<?xml version="1.0" encoding="utf-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Ticket search results</title><link>https://forge-allura.apache.org/p/allura/tickets/</link><description>You searched for labels:"cleanup"</description><language>en</language><lastBuildDate>Mon, 17 May 2021 15:07:00 -0000</lastBuildDate><item><title>Fix pep8 and pyflakes violations</title><link>https://forge-allura.apache.org/p/allura/tickets/7980/</link><description>At this time, we have 502 pep8 violations (at line length 119) and 346 pyflakes violations.  Would be good to clean those up some.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Dave Brondsema</dc:creator><pubDate>Tue, 06 Oct 2015 10:35:25 -0000</pubDate><guid>https://forge-allura.apache.org/p/allura/tickets/7980/</guid></item><item><title>Clean up old project-level permission management page</title><link>https://forge-allura.apache.org/p/allura/tickets/7958/</link><description>A long time ago we made a new permissions page for projects, which combined group management and permission management.  It is at the '/groups' url, but we never removed the old project permission management page at '/permissions'

I came across it recently because in a subproject we still link to /permissions and not /groups

We should update links, and remove the /permissions page for projects.  (Take care that the /permissions pages for tools is not affected, it is still needed!)</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Dave Brondsema</dc:creator><pubDate>Thu, 16 Mar 2017 19:43:18 -0000</pubDate><guid>https://forge-allura.apache.org/p/allura/tickets/7958/</guid></item><item><title>Update beautifulsoup usage to v4</title><link>https://forge-allura.apache.org/p/allura/tickets/7952/</link><description>[#7947] introduced beautifulsoup4, which can be installed side-by-side with beautifulsoup.  We should migrate our existing usage of beatifulsoup 3 over to 4.  

http://www.crummy.com/software/BeautifulSoup/bs4/doc/#porting-code-to-bs4

Note that when I started using bs4 in [e0e2f0] I had to choose a parser to use.  I went with html5lib because it seemed to work better ('html.parser' had some strange `&lt;/li&gt;` behavior in a test).  But html5lib spits out html/head/body tags so we'll have to strip those out again.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Dave Brondsema</dc:creator><pubDate>Thu, 06 Aug 2015 21:42:42 -0000</pubDate><guid>https://forge-allura.apache.org/p/allura/tickets/7952/</guid></item><item><title>monq_task 'args' index NEEDS ENSURE_INDEX</title><link>https://forge-allura.apache.org/p/allura/tickets/7562/</link><description>Commit [462bcfb5] added a few indexes on the `monq_task` model collection.  One was removed later in [62f4b4fe] but the 'args' index still remains.  It's not used in any queries and can be quite large for big `monq_task` collections.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Dave Brondsema</dc:creator><pubDate>Thu, 20 Aug 2015 22:07:00 -0000</pubDate><guid>https://forge-allura.apache.org/p/allura/tickets/7562/</guid></item><item><title>API cleanup: rename `app.has_access` to indicate it is mail-specific </title><link>https://forge-allura.apache.org/p/allura/tickets/7326/</link><description>Almost all `has_access` usage is from helpers and security packages.  The `app.has_access` method is used for mail tasks:

`Allura/allura/tasks/mail_tasks.py:59:                    if not app.has_access(c.user, userpart):`

And documented in the first paragraph of https://forge-allura.apache.org/docs/platform_tour.html#email-integration

I think it should be named something different (and fall back to `has_access` for backwards compatibility)</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Dave Brondsema</dc:creator><pubDate>Mon, 09 Mar 2015 16:00:48 -0000</pubDate><guid>https://forge-allura.apache.org/p/allura/tickets/7326/</guid></item><item><title>Make index page better</title><link>https://forge-allura.apache.org/p/allura/tickets/7308/</link><description>We should remove from the index page the non-functional category browsing on the left sidebar (and all the supporting code behind it - `ProjectCategory` iirc).

It'd also be good to have some welcoming content about how neighborhoods work, and pointers to adding/removing them, etc.  Se also [#5943]</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Dave Brondsema</dc:creator><pubDate>Mon, 15 Feb 2016 18:31:52 -0000</pubDate><guid>https://forge-allura.apache.org/p/allura/tickets/7308/</guid></item><item><title>Deprecate subforums</title><link>https://forge-allura.apache.org/p/allura/tickets/7187/</link><description>Subforums aren't fully functional.  Seems like you can only add them via Admin's 'add' form and not via the other 'add' form from main sidebar.  And then there's no way to view subforums (except admin link) nor post a new topic to the subforum.  Only can edit an existing post to move it to a subforum.

