<?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:"v2"</description><language>en</language><lastBuildDate>Thu, 20 Aug 2015 22:07:57 -0000</lastBuildDate><item><title>New extensions point for custom project admin pages</title><link>https://forge-allura.apache.org/p/allura/tickets/4818/</link><description>There should be an extension mechanism for admin pages that show up in the admin sidebar for all projects.  Currently we are limited to making custom settings in a tool.  That requires the user to install the "foobar" tool just to get to foobar options.  It should be accessible immediately.

The extension mechanism eventually will need deployment configuration options to control which projects get the custom menu item &amp; page.  For example, only a certain neighborhood; not neighborhood --init-- projects, etc.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Dave Brondsema</dc:creator><pubDate>Thu, 20 Aug 2015 22:07:20 -0000</pubDate><guid>https://forge-allura.apache.org/p/allura/tickets/4818/</guid></item><item><title>Make activity_url for comments show them in context</title><link>https://forge-allura.apache.org/p/allura/tickets/4669/</link><description>Not sure if this is possible or not, but worth investigating since the "lone comment on the page" default is not very useful.

&gt; Cory Johns
&gt; It'd be nice if the URL for a ticket comment showed you the comment in the context of the ticket, but I think that might be hard to do, since it might be a comment on a ticket or wiki or, etc, and might not be on the first page

&gt; Tim Van Steenburgh
&gt; might be doable since you can get the artifact that a comment is attached to

&gt; Cory Johns
&gt; There's still the issue of paging.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Tim Van Steenburgh</dc:creator><pubDate>Thu, 20 Aug 2015 22:07:10 -0000</pubDate><guid>https://forge-allura.apache.org/p/allura/tickets/4669/</guid></item><item><title>Improve activity listing for Discussion</title><link>https://forge-allura.apache.org/p/allura/tickets/4661/</link><description>When a topic is created in Discussion, it creates an activity saying that a comment was added to the forum (not the topic).  Additionally, all comment activities are against the entire forum, not the individual topics.

There should be a specific "topic created" activity.  It might also make sense to suppress the "added a comment" for the initial post on a topic, since you can't create a topic without creating an initial post, even if you leave it blank.

For active forums with lots of topics, having all comment activities listed against the entire forum will get confusing and useless very quickly.  The comments are on a specific topic and should be associated with that topic.  It would make sense for an aggregator to show "X posts in forum Y" for very active forums / users, but the activities themselves should be against the most specific container.

Replies should be presented as such as well, though they should probably still target the topic.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Cory Johns</dc:creator><pubDate>Tue, 14 Apr 2015 14:14:58 -0000</pubDate><guid>https://forge-allura.apache.org/p/allura/tickets/4661/</guid></item><item><title>Specify activitystream mongo connections</title><link>https://forge-allura.apache.org/p/allura/tickets/4553/</link><description>Our .ini files need to specify activitystream connections.  Specifically, test.ini needs to use mim:// and production needs to not be localhost.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Dave Brondsema</dc:creator><pubDate>Thu, 20 Aug 2015 22:07:56 -0000</pubDate><guid>https://forge-allura.apache.org/p/allura/tickets/4553/</guid></item><item><title>Tweaks to scm getting started text</title><link>https://forge-allura.apache.org/p/allura/tickets/4545/</link><description>* SVN needs a new section: importing an existing SVN repo.  Have a link to /p/PROJECT/admin/TOOL/importer/
* The message boxes have cursor:pointer.  Perhaps we need to change/refactor that SCSS class a bit.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Dave Brondsema</dc:creator><pubDate>Thu, 20 Aug 2015 22:07:57 -0000</pubDate><guid>https://forge-allura.apache.org/p/allura/tickets/4545/</guid></item><item><title>Runtime override to activity streams config flag</title><link>https://forge-allura.apache.org/p/allura/tickets/4544/</link><description>Determine and implement a way to test activity stream pages despite the config flag being off.  Is it as simple as installing the the tool on a project or two?  And a url param for user profiles?</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Dave Brondsema</dc:creator><pubDate>Thu, 20 Aug 2015 22:07:53 -0000</pubDate><guid>https://forge-allura.apache.org/p/allura/tickets/4544/</guid></item><item><title>Create activitystream package</title><link>https://forge-allura.apache.org/p/allura/tickets/4458/</link><description>A few things need to happen before we can merge tv/timelines down to dev (which needs to happen soon).

