GBrowse
view release on metacpan or search on metacpan
* Add option "track listing style" to select style of the track listing. Options are "categories" (default)
and "facets". Currently only "categories" works.
2.36
* Make semantic zooming work again when database option changed.
2.35
* Fix internal server error caused by certain semantic zooming declarations.
* Preparations for faceted track searching
* deprecate "subtrack select" and "subtrack table" options in favor of "subtrack facets",
which introspects values from database (will require updates to Bio::DB::SeqFeature and friends)
* add "facet labels" and "facet values" options
* allow options to be tagged with "ex:option_name" to identify facets
* Fix subtracks to prevent features from appearing in the subtrack manager that won't be
displayed in the track panel because the feature type doesn't match the configured "feature"
option.
* Add option "track listing style" to select style of the track listing. Options are "categories" (default)
and "facets". Currently only "categories" works.
2.34
* Fix broken https:// links in karyotype summary table; patch courtesy Dan Bolser
* Fix behavior of "q" parameter to match the documented behavior.
* Fix RemoteSet to avoid failures due to nonexistence of download directory.
* Fix bug that allowed unauthorized users to view restricted tracks under
following circumstances: FastCGI activated, track configuration changed, server
not restarted.
* Added an LDAP authentication plugin (LDAPAuthenticate).
* Clean up appearance of favorite stars and checkboxes to deconfuse users.
2.33
* Fixed plugin authentication system to prevent an all-dark screen during authentication
dialog on Firefox and Opera platforms.
* Fixed fasta and genbank track dumping so that "download all data" option works properly
rather than returning an empty file.
* Fixed sigPIPE warnings appearing in error log after an openid authentication.
* Removed some stray debugging messages.
2.32
* Fixed help menu so that "About GBrowse" works correctly again.
* Worked around issue of disappearing balloons under IE8.
* Fixed issue of malpositioned balloons under IE8.
* Fixed off-by-one error in decorated FASTA dumper when displaying borders
of minus strand features.
2.31
* Fixed "testdata/conf/languages/POSIX.pm: Permission denied" error when running Build test.
* Removed outdated instructions to install bioperl 1.6.9 from source.
2.30
* Improved behavior of track dumper so that when the user aborts the download the server
exits gracefully.
2.29
* Updated mod_fcgid "FcgidInitialEnv" directive to current recommended name.
* You can now narrow the search to features with a desired primary tag using the syntax
"desired_tag:feature_name"
* You can exclude a list of primary tags from search results in the database definition stanza
using the syntax:
[foo:database]
exclude types = type1 type2 type3
* The appearance of the track selection table has been updated to be less "busy".
* Uploaded track names assigned automatically (e.g. "upload_6842") can now be edited by
shift-clicking on the track name.
* You can now select the format for track data generated by clicking on the "disk" icon in
the track title bar.
* Fixed unwanted behavior in which searches that returned multiple features would open up all
tracks corresponding to those features even if user did not select them.
* Fixed "permission denied" error opening cgisess.db when installing on a fresh system.
2.28
* Added compile-time warnings about the bioperl version.
2.27
* THIS VERSION REQUIRES BIOPERL 1.0069 IN ORDER FOR SOME FEATURES TO WORK PROPERLY.
* Fixed account registration confirm code URL so that it doesn't get mangled by GBrowse redirection.
* Fixed the "restrict" argument so that you can restrict access to a track even if the datasource
as a whole does not require authentication.
* Changed default location of uploaded userdata to /var/lib/gbrowse2/userdata. This
avoids destruction of userdata on systems that clear /tmp on reboots. You may wish to copy existing
userdata to this location, or use symbolic links.
* The apache configuration file, /etc/apache2/conf.d/gbrowse2.conf, now contains a stanza that
makes the databases directory inaccessible to the web server unless explicitly enabled.
* All newly-added tracks now appear at top of panel rather than at bottom, as per user request.
* "remote renderer" option can now appear in GBrowse.conf as a global default
* Fixed handling of uploaded .gz files under FastCGI so that they don't result in a server error message.
* Changed handling of uploaded wiggle files so that autoscaling is set to per-chromosome by default.
