12 Apr, 2010
1 commit
09 Apr, 2010
2 commits
-
(ActionItem1474)
08 Apr, 2010
5 commits
-
This commit introduces a lot of changes in the HTML generated by Nosfero with the intent of making it load faster on users' browsers. This includes: * Adding apache configuration for compression of text content and aggresive cache policy (caching assets for 1 month) * Drastically reducing the number of stylesheets and javascript files referenced in the HTML. * The remaining CSS and Javascript files are now cached in a single file (one CSS file and one Javascript file) so that clients will only request a single file instead of several ones (and thus requiring a single HTTP connection instead of several ones). * Making sure that all assets referenced in the code have a timestamp (?21083219837) in the end of the URL so that when they are updated the client will fetch them again. During the course of the implemeantion, some other changes were made: * Updated INSTALL file with documentation about apache configuration for production environments. * Marked as deprecated all methods that use stylesheet_import, as well as stylesheet_import itself. * Removed "better browser promotion" code. It sucks. * Unifyed Noosfero stylesheets) * Unifyed stylesheets in the base theme * Enhanced the sample-data script with the creation of random events. in a range of 60 days around the current date, so that there will be events in the previous month, in the current month and in the next month. * Removed unused cms.js * Updated .gitignore example to ignore cached CSS and Javascript files. * Updated git-upgrade script to remove cached CSS and Javascript files. (ActionItem1425)
07 Apr, 2010
2 commits
-
This way the output of our green CI builds gets better.
-
It seems that the problem happens when selenium steps run before non-selenium steps (e.g. settind up data for the test). Doing this also seems to remove the need for the sleeps added to avoid the timeout.
05 Apr, 2010
6 commits
-
Besides being faster, consumming less memory, and being thread-safe, fast_gettext's approach is cleaner than Ruby-GetText's because it does not mess with the Rails internals. That's probably due to the fact that fast_gettext was designed after Rails had proper I18N support, so that's not exactly Ruby-GetText's fault. Current versions of Ruby-GetText are claimed to be thread-safe as well, but I decided to go with fast_gettext regardless. I am messing with the Rails internals myself by copying some code from Ruby-Gettext, but that code will be dropped when we upgrade to a more recent Rails version with proper I18N. Code was copied from Ruby-GetText to implement: * per-language cache * validation error messages translation During initialization, the needed .mo files installed system-wide are symlinked locally. By doing this we can take "similar" locales locally since fast_gettext does not seem to support loading of files from similar locales (e.g. loading pt_BR/LC_MESSAGES/domain.mo when pt/LC_MESSAGES/domain.mo is not available). This hopefully will fix the long-standing bug with messed up translations due to high concurrency and non-thread-safety of the version of Ruby-GetText in Debian Lenny. (ActionItem1315) -
(ActionItem1315)
-
They are need in DocControlller tests as well in order to not depend on the docs being built during tests
03 Apr, 2010
3 commits
31 Mar, 2010
1 commit
30 Mar, 2010
1 commit
-
This reverts commit d0f8d245c7304e20f86be76e84de59e62f30c889.
29 Mar, 2010
5 commits
-
Everybody loves dots. This way the upgrade process gets less verbose, and we can actually follow what's going on.
26 Mar, 2010
11 commits
-
(ActionItem1451)
-
Using a constant instead of a method call that instantiates that array at every call; besides, using symbols instead of strings.
-
This way they will actually be displayed using the environment's theme instead of the theme called "default". In multi-environment setups the environments do not use the "default" theme. (and hell yeah from now on we are going to use timestamp in migration file names) (ActionItem1432)
-
* Profile theme is nil by default. * Profile#theme was removed. (ActionItem1432)
-
* Used xss_terminate with the option 'white_list' that only filters the bad html codes. (ActionItem1437)
-
* Checking if each children of the folder should be displayed to the current user by using the 'display_to?' method. * Moving some test from the folder test to the folder_helper test. (ActionItem1396)
-
* This patch will centralize the logic of public/private article in the 'published' attribute. * Removing the public_article attribute from the code replacing it by published. * Creating two new migrations: 1. Setting the published attribute to false when the article has the public_article attribute false. This will grant that all the articles that aren't public now continue unpublished. 2. Removing the public_article attribute from the schema.
24 Mar, 2010
3 commits
-
(ActionItem1449)
-
(ActionItem1448)