*** ignas has quit IRC | 00:17 | |
*** ignas has joined #zope3-dev | 00:17 | |
*** vlado has quit IRC | 00:51 | |
*** yota has quit IRC | 00:53 | |
*** loreto has joined #zope3-dev | 01:03 | |
*** benji_york has joined #zope3-dev | 01:18 | |
*** mgedmin has quit IRC | 01:43 | |
*** projekt01 has quit IRC | 02:08 | |
*** xenru has quit IRC | 02:16 | |
*** ignas has quit IRC | 02:33 | |
*** loreto has quit IRC | 02:51 | |
roym | if anyone is awake - I have declared a resourceDirectory; what syntax would I use to reference a css file present in it from elsewhere in my hierarchy? Is there an equivalent of the good old ++resource++RES_NAME syntax? | 03:12 |
---|---|---|
roym | How would I use absolute_url in a zpt page to do this? | 03:16 |
*** roym has quit IRC | 03:33 | |
*** stub has joined #zope3-dev | 04:20 | |
*** tiredbones has left #zope3-dev | 04:29 | |
*** stub has quit IRC | 08:04 | |
*** sashav has quit IRC | 08:45 | |
*** j-w has joined #zope3-dev | 09:09 | |
*** Aiste has quit IRC | 09:20 | |
*** hdima has joined #zope3-dev | 09:24 | |
*** MrTopf has joined #zope3-dev | 09:30 | |
*** yota has joined #zope3-dev | 09:51 | |
*** Theuni has joined #zope3-dev | 09:56 | |
*** stub has joined #zope3-dev | 09:59 | |
*** Aiste has joined #zope3-dev | 10:00 | |
*** MJ has quit IRC | 10:01 | |
*** d2m has quit IRC | 10:21 | |
*** stub has quit IRC | 10:28 | |
*** mgedmin has joined #zope3-dev | 10:36 | |
*** sashav has joined #zope3-dev | 10:37 | |
*** __gotcha__ has joined #zope3-dev | 10:38 | |
*** __gotcha__ is now known as __gotcha | 10:38 | |
*** mgedmin has quit IRC | 10:40 | |
*** sashav_ has joined #zope3-dev | 10:43 | |
*** mgedmin has joined #zope3-dev | 10:44 | |
*** sashav has quit IRC | 10:57 | |
*** MJ has joined #zope3-dev | 10:57 | |
*** stub has joined #zope3-dev | 11:04 | |
*** Alef has joined #zope3-dev | 11:13 | |
*** vlado has joined #zope3-dev | 11:34 | |
*** projekt01 has joined #zope3-dev | 11:42 | |
*** Alef has quit IRC | 12:07 | |
*** jinty has joined #zope3-dev | 12:09 | |
projekt01 | I have some wired behavior on a apache http GET request to a z3 server. | 12:11 |
projekt01 | The z3 server ends in a error like: | 12:12 |
projekt01 | The proxy server could not handle the request GET / </>. | 12:12 |
projekt01 | But this happens only from a link on webpage where is running under SSL | 12:12 |
projekt01 | Why is the "/" different intepreted from different links on pages if you klick on it? | 12:16 |
projekt01 | Could it be possible that it is a difference if you click on the same link listed on a https:// page or a http:// page? | 12:18 |
projekt01 | I guess there is no apache guru here ;-) | 12:19 |
projekt01 | mgedmin? | 12:19 |
*** BjornT has quit IRC | 12:24 | |
mgedmin | ? | 12:26 |
mgedmin | do you use apache rewrite rules? | 12:26 |
mgedmin | what do they look like? | 12:26 |
projekt01 | Yup | 12:26 |
projekt01 | RewriteRule (/?.*) http://localhost:8080/++skin++Admin/++vh++http:admin.amadeus.demo:80/++$1 [P,L] | 12:27 |
mgedmin | and for https? | 12:27 |
projekt01 | I don't run https. The problem is the link is listed on a webpage where is running on SSL | 12:28 |
mgedmin | strange | 12:29 |
projekt01 | Yup | 12:29 |
mgedmin | does error.log say anything interesting? | 12:29 |
projekt01 | [Mon Sep 05 11:21:48 2005] [error] [client 212.4.83.9] proxy: error reading status line from remote server localhost | 12:30 |
projekt01 | [Mon Sep 05 11:21:48 2005] [error] [client 212.4.83.9] proxy: Error reading from remote server returned by / | 12:30 |
mgedmin | do zope's logs show the request? | 12:31 |
projekt01 | I guess apache is calling a GET request for "/" but only on link listed on SSL driven sites | 12:31 |
projekt01 | I'll take a look | 12:31 |
mgedmin | BTW (/?.*) is not a very nice regexp -- it is completely equivalent to (.*) | 12:32 |
