Regressions caused by bad refactoring in commit 21feac3bea.
Test cases in tests/CommandInterpreterTest.php were made against the pre-refactoring code, and now check out with the fixed code.
Failures were caused by not changing logic structure when moving from multiple exit points (each if point would return directly with a null or an object) to setting a result variable and then falling through to a common exit point. Without the if statements being restructured, the result variable would just get overridden by the next case.
I've extended the rights framework (centering on the Right class and Profile::hasRight()) to cover
Web login and API use. This will make it possible to prevent login and API use by users.
I added two new Right constants to the Right class: WEBLOGIN and API. I check these rights using
Profile::hasRight() when initializing users. If the rights check fails, I throw an exception.
I created a new AuthorizationException class for this particular
exception, in order to allow a different UI for these kinds of exceptions (or whatever).
* 0.9.x:
* update translator documentation. * remove superfluous whitespace. * tab to spaces. * add FIXME for undocumented class.
* update translator documentation. * remove superfluous whitespace. * L10n updates. * small refactoring in publicrss.php. * remove PHP4-isms
Localisation updates from http://translatewiki.net.
Double quotes to single quotes.
* improve L10n consistency for English. For example proper punctuation for all button and label titles. * fix some i18n bugs (in-message variables). * update/add translator documentation. * remove superfluous whitespace.
add path separators for Plugin::path()
argument to send email summary to all users on all sites
fix indentation in siteemailsummaryhandler
fix indentation in sendemailsummary.php
fix indentation in Email_summary_status.php
fix indentation in EmailSummaryPlugin.php
fix indentation in usermailsummaryhandler.php
Our mailbox actions (inbox and outbox) were doing their own display of
messages. This was causing issues with especially showmessage, which
since the more rigourous nickname checks were added, no longer works as
a mailbox subclass.
I've taken the time to rip out the message listing code from MailboxAction
and moved it to a MessageList widget. The different mailboxes now have their
own subclasses that show the correct profile in the list.
This entry was using the strings 'true' and 'false' instead of literal booleans, which could confuse clients expecting literal booleans as in other places and on Twitter in this place.
Format's already available as a member variable, so use it!
Fixes some error reponses in api/statusnet/groups/leave.json which were coming through as XML.
May fix some others as well.
$config['site']['logperf'] = true; // to record & dump total hits of each type and the runtime to syslog
$config['site']['logperf_detail'] = true; // very verbose -- dump the individual cache keys and queries as they get used (may contain private info in some queries)
Seeing 180 cache gets on a timeline page seems not unusual currently; since these run in serial, even relatively small roundtrip times can add up heavily.
We should consider ways to reduce the number of round trips, such as more frequently storing compound objects or the output of processing in memcached.
Doing parallel multi-key lookups could also help by collapsing round-trip times, but might not be easy to fit into SN's object model. (For things like streams this should actually work pretty well -- grab the list, then when it's returned go grab all the individual items in parallel and return the list)
* dropped unnecessary join on notice table
* made the function actually static, since it makes no sense as an instance variable. The only caller (in AttachmentList) is updated.
In order to apply to PHP's POST processing, the MAX_FILE_SIZE field must appear *before* the file upload field. They were incorrectly placed after, where they had no effect on POST processing.
Part of the reported issue was previuosly fixed by dc497ed0 (smaller size images being blanked).
This commit fixes the remaining bug with original-size avatars being left as BMP (which could include the 96px size for instance, which could cause problems in browsers not supporting BMP natively)
Added ImageFile::copyTo() as a convenient alias for resizeTo() when not resizing; this performs the BMP/XPM/XBM->PNG conversion if needed, or copies the original file.
Copying instead of using move_uploaded_file() is fine here since:
a) the files are cleaned up on script completion anyway (vs moving to remove it)
b) we're already performing getimagesize() and possibly load/resize on the file before this point (vs needing to move the file into a usable area to work with open_basedir restrictions that prevent working directly with uploaded files in the temp dir; since this would fail anyway, we lose nothing)
ImageFile::preferredType() now works on $this->type instead of asking for one, to make it handier to use from outside. (This is still needed in order for calling code to generate a target filename.)
Recommended for future:
* additional consolidation between the various ways of uploading avatars (touched avatarsettings, grouplogo, and apiaccountupdateprofileimage with similar minor changes)
* consolidate type checks and file naming into Avatar class
There's a new menu layout in this version of the software. It was
implemented as a plugin in 0.9.x to avoid clashes with existing themes,
but we're going to break that compatibility in this version, so we're just going for it.
This change involved moving all the changes in NewMenuPlugin into the
default code that was calling it. In addition, since
accountsettingsaction and connectsettingsaction differed only by menu,
I removed them, changed all references to them to the settingsmenu, and moved
the combined nav to its own class.
Let's put that episode behind us.
The CSS shim that was loaded by NewMenuPlugin for certain themes and certain actions
was removed.
'admin' is a pretty common username that people try when installing;
it was blacklisted because all of our admin panels were at /admin/*,
which would conflict with the admin user's namespace.
Changed the location of all admin panels to /panel/*, blacklisted the
nickname 'panel', and allowed 'admin'. Tested with a fresh install;
seems to work great.
* 'testing' of gitorious.org:statusnet/mainline: (63 commits)
Add a scary 'experimental feture' warning & are-you-sure prompt on moveuser.php
fix wrong datatypes (saving string instead of array) in AtomPub notice processing
Account moving is a background activity
return a 409 Conflict when subscription already exists
OStatusPlugin does discovery in Profile::fromURI()
considerably more logging and error checking in AccountMover
add a log method to AccountMover
normalize accounts and check for return in HTTP for moving
move account-moving classes to their own libraries
execution protection on discovery.php
PHPCS discovery.php
Move discovery library from OStatus plugin to core
Revert "Revert "0.9.7alpha1""
first example of moving a user
Parse properties of links in XRD files
Add the Atom username to the XRD output
preserve activities in object
let callers pass in an XMLOutputter to output to
execution protection on discovery.php
PHPCS linkheader.php
...
Removed the free calls (unneeded since destructors now work), and added an error check w/ logging & an exception for future attempts to forward calls to nonexistent object.
We were checking the list as comma-delimited (per the description of it as comma-delimited), but in fact spaces are also accepted, and who knows what else.
We can make a lot of HTTP requests from the server side. This change
adds some configuration options for using an HTTP proxy, which can
cache hits from multiple sites (good for status.net-like services, for example).
Version 0.9.6 and below of StatusNet assume anything in <author> is a
Person. So, we include an <activity:subject> element, which will be
checked first by those versions of the code, only for group feeds.
At some point we'll take this out, but it's useful for now.
* adds Right::CREATEGROUP
* logic in Profile::hasRight() checks for silencing
* NewgroupAction checks for the permission before letting you see or process the form in the UI
* User_group::register() logic does a low-level check on the specified initial group admin, and rejects creation if that user doesn't have the right; guaranteeing that API methods etc will also have this restriction applied sensibly.