* build an activitystream package, get it added to pythontree, include it in our allura requirements
* add activtystream configuration to ini file(s)
* make sure there is no activitystream functionality exposed in the allura ui that we don't want to expose yet; wrap stuff with the activitystream.enabled config check if necessary</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Tim Van Steenburgh</dc:creator><pubDate>Thu, 20 Aug 2015 22:07:56 -0000</pubDate><guid>https://forge-allura.apache.org/p/allura/tickets/4458/</guid></item><item><title>Activity stream indexes - NEEDS ENSURE_INDEX</title><link>https://forge-allura.apache.org/p/allura/tickets/4452/</link><description>Figure out which indexes are needed, and how we're going to keep them up-to-date (since we're not using ming). </description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Tim Van Steenburgh</dc:creator><pubDate>Thu, 20 Aug 2015 22:07:10 -0000</pubDate><guid>https://forge-allura.apache.org/p/allura/tickets/4452/</guid></item><item><title>Run timeline aggregations in the background with taskd</title><link>https://forge-allura.apache.org/p/allura/tickets/4397/</link><description>Once the volume ramps up, we probably won't want to be doing timeline aggregations on demand if we can help it. Find good spots to fire off aggregations in the background using taskd, so that when an activitystream page is requested, the cached timeline can just be pulled from mongo w/o doing an aggregation.

For users, a good spot to do this might be on login. For projects, not sure...needs more thought.

Do we need to worry about two aggregations for the same node running at the same time?

</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Tim Van Steenburgh</dc:creator><pubDate>Thu, 20 Aug 2015 22:07:10 -0000</pubDate><guid>https://forge-allura.apache.org/p/allura/tickets/4397/</guid></item><item><title>Implement new class(es) for timeline retrieval/curation - NEEDS MONGO CMDS</title><link>https://forge-allura.apache.org/p/allura/tickets/4396/</link><description>Right now, the only way to get a timeline is to call `director.create_timeline()`, which does an actual aggregation. We need something higher-level than this. Something like:

    def get_timeline(node, page=0, limit=100, subject_only=True, check_perms=True):
        # do stuff ...
        return timeline

Things encapsulated by this method or the class that implements it:

* figuring our whether a new aggregation needs to happen and doing it if necessary, or else fetching a cached timeline
* filtering activities based on permissions
* filtering activities based on whether `subject_only=True` or not; if True, only return activities where node is the subject (actor) in the activity; this will be used for the Your Activity page
* converting raw activity dicts from mongo into nice object that are easy to use for rendering in the html templates</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Tim Van Steenburgh</dc:creator><pubDate>Thu, 20 Aug 2015 22:07:54 -0000</pubDate><guid>https://forge-allura.apache.org/p/allura/tickets/4396/</guid></item><item><title>Expose atom feed for activity streams</title><link>https://forge-allura.apache.org/p/allura/tickets/4395/</link><description>There's a spec here: http://activitystrea.ms/specs/atom/1.0/

We may need to flesh out the activitystream object model a bit to conform to the spec (not sure).