* Added the ability to designate certain tracks as "favorites."
2.26
* Fixed account registration confirmation screen, which was blacked out on all
platforms.
* Fixed account registration error messages to be more informative.
* Removed Ace as a recommended library. Was only used for process_wormbase.pl,
which is no longer relevant anyway.
* Fixed problem calculating maximum length of displayed segment when details_multiplier>1.
2.25
* Fixed broken image generation.
* Made it easier to install GBrowse as an unprivileged user.
* Allow searching for features with pipe symbols in their names.
* Allow searching for features defined in uploaded tracks.
2.24
* Fixed gbrowse_metadb_config.pl script so that it no longer fails when creating a user database
from scratch.
2.23
* Added an "example_scripts" directory that contains a perl scripts that show how to manipulate
sessions, upload custom tracks, and get basic information about services. This directory accompanies
new REST API documentation located at http://gmod.oicr.on.ca/wiki/GBrowse2_REST_API.
* Fixed missing "registration" link in login dialog. This occurred when openid was unavailable.
* Fixed popup dialogs in the "Help" menu so that they appear again.
* Fixed file upload status display to prevent upload details section from disappearing when
uploading very small files.
* Tinkered with session settings to minimize session lock timeouts.
* Removed some dangling warnings in the track sharing module.
* Made SQLite the default user account database.
* Fixed default SQLite user account database to be created in /var/www/gbrowse2/databases rather
than in / (root) directory.
2.22
* Added a datasource "seqid_prefix" option that allows you to add and strip "chr" style prefixes from all seqids;
this allows for interoperability with UCSC, which always wants a "chr" to begin each chromosome.
* Fixed loading of SAM files to avoid loading of incorrect fasta file when both .fa and .fa.fai files
are present in the scaffolds directory.
2.13
* Fixed SAM file processing so as not to die on first line of SAM @SQ header.
This requires Bio::DB::Sam 1.20 to work properly.
* Fixed upload listing table so as not to duplicate track labels.
* Fixed track importation via the eurl parameter so that it works when the
user's uploadid is empty (due to clearing cache of cookies).
* Fixed DAS server handling of the link option.
2.12 Wed Jun 30 15:28:16 EDT 2010
* Fixed FASTA dumping due to typo introduced just prior to 2.11 release.
2.11 Mon Jun 28 11:42:02 EDT 2010
* Fixed generation of mysql database names for user upload sets so that names do
not exceed mysql's limits.
* Speeded up execution speed when configuration files get large (>1000 tracks).
* DAS server and client now functional.
* Reenabled the "eurl" parameter to link to a data set from the URL.
* Added "Get chromosome sizes" entry to the File menu (requires a developer's snapshot of
bioperl to work with all data sources).
* Fasta and GFF3 dumps working ok now.
* You can now import a BigWig file by reference.
* Uploaded WIG files will generate bigwig indexes if library is available.
* action=scan now returns the category for each track.
2.10 Tue Jun 15 10:58:32 EDT 2010
* Fully generalized subtrack creation.
2.09 Mon May 24 18:45:48 PDT 2010
* To turn off caching of the config file (which is causing deparse errors for some
callbacks) set the environment variable GBROWSE_NOCACHE to a true value
(like "1" or "yes"). This must be done in the Apache config.
* Fixed various bugs in the processing of SAM and BAM files.
* BAM parser now creates BigWig indexes if library is available.
* Fixed "no overloaded magic in Bio::DAS::Segment" error.
* Made Menus and tabbed section labels larger (if you don't like,
edit htdocs/css/dropdown/default-theme.css).
* Moved track table into its own tabbed section.
* TrackFinder plugin now hilights the matches in the track table.
* New "subtrack table" and "subtrack select" options allow you to create nice
sortable tables of potential subtracks. The old "select" option is still
supported, but deprecated.
* Added "kewl" blinds down effect when changing tabs.
* Greatly improved track configuration.
* From E.O. Stinson: added fixes for embedding GBrowse in other web pages.
* From E.O. Stinson: Added support for regex-style datasources that let you define
one data source that will pass different options to a script-based config depending
on the URL.