Moved most of the heavy-lifting for account restoration out of
restoreuser.php and into its own class, with the hope that we'll do
the work from the Web eventually.
The new DeleteaccountAction lets a user delete their own account
(subject to global rights set by the admin). It presents a form to
delete the account, with an "I am sure." text entry box.
It then schedules the account for deletion and logs the user out.
If a cache entry is dependent on the code that's running, upgrading
(or enabling/disabling plugins) can generate hard-to-track
inconsistencies.
This change adds a close-to-unique fingerprint of the running code to
some cache keys, so that if the fingerprint changes, the old values
are ignored and new values are used.
If the automated uniqueness fails, an administrator can add an extra
config value, $config['site']['build'], that's thrown into the key also.
If a cache entry is dependent on the code that's running, upgrading
(or enabling/disabling plugins) can generate hard-to-track
inconsistencies.
This change adds a close-to-unique fingerprint of the running code to
some cache keys, so that if the fingerprint changes, the old values
are ignored and new values are used.
If the automated uniqueness fails, an administrator can add an extra
config value, $config['site']['build'], that's thrown into the key also.
We had two ways to generate an activity entry from a notice; one through
Notice::asAtomEntry() and one through Notice::asActivity() and
Activity::asString(). The code paths had already diverged somewhat. I
took the conditions that were in Notice::asAtomEntry() and made sure
they were replicated in the other two functions. Then, I rewrote
Notice::asAtomEntry() to use the other two functions instead.
This change passes the ActivityGenerationTests unit tests, but there
may be some other stuff that's not getting covered.
common_shorten_links() can only access the web session's logged-in user, so never properly took user options into effect for posting via XMPP, API, mail, etc.
Adds an optional $user parameter on common_shorten_links(), and a $user->shortenLinks() as a clearer interface for that.
Tweaked some lower-level functions so $user gets passed down -- making the $notice_id param previously there for saving URLs at notice save time generalized a little.
Note also ticket #2919: there's a lot of duplicate code calling the shortening, checking the length, and reporting near-identical error messages. These should be consolidated to aid in code and translation maintenance.
commit 39fdd181d95d2c39a3ea1ca330b10a99a92b961f
Author: Evan Prodromou <evan@status.net>
Date: Mon Nov 29 10:37:49 2010 -0500
use cache key prefix for router cache key
commit 4cb9e56941922489b83d6425c059cf770991e68f
Author: Evan Prodromou <evan@status.net>
Date: Mon Nov 29 10:31:21 2010 -0500
use a unique hashkey based on the software version and loaded plugins
commit 44458b48aef719543e11f83b41fded65cbcb8be9
Author: Evan Prodromou <evan@status.net>
Date: Sat Nov 27 17:04:15 2010 -0500
cache the NUM object
commit 809c188307a9b4ada15f3d7fa573a6034341efef
Author: Evan Prodromou <evan@status.net>
Date: Sat Nov 27 15:44:12 2010 -0500
accelerate routing by pivoting paths on actions
commit 39fdd181d95d2c39a3ea1ca330b10a99a92b961f
Author: Evan Prodromou <evan@status.net>
Date: Mon Nov 29 10:37:49 2010 -0500
use cache key prefix for router cache key
commit 4cb9e56941922489b83d6425c059cf770991e68f
Author: Evan Prodromou <evan@status.net>
Date: Mon Nov 29 10:31:21 2010 -0500
use a unique hashkey based on the software version and loaded plugins
commit 44458b48aef719543e11f83b41fded65cbcb8be9
Author: Evan Prodromou <evan@status.net>
Date: Sat Nov 27 17:04:15 2010 -0500
cache the NUM object
commit 809c188307a9b4ada15f3d7fa573a6034341efef
Author: Evan Prodromou <evan@status.net>
Date: Sat Nov 27 15:44:12 2010 -0500
accelerate routing by pivoting paths on actions
This provides initial infrastructure for decoupling display names from internal canonical names, but continues to have us storing and using the canonical forms.
It should be/become possible to provide mixed-case and underscore-containing names in links, @-mention, !-group, etc, but we don't store those alternate forms generally.
max_id is not yet implemented, as it'll need support added to the search backends. (since_id we get 'for free' by just cropping off the list, it'll do for now)
- "To" drop-down list now defaults to showing "Select recipient:" instead of the first person on your list, reducing liklihood of accidentally sending a message to the wrong person.
- When there are no mutual subscribers to send to, instead of an empty list the list now shows 'No mutual subscribers.'
In both cases, attempting to send when the default is selected displays an error message.
I'm not disabling form elements in part because our themes right now don't show disabled button state correctly; we might want to tighten that up a bit more once fixed.
This option may be useful for intranet sites that don't have direct access to the internet, as they may be unable to successfully fetch those resources.
- workaround for providers that are skimpy on their data, such as missing width/height or thumbnail_width/thumbnail_height
- workaround for YFrog listing "image" instead of "photo" type
- generally more lax about formatting: if it comes back and looks kinda ok, we'll take it.
- discovery uses system HTML parser, should be more robust if the links include things like ampersands with proper HTML-level escaping
Newly supported:
- TwitPic: added a local function using TwitPic's API, since the oohembed implementation for TwitPic produced invalid output which Services_oEmbed rejects. (bug filed upstream)
Tweaked...
- Flickr: works, now using whitelist to use their endpoint directly instead of going through oohembed
- Youtube: worked around a bug in Services_oEmbed which broke the direct use of API discovery info, so we don't have to use oohembed.
Not currently working...
- YFrog: whitelisting their endpoint directly as the oohembed output is broken, but this doesn't appear to work currently as I think things are confused by YFrog's servers giving a '204 No Content' response on our HEAD checks on the original link.
StatusNet core code now sets the tooltip text on .attachment.more links when they receive their attachment-expansion magic; this will override the hardcoded tooltip text saved from OStatus plugin when displaying timelines in the web UI.
The router settings weren't quite right so we ended up with bogus regex values passed in as the 'id' parameter, which broke the regular fallback ordering of parameter checks.
* moved some translator comments that were not directly above the line with the message to the correct location.
* i18n for UI text.
* superfluous whitespace removed.
I've consolidated the checks for which user to use for single-user mode into User::singleUser(), which now uses the configured nickname by preference, falling back to the site owner if it's unset.
This is now called consistently from the places that needed to use the primary user's nickname in routing setup.
Setting $config['singleuser']['nickname'] should now work again as expected.
* skip 0-byte config files when initializing, go ahead and redirect to installer if no non-0-byte files
* tweak warning on installer.php to let you know if you have a 0-byte config.php that's not writable, as opposed to generally already having a config.php with data in it
Fetching basic column data and unique indexes. Still needs detail work, multi-value indexes, foreign keys, and distinguishing the primary key.
