| C8N | yup | 00:00 |
|---|---|---|
| srichter | of course, besides showing the directly provided interfaces, it actually inspects the object and displays the actual attribute values and allows you even to explore those values, if they are locatable | 00:02 |
| lifeless | back | 00:04 |
| sm | I like it | 00:09 |
| sm | an object explorer | 00:09 |
| srichter | yes, we had this before and it was not maintained | 00:12 |
| srichter | the one I wrote depends on the apidoc API, which minimized the code needed | 00:12 |
| *** projekt01 has quit IRC | 00:17 | |
| lifeless | ./win 24 | 00:17 |
| *** lifeless has left #zope3-dev | 00:18 | |
| C8N | ? | 00:21 |
| *** douglasc has quit IRC | 00:21 | |
| *** niemeyer has quit IRC | 00:25 | |
| *** GaryPoster has quit IRC | 00:46 | |
| *** bskahan has joined #zope3-dev | 00:48 | |
| *** MrTopf has quit IRC | 00:55 | |
| *** alienoid has quit IRC | 00:56 | |
| *** alga has quit IRC | 00:57 | |
| *** gintas has quit IRC | 01:08 | |
| benji_york | buy all | 01:09 |
| *** benji_york has quit IRC | 01:10 | |
| *** bskahan has quit IRC | 01:12 | |
| philiKON | srichter, rc2 on friday, right? | 01:16 |
| srichter | that's what I am thinking right now | 01:16 |
| srichter | or will be monday better ? | 01:16 |
| *** douglasc has joined #zope3-dev | 01:16 | |
| philiKON | i don't care, as long as i have time to teest and check in my zeo script fix | 01:17 |
| philiKON | which i'm planning to do tomorrow | 01:17 |
| srichter | ok | 01:17 |
| *** pcardune has quit IRC | 01:17 | |
| philiKON | g'night then | 01:17 |
| srichter | bye | 01:18 |
| *** guido_g has quit IRC | 01:43 | |
| *** yota has quit IRC | 02:05 | |
| *** douglasc has quit IRC | 02:14 | |
| *** BjornT has quit IRC | 02:28 | |
| *** bradb has quit IRC | 02:29 | |
| *** SteveA has quit IRC | 02:30 | |
| *** niemeyer has joined #zope3-dev | 02:37 | |
| *** tvon has quit IRC | 03:18 | |
| *** tvon has joined #zope3-dev | 03:46 | |
| *** C8N has left #zope3-dev | 04:10 | |
| *** stub has quit IRC | 04:18 | |
| *** RaFromBRC has joined #zope3-dev | 04:33 | |
| *** roym` has joined #zope3-dev | 05:19 | |
| *** roym` has quit IRC | 05:19 | |
| *** RaFromBRC has quit IRC | 05:54 | |
| *** RaFromBRC has joined #zope3-dev | 06:22 | |
| *** tvon has quit IRC | 06:31 | |
| *** niemeyer has quit IRC | 07:44 | |
| *** mexiKON has joined #zope3-dev | 07:55 | |
| *** philiKON has quit IRC | 08:05 | |
| *** AJC has joined #zope3-dev | 08:55 | |
| *** zagy has quit IRC | 08:57 | |
| *** zagy has joined #zope3-dev | 09:01 | |
| *** sashav has quit IRC | 09:05 | |
| *** hdima has joined #zope3-dev | 09:11 | |
| AJC | what's the difference between skin_macros (template.pt) and dialog_macros.pt? in the message board example, they are identical. | 09:13 |
| *** timte has joined #zope3-dev | 09:19 | |
| *** douglasc has joined #zope3-dev | 09:52 | |
| *** zagy_ has joined #zope3-dev | 10:15 | |
| *** The|uni has joined #zope3-dev | 10:17 | |
| *** MrTopf has joined #zope3-dev | 10:22 | |
| *** yota has joined #zope3-dev | 10:27 | |
| *** zagy has quit IRC | 10:30 | |
| *** sashav has joined #zope3-dev | 10:39 | |
| *** RaFromBRC has quit IRC | 10:51 | |
| bob2 | hm, what repaced NotFoundError in the trunk? | 11:01 |
| bob2 | the deprecation warning doesn't say | 11:01 |
| *** douglasc has quit IRC | 11:03 | |
| wiggy | src/BTrees/BucketTemplate.c:905: warning: pointer targets in passing argument 1 of ‘PyString_FromStringAndSize’ differ in signedness | 11:13 |
| wiggy | is anyone working on those? | 11:13 |
| MrTopf | bob2: actually I don't know where NotFoundError was removed but IIRC someone wrote/said KeyError | 11:16 |
| bob2 | ah, thanks | 11:17 |
| bob2 | erk, I suspect my tests aren't complete enough to see if that's right | 11:17 |
| mexiKON | bob2, well, KeyError and other more specific errors | 11:17 |
| mexiKON | KeyError for a particular case of traversal | 11:17 |
| mexiKON | and FoobarNotFoundError for more specific cases | 11:18 |
| MrTopf | YouDidSomethingWrongError | 11:18 |
| wiggy | hm, unittest failure in svn trunk | 11:18 |
| bob2 | hm, this is for an authentication service | 11:18 |
| mexiKON | wiggy, ouch | 11:18 |
| bob2 | (need to convert it to a utility, too) | 11:18 |
| mexiKON | bob2, PrincipalNotFoundError | 11:18 |
| mexiKON | ? | 11:19 |