We should remove them if we're not going to support/maintain them properly.  Forum.subforums also runs an unindexed query, scanning all forum documents.  Its called via a chain of forum widgets.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Dave Brondsema</dc:creator><pubDate>Fri, 21 Feb 2014 17:01:25 -0000</pubDate><guid>https://forge-allura.apache.org/p/allura/tickets/7187/</guid></item><item><title>Remove support for Neighborhoods having absolute url prefixes</title><link>https://forge-allura.apache.org/p/allura/tickets/6756/</link><description>See `Neighborhood.url()` and `Project.url()` implementations. They imply that a neighborhood url prefix can be absolute (by starting with `//`), but this doesn't seem useful at all. Seems like any useful application of this feature could be achieved with a web server. Furthermore, we have other code that assume the Neighborhood url prefix will always begin with a `/` (see `Ticket.email_address`).

Remove support for Neighborhood absolute urls, or provide a good reason to keep it.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Tim Van Steenburgh</dc:creator><pubDate>Sun, 30 Nov 2014 13:52:10 -0000</pubDate><guid>https://forge-allura.apache.org/p/allura/tickets/6756/</guid></item><item><title>__json__ should return plain dicts</title><link>https://forge-allura.apache.org/p/allura/tickets/6716/</link><description>All `__json__` methods should return "plain" data structures.  A particular example is that lists and dicts shouldn't be ming instrumented containers.  Then it'll always be safe to modify the results without side effects (i.e. changing the database).

Some sort of helper function seems like it would be useful.  It should preserve certain types (e.g. datetime, int, float, etc) since those aren't any risk and are converted to json just fine by TurboGears</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Dave Brondsema</dc:creator><pubDate>Thu, 20 Aug 2015 22:07:10 -0000</pubDate><guid>https://forge-allura.apache.org/p/allura/tickets/6716/</guid></item><item><title>Change has_access() handling of DENY</title><link>https://forge-allura.apache.org/p/allura/tickets/6715/</link><description>We currently have deny checks pretty early in the has_access() logic.  It would be better to have the ACL order be respected, and have a DENY return false.

Need to make sure that tool-level user blocks still work properly, as do private tickets and developer-only forums.

Make sure tests &amp; docstrings are updated, since this is important functionality.  See particularly `test_weird_allow_vs_deny`

`all_allowed()` will need to reflect these changes too</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Dave Brondsema</dc:creator><pubDate>Thu, 27 Nov 2014 01:11:08 -0000</pubDate><guid>https://forge-allura.apache.org/p/allura/tickets/6715/</guid></item><item><title>Rename &amp; move User.project_role() </title><link>https://forge-allura.apache.org/p/allura/tickets/6714/</link><description>The User.project_role method creates roles, which may not be what you want (and may cause a large number of roles to be added to mongo).  Rename it to something that makes its "upsert" functionality clear.

Also move it to the ProjectRole class, as that is more closely related to its functionality.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Dave Brondsema</dc:creator><pubDate>Thu, 20 Aug 2015 22:07:10 -0000</pubDate><guid>https://forge-allura.apache.org/p/allura/tickets/6714/</guid></item><item><title>Refactor Tracker Globals field into app.config.options</title><link>https://forge-allura.apache.org/p/allura/tickets/6582/</link><description>Most of the fields on ForgeTracker's Globals model would be better suited as options in the tracker's AppConfig.  Additionally, the unqualified collection name `"globals"` is confusing / misleading.

The bin count caches might need to stay on a separate collection, but that collection should be better named.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Cory Johns</dc:creator><pubDate>Wed, 12 Nov 2014 17:11:11 -0000</pubDate><guid>https://forge-allura.apache.org/p/allura/tickets/6582/</guid></item><item><title>Refactor custom fields</title><link>https://forge-allura.apache.org/p/allura/tickets/6190/</link><description>Symptoms:

* It's depressing how many different places the code has to change if you were to add a new type, for example
* `if` stmts around custom field types all over the place

The root problem I think is that custom fields are just dicts, so the logic ends up spread all over the code instead of encapsulated in the object.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Tim Van Steenburgh</dc:creator><pubDate>Sun, 08 Mar 2015 09:01:56 -0000</pubDate><guid>https://forge-allura.apache.org/p/allura/tickets/6190/</guid></item><item><title>Fix c (tmpl_context) and g (app_globals) imports</title><link>https://forge-allura.apache.org/p/allura/tickets/5837/</link><description>The `tmpl_context` and `app_globals` objects are no longer referenced as `c` and `g` in pylons (and haven't been for some time), but a lot of our code still tries to import them as such.  We have some code that imports them correctly and then monkey patches `c` and `g` as aliases in pylons, but it doesn't always work when running tests.