Since we don't get comments and such, for cleaner comparisons we should probably do a filtering on supported features.
This will apply to *ALL* plugins in *ALL* languages, so should probably only be used when doing site customization...
You'd probably do:
$config['site']['locale_path'] = '/srv/awesome/data/locale';
$config['plugins']['locale_path'] = '/srv/awesome/data/locale';
with a structure like:
srv/
awesome/
data/
locale/
en/
LC_MESSAGES/
statusnet.po
OpenID.po
AnonymousFave.po
etc, all alongside each other. You could separate plugins from the core if you like.
Where locale files have not already been generated, you can build one for a plugin like so:
php scripts/update_po_templates.php --plugin=MyPlugin
and pull out the template file:
plugins/MyPlugin/locale/MyPlugin.pot
Edit that (make sure you at least set the CHARSET, probably to UTF-8) and save your customized .po
files into the structure as above, and use msgfmt to generate .mo files for final output.
This will apply to *ALL* plugins in *ALL* languages, so should probably only be used when doing site customization...
You'd probably do:
$config['site']['locale_path'] = '/srv/awesome/data/locale';
$config['plugins']['locale_path'] = '/srv/awesome/data/locale';
with a structure like:
srv/
awesome/
data/
locale/
en/
LC_MESSAGES/
statusnet.po
OpenID.po
AnonymousFave.po
etc, all alongside each other. You could separate plugins from the core if you like.
Where locale files have not already been generated, you can build one for a plugin like so:
php scripts/update_po_templates.php --plugin=MyPlugin
and pull out the template file:
plugins/MyPlugin/locale/MyPlugin.pot
Edit that (make sure you at least set the CHARSET, probably to UTF-8) and save your customized .po
files into the structure as above, and use msgfmt to generate .mo files for final output.
* now ignoring if-modified-since if we failed an etag if-none-match comparison, per spec
* now including a hash of user id/nickname in most etags, so we'll update the view properly after login/logout
For API methods, checking the API-auth'ed user. (Many change results to include things like 'you're subscribed to this user' or 'this is one of your favorites', so user info is again needed)
There'll still be some last-modified stamps that aren't including user info properly, probably.
* skip more files (.xcf image sources, .html docs)
* skip files before rejecting them for funky filenames!
* allow period in filenames (eg foo-1.4.ttf) but blacklist some unsafe extensions-within-extensions
* fix double space to single space
* use consistent punctuation in messages
* complete translation hints for all commands
* add an i18n fixme for missing plural support
* trailing whitespace removed
When bogus SSL sites etc were hit through a shortening redirect, sometimes link resolution kinda blew up and the user would get a "Can't linkify" error, aborting their post.
Now catching this case and just passing through the URL without attempting to resolve it. Could benefit from an overall scrubbing of the freaky link/attachment code though...! :)
http://status.net/open-source/issues/2513
Two prongs here:
* We attempt to enable SNI on the SSL stream context with the appropriate hostname... This requires PHP 5.3.2 and OpenSSL that supports the TLS extensions. Unfortunately this doesn't seem to be working in my testing.
* If set $config['http']['curl'] = true, we'll use the CURL backend if available. In my testing on Ubuntu 10.04, this works. No guarantees on other systems.
I'm not enabling CURL mode by default just yet; want to make sure there's no other surprises.
File extensions can also be added to the upload type whitelist; they'll be normalized to types for the actual comparison, so only known extensions will work.
Warning: the format strings aren't actually being enforced here which is probably why they weren't caught earlier. Not quite sure why, it should be looked at!
Currently only one custom theme may be uploaded per site, saved with the name 'custom' and stored into the local/themes subdirectory.
Administrators can upload a .ZIP archive containing a theme through the design admin panel; its contents are validated to ensure that only legit files are saved, and a 5M size quota is enforced.
Theme upload requires the zip extension for PHP; if not present, theme uploading is disabled by default.
Uploading and the custom CSS can be controlled via $config['theme_upload']['enabled'] and $config['custom_css']['enabled'].
Configurable directory/path/server for 'local' subdirectory (currently only as used for themes; local plugins not yet switched over)
Can set $config['local']['dir'] etc; not currently exposed in the admin panels.
Per-site directories on a separate themes server could be set up such as:
$config['local']['dir'] = '/path/to/themes/local/' . $_nickname;
$config['local']['server'] = 'themes.example.com';
$config['local']['path'] = '/local/' . $_nickname;
$config['local']['ssl'] = 'never';
Fixed typo in RedirectingAction when no return-to data provided in form submission.
RedirectingAction::returnToArgs() has been renamed to returnToPrevious() to avoid conflict with Action::returnToArgs() which returns arguments to be passed to other actions as return-to arguments. All callers should now be updated.
More profile settings actions will now redirect through a login form if visited as a GET request, as would be expected from a bookmark, link sent in e-mail etc.
Added a 2-second default timeout for XMLRPC/extended pings, configurable as [ping,timeout].
No longer repeating the entire ping section if we had an HTTP error during a submission.
For now, dropping the bad item and continuing on with others. (Todo: individual retry and cleaner discards of blacklisted broken-for-now sites.)
* Fake_XMPP back to Queued_XMPP, refactor how we use it and don't create objects and load classes until we need them.
* fix fatal error in IM settings while waiting for a Jabber confirmation.
* Caching fix for user_im_prefs
* fix for saving multiple transport settings
* some fixes for AIM & using normalized addresses for lookups
Users and administrators can set how long an URL can be before it's
shortened, and how long a notice can be before all its URLs are
shortened. They can also turn off shortening altogether.
Squashed commit of the following:
commit d136b39011
Author: Evan Prodromou <evan@status.net>
Date: Mon Apr 26 02:39:00 2010 -0400
use site and user settings to determine when to shorten URLs
commit 1e1c851ff3
Author: Evan Prodromou <evan@status.net>
Date: Mon Apr 26 02:38:40 2010 -0400
add a method to force shortening URLs
commit 4d29ca0b91
Author: Evan Prodromou <evan@status.net>
Date: Mon Apr 26 02:37:41 2010 -0400
static method for getting best URL shortening service
commit a9c6a3bace
Author: Evan Prodromou <evan@status.net>
Date: Mon Apr 26 02:37:11 2010 -0400
allow 0 in numeric entries in othersettings
commit 767ff2f7ec
Author: Evan Prodromou <evan@status.net>
Date: Mon Apr 26 02:36:46 2010 -0400
allow 0 or blank string in inputs
commit 1e21af42a6
Author: Evan Prodromou <evan@status.net>
Date: Mon Apr 26 02:01:11 2010 -0400
add more URL-shortening options to othersettings
commit 869a6be0f5
Author: Evan Prodromou <evan@status.net>
Date: Sat Apr 24 14:22:51 2010 -0400
move url shortener superclass to lib from plugin
commit 9c0c9863d5
Author: Evan Prodromou <evan@status.net>
Date: Sat Apr 24 14:20:28 2010 -0400
documentation and whitespace on UrlShortenerPlugin
commit 7a1dd5798f
Author: Evan Prodromou <evan@status.net>
Date: Sat Apr 24 14:05:46 2010 -0400
add defaults for URL shortening
commit d259c37ad2
Author: Evan Prodromou <evan@status.net>
Date: Sat Apr 24 13:40:10 2010 -0400
Add User_urlshortener_prefs
Add a table for URL shortener prefs, a corresponding class, and the
correct mumbo-jumbo in statusnet.ini to make everything work.