mgedmin | I use RewriteRule ^/(.*) http://localhost:.../...++vh++.../++/$1 [P,L] | 12:33 |
projekt01 | Ok, I see what's wrong. The SSL page sends cookie information via the request. Uhhh | 12:33 |
projekt01 | File "C:\Python24\lib\Cookie.py", line 453, in set | 12:33 |
projekt01 | raise CookieError("Illegal key value: %s" % key) | 12:33 |
projekt01 | CookieError: Illegal key value: Williams/OU | 12:33 |
mgedmin | whoa, that's not nice | 12:33 |
mgedmin | z3 could at least render an error page | 12:33 |
mgedmin | 400 Bad Request or something | 12:34 |
projekt01 | hm, I'm not sure, I just will check this again and compare the exact time of the requests | 12:34 |
projekt01 | mgedmin yes, it's for sure the SSL server sends cookie information. | 12:37 |
projekt01 | perhaps this is because the SSL page list the page runs at URL: https://opera.amadeus.com | 12:38 |
projekt01 | and the page where the link is pointing to is http://www.ch.amadeus.com | 12:38 |
projekt01 | (same domain, different subdomain) | 12:39 |
projekt01 | What do you think is it a bug in z3, does apache do this everytime on same domains? | 12:39 |
projekt01 | and we have to make sure that such requests can be handled in z3? | 12:40 |
projekt01 | mgedmin, here is the error traceback: | 12:42 |
projekt01 | http://paste.plone.org/3646 | 12:42 |
projekt01 | Who is the z3 webserver guru at ZC? | 12:43 |
mgedmin | I'd say this is a zope 3 bug | 12:53 |
mgedmin | given an invalid Cookie request header, z3 should not just fail to return a response | 12:54 |
mgedmin | it should return a 400 response | 12:54 |
mgedmin | imho | 12:54 |
*** Theuni has quit IRC | 12:57 | |
*** anguenot has joined #zope3-dev | 13:27 | |
*** ignas has joined #zope3-dev | 13:30 | |
*** Theuni has joined #zope3-dev | 13:47 | |
projekt01 | mgedmin, which component is responsible to send a cookie with the request to a server if you click on a link? the browser or the webpage? | 14:14 |
projekt01 | Btw, there is no javascript involved in the link. | 14:15 |
bob2 | cookies aren't set by web pages | 14:24 |
*** vlado has quit IRC | 14:25 | |
*** d2m has joined #zope3-dev | 14:30 | |
projekt01 | bob2, Ok, where sends a cookie if you click on a link. | 14:43 |
projekt01 | Or perhaps I have to ask where is responsible that the Request Header contains a key "Cookie" if I click on a link? | 14:44 |
*** MrTopf has quit IRC | 14:44 | |
*** niemeyer has joined #zope3-dev | 14:45 | |
projekt01 | where/who | 14:47 |
*** bskahan has joined #zope3-dev | 15:05 | |
*** Theuni has quit IRC | 15:08 | |
*** MrTopf has joined #zope3-dev | 15:19 | |
*** BjornT has joined #zope3-dev | 15:23 | |
*** vlado has joined #zope3-dev | 15:41 | |
*** fcorrea has joined #zope3-dev | 15:46 | |
*** roym has joined #zope3-dev | 15:58 | |
roym | hi, does anyone know if using sql: expressions in zpt pages (w/sql_conn) is meant to be multi-threaded - or will requests be queued. | 15:59 |
*** tiredbones has joined #zope3-dev | 16:04 | |
projekt01 | romy, srichter can answer this question, but he isn't only yet. | 16:05 |
roym | ok, thanks... looking at the code, it seems to be a straightforward use of the ZopeDatabaseAdapter utility. Would you happen to know if ZopeDatabaseAdapter conns are pooled? | 16:11 |
benji_york | projekt01, I'm looking at your test failure on XP | 16:12 |
mgedmin | projekt01, the browser sends the Cookie: header | 16:17 |
mgedmin | if it had previously received a Set-Cookie: response header for a given domain | 16:17 |
projekt01 | mgedmin, I just googled for this. It's not exactly this way. | 16:19 |
projekt01 | A cookie is sent back just for the given domain | 16:19 |
projekt01 | But, the server can configure that the cookie is also sent back to servers on the same subdomain | 16:20 |