* From E.O. Stinson: Added a move_to_name function for navigating to a region in one step
via a URL call.
2.08 Thu May 20 22:46:13 EDT 2010
* Added the ability to control the toggle state of each category and subcategory at
startup time by using the "category state" and "category default state" options
(see the GBROWSE2 HOWTO at http://gmod.org/wiki/GBrowse_2.0_HOWTO#Appearance_Settings)
* Fixed searches on Bio::DB::GFF databases so that exact searches work again.
* Fixed the busy spinner so that it doesn't stick in "on" state when region is empty
and user turns on tracks.
2.07 Mon May 17 10:27:56 EDT 2010
` * Tracked down the "let Lincoln know bug is not fixed" bug, and found a real solution.
This was happening when user presses "back" button on browser after an unsuccessful
search. This still triggers a log message until we know the bug is really and truly
fixed.
* BAM track download now works properly under both FastCGI and mod_perl. mod_perl
download requires Bio-SamTools 1.19.
* Requires Bio::Graphics version 2.09, which addresses a bug in which subroutines
defined in init_code could not be referenced within callbacks.
2.06 Thu May 13 17:10:59 EDT 2010
* Fixed dangling references to GFFPrinter.pm module, which was renamed TrackDumper.pm
* Requires Bio::Graphics version 2.08, in order to fix the init_code option, which stopped
working when config file caching was introduced in Bio::Graphics 2.07.
2.05 * Implemented a new caching scheme for communicating with renderfarm slaves.
This will improve performance, but requires Bio::Graphics version 1.05.
* Experimental local rendering code -- each track is rendered in a separate process.
* Reverted to older versions of prototype and scriptaculous due to Google
Chrome incompatibilities.
* Preferred way to separate labels is now to use the "l" option and separate
labels with "%1E", as in "l=mdc-1%1Esma-23%1Eunknown genes"
* Work around race conditions in the 05.deferredrendering regression tests.
May still fail if system is under lots of load during test.
* Fixed bug in change_track_order javascript that prevented track order change from
"sticking" when track labels contained underscores.
* Download icons now selectively download GFF3, SAM or WIG data.
* Early support for a "show_summary" option in [TRACK DEFAULTS] and [TRACK] stanzas. If this
is present and set to a length in bases, then when the user tries to display a region
larger than the length the track will display feature density instead. For this to
work, the database adaptor must support a feature_summary() method. This is being actively
worked on in bioperl, but is not currently available.
2.04 * Can set the subtracks to be displayed by default using the new "select default" option.
* New syntax for select option:
select = type ;
dpy-27 'dpy-27 chip' =101;
htz-1 'htz-1 chip' =102;
pol2 'polymerase II' =103;
sdc-3 'sdc-3 chip' =104;
Everything following the = sign is an optional comment (modencode uses it for the submission ID).
The old syntax for this still works.
* Added a TrackFinder plugin which will limit tracks according to their key,
citation, or "keywords" option, based on keyword matches the user types in.
* Prevent gbrowse_img zombie processes from accumulating under FastCGI.
* Generate correct Galaxy callback URL and inhibit menu item when Galaxy not configured.
* OligoFinder plugin now works with gbrowse2.
* Updated js libraries to prototype 1.6.1 and scriptaculous 1.8.3.
2.03 * GBrowse_details: display of aligned sequences improved.
* Can now paste remote BAM URLs into "import a track" field.
* Private (authenticated) sessions never expire.
* Fixed bug that led to a keyword search always returning multiple hits
* Added a scalebar on top of the details ruler.
* Fixed bug in installation paths of example databases (rbuels)
* Bugfix to support for subroutines in user-defined 'head' configuration options (rbuels)
* Bug fix to avoid hitting database each time the client checks on the status of a render request.
* Bug fix to treat timed-out cache operations the same as timed-out render operations.
* Bug fix to interrupt long system operations, like MySQL queries.
2.02 * Sheldon McKay rolled gbrowse_syn back into gbrowse2
( run in 0.524 second using v1.01-cache-2.11-cpan-600a1bdf6e4 )