Re: [Trac] 0.13 stable release?
On Tue, Mar 13, 2012 at 3:38 PM, Remy Blank wrote: > Philip Semanchuk wrote: >> On Tue, Mar 13, 2012 at 11:40 AM, Benjamin Lau >> wrote: >>> You should probably add your voice to the ticket increasing the >>> pressure to actually get this patch included in the core as well. :-) >> >> Good advice. > > Nah, won't work ;) > > That patch actually looks pretty good, and I have had some more exposure > to HTTP-only cookies since the ticket was filed, so I think it's a good > idea. I'll apply it, probably later tonight or tomorrow. > > One thing I'm not quite sure, should the attribute be applied to all > cookies, or only to the auth cookie? I saw the discussion in the ticket on this subject. IMHO it should apply to all. But I say that only from the principle of "don't expose more than you have to". I don't know what cookies Trac uses or how plugins might take advantage of them, so my opinion is pretty naive. > And about 0.13, you'll have to wait and see. I don't give any estimates > anymore, as I have lost all credibility on the subject anyway. Fair enough. Thanks! bye Philip -- You received this message because you are subscribed to the Google Groups "Trac Users" group. To post to this group, send email to trac-users@googlegroups.com. To unsubscribe from this group, send email to trac-users+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/trac-users?hl=en.
Re: [Trac] 0.13 stable release?
On Tue, Mar 13, 2012 at 11:40 AM, Benjamin Lau wrote: >> We would like to get the Powers that Be off our backs. Right now, it >> seems like our options are -- >> 1) Install & monkey patch Trac 0.12.3 > > I'll leave the conversation about the 0.13 release to others... but > why don't you just apply the existing patch attached to that ticket to > the version of Trac you're already running? I tried applying the > patch to 0.11.7 and 0.12.3 and it seems to apply cleanly (the change > is pretty simple)... Hi Ben, I agree; I looked at the patch and it seems quite straightforward. I'd rather avoid patching Trac, though. It's not that I fear breakage. It's because my time on this project is limited and I'm afraid that (despite documentation) when I leave our Trac install will be regarded as "something that Philip had to hack, better not touch it". This is, of course, an institutional problem, not a Trac problem, but it's a problem I have to consider nonetheless. > You should probably add your voice to the ticket increasing the > pressure to actually get this patch included in the core as well. :-) Good advice. Thanks Philip -- You received this message because you are subscribed to the Google Groups "Trac Users" group. To post to this group, send email to trac-users@googlegroups.com. To unsubscribe from this group, send email to trac-users+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/trac-users?hl=en.
[Trac] 0.13 stable release?
Hi all, We have been using and appreciating Trac 0.11.5 for a few years now. Our project is hosted at a large university. The Powers That Be on our network run an automated 3rd party vulnerability scanner, and our Trac instance gets a red flag because it fails to use the HttpOnly flag when setting cookies. A patch for HttpOnly was offered in ticket 10453 , but the patch has not yet been integrated. The ticket has a milestone of Trac 0.13. I'm aware that 0.13dev exists, but there is no officially stable 0.13. I scanned trac- dev and didn't see any conversation about an imminent 0.13 release. We would like to get the Powers that Be off our backs. Right now, it seems like our options are -- 1) Install & monkey patch Trac 0.12.3 2) Follow these somewhat intimidating instructions to get Apache to automatically add the HttpOnly flag to all cookie settings: http://blog.modsecurity.org/2008/12/helping-protect-cookies-with-httponly-flag.html Obviously an upgrade to Trac 0.13 stable is our favorite solution, but the less elegant solutions have the advantage of being available right now. Does anyone have a sense of when Trac 0.13 (with the 10453 patch applied) might be out? That will help us to decide whether to wait or act. Thanks Philip -- You received this message because you are subscribed to the Google Groups "Trac Users" group. To post to this group, send email to trac-users@googlegroups.com. To unsubscribe from this group, send email to trac-users+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/trac-users?hl=en.
Re: [Trac] Access the SQLite database outside Trac
On Jan 10, 2010, at 4:03 AM, Scott Mead wrote: On Sat, Jan 9, 2010 at 2:54 PM, Philip Semanchuk wrote: But the OP was asking about accessing the database from a external process (i.e. not Trac), so how does SQLite's threading model enter into it? Sorry, by 'thread' I was referring to the issue that you're talking about below, not internal threading, but locking at the application -> DB layer. Sorry for the confusion. Gotcha. I rarely fail to take the most literal interpretation of what I read. It might help me to broaden my perspective once in a while. =) Cheers P -- You received this message because you are subscribed to the Google Groups "Trac Users" group. To post to this group, send email to trac-us...@googlegroups.com. To unsubscribe from this group, send email to trac-users+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/trac-users?hl=en.