The imports should be cleaned up and changed to:

    from pylons import tmpl_context as c
    from pylons import app_globals as g

The bad imports can be found with:

    ack 'from pylons import ([cg]|c, *g|g, *c)$'</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Cory Johns</dc:creator><pubDate>Thu, 20 Aug 2015 22:07:57 -0000</pubDate><guid>https://forge-allura.apache.org/p/allura/tickets/5837/</guid></item><item><title>Factor out SourceForge-specific bits of Allura</title><link>https://forge-allura.apache.org/p/allura/tickets/4808/</link><description>grep for these keywords that are specific to the SourceForge environment:

* sourceforge
* sf.net
* sfx
* sftheme
* sfbot
* geek.net
* /var/local/env-allura/bin/activate
* soghost
* sb-forge

These should all turn into generic configuration options or moved into the SourceForge internal repo and hooked into extensions points.

We should also provide a way for 3rd-party macros to include their own help text, which would be displayed in the "Formatting Help" section of Allura.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Dave Brondsema</dc:creator><pubDate>Thu, 20 Aug 2015 22:07:10 -0000</pubDate><guid>https://forge-allura.apache.org/p/allura/tickets/4808/</guid></item><item><title>Stats logging should not go to the "console" handler; remove it NEEDS INI CHANGE</title><link>https://forge-allura.apache.org/p/allura/tickets/3938/</link><description>We use `h.log_action` in many places to log details about actions that happen.  In production.ini, `handler_stats` is configured to write those details to a separate log file.  But they still go to handler_console too.  To have them not go there, I think we need a the log_action events to go to a different base logger (e.g. allura.stats) instead of whatever package logger is there (e.g. allura.model.repo_refresh).  `class StatsHandler` filters by checking for an "action" attribute on the event, but we need a simple way to separate them so that it can be adjusted via the logging ini settings.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Dave Brondsema</dc:creator><pubDate>Mon, 17 May 2021 15:07:00 -0000</pubDate><guid>https://forge-allura.apache.org/p/allura/tickets/3938/</guid></item><item><title>Depend on zmq and not all of zarkov</title><link>https://forge-allura.apache.org/p/allura/tickets/3067/</link><description>Zarkov pulls in a lot of deps but Allura only needs the zarkov client which is a very thin wrapper around zmq.  We should just use zmq directly.  We can also remove the ImportError handling and always require zmq.  The ini config option can stay.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Dave Brondsema</dc:creator><pubDate>Thu, 20 Aug 2015 22:07:52 -0000</pubDate><guid>https://forge-allura.apache.org/p/allura/tickets/3067/</guid></item><item><title>Delete old scm logic - NEEDS MONGO CMD</title><link>https://forge-allura.apache.org/p/allura/tickets/3060/</link><description>After [#3059], delete the "old" repo data structures from repository.py.  Check carefully though, some stuff may still be used.  The new models is in repo.py.   Also identify the mongo collections that are unused so we can drop them.

The `scm.new_refresh` setting should become irrelevant after this.  Everything should be the "new" logic.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Dave Brondsema</dc:creator><pubDate>Thu, 20 Aug 2015 22:07:55 -0000</pubDate><guid>https://forge-allura.apache.org/p/allura/tickets/3060/</guid></item><item><title>Move thread-level subscription UI</title><link>https://forge-allura.apache.org/p/allura/tickets/2996/</link><description>With forum-level subscriptions working now, I'm thinking we should remove this thread-level subscription UI http://screencast.com/t/5KuVWoc16  It's ugly.  If we want to keep thread-level at all, we could have the standard subscribe icon in the header of an individual thread, but no mass subscribe UI.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Dave Brondsema</dc:creator><pubDate>Mon, 31 Aug 2015 21:58:03 -0000</pubDate><guid>https://forge-allura.apache.org/p/allura/tickets/2996/</guid></item><item><title>remove ProjectCategory code</title><link>https://forge-allura.apache.org/p/allura/tickets/2601/</link><description>"ProjectCategory" was replaced by trove category implementation.  The UI is gone, we should remove the code.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Dave Brondsema</dc:creator><pubDate>Mon, 12 Sep 2011 17:44:51 -0000</pubDate><guid>https://forge-allura.apache.org/p/allura/tickets/2601/</guid></item><item><title>Remove the home app code</title><link>https://forge-allura.apache.org/p/allura/tickets/2217/</link><description>After the script in [#1507] is run and the home app isn't used anywhere, delete it.  We should double-check neighborhoods and their home pages first.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Dave Brondsema</dc:creator><pubDate>Thu, 20 Aug 2015 22:07:55 -0000</pubDate><guid>https://forge-allura.apache.org/p/allura/tickets/2217/</guid></item><item><title>Move trove categories into a nested structure in the project document</title><link>https://forge-allura.apache.org/p/allura/tickets/2199/</link><description>Right now we have several properties in the Project document starting with trove_, and we have lots of places in the code where we generate the property name with string pasting/interpolation. Better would be to have a single property 'trove' with sub-properties for the various categories. Refactor the model and update the controllers to reflect this.