Fix extraction of Atom <content type="text"> and <content type="html">; we were failing to escape plaintext source data to HTML, and doing an extraneous double-deescape on HTML source resulting in breakage of notices containing text that looks like HTML. Only <content type="xhtml"> was working correctly previously.
Fixes for RSS2 content processing: we were failing to load <content:encoded> at all due to using wrong element name, and were applying an extraneous de-escape for <description> rather than the escaping that is required to turn plaintext into HTML. (Per spec, <description> must be plaintext.)
I added the statusnet: prefix to the xml output. This prefix should
be declared on the root element of all relevant XML output.
I also added two StatusNet-specific fields:
* statusnet:html - rendered HTML. Clients shouldn't have to guess
at the correct HTML rendering for notices, especially since some
of the links depend on context.
* statusnet:profile_url - profile URL for a user. You can't count
on a user being a local user in a distributed microblogging world.
So, this shows the explicit profile_url.
This reverts commit 98f0d970da.
Per xopher we're not yet ready to push 0.9.2 theme directories live; we also haven't merged down various things from testing that need to be in the release such as installer fixes.
* Moved notification sending from Notice::saveReplies to distrib queue handler, so it'll pull from the reply set we've saved regardless of how we got it.
* Set up gettext infrastructure for command-line scripts; gets localization mail notifications etc working from background queues.
* Adjusted locale switching: common_switch_locale() works at runtime for bg scripts, forces a message catalog update
* added a few XXX (that's StatusNet for FIXME, right?)
** proposing de-duplication of a message appearing ~50 times
** marking bad pagination implementation
* moved the TRANS: comments in action.php down to exactly before the line in which the message appears. Otherwise gettext does not put them in the pot file
* fixed a XXX in action.php removing double spaces
Because N/S/E/W were on the same lines, and the "// TRANS:" comment must be directly above the line in which the message is located to work, I added the messages to a variable first.
Refactored some of the returnto handling code. It looks like we have several different ways of handling this in the software, icky!
Marked the session-based functions with fixmes (they'll stomp on other forms when multiple tabs/windows are used) and combined some commonish bits of code between ProfileFormAction and the group block & makeadmin actions where they're using hidden form parameters. Extended that to allow passing dynamic parameters (eg 'page') as well as static ones (action, target user/group).
May be slow or run out of memory if run on particularly prolific posters -- not yet optimized for that case.
Note that geodata that has already been sent out to other services (via ostatus, omb, twitter, etc) will not be removed from them.
Conflicts:
actions/imsettings.php
lib/jabber.php
Made a quick attempt to merge the new JID validation into the XmppPlugin, have not had a chance to test that version live yet.
Should also move over the test cases.
This reverts commit 260f00d60b.
As mentioned in
260f00d60b (comment_8367)
Reverting this merge until the bugs are fixed and there is a general
agreement on the need for this enhancement.
Basic splitting/validation code submitted via http://status.net/wiki/XMPP/JID_validation -- Copyright 2009 Patrick Georgi <patrick@georgi-clan.de> Licensed under ISC-L, which is compatible with everything else that keeps the copyright notice intact.
Added PEAR Net_IDNA package to extlib to handle IDN normalization (also used by Validate's email verifier if present).
* added test suite, supplemented my own test cases with JID validation and normalization test cases from libpurple
* follows XMPP rules for validation of name part
* fixes for normalization with non-ASCII names
* will do domain checks if $config['email']['check_domain'] is on, checking for an XMPP-server SRV record or any lookup. (We don't actually need to ping those direct though.)
* some more obscure stringprep validation rules aren't quite followed yet, but we err on the side of permissiveness.
* we still don't actually let you save your address with a resource on it, as we strip resources when looking up users who've sent us presence or message updates. I would recommend saving the outgoing resource as a separate field if/when we add that..?
* respect count instead of listing all groups
* respect page parameter
* don't spew notice on undefined $id
* don't spew notice on undefined $group->homepage_url (dropped the element since there's nothing to go in it)
* respect count instead of listing all groups
* respect page parameter
* don't spew notice on undefined $id
* don't spew notice on undefined $group->homepage_url (dropped the element since there's nothing to go in it)
I swapped the settings from negative to positive ($config['queue']['stomp_transactions'] = false rather than $config['queue']['stomp_no_transactions'] = true), gave them defaults (both on for best ActiveMQ experience), and added notes to the README about configuring them.
Conflicts:
lib/attachmentlist.php
plugins/OStatus/classes/Ostatus_profile.php
Merge tried to delete things that it seems it shouldn't, very confusing order. Hope rest of the cherry-picking isn't a problem.
Moved the various classes used by the Activity class to their own
files. There were >10 classes in the same file, with around 1500 lines
in the file. Just too big.
This change makes autoloading work for these classes, so also removed
the hard require in lib/common.php.
RSS feeds have the format
<rss><channel><item/><item/><item/></channel></rss>. The element named
$rss was actually the <channel> element, so I renamed the variable so
I wouldn't hurt my head.
First steps to parsing RSS items as activities. RSS feeds don't seem
to have enough data to make good remote profiles, but this may work
with some "hints".
* Subscription::start was sometimes passing users instead of profiles to hooks, which broke OStatus subscription notifications; now normalizing to profiles for processing.
* H-card parsing would trigger a lot of PHP warnings and notices in hKit. Now suppressing warnings and notices for the duration of the call to keep them out of output when display_errors is on.
* H-card parsing would trigger a PHP fatal error if the source page was not well-formed XML and Tidy was not present on the system. Switched normalization to use the PHP DOM module which is always present, as we have no need for Tidy's extra features here.
* Trying to fetch avatars from Google profiles failed and triggered a PHP warning due to the relative URL not being resolved during h-card parsing. Now passing profile page URL into hKit by sneaking a <base> tag in while we normalize the HTML source.
* Profile pages without a "Link" header could trigger PHP notices due to a bad NULL -> array(NULL) conversion in LinkHeader::getLink(). Now checking that there was a return value before converting single return value into array.
We were double-unescaping for <content type="html">, turning <b> escaped chars into literal tags (which then may get removed entirely by the HTML scrubber).