| bob2 | ah, thanks | 11:19 |
| bob2 | are the new ones in zope.exceptions still? | 11:19 |
| wiggy | http://paste.plone.org/3093 | 11:19 |
| mexiKON | bob2, don't think so | 11:19 |
| bob2 | and would it be feasible to fix pydoc to work with zope source? | 11:19 |
| mexiKON | sure, you could give it a try | 11:19 |
| mexiKON | or just use the code browser of apidoc | 11:19 |
| mexiKON | wiggy, yeah, the zdaemon stuff | 11:20 |
| bob2 | yeah, apidoc is more mousing, tho :) | 11:20 |
| mexiKON | wiggy, i never dare to touch that | 11:20 |
| wiggy | seems to be the only failure | 11:20 |
| bob2 | wiggy: is that with gcc 4.0 -Wall? | 11:21 |
| wiggy | gcc 4.0.2 | 11:21 |
| bob2 | hah | 11:21 |
| wiggy | not sure if it passes -Wall in | 11:21 |
| mexiKON | i don't think zdaemon has C code | 11:24 |
| bob2 | hrm, PrincipalNotFoundError isn't mentioned in the zope3 source at all | 11:24 |
| mexiKON | bob2, sorry, I meant PrincipalLookupError | 11:26 |
| mexiKON | in zope.app.security.interfaces | 11:26 |
| wiggy | mexiKON: no, but the BTree stuff gave compile errors | 11:26 |
| mexiKON | ah | 11:27 |
| mexiKON | oh | 11:27 |
| bob2 | ah, thank you | 11:27 |
| bob2 | I was just reading the source and notice zope.exceptions is nearly empty | 11:28 |
| *** mexiKON is now known as philiKON | 12:01 | |
| *** tarek has quit IRC | 12:10 | |
| *** faassen has joined #zope3-dev | 12:25 | |
| *** philiKON has quit IRC | 12:32 | |
| *** MJ has joined #zope3-dev | 13:17 | |
| *** MrTopf has quit IRC | 13:19 | |
| *** ignas has joined #zope3-dev | 13:21 | |
| *** tvon has joined #zope3-dev | 13:32 | |
| *** philiKON has joined #zope3-dev | 13:35 | |
| srichter | bob2: Adjusting pydoc to work with zope source would be a lot of workalso, it would never be as useful as apidoc | 13:37 |
| srichter | why do you say apidoc requires a lot of mousing? | 13:38 |
| *** philiKON has quit IRC | 14:02 | |
| *** BjornT has joined #zope3-dev | 14:08 | |
| *** gintas has joined #zope3-dev | 14:18 | |
| *** stub has joined #zope3-dev | 14:57 | |
| *** bskahan has joined #zope3-dev | 15:08 | |
| *** anguenot has joined #zope3-dev | 15:09 | |
| *** bradb has joined #zope3-dev | 15:16 | |
| *** mkerrin has joined #zope3-dev | 15:16 | |
| *** SteveA has joined #zope3-dev | 15:20 | |
| *** kakella has joined #zope3-dev | 15:28 | |
| *** roym has quit IRC | 15:29 | |
| *** guido_g has joined #zope3-dev | 15:30 | |
| *** ignas has quit IRC | 15:35 | |
| *** philiKON has joined #zope3-dev | 15:40 | |
| *** niemeyer has joined #zope3-dev | 15:47 | |
| *** MrTopf has joined #zope3-dev | 15:57 | |
| *** tav has quit IRC | 15:58 | |
| srichter | philiKON: Don't forget CHANGES.txt | 16:06 |
| philiKON | yes | 16:06 |
| philiKON | i'm not done yet :) | 16:06 |
| philiKON | srichter, tim is not on IRC, is he? | 16:06 |
| srichter | usually not | 16:07 |
| philiKON | gonna write him an email then | 16:07 |
| srichter | but he is very responsive to emails | 16:07 |
| *** The|uni is now known as theuni | 16:12 | |
| *** SureshZ has joined #zope3-dev | 16:17 | |
| *** lunatik has joined #zope3-dev | 16:34 | |
| *** lunatik has left #zope3-dev | 16:34 | |
| *** ignas has joined #zope3-dev | 16:35 | |
| *** AJC has quit IRC | 16:44 | |
| *** AJC has joined #zope3-dev | 16:44 | |
| *** sashav has quit IRC | 16:47 | |
| bob2 | srichter: using pydoc is a matter of one keystroke to switch to a terminal, then pasting the name of the module | 16:49 |
| bob2 | srichter: apidoc means switch to firefox, finding the right tab, pasting the module, converting . to / | 16:49 |
| bob2 | and means having a zope instance running | 16:50 |
| bob2 | not a huge deal, but overall less convenient | 16:50 |
| srichter | no, just search for it | 16:50 |
| srichter | also, a module does nto have much documentation in Zope | 16:50 |
| srichter | but in Zope trying to look at the module documentation is often totally useless | 16:51 |
| srichter | just have the menu with the codebrowser search menu always open | 16:51 |
| srichter | or write a small script that will start firefox and insert the right query | 16:51 |
| srichter | btw, the browsability of the documentation with its many inter-documentation links makes it so useful, otherwise you could just read the interfaces Py files and know the same | 16:52 |