projekt01 | Cookies get never sent to another server if they not at least in the same subdomain via a link. | 16:21 |
bob2 | yes | 16:21 |
mgedmin | afaiu links don't come into this | 16:22 |
*** alga has joined #zope3-dev | 16:23 | |
bob2 | only in the sense that they control who the request goes to | 16:23 |
bob2 | stop thinking of it in terms of web pages or links, it's all about http requests | 16:23 |
projekt01 | bob2, is this correct, that the server can define to which subdomain server a cookie will be sent via the browser if you click on a link pointing to the subdomain? | 16:26 |
projekt01 | bob2, mgedmin, found a documentaion about that this at: | 16:29 |
projekt01 | http://wp.netscape.com/newsref/std/cookie_spec.html | 16:29 |
projekt01 | Seems that this is defined in the previous request header from the server with. | 16:30 |
projekt01 | domain=DOMAIN_NAME | 16:30 |
projekt01 | e.g. Cookie: ... domain=DOMAIN_NAME ... | 16:31 |
projekt01 | But, I'm not sure what should happen if z3 recieves such a cookie from such a friendly subdomain server. | 16:32 |
bob2 | igore the link | 16:35 |
bob2 | but, yes, the cookie includes information about who should be able to retrieve it | 16:35 |
bob2 | doesn't zope's request object have cookie handling methofs? | 16:35 |
*** Theuni has joined #zope3-dev | 16:37 | |
mgedmin | yes -- and it raises an exception that is not caught | 16:37 |
mgedmin | look at projekt01's traceback: http://paste.plone.org/3646 | 16:37 |
projekt01 | bob2, I'm not sure right now. | 16:37 |
projekt01 | mgedmin, Yes, there is no explicit error handling for this part. | 16:38 |
projekt01 | who is responsible or the master for the zope.publisher.http.py webserver part at ZC? | 16:39 |
mgedmin | suggestion: zope 3 should complain loudly in a log file somewhere, and then ignore ill-firmed cookies | 16:39 |
mgedmin | projekt01, perhaps you can file a bug report in the collector? | 16:39 |
*** vlado has quit IRC | 16:39 | |
mgedmin | (if there isn't one already) | 16:39 |
*** vlado has joined #zope3-dev | 16:40 | |
projekt01 | mgedmin, I like to fix it, but I don't like to do this without contact the big boss. | 16:40 |
projekt01 | Ok, I take a look at the checkin logs. | 16:41 |
projekt01 | mgedmin, bob2, thanks for your help, just reported the issue at http://www.zope.org/Collectors/Zope3-dev/450 | 17:01 |
bob2 | np | 17:01 |
*** hdima has quit IRC | 17:20 | |
*** sashav_ has quit IRC | 17:23 | |
benji_york | projekt01, I just checked in a fix for the syntax error when running the tests on XP, I haven't looked at the other exceptions yet | 17:29 |
*** alga has quit IRC | 17:38 | |
*** Theuni has quit IRC | 17:39 | |
roym | \w | 17:43 |
roym | \? | 17:43 |
projekt01 | benji_york, thanks a lot, the test_textwidget.py tests are Ok now. | 17:47 |
benji_york | np | 17:47 |
benji_york | I took a look at the fsmerger tests, but don't have a clue without digging in a bit more, so I'll leave those for someone who knows what's going on | 17:48 |
projekt01 | Ok, I also didn't deep into this part, perhaps we can do it if we have more time and if it's get relevant. | 17:50 |
projekt01 | btw, does the fssync work at all? Does somebody use this package? | 17:52 |
benji_york | I don't really know, but I've never seen these tests fail before | 17:54 |
*** MrTopf has quit IRC | 17:55 | |
projekt01 | But, I remember a long time before the tests for diff3 where disabled on win. or I'm worng? | 17:55 |
projekt01 | worng/wrong | 17:55 |
benji_york | you're right, the tests check for os.popen, and if found check for diff3 and if found run the tests | 18:00 |
*** efge has joined #zope3-dev | 18:07 | |
*** sashav has joined #zope3-dev | 18:16 | |
*** niemeyer is now known as nie_lunch | 18:21 | |
SteveA | anyone know if there is a particular release date planned for 3.1.0 ? | 18:27 |