Re: [Trac] Access the SQLite database outside Trac
On Jan 9, 2010, at 1:29 PM, Scott Mead wrote: On Sat, Jan 9, 2010 at 11:08 AM, David Champion wrote: Certainly, it's at $TRAC_ENV/db/trac.db. But you might find yourself more productive as well as upward-compatible to use the Trac API instead. Not only that, but sqlite is a single threaded database, so if you got your reporting app going you either have access from trac, or your app. But the OP was asking about accessing the database from a external process (i.e. not Trac), so how does SQLite's threading model enter into it? To the OP - multiple processes can comfortably read from a SQLite database, but SQLite locks out writes while another query is reading. The doc says: The first step toward reading from the database file is obtaining a shared lock on the database file. A "shared" lock allows two or more database connections to read from the database file at the same time. But a shared lock prevents another database connection from writing to the database file while we are reading it. http://www.sqlite.org/atomiccommit.html So if you have long-running queries (as a reporting app might), it could wreak havoc with Trac trying to write to the DB. If your reports are short, you might be OK, but that sounds like an assumption one might eventually regret. If you don't want to move to something like Postgres, you might want to look into making a copy of the database, or exporting relevant tables before running your reporting app. Good luck Philip -- You received this message because you are subscribed to the Google Groups "Trac Users" group. To post to this group, send email to trac-us...@googlegroups.com. To unsubscribe from this group, send email to trac-users+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/trac-users?hl=en.
Re: [Trac] Full text search performance with Postgres
On Jan 4, 2010, at 6:42 PM, David Huang wrote:
>
> On Jan 4, 2010, at 2:34 PM, Philip Semanchuk wrote:
>> Hi David,
>> I could be wrong, but this is more of a generic SQL problem than a
>> Postgres problem. You need to find out what columns are being
>> accessed
>> during a full text search in order to decide where an index might
>> help. You might find that easiest to do with Trac.
>
> The SQL being executed is in the Trac source at:
> http://trac.edgewall.org/browser/branches/0.11-stable/trac/ticket/web_ui.py?rev=7569#L202
>
> Enabling Postgres logging gives me this as the actual statement (for
> a search for "added"):
Good on ya for finding the query and digging into the Postgres
logging. The explain analyze is helpful too, although reading the
output is an art at which I never surpassed the level of novice.
However I can make a few stabs at what you sent, but take my
observations with the same credibility that you'd lend to Sir
Bedevere's observation, "...and that, my liege, is how we know the
Earth to be banana shaped".
First of all, your sort is happening on disk ("Sort Method: external
merge Disk: 18072kB"). It looks like that's about 11% of your time.
You can increase the work_mem parameter (I think) to try to get that
sort to happen in memory. The disk sort might have been happening
before, though, so we can't point to that as the perpetrator to the
sudden performance decrease.
I'm not sure about the use of the index on ticket_custom.ticket.
Postgres is using a hash or building one on the fly ("Hash
(cost=475.34..475.34 rows=30134 width=6...") which I guess would allow
for O(1) lookups in ticket_custom. Postgres might decide that's
cheaper than using the index, especially if the hash can be
constructed in memory rather than loading from disk the pages that
describe the index on that table.
Nevertheless, if I read this correctly (and I'm really not confident
of that, in case I haven't already overstated that point), the hash/
index on ticket_custom doesn't really matter. The query is spending
most of the time that's not lost to the sort in the hash left join
("Hash Left Join (cost=852.02..15159.20 rows=11190 width=414) (actual
time=125.952..56215.314"). If you add up the "actual time" of the
items after that, they're a pittance compared to the 56215ms. I think
"actual time" is supposed to be multiplied by the number of loops, but
in each case loops==1.
My guess, then, is that the Filter step chews up the lion's share of
the time, and to me that makes intuitive sense. I don't think there's
a way for any database to shortcut this -- the rows have to be loaded
from disk and searched for the string in question. Note that six
fields from two tables are ORed:
(a.summary ~~* '%added%'::text) OR
(a.keywords ~~* '%added%'::text) OR
(a.description ~~* '%added%'::text) OR
(a.reporter ~~* '%added%'::text) OR
(a.cc ~~* '%added%'::text) OR
((a.id)::text ~~* '%added%'::text) OR (c.value ~~* '%added
%'::text))
Postgres has to search each of these columns (from two different
tables) for each row until it finds a hit. That's a lot of disk
traffic, especially if these fields are large.
So that's my hypothesis. One way to test it would be to hack
web_ui.py, revert the query in question to a code revision where you
didn't see the big performance hit, and run this again with the
EXPLAIN ANALYZE output and see how it differs from this.