For instance this was throwing an exception for DB_DataObject::staticGet when there's no match... definitely not what we want when all our code expects to get a nice null.
Example of this causing trouble: http://gitorious.org/statusnet/mainline/merge_requests/131
Revert "Don't attempt to retrieve the current user from the DB while processing a DB error"
This reverts commit 68347691b0.
Revert "Use PHP exceptions for PEAR error handling."
This reverts commit d8212977ce.
For instance this was throwing an exception for DB_DataObject::staticGet when there's no match... definitely not what we want when all our code expects to get a nice null.
Example of this causing trouble: http://gitorious.org/statusnet/mainline/merge_requests/131
Revert "Don't attempt to retrieve the current user from the DB while processing a DB error"
This reverts commit 68347691b0.
Revert "Use PHP exceptions for PEAR error handling."
This reverts commit d8212977ce.
While deletion is in progress, the account is locked with the 'deleted' role, which disables all actions with rights control.
Todo:
* Pretty up the notice on the profile page about the pending delete. Show status?
* Possibly more thorough account disabling, such as disallowing all use for login and access.
* Improve error recovery; worst case is that an account gets left locked in 'deleted' state but the queue jobs have gotten dropped out. This would leave the username in use and any undeleted notices in place.
Under MySQL, new tables will be created as InnoDB with UTF-8 (utf8/utf8_bin) same as core tables.
Existing plugin tables will have table engine and default charset/collation updated, and string columns will have charset updated, at checkschema time.
Switched from 'DESCRIBE' to INFORMATION_SCHEMA for pulling column information in order to get charset. A second hit to INFORMATION_SCHEMA is also needed to get table properties.
Indices were only being created at table creation time, which ain't so hot. Now also adding/dropping indices when they change.
Fixed up some schema defs in OStatus plugin that were a bit flaky, causing extra alter tables to be run.
TODO: Generalize this infrastructure a bit more up to base schema & pg schema classes.
I think this is a bug in Youtube's web server (sending chunked encoding of an empty body with a HEAD response, leaving the connection out of sync when it doesn't attempt to read a body) but the HTTP_Request2 library may need to be adjusted to watch out for that.
Keeps stray daemon subprocesses from floating around when we kill the parents via a signal!
Accomplished by opening a bidirectional pipe in the parent process; the children close out the writer end and keep the reader in their open sockets list. When the parent dies, the children see that the socket's been closed out and can perform an orderly shutdown.
The final whitespace should be dropped from the source messages after we've stabilized; trailing space is pretty unreliable to keep through translation tools and should be avoided. Use separator strings outside the messages!
OStatus plugin overrides these to allow using webfinger (user@example.com), profile URL (http://example.com/user) and bare profile URL (example.com/user) as arguments.
Gets Spanish, French, Russian etc UI localization working on Debian Lenny fresh installation set up in Spanish (so es_ES.UTF-8 is available but en_US.UTF-8 isn't).
The upstream class sets $this->basejid with host unconditionally, which wasn't previously an issue as the fulljid would always be filled in by the server at connect time before sending messages. With the new queued messaging, we need to make sure we've filled out $this->fulljid correctly without making a connection.
Now using $server if provided to build $this->basejid and $this->fulljid in the queued XMPP proxy class, so queued messages are sent correctly.
- switch 'en_US' to 'en', fixes the "admin panel switches to Arabic" bug
- tweak setting descriptions to clarify that most of the time we'll be using browser language
- add a backend switch to disable language detection (should this be exposed to ui?)
* 'testing' of gitorious.org:statusnet/mainline:
Using position relative only for the remote subscription in section
Added group subscription button to groups mini list
Added event hooks at the start and end of groups mini list
All 'connect' menu panels used to be optional, so Action tried to
figure out what the first item on the 'connect' menu should be.
This is no longer necessary because we have the non-optional OAuth
client connections panel now, which is not optional and can't be
turned off.
Also stripping id from foreign HTML messages (could interfere with UI) and disabled failing attachment popup for a.attachment links that don't have a proper id, so you can click through instead of getting an error.
Issues:
* any other links aren't marked and saved
* inconsistent behavior between local and remote attachments (local displays in lightbox, remote doesn't)
* if the enclosure'd object isn't referenced in the content, you won't be offered a link to it in our UI
We only need one author for user feeds: the user themselves. So, show
the user as the activity:subject, and don't repeat the same
activity:actor for every notice unnecessarily.
* 'testing' of gitorious.org:statusnet/mainline: (25 commits)
Fix a bunch of notice & warning-level messages that were breaking my inter-instance communications
more output in updateostatus.php
lost important fields when switching queries
show service debug info
pass listener URI into consumer for OMB
remove strict check on OMB exception strings
return correct HTTP status code for OMB errors
send smaller error pages for OMB API endpoints
Remove check for secret in token deletion on Subscription::cancel()
Better logging on bad token in subscription
Return empty array when no subscriptions to remote
drop tokens for OMB on unsubscribe
fix path for updateostatus.php
Script to convert OMB subscriptions to OStatus subscriptions
show service debug info
pass listener URI into consumer for OMB
remove strict check on OMB exception strings
return correct HTTP status code for OMB errors
send smaller error pages for OMB API endpoints
Remove check for secret in token deletion on Subscription::cancel()
...
In a federated system, "@nickname" is insufficient to uniquely
identify a user. However, it's a very convenient idiom. We need to
guess from context who 'nickname' refers to.
Previously, we were using the sender's profile (or what we knew about
them) as the only context. So, we assumed that they'd be mentioning to
someone they followed, or someone who followed them, or someone on
their own server.
Now, we include the notice information for context. We check to see if
the notice is a reply to another notice, and if the author of the
original notice has the nickname 'nickname', then the mention is
probably for them. Alternately, if the original notice mentions someone
with nickname 'nickname', then this notice is probably referring to
_them_.
Doing this kind of context sleuthing means we have to render the
content very late in the notice-saving process.
* 'testing' of gitorious.org:statusnet/mainline:
init_conversation.php script to copy old notice conversations into the conversation table
cache results of webfinger lookups
Parse an hcard for hints, if available
use new hcard method for webfinger
add hkit for hCard parsing
use new dedicated hcard method for Webfinger profile
Add an hcard action
* 'testing' of gitorious.org:statusnet/mainline:
OStatus: if no <link> available and no profileurl hint passed in, use object id if it's an HTTP(S) URL. Fixes profile link for Google accounts.
Tweak common_url_to_nickname to take the last path component; fixes pulling nicks from Google profile pages (path is "/profile/<nickname>")
Fix bug on subscribe/unsubscribe in profile lists. Bogus call to nonexisting profile->getProfile() was masked by DB_DataObject
A dedicated hcard action for users. Our profile page includes
an hcard, but it's so full of other hcards that it's ambiguous which
one is the "real" one. So, this one make sense for meaning, "This is
my hcard."