| hdima | bob2: try 'python -m pydoc -p 8082' and then go to http://localhost:8082 | 16:54 |
| bob2 | hdima: I'm trying to avoid using a web brwoser :) | 16:54 |
| hdima | ah :) | 16:56 |
| srichter | bob2: you can always write a CLI for apidoc | 16:57 |
| srichter | other people like you might find this helpful | 16:57 |
| bob2 | yeah, that just occured to me | 16:57 |
| bob2 | I do agree the links are useful, and I do use apidoc | 16:57 |
| bob2 | but sometimes I just want to very quickly find the spelling for that class or whatever | 16:57 |
| srichter | zope.app.apidoc provides a solid API for Zope 3 code inspection | 16:58 |
| srichter | you could effectively have the same info as on the Web page | 16:58 |
| srichter | (especially when mixed with colors) | 16:58 |
| bob2 | ah, awesome | 16:58 |
| bob2 | I'll have a look at doing a cli viewer in the morning, thanks again :) | 16:59 |
| srichter | bob2: let me know if you need help | 16:59 |
| bob2 | will do, thanks | 17:00 |
| bob2 | ah, you're the maintainer for that module | 17:00 |
| srichter | yes, I wrote it | 17:00 |
| *** roym has joined #zope3-dev | 17:00 | |
| srichter | I have come to the conclusion that general-purpose API documentation tools must suck | 17:00 |
| bob2 | as in general input or general output? | 17:01 |
| srichter | both | 17:01 |
| srichter | the value of a documentation tool comes from handling very specific cases and provide info that is usually hard to reach | 17:02 |
| srichter | for example, apidoc will tell you the permission required to get/set any attribute in a class | 17:02 |
| srichter | that info would be impossible for pydoc to ever generate | 17:02 |
| srichter | unless they provide very generic plugins for this sort of thing | 17:03 |
| bob2 | ahh, right | 17:03 |
| *** benji_york has joined #zope3-dev | 17:08 | |
| *** d2m has quit IRC | 17:13 | |
| *** d2m has joined #zope3-dev | 17:24 | |
| benji_york | srichter, the author of mechanize respectfully declined the offer of testbrowser, he (understanably) said that he doesn't have enough time to maintain the code there now, so doesn't want to add more | 17:27 |
| srichter | ok | 17:27 |
| srichter | at least he is aware of it now | 17:28 |
| benji_york | there was some interest from other people, so I may put together a zope-free distribution | 17:28 |
| philiKON | maybe we should offer to maintain it as part of zope.testing...? | 17:28 |
| philiKON | all zope.* packages (except for zope.app) should more or less be usable outside of Zope-the-application-server (which is zope.app) | 17:28 |
| * wiggy zooms in on the zdaemon error | 17:28 | |
| wiggy | it looks like it always breaks on python 2.3 | 17:29 |
| wiggy | and always works on python 2.4 | 17:29 |
| benji_york | philiKon, now that you mention it, if you import the right thing, it already works outside of zope, I'll have to look at it a bit more | 17:30 |
| srichter | wiggy: I test with 2.4 only | 17:30 |
| wiggy | that explains nicely why you don't see it | 17:30 |
| benji_york | also, philiKON, the meaning of being a top-level zope.whatever package is changing, Jim wants to start using that namespace for new packages and slowly move things out of zope.app | 17:30 |
| philiKON | benji_york, basically, zope.* packages could depend on zope.interface or so, so any distribution would also have to include zope.interface. but other than that they shoudl really work outside of zope-the-app-server. | 17:30 |
| philiKON | aha | 17:31 |
| philiKON | makes sense for self-contained things | 17:31 |
| *** hdima has quit IRC | 17:31 | |
| philiKON | but the advantage of zope.app is that you can basically rely on a lot of functionality without worrying about dependencies | 17:31 |
| philiKON | some zope.* packages are already being distributed separately (e.g. in debian) | 17:32 |
| bob2 | ahh, I was wondering what the significance of things being in zope.app was | 17:32 |
| benji_york | true, that's where I see more reliance on ZPackage's dependency code, etc. | 17:32 |
| philiKON | bob2, basically zope.* packages are standard python packages you could use in any kind of project; zope.app is Zope-the-application-server | 17:33 |
| bob2 | that makessense | 17:33 |
| philiKON | at least that's the old-school interpretation | 17:34 |