Hope this helps
Philip
PS - let me know if you want to know how sheep's bladders can be used
to prevent earthquakes
> SELECT DISTINCT a.summary,a.description,a.reporter,
> a.type,a.id,a.time,a.status,a.resolution
> FROM ticket a
> LEFT JOIN ticket_change b ON a.id = b.ticket
> LEFT OUTER JOIN ticket_custom c ON (a.id = c.ticket)
> WHERE (b.field='comment' AND (b.newvalue ILIKE E'%added%' ESCAPE
> '/')) OR
> (summary ILIKE E'%added%' ESCAPE '/' OR keywords ILIKE E'%added%'
> ESCAPE '/' OR description ILIKE E'%added%' ESCAPE '/' OR reporter
> ILIKE E'%added%' ESCAPE '/' OR cc ILIKE E'%added%' ESCAPE '/' OR
> CAST(id AS text) ILIKE E'%added%' ESCAPE '/') OR
> (c.value ILIKE E'%added%' ESCAPE '/')
>
> And an EXPLAIN ANALYZE of the query plan shows:
>
> Unique (cost=18018.72..18270.50 rows=6525 width=414) (actual
> time=63595.824..64390.470 rows=931 loops=1)
> -> Sort (cost=18018.72..18046.70 rows=11190 width=414) (actual
> time=6359
Re: [Trac] Full text search performance with Postgres
On Jan 4, 2010, at 2:42 PM, David Huang wrote: > Hi, I had been running Trac 0.11 with a PostgreSQL database for a > while, and recently upgraded it to 0.11.6. After the upgrade, the > full text search is much slower than it was in 0.11, and while it's > searching, Postgres is using all the CPU. I believe this is due to > the changes for ticket #2530, which changed the search to include > ticket custom fields. I noticed the ticket_custom table didn't have > any indexes, so I added one on the ticket column, but that didn't > seem to help any. I'm no Postgres guru, so adding the index is all I > could think of. Does anyone have any ideas on how to speed up the > search? A search for a single word used to take a few seconds with > 0.11; it now takes over a minute. Hi David, I could be wrong, but this is more of a generic SQL problem than a Postgres problem. You need to find out what columns are being accessed during a full text search in order to decide where an index might help. You might find that easiest to do with Trac. Alternatively (and here's where it gets Postgres-specific), you could turn on Postgres logging via postgresql.conf with the log_statement param set to "all". You'll need to set some other params, too, like log_destination, etc. My Postgres skills are a bit rusty and I don't have this stuff in my head anymore, so check the manual for details. Setting log_statement=all will cause every query executed to be written to the log. This will kill your performance, so you don't want to leave this on any longer than you have to. If you log all queries, you'll get to see the exact SQL that's executed when you do a full-text search. Copy and paste that SQL into a text file, and then use psql (Postgres' interactive command line) to play around with it. You can add an index to a column and re-run the query to see if the run time improves. (Beware of caching which can skew your results.) Hope this helps Philip -- You received this message because you are subscribed to the Google Groups "Trac Users" group. To post to this group, send email to trac-us...@googlegroups.com. To unsubscribe from this group, send email to trac-users+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/trac-users?hl=en.
[Trac] Re: TracNav plugin on 0.10.4 on linux not working
On Jul 8, 3:19 pm, "Thomas Moschny" <[EMAIL PROTECTED]> wrote: > 2008/7/8 Philip <[EMAIL PROTECTED]>: > > > Can anybody respond to my latest post? I can't even seem to get the > > plugin to load. I have 2 plugins, and they are in 3 locations: > > - And, does the TracHelloworld plugin work? > - Did you set up a plugin cache, as described > here:http://trac.edgewall.org/wiki/TracPlugins#SettingupthePluginCache? > > Best, > Thomas Finally found the problem! Even though the plugin egg file had the right permissions, the trac.ini was not readable by the web server id. After making this file readable, the TracNav "magically" appeared. :-) Thanks for all who replied. I removed the egg file from these 2 locations: /usr/local/lib64/python2.4/site-packages /trac/myproject/plugins and left it in: /usr/local/share/trac/plugins so it would be common to all of my projects. --~--~-~--~~~---~--~~ You received this message because you are subscribed to the Google Groups "Trac Users" group. To post to this group, send email to trac-users@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/trac-users?hl=en -~--~~~~--~~--~--~---
[Trac] Re: TracNav plugin on 0.10.4 on linux not working