It also could be nice to store the trove id number, rather than an object id.  Not sure how important that is, or ramifications.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Rick Copeland</dc:creator><pubDate>Tue, 31 Jan 2012 18:21:09 -0000</pubDate><guid>https://forge-allura.apache.org/p/allura/tickets/2199/</guid></item><item><title>Combine allura and project-data dbs</title><link>https://forge-allura.apache.org/p/allura/tickets/1792/</link><description>Project info goes into `project-data` db.  Top-level collections are in `allura`.  Might be nice to have all share one.

The `globals` collection would conflict.  ForgeTracker has a `globals` on `project-data` and ForgeChat has a `globals` on `allura`.  Both should be named better, "globals" is way too generic.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Dave Brondsema</dc:creator><pubDate>Thu, 30 Oct 2014 16:53:32 -0000</pubDate><guid>https://forge-allura.apache.org/p/allura/tickets/1792/</guid></item><item><title>MonQTask.get code cleanup</title><link>https://forge-allura.apache.org/p/allura/tickets/1780/</link><description>Change ming to use pymongo 1.10's Collection.find_and_modify command.  That returns None instead of raising an OperationFailure, so MonQTask.get can handle that more cleanly.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Dave Brondsema</dc:creator><pubDate>Sun, 08 Mar 2015 12:42:05 -0000</pubDate><guid>https://forge-allura.apache.org/p/allura/tickets/1780/</guid></item><item><title>Core tests should not require all the tools to be installed</title><link>https://forge-allura.apache.org/p/allura/tickets/1742/</link><description>ForgeTracker, ForgeHg, etc, and even their respective dependencies are required just to run a core test.  Only Allura and AlluraTesting should be required, so it's easier for people to develop locally without setting up the scm deps.

~~~~
  File "/Users/brondsem/sf/forge/AlluraTesting/alluratest/controller.py", line 51, in setup_functional_test
    setup_basic_test(config, app_name)
  File "/Users/brondsem/sf/forge/AlluraTesting/alluratest/controller.py", line 46, in setup_basic_test
    cmd.run([test_file])
  File "/Users/brondsem/sf/forge_env/lib/python2.6/site-packages/paste/script/appinstall.py", line 68, in run
    return super(AbstractInstallCommand, self).run(new_args)
  File "/Users/brondsem/sf/forge_env/lib/python2.6/site-packages/paste/script/command.py", line 218, in run
    result = self.command()
  File "/Users/brondsem/sf/forge_env/lib/python2.6/site-packages/paste/script/appinstall.py", line 456, in command
    self, config_file, section, self.sysconfig_install_vars(installer))
  File "/Users/brondsem/sf/forge_env/lib/python2.6/site-packages/paste/script/appinstall.py", line 598, in setup_config
    mod.setup_app, command, filename, section, vars)
  File "/Users/brondsem/sf/forge_env/lib/python2.6/site-packages/paste/script/appinstall.py", line 612, in _call_setup_app
    func(command, conf, vars)
  File "/Users/brondsem/sf/forge/Allura/allura/websetup/__init__.py", line 19, in setup_app
    bootstrap.bootstrap(command, conf, vars)
  File "/Users/brondsem/sf/forge/Allura/allura/websetup/bootstrap.py", line 65, in bootstrap
    wipe_database()
  File "/Users/brondsem/sf/forge/Allura/allura/websetup/bootstrap.py", line 205, in wipe_database
    flyway.run(['--force', '-u', 'mim:///'+db.name])
  File "/Users/brondsem/sf/forge_env/lib/python2.6/site-packages/paste/script/command.py", line 218, in run
    result = self.command()
  File "/Users/brondsem/sf/forge_env/lib/python2.6/site-packages/flyway/command.py", line 31, in command
    self._load_migrations()
  File "/Users/brondsem/sf/forge_env/lib/python2.6/site-packages/flyway/command.py", line 89, in _load_migrations
    reload(ep.load())
  File "build/bdist.linux-i686/egg/pkg_resources.py", line 1954, in load
    entry = __import__(self.module_name, globals(),globals(), ['__name__'])
  File "/Users/brondsem/sf/forge/Allura/allura/migrations.py", line 15, in &lt;module&gt;
    from forgesvn import model as SVNM
  File "/Users/brondsem/sf/forge/ForgeSVN/forgesvn/model/__init__.py", line 1, in &lt;module&gt;
    from svn import Repository
  File "/Users/brondsem/sf/forge/ForgeSVN/forgesvn/model/svn.py", line 12, in &lt;module&gt;
    import pysvn
ImportError: No module named pysvn
~~~~</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Dave Brondsema</dc:creator><pubDate>Thu, 05 Feb 2015 01:53:10 -0000</pubDate><guid>https://forge-allura.apache.org/p/allura/tickets/1742/</guid></item></channel></rss>