We'll want a feed link on any page that shows an activity stream.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Tim Van Steenburgh</dc:creator><pubDate>Thu, 20 Aug 2015 22:07:10 -0000</pubDate><guid>https://forge-allura.apache.org/p/allura/tickets/4395/</guid></item><item><title>switch activitystreams to use ming</title><link>https://forge-allura.apache.org/p/allura/tickets/4394/</link><description>Change activitystreams to use ming instead of pymongo.  This will make [#4452] more direct</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Tim Van Steenburgh</dc:creator><pubDate>Fri, 20 Mar 2015 14:03:07 -0000</pubDate><guid>https://forge-allura.apache.org/p/allura/tickets/4394/</guid></item><item><title>Make timeline aggregation smarter</title><link>https://forge-allura.apache.org/p/allura/tickets/4392/</link><description>Right now the aggregator will pull in all activities from connected Nodes indiscriminately. It should only pull in activities that have happened since the last aggregation for the given Node. This probably means adding a `last_timeline_aggregation` timestamp to each Node.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Tim Van Steenburgh</dc:creator><pubDate>Thu, 20 Aug 2015 22:07:55 -0000</pubDate><guid>https://forge-allura.apache.org/p/allura/tickets/4392/</guid></item><item><title>Create News Feed and Your Activity pages for user profile</title><link>https://forge-allura.apache.org/p/allura/tickets/4390/</link><description>Figure out how to work them into the current UI (replace main content with News Feed? links in sidebar?)

News Feed will be an aggregated view of all activities you're interested in. Your Activity will be a list of stuff you've done recently.

Pages can be blank for now until timeline rendering is fleshed out. Or, for the News Feed page you could use the simple prototype from the `timeline()` method in `Allura/allura/ext/user_profile/user_main.py` on forge:tv/timelines.

This should check the config setting for showing activity streams.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Tim Van Steenburgh</dc:creator><pubDate>Thu, 20 Aug 2015 22:07:56 -0000</pubDate><guid>https://forge-allura.apache.org/p/allura/tickets/4390/</guid></item><item><title>Expose, Follow, and Watch buttons in the UI</title><link>https://forge-allura.apache.org/p/allura/tickets/4388/</link><description>A simple `follow()` controller method is implemented in `Allura/allura/ext/user_profile/user_main.py` on forge:tv/timelines. Implement a similar `watch()` method for projects, and hook up the follow and watch actions to buttons in the UI (someplace that makes sense).</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Tim Van Steenburgh</dc:creator><pubDate>Thu, 20 Aug 2015 22:07:56 -0000</pubDate><guid>https://forge-allura.apache.org/p/allura/tickets/4388/</guid></item><item><title>Add director.create_activity() calls for more events</title><link>https://forge-allura.apache.org/p/allura/tickets/4385/</link><description>Any time an event happens that we want to appear in the activity stream, we need to call `director.create_activity(...)`. See the `save_ticket()` method in `tracker_main.py` on forge:tv/timelines for an example:

    director.create_activity(c.user, 'created', ticket, target=c.project)

See also FeedItem.add() lines for re-use/refactor</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Tim Van Steenburgh</dc:creator><pubDate>Thu, 20 Aug 2015 22:07:53 -0000</pubDate><guid>https://forge-allura.apache.org/p/allura/tickets/4385/</guid></item><item><title>Add ActivityObject mixin to Artifact classes</title><link>https://forge-allura.apache.org/p/allura/tickets/4384/</link><description>Any object that can participate in an activity needs to subclass from ActivityObject and implement an `activity_name` property. See the `Ticket` class on forge:tv/timelines for an example.

We'll probably want to do this for most, if not all of our Artifacts.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Tim Van Steenburgh</dc:creator><pubDate>Thu, 20 Aug 2015 22:07:53 -0000</pubDate><guid>https://forge-allura.apache.org/p/allura/tickets/4384/</guid></item><item><title>Getting started text for scms</title><link>https://forge-allura.apache.org/p/allura/tickets/4382/</link><description>For each of Git, SVN, and Hg we should have instructions for how to get started with a brand new (empty) repo, and for pushing existing code to the new repo.  Provide specific commands.  Show the text/commands on the repo page to developers/admins if the repo is empty (or r1 for SVN)</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Dave Brondsema</dc:creator><pubDate>Thu, 20 Aug 2015 22:07:54 -0000</pubDate><guid>https://forge-allura.apache.org/p/allura/tickets/4382/</guid></item><item><title>Remove unnecessary project lookups from app config setup</title><link>https://forge-allura.apache.org/p/allura/tickets/4358/</link><description>When building the top nav of a project, each app config entry queries mongo to get the project.  That isn't necessary since the project is already known.  This can be seen with the timing middleware.  Set production.ini stats.sample_rate to 1, and increase the logging:

~~~~
[logger_timermiddleware]
level = DEBUG
handlers =
qualname = timermiddleware
~~~~