On Jun 25, 1:04 pm, Philip <[EMAIL PROTECTED]> wrote: > On Jun 24, 5:07 pm, "Thomas Moschny" <[EMAIL PROTECTED]> wrote: > > > > > > > 2008/6/24 Philip <[EMAIL PROTECTED]>: > > > > I tried that, and I still get the same thing. Plus, I still don't see > > > the "Loading..." message in the log, even after I attempt to access a > > > wiki page containing [[TracNav]]. If I google "No macro or processor > > > named 'TracNav' found", I get several other pages that seem to have > > > similar problem. So, maybe it's a fairly common problem. Still > > > looking for a solution. > > > Not sure this is specific to the TracNav plugin though (but if it was, > > I'd be glad to fix it!), merely an installation issue. Do you have any > > other plugins working? Did you check permissions of the egg? The web > > server (running e.g. as wwwrun) must be allowed to access it. > > > Regards, > > Thomas > > I followed the instructions from this > linkhttp://trac-hacks.org/wiki/EggCookingTutorial/BasicEggCookingto > create and install a helloworld plugin into the same project. After > copying the TracHelloworld-0.1-py2.4.egg to the project plugins > directory, setting permissions to 777, and restarting web server, I > see no difference in wiki pages and no "loading" messages in log > file. I'm guessing I need to concentrate on getting the "loading" > message again, but not sure what I need to do. Seemed random at best > when I was seeing it before. > > Thanks, > Philip- Hide quoted text - > > - Show quoted text - Can anybody respond to my latest post? I can't even seem to get the plugin to load. I have 2 plugins, and they are in 3 locations: /usr/local/share/trac/plugins # ls -l total 16 -rwxrwxrwx 1 root root 2943 Jun 25 17:05 TracHelloworld-0.1-py2.4.egg -rwxrwxrwx 1 root root 11623 Jun 25 17:05 TracNav-3.92-py2.4.egg /usr/local/lib64/python2.4/site-packages # ls -l total 32 -rwxrwxrwx 1 root root 2943 Jun 25 17:11 TracHelloworld-0.1- py2.4.egg -rwxrwxrwx 1 root root 11623 Jun 25 17:11 TracNav-3.92-py2.4.egg -rw-r--r-- 1 root root 239 Jun 23 19:26 easy-install.pth drwxrwxrwx 4 root root 4096 Jul 8 18:05 setuptools-0.6c8-py2.4.egg -rw-r--r-- 1 root root29 Jun 23 17:07 setuptools.pth drwxr-xr-x 11 root root 4096 Apr 17 15:01 trac /trac/myproject/plugins # ls -l total 16 -rwxrwxrwx 1 websrvr webgrp 2943 Jun 24 21:23 TracHelloworld-0.1- py2.4.egg -rwxrwxrwx 1 websrvr webgrp 11623 Jun 24 20:36 TracNav-3.92-py2.4.egg I have logging set to debug, and after restarting apache2, I don't see anything in the log. When I bring up a page that has something like [[TracNav(FAQ/TOC)]], I get an error: Error: Failed to load processor TracNav No macro or processor named 'TracNav' found Please help! Thanks. --~--~-~--~~~---~--~~ You received this message because you are subscribed to the Google Groups "Trac Users" group. To post to this group, send email to trac-users@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/trac-users?hl=en -~--~~~~--~~--~--~---
[Trac] Re: TracNav plugin on 0.10.4 on linux not working
On Jun 24, 5:07 pm, "Thomas Moschny" <[EMAIL PROTECTED]> wrote: > 2008/6/24 Philip <[EMAIL PROTECTED]>: > > > I tried that, and I still get the same thing. Plus, I still don't see > > the "Loading..." message in the log, even after I attempt to access a > > wiki page containing [[TracNav]]. If I google "No macro or processor > > named 'TracNav' found", I get several other pages that seem to have > > similar problem. So, maybe it's a fairly common problem. Still > > looking for a solution. > > Not sure this is specific to the TracNav plugin though (but if it was, > I'd be glad to fix it!), merely an installation issue. Do you have any > other plugins working? Did you check permissions of the egg? The web > server (running e.g. as wwwrun) must be allowed to access it. > > Regards, > Thomas I followed the instructions from this link http://trac-hacks.org/wiki/EggCookingTutorial/BasicEggCooking to create and install a helloworld plugin into the same project. After copying the TracHelloworld-0.1-py2.4.egg to the project plugins directory, setting permissions to 777, and restarting web server, I see no difference in wiki pages and no "loading" messages in log file. I'm guessing I need to concentrate on getting the "loading" message again, but not sure what I need to do. Seemed random at best when I was seeing it before. Thanks, Philip --~--~-~--~~~---~--~~ You received this message because you are subscribed to the Google Groups "Trac Users" group. To post to this group, send email to trac-users@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/trac-users?hl=en -~--~~~~--~~--~--~---
[Trac] Re: TracNav plugin on 0.10.4 on linux not working
