workspace checkouts to be stored as relations inside a relation tool17:47
J1mWe *are* using z.a.versioncontrol for our current z3 project, so I know it works. :)17:47
efgeInstead of storing histories & such, and of annotating the checkout, put this as relations in the realtion tool17:47
efgeThat'll be my sprinting these few days :)17:48
*** andrew_m has left #zope3-dev17:52
*** mooded has quit IRC17:58
*** AJC has quit IRC18:06
*** bskahan has quit IRC18:30
*** ciphergoth has joined #zope3-dev18:36
ciphergothI'm using a Zope3 application called "SchoolTool" and I've tried to use Apache as a reverse proxy to expose it to the world18:36
ciphergothbut it seems to use absolute URLs to refer to its own resources, and those URLs don't work elsewhere...18:37
ciphergothis there a way to make Zope3 applications use relative URLs instead (ie leave off the host and protocol)?18:37
VladDracthere is, check srichters book on apache / virtual hosting18:38
VladDracit requires inserting appropriate ++vh strings18:38
ciphergothah18:40
ciphergothso I redirect http://my.long.url/this/that to http://localhost:port/++vh/this/that18:40
VladDracI said "appropriate" :)18:41
VladDrachere's an example, you need to consult the documentation to get it right18:41
VladDrachttp://viewcvs.amazesolutions.com/zope3/z3alot/trunk/doc/README.txt?rev=133&view=markup18:41
ciphergothie RewriteRule ^/(.*) http://localhost:7180/+vh/++$1 [P,L]18:41
ciphergothoh, I have to specify what absolute thing to use instead,  OK.18:42
VladDracmore like http://localhost:7180/++vh++http:yourdomain:80/++$i18:42
VladDracoffline now, good luck :)18:42
*** regebro has joined #zope3-dev18:47
*** bradb-bbl is now known as bradb18:55
*** alga has joined #zope3-dev19:24
ciphergothgot it working - rah!19:31
*** regebro has quit IRC19:35
efgeCan I remove INonVersionable from zope.app.versioncontrol.interfaces ? It's used nowhere I can find.19:54
J1myup19:58
efgezope.app.versioncontrol.nonversioned.ObjectManagerNonVersionedDataAdapter seems to be dead code too, it even contains code with obj._delOb(name), kill it ?20:07
efgeOTOH I can imagine it updated to be useful. I'll see if I can fix it and use it.20:08
*** yota_ has joined #zope3-dev20:09
J1mkill it and fix it if and when you need it20:10
*** jbb666 has quit IRC20:10
J1mImagination is a dangerous tool, at least wrt requirements. :)20:10
efgeok :)20:12
*** yota has quit IRC20:22
*** efge has quit IRC20:25
*** tarek has quit IRC20:29
*** __gotcha has quit IRC20:45
*** tarek has joined #zope3-dev20:46
*** anguenot has left #zope3-dev20:52
*** mkerrin has quit IRC20:56
*** tarek has quit IRC21:00
*** ignas_ has quit IRC21:01
*** mkerrin has joined #zope3-dev21:19
*** bradb is now known as bradb-away21:26
*** Aiste has quit IRC21:34
*** jhauser_ has joined #zope3-dev21:54
*** jhauser has quit IRC22:02
*** projekt01 has joined #zope3-dev22:11
*** alga has quit IRC22:29
*** projekt01 has quit IRC23:05
*** mkerrin has quit IRC23:08
*** Jim7J1AJH has quit IRC23:22
*** Jim7J1AJH has joined #zope3-dev23:22
*** RaFromBRC has joined #zope3-dev23:25
*** Aiste has joined #zope3-dev23:30
*** Jim7J1AJ1 has joined #zope3-dev23:33
*** Jim7J1AJH has quit IRC23:33
srichterJ1m: can you give Corey, who just got his access today, access to the special projects repository, please?23:40
srichterthanks23:40
J1mplease remind me what his zope.org id is23:55
J1mnever mind23:55
J1mdone23:56
srichterthanks23:57
« Thursday, 2005-06-23 Index Saturday, 2005-06-25 »

Generated by irclog2html.py 2.15.1 by Marius Gedminas - find it at mg.pov.lt!