* 'testing' of gitorious.org:statusnet/mainline:
Not sure how this ended up in wrong-cased dir...
OStatus: fix remote groups to work with new user_groups/local_groups split.
Conflicts:
classes/User_group.php
- fix <activity:subject> generation so we get the profile info (what's available so far)
- use id instead of nickname for group join/leave forms so we can join/leave remote groups
while the rest of the groups UI remains limited to local groups
(plugins are responsible for making sure remote notifications and permission checks are done)
- fix remote notification when joining group through OStatus's remote subscribe form
* 'testing' of gitorious.org:statusnet/mainline:
Merge StatusNet core localization updates from 0.9.x branch
Fix update_po_templates.php to support the plural and context variants of _m() in plugins
Drop HTMLPurifier; we don't need its extra capabilities and we're already using htmLawed which is lighter-weight.
OStatus: handle update-profile Salmon pings
Revert "Updated jQuery Form Plugin from v2.17 to v2.36"
OStatus: disable HTMLPurify cache unless we've configured a writable path for it.
- add event hooks to profile update pings
- send Salmon pings with custom update-profile event to OStatus subscribees and groups (subscribers will see it on your next post)
- fix OStatus queues with overlong transport names, should work on DB queues now
- Ostatus_profile::notifyActivity() and ::notifyDeferred() now can take XML, Notice, or Activity for convenience
the shorthand that I've proposed at
http://microformats.org/wiki/geo-brainstorming#latitude_longitude_shorthand_and_geo_link
If anyone wants to pick up on where the discussion was left off or
get more implementation support by other sites and software, and be
recognized by parsers, I'd be happy to go back to the shorthand.
Because you know, it actually makes a lot of sense.
- added rel="ostatus:attention" links for group delivery
- added events for plugins to override group profile/permalink pages
- pulled Notice::saveGroups up to save-time so we can override;
it's relatively cheap and gives us a clean list of target
groups for distrib time even with customized delivery.
- fixed notice::getGroups to return group objects as expected
- added some doc on new parameters to Notice::saveNew
- 'groups' list of group IDs to push to in place of parsing
- messages that come in via PuSH and contain local group targets
are delivered to local group members
- messages that come in via PuSH and contain remote group targets
are delivered to local members of the remote group
Todo:
- handle group posts that only come through Salmon
- handle conflicts in case something comes in both through Salmon and PuSH
- better source verification
- need a cleaner interface to look up groups by URI
- need a way to handle remote groups with conflicting names
Combined the code that finds mentions of other profiles into one place.
common_find_mentions() finds mentions and calls hooks to allow
supplemental syntax for mentions (like OStatus).
common_linkify_mentions() links mentions.
common_linkify_mention() links a mention.
Notice::saveReplies() now uses common_find_mentions() instead of
trying to parse everything again.
I changed the way that tag: URIs are minted, so we now use the right
base. Ideally most of these would use HTTP URIs instead, but for
now at least they use the right base.
We've been making pretty crummy tag: URIs for a while. We should
continue to favor HTTP URIs, since it's nice to be able to discover
things about an object you've shared the ID of. Where that's not
possible, this makes nicer tag URIs.
The subs_* functions in subs.php have made a lot of assumptions
about users versus profiles. I've refactored the functions to
be methods of the Subscription class instead, and to use Profile
objects throughout.
Some of the checks for blocks or existing subscriptions depended
on users or profiles, so I've moved those methods around a bit.
I've left stubs for the subs_* functions until we get time to replace
them.
All breakout queues that we're going to need to listen to now need to be explicitly listed in $config['queue']['breakout'].
Until XMPP is moved to component model, this setting will let the individual processes work with their own queues:
$config['queue']['breakout'][] = 'xmpp/xmppout/' . $config['site']['nickname'];
- Multiplexing queues into groups and for multiple sites.
- Sharing vs breakout configurable per site and per queue via $config['queue']['breakout']
- Detect how many times a message is redelivered, discard if it's killed too many daemons
- count configurable with $config['queue']['max_retries']
- can dump the items to files in $config['queue']['dead_letter_dir']
Queue daemon memory & resource leak fixes:
- avoid unnecessary reconnections to memcached server (switch persistent connections back in on second initialization, assuming it's child process)
- monkey-patch for leaky .ini loads in DB_DataObject::databaseStructure() - was leaking 200k per active switch
- applied leak fixes to Status_network as well, using intermediate base Safe_DataObject for both it and Memcache_DataObject
Misc queue fixes:
- correct handling of child processes exiting due to signal termination instead of regular exit
- shutdown instead of infinite respawn loop if we're already past the soft memory limit at startup
- Added --all option for xmppdaemon... still opens one xmpp connection per site that has xmpp active
Cache updates:
- add Cache::increment() method with native support for memcached atomic increment
* skip unnecessary unsubscribes on graceful shutdown -- takes a long time for many queues, slows down our restarts when hitting graceful mem limit
* fix control channel (was broken when we switched to support multiple queue servers)
* detection of group feeds is currently a nasty hack based on presence of '/groups/' in URL -- should use some property on the feed?
* listing for the remote group is kinda cruddy; needs to be named more cleanly
* still need to establish per-author profiles (easier once we have the updated Atom code in)
* group delivery probably not right yet
* saving of group messages still triggering some weird behavior
Added support for since_id and max_id on group timeline feeds as a free extra. Enjoy!
* Treat linkless feed posts as status updates; drop the "New post:" prefix and quotes on them.
* Use stable user IDs for atom/rss2 feed links instead of unstable nicknames
* Pull Atom feed preferentially when subscribing -- can now put the remote user's profile page straight into the feed subscription form and get to the right place.
* Clean up naming for push endpoints
* renamed FeedSub plugin to OStatus
* now setting avatar on subscriptions
* general fixes for subscription
* integrated PuSH hub to handle only user timelines on canonical ID url; sends updates directly
* set $config['feedsub']['nohub'] = true to test w/ foreign feeds that don't have hubs (won't actually receive updates though)
* a few bits of code documentation
* HMAC support for verified distributions (safest if sub setup is on HTTPS)
And a couple core changes:
* minimizing HTML output for exceptions in API requests to aid in debugging
* fix for rel=self link in apitimelineuser when id given
This does not not yet include any of the individual subscription management (Salmon notifications for sub/unsub, etc) nor a nice UI for user subscriptions.
Needs some further cleanup to treat posts as status updates instead of link references.
Defaulting to only looking at last 90 days of activity, can be adjusted up or down.
$config['tag']['cutoff'] = 86400 * 90;
$config['popular']['cutoff'] = 86400 * 90;
Per-user and per-group tag clouds do not use the cutoff (and it doesn't help with indexing on them).