On Jun 24, 3:03 pm, "Thomas Moschny" <[EMAIL PROTECTED]> wrote: > 2008/6/24 Philip <[EMAIL PROTECTED]>: > > > You are right -- I guess the install put the egg into /usr/local/lib64/ > > python2.4/site-packages. I deleted the egg file from /usr/local/share/ > > trac/plugins and restarted web server. I don't automatically see a > > mesage like you described when I recycle apache2, BUT if I look at the > > log, I do see one from earlier in the day yesterday. So, sounds like > > it was loading at some point anyway. > > For the beginning, it might be easier to run "./setup bdist_egg" and > copy the egg from 'dist' to the 'plugins' subdirectory of your trac > project, seehttp://trac.edgewall.org/wiki/TracPlugins#ForaSingleProject. > > Note that plugin loading might be delayed until someone accesses your > trac instance, so you might not see something in the logs immediately > after restarting apache. > > Regards, > Thomas I tried that, and I still get the same thing. Plus, I still don't see the "Loading..." message in the log, even after I attempt to access a wiki page containing [[TracNav]]. If I google "No macro or processor named 'TracNav' found", I get several other pages that seem to have similar problem. So, maybe it's a fairly common problem. Still looking for a solution. Thanks, Philip --~--~-~--~~~---~--~~ You received this message because you are subscribed to the Google Groups "Trac Users" group. To post to this group, send email to trac-users@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/trac-users?hl=en -~--~~~~--~~--~--~---
[Trac] Re: TracNav plugin on 0.10.4 on linux not working
On Jun 24, 12:00 pm, "Thomas Moschny" <[EMAIL PROTECTED]> wrote: > 2008/6/24 Philip <[EMAIL PROTECTED]>: > > > I deleted the trac.ini from /usr/local/share/trac/conf and added these > > 2 lines to the project-specific trac.ini file. After recycling web > > server, I still see no difference. Any other ideas? > > Maybe the plugin is already loaded? What happens if you put > [[TracNav]] on a wiki page? > > Why do you manually copy the egg after running ./setup.py install, > which already copies the egg into a system-wide dir? > > Do any other plugins work for you? > > Normally you should see a number of messages like this in your log: > 2008-05-31 11:15:24,220 Trac[loader] DEBUG: Loading egg plugin XXX > from /srv/trac/myproject/plugins/XXX-N.N-py2.5.egg > > - Thomas You are right -- I guess the install put the egg into /usr/local/lib64/ python2.4/site-packages. I deleted the egg file from /usr/local/share/ trac/plugins and restarted web server. I don't automatically see a mesage like you described when I recycle apache2, BUT if I look at the log, I do see one from earlier in the day yesterday. So, sounds like it was loading at some point anyway. However, when I go to a wiki page that has [[TracNav]] on it, I see this message: Error: Failed to load processor TracNav No macro or processor named 'TracNav' found Philip --~--~-~--~~~---~--~~ You received this message because you are subscribed to the Google Groups "Trac Users" group. To post to this group, send email to trac-users@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/trac-users?hl=en -~--~~~~--~~--~--~---
[Trac] Re: TracNav plugin on 0.10.4 on linux not working
On Jun 23, 5:20 pm, "Thomas Moschny" <[EMAIL PROTECTED]> wrote: > 2008/6/23 Philip <[EMAIL PROTECTED]>: > > > Trying to get the TracNav plugin to work on linux (zLinux, actually), > > but it's not showing the navigation or giving any messages in the > > log. > > [...] > > - there was no /usr/local/share/trac/conf/trac.ini file, so I created > > one, and added the following: > > [components] > > tracnav.* = enabled > > A globally installed plugin must be enabled separately for each trac > project. So you have to add these lines to the appropriate project's > trac.ini. Alternatively, you could use the WebAdmin plugin's gui to > enable other plugins. > > - Thomas I deleted the trac.ini from /usr/local/share/trac/conf and added these 2 lines to the project-specific trac.ini file. After recycling web server, I still see no difference. Any other ideas? Thanks, Philip --~--~-~--~~~---~--~~ You received this message because you are subscribed to the Google Groups "Trac Users" group. To post to this group, send email to trac-users@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/trac-users?hl=en -~--~~~~--~~--~--~---
[Trac] TracNav plugin on 0.10.4 on linux not working
Trying to get the TracNav plugin to work on linux (zLinux, actually), but it's not showing the navigation or giving any messages in the log. Here's what I've done: - downloaded TracNav-3.92.zip (8812 bytes) from http://svn.ipd.uka.de/trac/javaparty/attachment/wiki/TracNav/DownloadAndInstall/TracNav-3.92.zip - uploaded to server - unzipped - cd TracNav-3.92 - ran install... python setup.py install running install Checking .pth file support in /usr/local/lib64/python2.4/site- packages/ /usr/bin/python -E -c pass TEST PASSED: /usr/local/lib64/python2.4/site-packages/ appears to support .pth files running bdist_egg running egg_info writing TracNav.egg-info/PKG-INFO writing top-level names to TracNav.egg-info/top_level.txt writing dependency_links to TracNav.egg-info/dependency_links.txt writing entry points to TracNav.egg-info/entry_points.txt reading manifest file 'TracNav.egg-info/SOURCES.txt' writing manifest file 'TracNav.egg-info/SOURCES.txt' installing library code to build/bdist.linux-s390x/egg running install_lib running build_py creating build creating build/lib creating build/lib/tracnav copying tracnav/__init__.py -> build/lib/tracnav copying