| philiKON | jim might have crazy new ideas :) | 17:34 |
| benji_york | he does :) | 17:34 |
| philiKON | always has and always will :) | 17:34 |
| bob2 | good to know I can just drop z.interfaces in other projects | 17:34 |
| bob2 | until python just includes it in the standard lib, anyway ;) | 17:34 |
| philiKON | bob2, yup. twisted is doing that | 17:34 |
| benji_york | bob2, there is even a stand-alone distribution, somewhere on zope.org | 17:35 |
| bob2 | yeah, I heard they went with zope's instead of writing their own | 17:35 |
| bob2 | and it's awesome zope is moving to using twisted's core | 17:35 |
| *** sashav has joined #zope3-dev | 17:35 | |
| MrTopf | so will zi be in the standard lib sometime? or will it more likely be some PyProtocols stuff? | 17:35 |
| philiKON | wiggy, you're managing the zope 3 debian package, right? | 17:35 |
| MrTopf | hi, btw :-) | 17:35 |
| philiKON | hi MrTopf | 17:36 |
| philiKON | MrTopf, who knows | 17:36 |
| wiggy | philiKON: it's a team effort | 17:36 |
| benji_york | MrTopf, Guido made overtures about 6 months ago, but nothing since | 17:36 |
| wiggy | and I haven't done anything yet | 17:36 |
| MrTopf | looked only once at PyProtocols some very long time ago | 17:36 |
| wiggy | but otherwise, yes | 17:36 |
| philiKON | wiggy, who has done the package(s) that is(are) in so far? | 17:36 |
| MrTopf | benji_york: well, heard him talking about interfaces last year on EP | 17:36 |
| philiKON | the 3.0.91etc. stuff | 17:36 |
| * wiggy checks | 17:37 | |
| MrTopf | back then he wanted to see what's going to work out what way.. | 17:37 |
| philiKON | wiggy, ah found it | 17:37 |
| philiKON | wiggy, matthias klose | 17:37 |
| philiKON | it seems | 17:37 |
| wiggy | ah, yes | 17:37 |
| philiKON | is he on irc? | 17:37 |
| bob2 | wiggy: are they based on the one's doko did for ubuntu? | 17:37 |
| wiggy | yes, doko | 17:37 |
| bob2 | hah | 17:37 |
| wiggy | the idea is to have a single source of zope3 packages for both ubuntu and debian | 17:37 |
| philiKON | yes | 17:38 |
| philiKON | i think there's a bug in z3 that prevents it from running properly the way debian sets up zope | 17:38 |
| wiggy | oh? | 17:39 |
| *** kakella has quit IRC | 17:40 | |
| bob2 | so, zope 3.1 is the first release I've been around for...has the trunk historically been stable enough to develop code against, or do people only develop apps against stable releases unless they have a very good reason? | 17:40 |
| philiKON | wiggy, does /usr/lib/zope3/bin/mkzopeinstance work for you? | 17:41 |
| philiKON | bob2, the trunk should never be broken | 17:42 |
| benji_york | bob2, I've also only been around for 3.1, but we (ZC) have been developing against the head | 17:42 |
| philiKON | bob2, that's why we have a proposal- and test-driven development philosophy | 17:42 |
| bob2 | right, I've picked that up...I guess I'm wondering how well it works in practice :) | 17:43 |
| philiKON | of course, refactorings etc. are still allowed | 17:43 |
| philiKON | when you're developing against the trunk you can never be sure | 17:43 |
| philiKON | Before the X3.0 release we lots of geddons | 17:43 |
| philiKON | and even since then we had srichter's blowservices geddon | 17:43 |
| philiKON | only this time we have backward compat :) | 17:43 |
| srichter | well, sorta :-) | 17:44 |
| wiggy | philiKON: let me find a machine where I have that installed | 17:44 |
| philiKON | wiggy, essentially, the problem is that zope actually expects to live inside its own "world" (usually something like /usr/local/Zope-3.1.0/) | 17:44 |
| philiKON | wher eyou have /usr/local/Zope-3.1.0/lib/python | 17:44 |
| philiKON | and .../bin containing the scripts | 17:44 |
| philiKON | such as mkzopeinstance | 17:44 |
| wiggy | how lame, I forgot where the debs are | 17:44 |
| philiKON | now, these scripts actually rely on being in such a "world" | 17:45 |
| philiKON | they can't manage when being detached from their zope installation | 17:45 |
| philiKON | debian does this | 17:45 |
| philiKON | debian installs zope in site-packages | 17:45 |
| philiKON | and the scripts in /usr/lib/zope3/bin | 17:45 |
| philiKON | which I personally think is ok (Fred seems to disagree) | 17:45 |
| wiggy | it does fit the zope-can-be-used-without-zope-the-app-theme | 17:46 |