benji_york | I don't know of a date, but it should be Real Soon Now (tm) | 18:28 |
SteveA | thanks | 18:28 |
SteveA | the packager of zope for ubuntu was asking | 18:29 |
SteveA | it would be nice to get zope 3.1 into the breezy badger release of ubuntu | 18:29 |
benji_york | I would predict a release on Sunday or before, but that's just a guess, srichter may have a different plan | 18:30 |
benji_york | I don't know when he'll be back on IRC | 18:30 |
projekt01 | benji_york, could it be that the check for os.popen is incorrect, since the implementation of popen changed in py 2.4.1 | 18:36 |
projekt01 | and the tests are running now on win too, but shouldn't like before? | 18:37 |
benji_york | oh!, you just prompted me to think of the answer (I think :) | 18:38 |
benji_york | do you have a diff3 executable on your Windows box? | 18:38 |
benji_york | I do, and that's why it's running the tests | 18:39 |
projekt01 | No, I guess not, or is it included in GNU diff? | 18:39 |
benji_york | don't know, type diff3 at a command promt and find out | 18:39 |
projekt01 | Oh, well I have diff3 installed! | 18:40 |
benji_york | :) | 18:41 |
benji_york | that explains why we are seeing failures and noone else is | 18:41 |
projekt01 | It's hot here in switzerland today :-) | 18:41 |
projekt01 | Ok | 18:41 |
benji_york | so, apparently fssync (or its tests) doesn't work on Windows | 18:42 |
projekt01 | Yes, I guess it doesn't work since a looong time or didn't never work on windows. | 18:45 |
projekt01 | benji_york, perhaps we can contact fdrake and ask him about the state of fssync | 18:51 |
*** MJ has quit IRC | 18:53 | |
benji_york | I think just explicitly disabling the tests on windows would be best, they are obviously not ment to run on windows, so they shouldn't | 18:57 |
*** benji_york has quit IRC | 19:04 | |
*** j-w has quit IRC | 19:05 | |
*** vlado has quit IRC | 19:10 | |
*** nie_lunch is now known as niemeyer | 19:21 | |
*** faassen has joined #zope3-dev | 19:23 | |
*** stub has quit IRC | 19:29 | |
*** horizon5 has joined #zope3-dev | 19:29 | |
*** roym has quit IRC | 19:40 | |
*** tarek has joined #zope3-dev | 19:58 | |
*** niemeyer has quit IRC | 20:04 | |
*** niemeyer has joined #zope3-dev | 20:14 | |
*** tarek has quit IRC | 20:32 | |
*** MJ has joined #zope3-dev | 20:46 | |
*** faassen has quit IRC | 20:57 | |
*** efge has quit IRC | 21:14 | |
*** mgedmin has quit IRC | 21:18 | |
*** dman13 has quit IRC | 21:21 | |
*** dman13 has joined #zope3-dev | 21:23 | |
*** dman13_ has joined #zope3-dev | 21:24 | |
*** tarek has joined #zope3-dev | 21:29 | |
*** anguenot has quit IRC | 21:39 | |
*** sashav has quit IRC | 21:51 | |
*** ignas has quit IRC | 22:14 | |
*** dman13 has quit IRC | 22:17 | |
*** dman13_ has quit IRC | 22:17 | |
*** dman13 has joined #zope3-dev | 22:17 | |
*** yota_ has joined #zope3-dev | 22:27 | |
*** yota has quit IRC | 22:39 | |
*** bskahan has quit IRC | 22:49 | |
*** bskahan has joined #zope3-dev | 22:50 | |
*** vlado has joined #zope3-dev | 22:57 | |
*** vlado has quit IRC | 22:57 | |
*** vlado has joined #zope3-dev | 22:58 | |
*** roym has joined #zope3-dev | 23:23 | |
roym | folks, how woud I get info about the currently logged in user (role, id, etc)... there used to be something like session.getUser() in z2... | 23:24 |
mp | self.request.principal? | 23:24 |
*** jinty has left #zope3-dev | 23:25 | |
roym | mp: thanks. | 23:25 |
mp | At least I hope... I've never used it ;) | 23:27 |
roym | That gives me the login id; I don't see how I can get role information.. | 23:28 |
roym | (ie: I can do "request/principal/getLogin" in zpt) | 23:29 |
*** sashav has joined #zope3-dev | 23:31 | |
*** ignas has joined #zope3-dev | 23:40 | |
*** tarek has quit IRC | 23:55 | |
*** fcorrea has quit IRC | 23:57 |
Generated by irclog2html.py 2.15.1 by Marius Gedminas - find it at mg.pov.lt!