tracnav/tracnav.py -> build/lib/tracnav creating build/lib/tracnav/htdocs creating build/lib/tracnav/htdocs/css copying tracnav/htdocs/css/tracnav.css -> build/lib/tracnav/htdocs/css creating build/bdist.linux-s390x creating build/bdist.linux-s390x/egg creating build/bdist.linux-s390x/egg/tracnav copying build/lib/tracnav/__init__.py -> build/bdist.linux-s390x/egg/ tracnav copying build/lib/tracnav/tracnav.py -> build/bdist.linux-s390x/egg/ tracnav creating build/bdist.linux-s390x/egg/tracnav/htdocs creating build/bdist.linux-s390x/egg/tracnav/htdocs/css copying build/lib/tracnav/htdocs/css/tracnav.css -> build/bdist.linux- s390x/egg/tracnav/htdocs/css byte-compiling build/bdist.linux-s390x/egg/tracnav/__init__.py to __init__.pyc byte-compiling build/bdist.linux-s390x/egg/tracnav/tracnav.py to tracnav.pyc creating build/bdist.linux-s390x/egg/EGG-INFO copying TracNav.egg-info/PKG-INFO -> build/bdist.linux-s390x/egg/EGG- INFO copying TracNav.egg-info/SOURCES.txt -> build/bdist.linux-s390x/egg/ EGG-INFO copying TracNav.egg-info/dependency_links.txt -> build/bdist.linux- s390x/egg/EGG-INFO copying TracNav.egg-info/entry_points.txt -> build/bdist.linux-s390x/ egg/EGG-INFO copying TracNav.egg-info/top_level.txt -> build/bdist.linux-s390x/egg/ EGG-INFO zip_safe flag not set; analyzing archive contents... creating dist creating 'dist/TracNav-3.92-py2.4.egg' and adding 'build/bdist.linux- s390x/egg' to it removing 'build/bdist.linux-s390x/egg' (and everything under it) Processing TracNav-3.92-py2.4.egg Removing /usr/local/lib64/python2.4/site-packages/TracNav-3.92- py2.4.egg Copying TracNav-3.92-py2.4.egg to /usr/local/lib64/python2.4/site- packages TracNav 3.92 is already the active version in easy-install.pth Installed /usr/local/lib64/python2.4/site-packages/TracNav-3.92- py2.4.egg Processing dependencies for TracNav==3.92 Finished processing dependencies for TracNav==3.92 - copied dist/TracNav-3.92-py2.4.egg to /usr/local/share/trac/plugins/ TracNav-3.92-py2.4.egg - changed permissions to 777 for /usr/local/share/trac/plugins/ TracNav-3.92-py2.4.egg - there was no /usr/local/share/trac/conf/trac.ini file, so I created one, and added the following: [components] tracnav.* = enabled - changed permissions to 777 for /usr/local/share/trac/conf/trac.ini - recycled web server... /etc/init.d/apache2 restart - accessed project wiki page in web browser and no tracnav navigation appears - checked log file for specific project, and only shows this: 2008-06-23 19:34:26,736 Trac[api] DEBUG: Updating wiki page index Obviously, I'm missing something here (though no idea what). I have multiple projects, so I want to enable this TracNav plugin globally. Please help! Thanks in advance. --~--~-~--~~~---~--~~ You received this message because you are subscribed to the Google Groups "Trac Users" group. To post to this group, send email to trac-users@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/trac-users?hl=en -~--~~~~--~~--~--~---
[Trac] Re: Genshi0.5 and mod_python
On Apr 28, 1:05 pm, Noah Kantrowitz <[EMAIL PROTECTED]> wrote: > On Apr 28, 2008, at 2:37 PM, Philip Cooper wrote: > > Well now the httpd log claims it was denied permission to the new > > (correct) cache location. Even after I set the permissions to world > > writeable (777)! > > > The only other thing I can think of is to try to run a pylons app > > (with Genshi 0.5) through mod_python or try mod_wsgi. > > Sounds like a different, but similar error. The quick fix is to > install unzipped. Set zip_safe=False in setup.py. > Thanks Noah. I'll run that way for a while and report back if I do the "pylons test" That might point the finger away from trac and to the real culprit. On Apr 28, 1:15 pm, John Hampton <[EMAIL PROTECTED]> wrote: > Philip Cooper wrote: > > If anyone out there gets a mod_python/trac/Genshi0.5 working ... give > > a shout. > > I have gotten mod_python/trac/Genshi0.5 working fine with a non-default > egg_cache location. > Thanks John is that a linux machine? -- Phil --~--~-~--~~~---~--~~ You received this message because you are subscribed to the Google Groups "Trac Users" group. To post to this group, send email to trac-users@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/trac-users?hl=en -~--~~~~--~~--~--~---
[Trac] Re: Genshi0.5 and mod_python
I've recreated this error on a test machine. (with setuptools 0.6c8) The apache error log kept pointing the the egg cache of whomever started httpd. I thought I'd get a quick workaround by logging in as root, setting the PYTHON_EGG_CACHE, exporting it and then starting httpd. Well now the httpd log claims it was denied permission to the new (correct) cache location. Even after I set the permissions to world writeable (777)! The only other thing I can think of is to try to run a pylons app (with Genshi 0.5) through mod_python or try mod_wsgi. If anyone out there gets a mod_python/trac/Genshi0.5 working ... give a shout. -- Phil --~--~-~--~~~---~--~~ You received this message because you are subscribed to the Google Groups "Trac Users" group. To post to this group, send email to trac-users@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/trac-users?hl=en -~--~~~~--~~--~--~---
