summaryrefslogtreecommitdiffstats
AgeCommit message (Collapse)AuthorFilesLines
2012-11-27Don't cache package properties as aggressivelyDan McGee2-6/+3
For package signatures, it turns out it is way cheaper to just parse the signature again rather than going though all the decorator and cache_function_key business. This speeds up the mismatched signatures report significantly once this is removed. For base_package, given that we only call it once from our package details template, it makes little sense to cache the result. Signed-off-by: Dan McGee <dan@archlinux.org>
2012-11-21Fix new magnet link generationDan McGee1-8/+7
Apparently clients don't like urlencoded values in the magnet link, so %3A isn't treated the same as ':'. Signed-off-by: Dan McGee <dan@archlinux.org>
2012-11-21Bump django-countries requirementDan McGee2-2/+2
Signed-off-by: Dan McGee <dan@archlinux.org>
2012-11-21Fix mirror URL duplication in status viewDan McGee1-1/+2
We need to ensure we don't duplicate URLs in the status view, so add a distinct() call back in to the queryset when it was inadvertently dropped in commit a2cfa7edbb. This negates a lot of the performance gains we had, unfortunately, so it looks like a nested subquery might be more efficient. Disappointing the planner can't do this for us. Signed-off-by: Dan McGee <dan@archlinux.org>
2012-11-21Minor download template tweaksDan McGee1-10/+6
Signed-off-by: Dan McGee <dan@archlinux.org>
2012-11-21Show release notes on downloads pageDan McGee2-0/+12
Signed-off-by: Dan McGee <dan@archlinux.org>
2012-11-21Move some logic out of the templates to the Release modelDan McGee2-8/+26
This includes magnet URI generation, ISO paths, etc. Signed-off-by: Dan McGee <dan@archlinux.org>
2012-11-21Add Release model to relengDan McGee5-24/+167
This should prevent the need for monthly template updates from Pierre and Thomas; best to just let them enter the data themselves and have it show up on the website. Signed-off-by: Dan McGee <dan@archlinux.org>
2012-11-21Improve performance of todolists queryDan McGee1-15/+26
Use some standard SQL and split the query into two different parts to save a lot of unnecessary sorting and field retrieval at the database level. The `CASE WHEN complete THEN 1 ELSE 0 END` syntax should be accepted by any database that implements proper SQL; it was tested in PostgreSQL and sqlite3 without issues. Signed-off-by: Dan McGee <dan@archlinux.org>
2012-11-16Use python set comprehension syntax supported in 2.7Dan McGee9-18/+18
Signed-off-by: Dan McGee <dan@archlinux.org>
2012-11-16Use Python 2.7 dictionary comprehension syntaxDan McGee8-22/+16
Rather than the old idiom of dict((k, v) for <> in <>). Signed-off-by: Dan McGee <dan@archlinux.org>
2012-11-16Bump several virtualenv requirementsDan McGee2-7/+7
Signed-off-by: Dan McGee <dan@archlinux.org>
2012-11-16Optimize mirror status data fetchingDan McGee1-8/+16
Now that we have as many mirror URLs as we do, we can do a better job fetching and aggregating this data. The prior method resulted in a rather unwieldy query being pushed down to the database with a horrendously long GROUP BY clause. Instead of trying to group by everything at once so we can retrieve mirror URL info at the same time, separate the two queries- one for getting URL performance data, one for the qualitative data. The impetus behind fixing this is the PostgreSQL slow query log in production; this currently shows up the most of any queries we run in the system. Signed-off-by: Dan McGee <dan@archlinux.org>
2012-11-13Move JSON keys view to public/ appDan McGee5-50/+58
This seems like a more appropriate place, and now the visualization is done here anyway so we should move the data backing it. Signed-off-by: Dan McGee <dan@archlinux.org>
2012-11-13Update rss.pngDan McGee1-0/+0
Signed-off-by: Dan McGee <dan@archlinux.org>
2012-11-13Move PGP key visualizations to master keys pageDan McGee2-7/+21
Signed-off-by: Dan McGee <dan@archlinux.org>
2012-11-13Mirror details style and JS cleanupDan McGee2-3/+13
Signed-off-by: Dan McGee <dan@archlinux.org>
2012-11-13Slight master key template adjustmentDan McGee1-0/+4
This is in anticipation of moving the visualization stuff to this page rather than grouped with unrelated things. Signed-off-by: Dan McGee <dan@archlinux.org>
2012-11-13Prefetch the available protocols on the mirror overview pageDan McGee3-6/+9
Otherwise we are doing one query per mirror, which at this point is over 100 separate queries. Signed-off-by: Dan McGee <dan@archlinux.org>
2012-11-12Remove SevenL from sponsors listDan McGee3-19/+0
They've been good to us for several years, but our hardware requirements have changed and we've moved on to using different machines for the purposes this donated machine served. Thanks! Signed-off-by: Dan McGee <dan@archlinux.org>
2012-11-12Add a test high-resolution imageDan McGee1-0/+0
Just the simple RSS feed image for now, but want to see how complicated this will get. Signed-off-by: Dan McGee <dan@archlinux.org>
2012-11-12Add comments to visualize JSDan McGee1-0/+2
Signed-off-by: Dan McGee <dan@archlinux.org>
2012-11-12Add get_latest_by to MirrorLog Meta classDan McGee1-0/+1