| philiKON | point is if debian wants to keep doing it this way, we need to adjust the scripts in bin | 17:46 |
| philiKON | exactly | 17:46 |
| wiggy | so I suspect we'll want to keep it that way unless there is a good reason not to | 17:47 |
| philiKON | that's what i think | 17:47 |
| philiKON | (on windows, btw, zope3 already gets installed into site-packages) | 17:52 |
| philiKON | srichter, in SETUP.cfg there's a line saying script bin/*\ | 17:52 |
| philiKON | err, without the \ | 17:52 |
| philiKON | srichter, so that would mean that everthing in Zope3/bin is also in an installations' bin, right? | 17:53 |
| philiKON | that doesn't really happen, though | 17:53 |
| srichter | mmh, no | 17:54 |
| srichter | actually, yes | 17:54 |
| srichter | I don't know much about cfg files | 17:54 |
| philiKON | e.g. Zope3/bin/debugzope is not in my Zope-3.1.0/bin/ directory | 17:55 |
| philiKON | even though i would expect it to be judging from that SETUP.cfg entry | 17:55 |
| srichter | mmh, no clue why | 17:56 |
| philiKON | because PACKAGE.cfg seems to define that in this case | 17:57 |
| philiKON | releases/Zope/PACKAGE.cfg that is | 17:57 |
| philiKON | weird | 17:57 |
| philiKON | in either case I'll know how to hook up mkzeoinstance | 17:57 |
| srichter | cool | 17:58 |
| srichter | I think PACKAGE.cfg is the right one to add it | 17:58 |
| srichter | that's where I add version.txt all the time | 17:58 |
| philiKON | yup | 18:06 |
| *** tav has joined #zope3-dev | 18:14 | |
| *** alga has joined #zope3-dev | 18:23 | |
| roym | Gentlemen, I'm writing a doctest (unit test); How would I make a | 18:33 |
| roym | vocabulary available to my test harness. I have a vocabulary | 18:33 |
| roym | implementation, and I use in ZCML. | 18:33 |
| philiKON | execute the zcml | 18:35 |
| benji_york | srichter, the HTML spec doesn't allow for a form's action to be empty, I wish it did :( | 18:41 |
| srichter | arg, what if I do not specify the action? | 18:41 |
| srichter | isn't that the same as action=""? | 18:42 |
| benji_york | action is required | 18:42 |
| srichter | this is really dumb | 18:42 |
| philiKON | make it . or context/@@absolute_url | 18:42 |
| benji_york | yep, I wanted to do the same thing about a month ago | 18:42 |
| srichter | no, I don't want . | 18:42 |
| benji_york | nope, "." isn't allowed either :( | 18:42 |
| philiKON | argh | 18:42 |
| srichter | this brings me to the default view of the context | 18:42 |
| philiKON | request/URL | 18:43 |
| srichter | I really want request.url or something | 18:43 |
| benji_york | yep | 18:43 |
| srichter | ok, I'll change that, sigh | 18:43 |
| * srichter starts to hate the HTML standard | 18:43 | |
| * benji_york already hates the HTML standards :) | 18:44 | |
| *** alga has quit IRC | 18:45 | |
| roym | philiKON: excuse my ignorance: how do I execute the ZCML in a unit test framework? | 18:45 |
| srichter | zope.configuration.xmlconfig.[file|string](...) | 18:46 |
| philiKON | use zope.configuration.xmlconfig.file() | 18:46 |
| srichter | :-) | 18:46 |
| philiKON | right, or .string() | 18:46 |
| roym | thanks. | 18:46 |
| philiKON | roym, execute zope.app.schema/meta.zcml first | 18:46 |
| philiKON | so that you have the <vocabulary /> directive | 18:46 |
| philiKON | roym, look at other tests that do it | 18:47 |
| roym | ok - will browse. | 18:47 |
| philiKON | roym, e.g. http://codespeak.net/svn/z3/Five/trunk/tests | 18:47 |
| philiKON | the spelling is a bit different | 18:47 |
| philiKON | but the concept is the same | 18:47 |
| *** MJ has quit IRC | 18:48 | |
| srichter | benji_york: does mechanize/testbrowser handle frames well? | 18:49 |
| benji_york | I have no idea :) | 18:50 |
| benji_york | at worst, you'd have to "open" the target url of a frame, instead of opening the url of the frame set | 18:51 |
| srichter | yeah | 18:51 |
| philiKON | srichter, now i know the difference between PACKAGE.cfg and SETUP.cfg :) | 18:57 |
| philiKON | PACKAGE.cfg defines what the .tgz will look like | 18:58 |
| philiKON | SETUP.cfg defines what to do when you're inside the .tgz and calling setup.py build etc. | 18:58 |
| srichter | oh, I see | 18:58 |
| philiKON | so, basically, when you have stuff that isn't part of a package but you want it to be installed somehow, you need to add it to the PACKAGE.cfg *and* the global SETUP.cfg | 18:59 |