[Trac] Re: Genshi0.5 and mod_python
Noah thanks for the info. I'm not there yet. > > Do you know what the issue/resolution (ie trac changeset or ticket) > > was? > > http://trac.edgewall.org/changeset/6871 > I have mod_python 3.2.8 and have a system with lib64/python2.4/site- packages. Dam, lib64 dual site-packages bugs are over my head. An abbreviated version of the error / stacktrace is (all from frontend.py): PythonHandler trac.web.modpython_frontend: File "/usr/lib64/ python2.4/site-packages/mod_python/apache.py", line 287, in HandlerDispatch\n log=debug) PythonHandler trac.web.modpython_frontend: File "/usr/lib64/ python2.4/site-packages/mod_python/apache.py", line 464, in import_module\nmodule = imp.load_module(mname, f, p, d) then: /trac/web/__init__.py", line 1, from trac.web.api import * /trac/web/api.py", line 29 from trac.util import get_last_traceback /trac/util/__init__.py", line 33, from trac.util.html import escape, unescape, Markup, Deuglifier /trac/util/html.py", line 16, from genshi import Markup, escape, unescape Looks like importing mod_python fires line 287 then link 464 in apache.py At that point a load module is fired which cascades to an import genshi failure Can't track this down on my production machine. I'll have to wait to reproduce the errors on a similarly configured and versioned machine. -- Phil --~--~-~--~~~---~--~~ You received this message because you are subscribed to the Google Groups "Trac Users" group. To post to this group, send email to trac-users@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/trac-users?hl=en -~--~~~~--~~--~--~---
[Trac] Re: Genshi0.5 and mod_python
> Noah Kantrowitz wrote: > > Philip Cooper wrote: > >> problems with Genshi 0.5. > >> I can run trac with tracd but my normal usage is with mod_python. > >> Apache now throws an error that > >>ExtractionError: Can't extract file(s) to egg cache > > > This was fixed on trunk last week, please upgrade. > > Trac trunk that is, not Genshi. > Thanks for the quick response I updated Trac and have: * Trac r6895 * Genshi r847 *setuptools 0.6c7 Same error. Do you know any more about the original problem and/or fix? -- Phil --~--~-~--~~~---~--~~ You received this message because you are subscribed to the Google Groups "Trac Users" group. To post to this group, send email to trac-users@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/trac-users?hl=en -~--~~~~--~~--~--~---
[Trac] Re: Genshi0.5 and mod_python
On Apr 27, 11:22 am, Noah Kantrowitz <[EMAIL PROTECTED]> wrote: > Noah Kantrowitz wrote: > > Philip Cooper wrote: > >> problems with Genshi 0.5. > >> I can run trac with tracd but my normal usage is with mod_python. > >> Apache now throws an error that > >>ExtractionError: Can't extract file(s) to egg cache > > > This was fixed on trunk last week, please upgrade. > > Trac trunk that is, not Genshi. > Thanks for the quick response. Tried: Trac as of r6895. Genshi as of r847 setuptools as of 6.7 same behavior. Do you know what the issue/resolution (ie trac changeset or ticket) was? -- Phil --~--~-~--~~~---~--~~ You received this message because you are subscribed to the Google Groups "Trac Users" group. To post to this group, send email to trac-users@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/trac-users?hl=en -~--~~~~--~~--~--~---
[Trac] Genshi0.5 and mod_python
I have not figured this one out yet but I saw a previous post about problems with Genshi 0.5. My issue is, to gain some transform functionality I have upgraded to Genshi 0.5 (dev). I can run trac with tracd but my normal usage is with mod_python. Apache now throws an error that ExtractionError: Can't extract file(s) to egg cache The following error occurred while trying to extract file(s) to the Python egg cache: [Errno 13] Permission denied: '/home/phil/.python-eggs' The Python egg cache directory is currently set to: /home/phil/.python-eggs\n\n Problem is that I have set SetEnv PYTHON_EGG_CACHE /var/cache/trac in my httpd.conf the call path seems to be: mod_python calls trac trac eventually imports from genshi genshi calls speedups.py (some c speedups) when setuptools (0.6c7) gets around to looking for the module it's not getting the right stuff. This will probably become a more prevelant error as time goes on. I'll update if i find a workaround but for now I'll have drop out of mod_python in to another mode. -- Phil --~--~-~--~~~---~--~~ You received this message because you are subscribed to the Google Groups "Trac Users" group. To post to this group, send email to trac-users@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/trac-users?hl=en -~--~~~~--~~--~--~---
[Trac] Re: Which plugins for displaying LaTEx in Trac?