Adds a robots.txt file to the site root. Defaults defined by
'robotstxt' section of config. New events StartRobotsTxt and
EndRobotsTxt to let plugins add information. Probably not
useful if path is not /, but won't hurt anything, either.
If an array of multiple servers is put in $config['queue']['stomp_server'], enqueues will pick a random server to send to (failing over automatically if any are down).
Queue handling daemons connect all servers so they get events no matter where they were delivered.
In case of disconnection, daemons should now handle it gracefully and attempt to reconnect every 60 seconds or so, automatically resubscribing to all queues once it's back up.
Can put to 'native' failover for reads as well by disabling $config['stomp']['manual_failover'] = false; but this is untested and may explode in addition to requiring that your ActiveMQ cluster actually be set up to handle its own data distribution.
Additionally, can choose which queues to mark as persistent by setting $config['stomp']['persistent'] to an array of queue names.
Move definition of NICKNAME_FMT above plugin initialization but below loading of Validate package.
A merge error when refactoring setup lead to this not being defined yet when plugins were initialized, causing the router setup in MapstractionPlugin which tried to use this constant to fail.
Result was bogus links and if you hit the URL directly the action would be "unrecognized".
* testing: (130 commits)
HTTP auth provided is evaluated even if it's not required
Rename rc3to09.sql to rc3torc4.sql to avoid confusion if we add a last-minute change after this!
Add new oauth tables and modifications to 'consumer' table for rc4
Centred leaderboard ad
camelcase the uap param names
move leaderboard to after the header
Moved rectangle ad into aside and leaderboard to the right in header.
Aligning wide skyscraper to the right instead of left
CSS ids and classes fixed in UAPPlugin
wrong height for rectangle in BlankAd
Add the moved BlankAdPlugin
make BlankAd dir and change to use a 1x1 image
move BlankAdPlugin to its own dir
Add BlankAdPlugin to test ad layout in different themes
make uapplugin an abstract class
move UAP plugin to core
Lowercased switch cases in UAP Plugin
Plugin for Universal Ad Package. Outputs four most widely used ad types.
Add persistent:true property to Stomp messages so ActiveMQ doesn't decide to discard them even though persistence is enabled on the broker. :) (Thanks Aric!)
quick fix: use common_path() on realtime update JS so it works with the new JS path code (will pull from main server for now)
...
Conflicts:
actions/apioauthaccesstoken.php
actions/apioauthauthorize.php
actions/apioauthrequesttoken.php
actions/editapplication.php
actions/newapplication.php
lib/apiauth.php
lib/queuemanager.php
lib/router.php
We have about 10-12 JavaScript pages per Web page. They usually
are based on the same server as the Web pages, but since they're
static files, it makes sense to offload them to a lite server that
handles static files well.
This commit lets you set a separate Javascript server and path for the
default Javascript code in StatusNet.
Squashed commit of the following:
commit 139d1622fdafe5ad00c820224416d9021efc3234
Author: Evan Prodromou <evan@status.net>
Date: Wed Jan 27 11:30:24 2010 -0500
modules that call htmloutputter::script() don't prescribe js/ path
commit c6ca3174af73efed55eaed5ff1e2a3bdc77d2d87
Author: Evan Prodromou <evan@status.net>
Date: Wed Jan 27 11:28:07 2010 -0500
configurable server and path for javascript files
queuectl.php --update -s<site>
queuectl.php --stop
queuectl.php --restart
Default control channel is /topic/statusnet-control. For external utilities to send a site update ping direct to the queue server, connect via Stomp and send a message formatted thus:
update:<nickname>
(Nickname here, *not* server hostname! The rest of the queues will be updated to use nicknames later.)
Note that all currently-connected queue daemons will get these notifications, including both queuedaemon.php and xmppdaemon.php. (XMPP will ignore site update requests for sites that it's not handling.)
Limitations:
* only implemented for stomp queue manager so far
* --update may not yet handle a changed server name properly
* --restart won't reload PHP code files that were already loaded at startup. Still need to stop and restart the daemons from 'outside' when updating code base.
New configuration options to define a single-user mode. This hides
most of the "community" pages, like the public timeline and groups.
The main user's timeline becomes the main page, and most other URLs
are changed.
Switching back and forth between 1-user and multi-user mode is
probably hazardous.
Squashed commit of the following:
commit d814aa5c92d14a27a12baba7893f3f8bf63f1d08
Author: Evan Prodromou <evan@status.net>
Date: Tue Jan 26 00:17:27 2010 -0500
don't show inbox and outbox in single-user mode
commit 47f19b9523a7015d4c6e460b73ea32c839e00aa1
Author: Evan Prodromou <evan@status.net>
Date: Tue Jan 26 00:15:22 2010 -0500
show correct URL for logo in single-user mode
commit 552010cffc33eadbc512ec5a67619dbc2015239a
Author: Evan Prodromou <evan@status.net>
Date: Tue Jan 26 00:15:06 2010 -0500
make singleuser its own config section
commit 786ab260a3ca172e57b555c75ca10946d8f258a1
Author: Evan Prodromou <evan@status.net>
Date: Tue Jan 26 00:05:19 2010 -0500
make single-user mode work
commit 5b21d7309b3a8dd5a4e0f29aea76f7897f1818b1
Author: Evan Prodromou <evan@status.net>
Date: Mon Jan 25 23:45:55 2010 -0500
add single-user mode
Moved much of the writing that happens when posting a notice to a new
queuehandler, distribqueuehandler. This updates tags, groups, replies
and inboxes at queue time (or at Web time, if queues are disabled).
To make this work well, I had to break up the monolithic
Notice::blowCaches() and make cache blowing happen closer to where
data is updated.
Squashed commit of the following:
commit 5257626c62750ac4ac1db0ce2b71410c5711cfa3
Author: Evan Prodromou <evan@status.net>
Date: Mon Jan 25 14:56:41 2010 -0500
slightly better handling of blowing tag memory cache
commit 8a22a3cdf6ec28685da129a0313e7b2a0837c9ef
Author: Evan Prodromou <evan@status.net>
Date: Mon Jan 25 01:42:56 2010 -0500
change 'distribute' to 'distrib' so not too long for dbqueue
commit 7a063315b0f7fad27cb6fbd2bdd74e253af83e4f
Author: Evan Prodromou <evan@status.net>
Date: Mon Jan 25 01:39:15 2010 -0500
change handle_notice() to handle() in distributqueuehandler
commit 1a39ccd28b9994137d7bfd21bb4f230546938e77
Author: Evan Prodromou <evan@status.net>
Date: Mon Jan 25 16:05:25 2010 -0500
error with queuemanager
commit e6b3bb93f305cfd2de71a6340b8aa6fb890049b7
Author: Evan Prodromou <evan@status.net>
Date: Mon Jan 25 01:11:34 2010 -0500
Blow memcache at different point rather than one big function for Notice class
commit 94d557cdc016187d1d0647ae1794cd94d6fb8ac8
Author: Evan Prodromou <evan@status.net>
Date: Mon Jan 25 00:48:44 2010 -0500
Blow memcache at different point rather than one big function for Notice class
commit 1c781dd08c88a35dafc5c01230b4872fd6b95182
Author: Evan Prodromou <evan@status.net>
Date: Wed Jan 20 08:54:18 2010 -0500
move broadcasting and distributing to new queuehandler
commit da3e46d26b84e4f028f34a13fd2ee373e4c1b954
Author: Evan Prodromou <evan@status.net>
Date: Wed Jan 20 08:53:12 2010 -0500
Move distribution of notices to new distribute queue handler
Previously, messages once delivered would just get stuck in the queue seemingly forever if they never got ACKed.