And then:

`tail -F /var/log/allura/allura.log | grep ODMSession.*Project`</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Dave Brondsema</dc:creator><pubDate>Thu, 20 Aug 2015 22:07:53 -0000</pubDate><guid>https://forge-allura.apache.org/p/allura/tickets/4358/</guid></item><item><title>Build top nav using main_menu() instead of sitemap() methods</title><link>https://forge-allura.apache.org/p/allura/tickets/4357/</link><description>The top nav of a project is built using `project.sitemap()` which calls `sitemap()` for all the tools, which frequently call their `sidebar_menu` methods.  Those may do a lot of unnecessary extra work, since all that is needed is the top menu.

Each app should also define a `main_menu()` method.  Let's make sure all of those are defined properly in our apps, and then make the top nav render only use those.  Consider if `Project.sitemap` is used elsewhere - maybe it can be simplified or maybe a new method needs to be created.  Also see how `Project.menus` is used.</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/4357/</guid></item><item><title>Timermiddleware: categorize page types </title><link>https://forge-allura.apache.org/p/allura/tickets/4307/</link><description>Sfpy uses a regex to categorize page types for Timermiddleware.  Allura should categorize also.  Instead of using a regex, the current tool type should be used  (we do this for GA page type tagging too).</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Dave Brondsema</dc:creator><pubDate>Thu, 20 Aug 2015 22:07:53 -0000</pubDate><guid>https://forge-allura.apache.org/p/allura/tickets/4307/</guid></item><item><title>user-fields on ticket tracker function poorly [25856]</title><link>https://forge-allura.apache.org/p/allura/tickets/4299/</link><description>https://sourceforge.net/apps/trac/sourceforge/ticket/25856

&gt;the upgraded bug tracker's assigned-to field isn't any kind of drop-down selector, so I have to remember user names exactly to enter them in there. If I enter a nonexistent username, then it just silently fails to update. Then, after I enter a correct username, the ticket will display the assigned-to field as the name of the user, instead of their username, which is a bad idea, as it doesnt help me remember their usernames at all.

I agree with the user on both points.</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Chris Tsai</dc:creator><pubDate>Thu, 20 Aug 2015 22:07:56 -0000</pubDate><guid>https://forge-allura.apache.org/p/allura/tickets/4299/</guid></item><item><title>Attachments of content-type image/* should be displayed inline</title><link>https://forge-allura.apache.org/p/allura/tickets/4277/</link><description>It is safe to show image attachments directly, rather than serving them as downloads (like we must do for everything else, for security)</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Dave Brondsema</dc:creator><pubDate>Thu, 20 Aug 2015 22:07:57 -0000</pubDate><guid>https://forge-allura.apache.org/p/allura/tickets/4277/</guid></item><item><title>API to update ticket fields</title><link>https://forge-allura.apache.org/p/allura/tickets/4273/</link><description>We should have a REST API to update ticket milestones and values for select boxes (status, component, etc)</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Dave Brondsema</dc:creator><pubDate>Tue, 07 Apr 2015 02:47:16 -0000</pubDate><guid>https://forge-allura.apache.org/p/allura/tickets/4273/</guid></item><item><title>API for searching for tickets</title><link>https://forge-allura.apache.org/p/allura/tickets/4272/</link><description>It would be useful for the tracker's rest API to allow searching.  It should take the same input as /search/?q= already does.

A special case for milestones would be nice, although not necessary since you can search by _milestone:foo</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Dave Brondsema</dc:creator><pubDate>Thu, 20 Aug 2015 22:07:54 -0000</pubDate><guid>https://forge-allura.apache.org/p/allura/tickets/4272/</guid></item></channel></rss>