| philiKON | if stuff is part of a package, you just add it to that package's SETUP.cfg (it will be part of the .tgz through its package already) | 18:59 |
| srichter | ok, I see | 19:01 |
| *** tvon has quit IRC | 19:12 | |
| *** tvon has joined #zope3-dev | 19:13 | |
| *** BjornT has quit IRC | 19:19 | |
| *** bskahan has quit IRC | 19:20 | |
| *** projekt01 has joined #zope3-dev | 19:20 | |
| *** BjornT has joined #zope3-dev | 19:21 | |
| *** zagy has joined #zope3-dev | 19:27 | |
| *** theuni has quit IRC | 19:34 | |
| *** theuni has joined #zope3-dev | 19:34 | |
| *** th1a has joined #zope3-dev | 19:36 | |
| th1a | Does anyone know the status of running Zope 3 as a Windows Service? | 19:37 |
| srichter | benji_york: projekt01: do you know an answer to Tom's question? | 19:38 |
| philiKON | th1a, there's zope.app.winservice, but it's supposedly not in the same good shape that its Zope 2.8 equivalent is | 19:39 |
| th1a | philiKON: Thanks, taking a look. | 19:41 |
| *** zagy_ has quit IRC | 19:43 | |
| benji_york | thla, we have a product that runs z3 as a service. It works fine as long as you don't try to log to STDOUT or STDERR :) | 19:54 |
| th1a | benji_york: Cool. We might have some questions for you about getting SchoolTool to run as a service. | 19:56 |
| benji_york | ok | 19:56 |
| *** bskahan has joined #zope3-dev | 20:10 | |
| *** AJC has quit IRC | 20:31 | |
| *** bska|mobile has joined #zope3-dev | 20:39 | |
| *** gintas has quit IRC | 20:40 | |
| *** bskahan has quit IRC | 20:41 | |
| *** th1a has quit IRC | 20:49 | |
| *** tav_ has joined #zope3-dev | 21:00 | |
| *** MrTopf has quit IRC | 21:00 | |
| *** tav has quit IRC | 21:01 | |
| *** faassen has quit IRC | 21:08 | |
| *** bska|mobile has quit IRC | 21:16 | |
| *** bska|mobile has joined #zope3-dev | 21:18 | |
| *** BjornT_ has joined #zope3-dev | 21:19 | |
| *** BjornT has quit IRC | 21:20 | |
| *** RaFromBRC has joined #zope3-dev | 21:20 | |
| *** stub has quit IRC | 21:20 | |
| *** stub has joined #zope3-dev | 21:20 | |
| *** douglasc has joined #zope3-dev | 21:30 | |
| srichter | does anyone here know how to include a picture in a Wiki page? | 21:31 |
| benji_york | not on zwiki | 21:33 |
| srichter | ok | 21:33 |
| srichter | sm: how can I add a picture in ZWiki? | 21:35 |
| sm | hi.. on z2 ? | 21:36 |
| srichter | yeah, zope.org | 21:36 |
| * sm shudders | 21:36 | |
| sm | if you have add documents, images, files permission in that folder it should show an upload field in the edit form | 21:37 |
| srichter | ok, I can do that | 21:37 |
| d2m | srichter: upload the image (you dont need a cmf image) and add the img tag | 21:37 |
| sm | though zope.org may not as its edit form is f*d | 21:37 |
| srichter | I can handle the upload via the ZMI | 21:38 |
| srichter | is using the IMG tag the recommended wya? | 21:38 |
| srichter | d2m: thanks, btw | 21:38 |
| *** mkerrin has quit IRC | 21:39 | |
| *** douglasc has quit IRC | 21:42 | |
| *** tav_ is now known as tav | 21:46 | |
| *** gintas has joined #zope3-dev | 21:52 | |
| *** anguenot has quit IRC | 21:54 | |
| *** bskahan_ has joined #zope3-dev | 21:55 | |
| *** bska|mobile has quit IRC | 22:00 | |
| *** Jim7J1AJH has quit IRC | 22:04 | |
| *** bskahan has joined #zope3-dev | 22:04 | |
| *** bskahan_ has quit IRC | 22:17 | |
| *** Jim7J1AJH has joined #zope3-dev | 22:24 | |
| *** tvon is now known as tvon|afk | 22:30 | |
| *** bskahan has quit IRC | 22:33 | |
| *** bskahan has joined #zope3-dev | 22:34 | |
| *** bobessutio has joined #zope3-dev | 22:34 | |
| *** ignas has quit IRC | 22:36 | |
| *** bskahan has quit IRC | 22:42 | |
| *** bskahan has joined #zope3-dev | 22:42 | |
| *** faassen has joined #zope3-dev | 22:43 | |
| *** alienoid has joined #zope3-dev | 22:50 | |
| *** bskahan has quit IRC | 23:03 | |
| bobessutio | is zope3 slower or faster than zope2? | 23:08 |
| srichter | I am pretty sure Z3 is faster, though it is hard to say | 23:09 |
| srichter | Zope 2 is larger in size and thus provides more features | 23:10 |
| sm | good question | 23:13 |
| benji_york | bobessutio, also note that "slower" and "faster" don't mean much without a way to measure what you're talking about | 23:13 |
| sm | comparable sites in z2 & z3 on the same machine | 23:13 |