Signed-off-by: Dan McGee <dan@archlinux.org>
2012-11-11Add title to mirror status data pointsDan McGee1-1/+3
Signed-off-by: Dan McGee <dan@archlinux.org>
2012-11-11Exclude news.safe_mode on news edit screenDan McGee1-1/+1
Signed-off-by: Dan McGee <dan@archlinux.org>
2012-11-11Mirror graph tweaking after usage with real dataDan McGee3-4/+6
* Clamp y-axis minimum to 0. * Don't plot `is_success == false` values. * Ensure URLs are sorted predictably. Signed-off-by: Dan McGee <dan@archlinux.org>
2012-11-11Re-add nice() calls for mirror status axesDan McGee1-2/+2
Signed-off-by: Dan McGee <dan@archlinux.org>
2012-11-11Slight style tweaksDan McGee1-0/+1
Signed-off-by: Dan McGee <dan@archlinux.org>
2012-11-11Mirror status graph, now with points AND linesDan McGee3-18/+50
We might have to tweak the interpolation method once we see this with real data, but for now it looks really pretty locally. Signed-off-by: Dan McGee <dan@archlinux.org>
2012-11-10Enable mirror status graph resizingDan McGee1-15/+28
Signed-off-by: Dan McGee <dan@archlinux.org>
2012-11-10Add initial version of mirror status chartDan McGee3-0/+115
Still have some hardcoded stuff to rip out and replace to make this a bit more dynamic on things like sizing, but for now, this is a great start. Signed-off-by: Dan McGee <dan@archlinux.org>
2012-11-10Allow filtering retrieved mirror statuses by mirror_idDan McGee2-8/+19
When we don't need them all, no need to fetch them all. Let the database do the work for us, hopefully. Signed-off-by: Dan McGee <dan@archlinux.org>
2012-11-10Add mirror extended status JSON viewDan McGee2-3/+35
When asking for status for a single mirror, we can include logs from the past 24 hours in addition to the normal information we provide. This is slated for usage by a frontend graph still to come, similar to those on the NTP pool website. Signed-off-by: Dan McGee <dan@archlinux.org>
2012-11-06Remove explicit timezone handling from feedsDan McGee1-3/+2
Signed-off-by: Dan McGee <dan@archlinux.org>
2012-11-06Make guid feeds helper a @staticfunctionDan McGee1-9/+12
Signed-off-by: Dan McGee <dan@archlinux.org>
2012-11-01Allow editing news.safe_mode flag via admin screenDan McGee2-2/+3
We need to mark the property as editable, but you still don't have access to it through the normal non-admin views and edit screen. Signed-off-by: Dan McGee <dan@archlinux.org>
2012-11-01Signal attachment cleanupDan McGee1-4/+4
Signed-off-by: Dan McGee <dan@archlinux.org>
2012-11-01Update for November ISO releaseDan McGee1-1/+1
Signed-off-by: Dan McGee <dan@archlinux.org>
2012-11-01download.html: Provide a magnet linkThomas Bächler1-2/+3
Some people prefer these over torrent files, as they're easier to pass than the torrent files themselves. On updates, this only needs the new info hash. The tracker bits are optional, but ensure that the torrent client gets peers more quickly to receive the actual torrent file. Signed-off-by: Dan McGee <dan@archlinux.org>
2012-10-31Store 'safe_mode' attribute on news modelDan McGee3-1/+143
This lets us identify old news items that need to allow HTML through the markdown parser. For all new news items, we will disallow raw HTML. Signed-off-by: Dan McGee <dan@archlinux.org>
2012-10-31Disable markdown safe modeDan McGee2-2/+2
Unless we want older news items to look like [HTML_REMOVED]this[HTML_REMOVED] all over the place. I'm tempted to mark old items as non-safe but enforce safe mode for all new news postings. Signed-off-by: Dan McGee <dan@archlinux.org>
2012-10-30Remove information regarding test ISOsThomas Bächler2-15/+0
The test builds are dead and nobody is taking care of them. With our monthly releases, they serve no purpose. Signed-off-by: Dan McGee <dan@archlinux.org>
2012-10-27Fix news feed content displayDan McGee1-1/+1
Signed-off-by: Dan McGee <dan@archlinux.org>
2012-10-27Convert silhouette.png to an 8-bit colormapDan McGee1-0/+0
This cuts the size from 33KB to just 2KB. Signed-off-by: Dan McGee <dan@archlinux.org>
2012-10-27Remove usages of 'django.contrib.markup'Dan McGee6-9/+13
Switch to the news model being able to spit out the HTML version of the content, and don't use the markup contrib module. This is deprecated as of Django 1.5 so we can move off it now to save trouble down the road when it is fully removed. Signed-off-by: Dan McGee <dan@archlinux.org>
2012-10-27Cleanup meta model attributesDan McGee3-6/+8
Signed-off-by: Dan McGee <dan@archlinux.org>
2012-10-27Extract some common architecture grabbing logicDan McGee1-6/+9
Signed-off-by: Dan McGee <dan@archlinux.org>
2012-10-26Enable safe mode for markdown parsingDan McGee4-5/+5
Although we don't allow unauthenticated users to post content, we should still cover our bases here and ensure people can't inject stuff into the production website via an inadvertent XSS. Signed-off-by: Dan McGee <dan@archlinux.org>
2012-10-26Add .DS_Store to .gitignoreDan McGee1-0/+1
Thanks for your silly files, OS X. Signed-off-by: Dan McGee <dan@archlinux.org>
2012-10-21Remove 'table' tag specifier from '.arch-bio-entry'Dan McGee1-7/+7
Class-based selectors make sense here, we don't need the table tag too. Signed-off-by: Dan McGee <dan@archlinux.org>