I wrote one you can find at: http://www.openvest.com/trac/wiki/LatexEquation That page also has a link to another. Mine is a plugin socopy one .py file to your project plugin dir and you are done! It's pretty simple but works 4 me. One that handles more than equations would be cool...being able to draw picts with all of the latex graphics extension If you have an openid you can login to my site and play with it in the Sandbox: http://www.openvest.com/trac/wiki/SandBox...just b nice :-) -- Phil On Jan 19, 12:15 pm, Kamil Kisiel <[EMAIL PROTECTED]> wrote: > I didn't find any of them satisfactory either. > > > It seems that hre are 3 plugins for displaying LateX formula's in Trac: > > > TracMath > > Trac2Latex > > LatexFormula > --~--~-~--~~~---~--~~ You received this message because you are subscribed to the Google Groups "Trac Users" group. To post to this group, send email to trac-users@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/trac-users?hl=en -~--~~~~--~~--~--~---
[Trac] Re: SyntaxHighliting
dgh Missed it by that much!. Hacking the api.py did NOT work. I had also tried the pygments_modes in trac.ini: [mimeviewer] pygments_modes = application/x-turtle:n3:7 after your post, I added: [mimeviewer] mime_map = <...otherstuff...> ,application/x-turtle:n3 Didn't see anything in the ticket http://trac.edgewall.org/ticket/4523 that would tell me I need BOTH of th ini lines above to make it work ... but that's what it takes. Now I'll poke around my 10.4 version ... or maybe it's time for that to get upgraded. Thanks, -- Phil On Feb 27, 6:42 pm, "Emmanuel Blot" <[EMAIL PROTECTED]> wrote: > > Check outhttp://trac.edgewall.org/ticket/4523for an exemple about > how to map a file extension to a specific mimetype, without editing > the api.py file. > > Cheers, > Manu --~--~-~--~~~---~--~~ You received this message because you are subscribed to the Google Groups "Trac Users" group. To post to this group, send email to trac-users@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/trac-users?hl=en -~--~~~~--~~--~--~---
[Trac] SyntaxHighliting
I've spent a rediculous amount of time trying to something that should take 30 seconds--I think. I have written a Pygments lexer for a language not previously supported. I have run _mappers.py from the pygments lexers subdir. I have tested it and can generate output with the pygmentize script How do I get it to work in Trac? I have a .10.4 and .11 versions and no luck. I thought .11 would be easy since Pygments is the default. I've tried adding the language (n3/turtle) to api.py in the mimeview subdir and that gets me code in my pages without the coloring. I've even tried creating a WikiProcessor in the plugins dir and have incrementally added more code to get it to work but I've put WAY too much into the processor file for this to be right. How do you add a new language that will be rendered by Pygments? -- Phil --~--~-~--~~~---~--~~ You received this message because you are subscribed to the Google Groups "Trac Users" group. To post to this group, send email to trac-users@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/trac-users?hl=en -~--~~~~--~~--~--~---
[Trac] Re: Trac Google Groups woes....
Everything working fine here (Mac, using Mail). /Philip On Oct 5, 2006, at 7:58 AM, Christian Boos wrote: > > Hi, > > Since a few days, I noticed that I get only a fraction from the > mails I > should normally get from the traffic on Google groups... This affects > all my groups (genshi, genshi-tickets, trac-users, trac-dev, > trac-tickets). On average, only 1 out 5 mails makes it to my mail box. > Worst example: I got only one mail from the recent "Trac Add comments > macro" thread, which contains currently 11 mails... > > Am I the only one to see this? > > Otherwise, I guess I should take the habit of reading my mails through > Google Group's web interface, but then, I don't really get how it > detects the read vs. unread mails... > > Btw, I just noticed the Google Groups BETA interface (1), now that's > nice ;) > > -- Christian > > (1) - http://groups-beta.google.com/group/trac-users?lnk=li > > > --~--~-~--~~~---~--~~ You received this message because you are subscribed to the Google Groups "Trac Users" group. To post to this group, send email to trac-users@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/trac-users -~--~~~~--~~--~--~---