| benji_york | sm, so requests/second? | 23:14 |
| projekt01 | hi | 23:14 |
| projekt01 | srichter, You where asking about a answer for Toms question, what was the question? | 23:14 |
| philiKON | bobessutio, i think that question is too general to be answered in a simple yes or no | 23:14 |
| sm | I know launchpad.net feels fast, for a zope site doing cmf-like things | 23:14 |
| srichter | projekt01: whether Zope 3 can run as a service | 23:15 |
| projekt01 | I use firedeamon on some custom projects for z3 to run as a service | 23:15 |
| philiKON | grrr. zope.org is annoying the crap out of me | 23:16 |
| sm | what's new :) | 23:16 |
| srichter | yep, me too | 23:16 |
| srichter | projekt01: ok, I'll tell tom | 23:17 |
| projekt01 | srichter, It's cheap and works very well, http://www.firedaemon.com/ | 23:17 |
| benji_york | projekt01, are you talking about running z3 as a windows service? | 23:17 |
| projekt01 | Yup, I run the a start.bat file with the firedeamon, since more then a year, but perhaps is there a win service in zope3 now? or not? | 23:19 |
| benji_york | I think so. | 23:19 |
| *** tvon|afk is now known as tvon | 23:20 | |
| projekt01 | benji_york, I think I have to take a look at zope.app.winservice. I don't know if this works or it's just a take over from z2. | 23:22 |
| benji_york | something works because we're using it :) | 23:22 |
| benji_york | let me look at the installer for the product I'm thinking of and see how it sets up the service | 23:22 |
| projekt01 | Yeah, whould be cool | 23:23 |
| benji_york | it uses the zopeservice.py script, but I'm not sure where that comes from, looking more... | 23:24 |
| projekt01 | ;-) | 23:24 |
| benji_york | it's in zopeskel/bin/zopeservice.py.in | 23:26 |
| philiKON | projekt01, it's more or less a takeover from zope 2; but i think recent fixes to the zope 2 equivalent weren't carried over to zope.app.winservice | 23:26 |
| philiKON | someone should compare the zope3-checkins and zope-checkins list and see if fixes still need to be merge | 23:27 |
| philiKON | d | 23:27 |
| benji_york | it would be nice if someone woul | 23:27 |
| benji_york | d | 23:27 |
| benji_york | :) | 23:28 |
| *** MrTopf has joined #zope3-dev | 23:28 | |
| srichter | does anyone here know a nice little program to record a movie of the screen? | 23:30 |
| srichter | i.e. for doing demos | 23:30 |
| * sm would like to know too (on linux) | 23:31 | |
| projekt01 | benji_york, I see, I never used a built z3 version. I always build it in my workspace. The script zopeservice.py seems to use some variables from the installer. | 23:32 |
| projekt01 | srichter, wink | 23:32 |
| sm | http://www.oreillynet.com/pub/a/network/2004/11/11/primetime.html seems relevant | 23:32 |
| benji_york | http://www.debugmode.com/wink/ | 23:32 |
| sm | "Camtasia Studio" is the one paul uses | 23:33 |
| SteveA | srichter: in launchpad, we have several .conf files, for different deployments. | 23:33 |
| SteveA | srichter: not sure how this would work with zcml. | 23:33 |
| *** bskahan has joined #zope3-dev | 23:33 | |
| srichter | I am trying Wink :-) | 23:33 |
| projekt01 | srichter, wink is pretty cool for that's open source. | 23:33 |
| srichter | SteveA: so you agree with me to put it in zope.conf? | 23:33 |
| SteveA | i have only caught the end of the discussion | 23:34 |
| SteveA | so i don't know what is being proposed | 23:34 |
| srichter | I propose to put the devmode on/off switch into zope.conf | 23:34 |
| SteveA | yes, that is where it should go | 23:34 |
| SteveA | zcml is about representing the glue we use to stick the app together | 23:35 |
| srichter | All that will do is insert a feature into the ZCML context, so that certain packages only loaded selectively | 23:35 |
| benji_york | projekt01, I don't think wink is OS, but it is free | 23:35 |
| srichter | that's what I did for SchoolBell and works well and is logical | 23:35 |
| SteveA | sounds good to me | 23:35 |
| SteveA | we do this in launchpad | 23:35 |
| SteveA | by having several "zcml root files" | 23:35 |
| SteveA | like, site.zcml for the webapp | 23:35 |
| SteveA | scripts.zcml for scripts that load zcml before they run | 23:36 |
| SteveA | scripts.zcml omits zcml from the browser: namespace | 23:36 |
| SteveA | as well as a bunch of other stuff that scripts don't need | 23:36 |