Note this could lead to partial duplication, for instance if the OMB or Twitter queue handlers die after 1/2 of the outgoing sends.
Recommendations:
* catch exceptions more aggressively within queue handlers (so only PHP fatal errors are likely to kill in the middle)
* for processing that involves sending to multiple clients, consider a second queue similar to the XMPP output, eg for OMB:
- first queue gets delivery list and builds message data, enqueueing it for each target address
- second queue can handle each individual outgoing message (and attempt redelivery etc separately)
This would also protect better against a recurring error preventing delivery in the second part, and could spread out any slow sends over multiple threads.
Queue handlers for XMPP individual & firehose output now send their XML stanzas
to another output queue instead of connecting directly to the chat server. This
lets us have as many general processing threads as we need, while all actual
XMPP input and output go through a single daemon with a single connection open.
This avoids problems with multiple connected resources:
* multiple windows shown in some chat clients (psi, gajim, kopete)
* extra load on server
* incoming message delivery forwarding issues
Database changes:
* queue_item drops 'notice_id' in favor of a 'frame' blob.
This is based on Craig Andrews' work branch to generalize queues to take any
object, but conservatively leaving out the serialization for now.
Table updater (preserves any existing queued items) in db/rc3to09.sql
Code changes to watch out for:
* Queue handlers should now define a handle() method instead of handle_notice()
* QueueDaemon and XmppDaemon now share common i/o (IoMaster) and respawning
thread management (RespawningDaemon) infrastructure.
* The polling XmppConfirmManager has been dropped, as the message is queued
directly when saving IM settings.
* Enable $config['queue']['debug_memory'] to output current memory usage at
each run through the event loop to watch for memory leaks
To do:
* Adapt XMPP i/o to component connection mode for multi-site support.
* XMPP input can also be broken out to a queue, which would allow the actual
notice save etc to be handled by general queue threads.
* Make sure there are no problems with simply pushing serialized Notice objects
to queues.
* Find a way to improve interactive performance of the database-backed queue
handler; polling is pretty painful to XMPP.
* Possibly redo the way QueueHandlers are injected into a QueueManager. The
grouping used to split out the XMPP output queue is a bit awkward.
Conflicts:
scripts/xmppdaemon.php
Previously, messages once delivered would just get stuck in the queue seemingly forever if they never got ACKed.
Note this could lead to partial duplication, for instance if the OMB or Twitter queue handlers die after 1/2 of the outgoing sends.
Recommendations:
* catch exceptions more aggressively within queue handlers (so only PHP fatal errors are likely to kill in the middle)
* for processing that involves sending to multiple clients, consider a second queue similar to the XMPP output, eg for OMB:
- first queue gets delivery list and builds message data, enqueueing it for each target address
- second queue can handle each individual outgoing message (and attempt redelivery etc separately)
This would also protect better against a recurring error preventing delivery in the second part, and could spread out any slow sends over multiple threads.
Queue handlers for XMPP individual & firehose output now send their XML stanzas
to another output queue instead of connecting directly to the chat server. This
lets us have as many general processing threads as we need, while all actual
XMPP input and output go through a single daemon with a single connection open.
This avoids problems with multiple connected resources:
* multiple windows shown in some chat clients (psi, gajim, kopete)
* extra load on server
* incoming message delivery forwarding issues
Database changes:
* queue_item drops 'notice_id' in favor of a 'frame' blob.
This is based on Craig Andrews' work branch to generalize queues to take any
object, but conservatively leaving out the serialization for now.
Table updater (preserves any existing queued items) in db/rc3to09.sql
Code changes to watch out for:
* Queue handlers should now define a handle() method instead of handle_notice()
* QueueDaemon and XmppDaemon now share common i/o (IoMaster) and respawning
thread management (RespawningDaemon) infrastructure.
* The polling XmppConfirmManager has been dropped, as the message is queued
directly when saving IM settings.
* Enable $config['queue']['debug_memory'] to output current memory usage at
each run through the event loop to watch for memory leaks
To do:
* Adapt XMPP i/o to component connection mode for multi-site support.
* XMPP input can also be broken out to a queue, which would allow the actual
notice save etc to be handled by general queue threads.
* Make sure there are no problems with simply pushing serialized Notice objects
to queues.
* Find a way to improve interactive performance of the database-backed queue
handler; polling is pretty painful to XMPP.
* Possibly redo the way QueueHandlers are injected into a QueueManager. The
grouping used to split out the XMPP output queue is a bit awkward.
Queue handlers for XMPP individual & firehose output now send their XML stanzas
to another output queue instead of connecting directly to the chat server. This
lets us have as many general processing threads as we need, while all actual
XMPP input and output go through a single daemon with a single connection open.
This avoids problems with multiple connected resources:
* multiple windows shown in some chat clients (psi, gajim, kopete)
* extra load on server
* incoming message delivery forwarding issues
Database changes:
* queue_item drops 'notice_id' in favor of a 'frame' blob.
This is based on Craig Andrews' work branch to generalize queues to take any
object, but conservatively leaving out the serialization for now.
Table updater (preserves any existing queued items) in db/rc3to09.sql
Code changes to watch out for:
* Queue handlers should now define a handle() method instead of handle_notice()
* QueueDaemon and XmppDaemon now share common i/o (IoMaster) and respawning
thread management (RespawningDaemon) infrastructure.
* The polling XmppConfirmManager has been dropped, as the message is queued
directly when saving IM settings.
* Enable $config['queue']['debug_memory'] to output current memory usage at
each run through the event loop to watch for memory leaks
To do:
* Adapt XMPP i/o to component connection mode for multi-site support.
* XMPP input can also be broken out to a queue, which would allow the actual
notice save etc to be handled by general queue threads.
* Make sure there are no problems with simply pushing serialized Notice objects
to queues.
* Find a way to improve interactive performance of the database-backed queue
handler; polling is pretty painful to XMPP.
* Possibly redo the way QueueHandlers are injected into a QueueManager. The
grouping used to split out the XMPP output queue is a bit awkward.