| srichter | I see | 23:36 |
| SteveA | so, the way i'd do it is | 23:36 |
| SteveA | have "developer.zcml" that includes "site.zcml" plus some other things | 23:37 |
| SteveA | the switch already exists to say in the .conf file which root zcml to use | 23:37 |
| SteveA | so, i don't think it needs an extra switch or to do special things | 23:37 |
| SteveA | if anything, it should be just syntactic sugar for doing what i described | 23:37 |
| srichter | mmh, that's an intersting solution too | 23:37 |
| SteveA | think of the price of the different solutions, in terms of their complexity and extra code to maintain | 23:38 |
| projekt01 | benji_york, right | 23:38 |
| srichter | ok, I put this down as an option | 23:39 |
| benji_york | we use runzope's -X switch to change the top-level zcml to use | 23:39 |
| benji_york | for such things | 23:39 |
| SteveA | srichter: thanks | 23:39 |
| srichter | I really like it; even if developer.zcml just installs the "devmode" feature and then loads site.zcml | 23:39 |
| SteveA | it is discoverable | 23:40 |
| benji_york | srichter, one thing to consider: lets say you have a running site in production mode (non-dev), start having problems and want to diagnose the site without shutting it down or turning on dev-mode for anyone else | 23:42 |
| SteveA | benji_york: here's what we do for that with launchpad | 23:42 |
| SteveA | we have a layer called IDebugLayer | 23:42 |
| srichter | benji_york: the dev-mode should never run on a production site | 23:42 |
| SteveA | there's an http server running on a port that is accessible only to developers | 23:42 |
| SteveA | this http server decorates the request by marking it IDebugLayer | 23:43 |
| SteveA | development-only features are registered to appear only on the IDebugLayer | 23:43 |
| SteveA | we don't allow users to switch to the debug layer | 23:43 |
| srichter | it pokes a lot of security holes into the system | 23:43 |
| benji_york | SteveA, I like that | 23:43 |
| SteveA | also, in production, we actually turn these things off, generally | 23:43 |
| SteveA | but still, we would turn them on if needed, i guess | 23:44 |
| SteveA | benji_york: i don't like how the debug skin is typically used in zope3 right now | 23:44 |
| SteveA | it's kinda random :-/ | 23:44 |
| srichter | I guess we would need some sort of developer permission | 23:44 |
| benji_york | or block the ports from external access | 23:44 |
| SteveA | i think ++skin++ should be removed in general | 23:44 |
| benji_york | srichter, I'd go with a network security solution (differenc ports, etc.) | 23:45 |
| SteveA | or be only available on the developer port | 23:45 |
| SteveA | most of the ++ stuff is kinda crackful | 23:45 |
| benji_york | SteveA, right | 23:45 |
| SteveA | like, users should never be able to see ++etc++ | 23:45 |
| srichter | benji_york: but part of the big proposal is that the dev mode stuff is not even loaded when in non-dev mode | 23:45 |
| srichter | this is very important for security | 23:45 |
| SteveA | having the development view modular is important | 23:46 |
| SteveA | you could use a zcml namespace to register them | 23:46 |
| SteveA | dev-only:page etc. | 23:46 |
| SteveA | same as browser:page, but on the IDevelopmentOnly layer | 23:46 |
| benji_york | right, I'm saying that there are two sets of things, a production/dev mode that affects how the server is configured, and a seperate set of developer tools that can be enabled/disabled | 23:46 |
| SteveA | so, you can easily cut out all development only stuff at the zcml level | 23:46 |
| srichter | that's the way I went in SB | 23:46 |
| SteveA | and you can also choose to run the stuff out of a different port | 23:47 |
| srichter | ok, that would need a separate proposal; I am only going to worry about a specific use case now | 23:47 |
| srichter | everything else would need a very careful security review | 23:47 |
| benji_york | for example, one of our projects uses Windows users/passwords for credentials in "production mode" and uses a set of fake users in "dev mode" | 23:48 |
| srichter | benji_york: thanks for wink, seems to work well | 23:54 |
| *** bskahan has quit IRC | 23:55 | |
Generated by irclog2html.py 2.15.1 by Marius Gedminas - find it at mg.pov.lt!