Skip to content


Subversion checkout URL

You can clone with
Download ZIP
Commits on Mar 1, 2012
  1. @LemonTVnz
  2. @LemonTVnz
Commits on Feb 28, 2012
  1. @daniel-kristjansson
  2. @gigem

    As noted in SHA1:9551bce2, MainServer doesn't know the actual

    gigem authored
    recording status because it defaults to rsRecorded when the
    recordinginfo is initialized from the recorded table.  MainServer
    doesn't account for this and tries to make any needed state change
    which always ends up staying rsRecorded.  MainServer then passes this
    bad status to the scheduler.  If the recording had already failed, the
    rsFailed status gets overwritten.
    Since the scheduler knows the correct status, change the status to
    rsUnknown in MainServer before updating the scheduler.  The scheduler
    now notices the special value and applies any needed state change
    itself.  In the case of a failed recording, the status now remains
    Also, remove the duplicate code in MS::DoHandleDeleteRecording() used
    to stop an in-progress recording and use MS::DoHandleStopRecording()
    instead.  Try to avoid unnecessary reschedules when deleting programs
    from the Deleted and LiveTV recording groups.
    Fixes #10282
  3. @nriendeau

    Spelling consistency fixes

    nriendeau authored
    Replace all (currently) known occurences of translatable strings using the British
    and/or Canadian spelling with the US spelling since this is the default spelling
    which should be used throughout the application.
    Mapping these back to the proper spelling for each variant of English will be done
    using the translation files like the rest of the strings.
    After applying this the only translation context which could possibly contain words
    with a spelling other than the US one will be the "ThemeUI" context were the theme
    maintainers will be free to use the spelling they prefer.
  4. @daniel-kristjansson

    Fixes #9233. Refs #7408. Re-enables WindowResized().

    daniel-kristjansson authored
    We traded a one small bug for two others just prior to 0.24.
Commits on Feb 27, 2012
  1. @daniel-kristjansson

    Fixes #10153. Guess at DVB-S vs DVB-S2 when not explicitly set.

    daniel-kristjansson authored
    The problem was discovered by Lawrence Rust. It has been decided not to fix the DVB API regression upstream, so this code attempts to guess which system is being used using the modulation scheme as a hint.
  2. @daniel-kristjansson

    Fixes #10377. Prevent getting "stuck on a mux".

    daniel-kristjansson authored
    When multiple digital cards share a video source we can get stuck to a mux in Live TV. This fixes that bug. See ticket for steps to reproduce.
  3. @daniel-kristjansson
  4. @stuartm

    Change the rsTuning UI state equivalent from 'tuning' to 'running'. T…

    stuartm authored
    …he UI states for the 'status' statetype are limited to 5 generic values (normal, running, warning, error, disabled) to allow some portability to scenarios other than recording. (This was supposed to save themers some time and effort although it also binds their hands and doesn't allow as much creativity with how various recording states are represented. For this reason it may be re-visited post-0.25)
  5. @nriendeau

    Updated Russian MythFrontend, MythArchive, MythBrowser, MythGallery,

    Alex Vasilyev authored nriendeau committed
    MythGame, MythMusic, MythNews, MythWeather and MythZoneminder.
    Closes #10387
    Signed-off-by: Nicolas Riendeau <>
  6. @stichnot

    Subtitles: fix an incorrect CEA-708 font mapping.

    stichnot authored
    SetPenStyle was mapping the style to the wrong font tag.  This
    generally only showed up as a slightly incorrect indent for subtitles
    with leading spaces.
  7. @Beirdo

    Fix the faulty exit code 139 issue in Ubuntu systems

    Beirdo authored
    I think it might be an oddity in eglibc, but the status return from waitpid()
    is buggered sometimes for signalled children.  Rather than showing up as signal
    11 for a segfault, it shows up as an exit value of 139.  This is the most
    common offender, but signal 6 (exit value 134) shows up sometimes as well.
    The workaround limits the valid exit values of the child to being 0-127.  An
    exit value above that is assumed to be a buggered up signal return and is
    As this is most prevalent with mythcommflag, the workaround is only active
    for mythcommflag (for now).  If other MythSystem calls need this added, we
    can add it later.
  8. @stichnot
Commits on Feb 26, 2012
  1. @stuartm

    Fixes display of position in the Program Lister. MythUIButtonList is …

    stuartm authored
    …zero indexed, to get a position for display we must add one.
  2. @gigem
  3. @gigem

    Fix the scheduler to not write incorrect oldrecorded entries.

    gigem authored
    Don't modify the recording status of programs that aren't eligible to
    be rescheduled.  Previously, the scheduler properly ignored most
    status changes caused by this oversight.  The improved missed
    recording detection changed that and allowed some status changes to
    get rewritten to oldrecorded.  In the particular case of stopping or
    deleting an in-progress recording, the status could change from
    rsRecorded and affect future duplicate detection.
  4. @stichnot

    Subtitles: improve the layout.

    stichnot authored
    With the existing PAD_WIDTH value, there was still some clipping going
    on, for example an italic "W" at the end of the line with the Droid
    Sans Mono font.  Simply increasing PAD_WIDTH would have made the extra
    spacing between adjacent but differently-formatted text chunks much
    more noticeable.
    This is improved by increasing PAD_WIDTH but no longer adding padding
    between chunks.  A consequence is that the background rectangle now
    has the possibility of clipping the text to its left, so there is a
    final step that moves all text objects to the front (rather than
    drawing background and text in separate passes).
    Refs #10194.
  5. @knowledgejunkie @nriendeau

    String consistency updated

    knowledgejunkie authored nriendeau committed
    Thank you Nick!
    PS: This is essentially what you submitted with a few minor modifications on my part.
    Closes #10374
    Signed-off-by: Nicolas Riendeau <>
  6. @nriendeau

    Updated Estonian MythFrontend and all plugins translation

    Marko Punnar authored nriendeau committed
    from Marko Punnar
    Thank you Marko!
    Closes #10385
    Signed-off-by: Nicolas Riendeau <>
  7. @stuartm
  8. @knowledgejunkie @nriendeau

    Use proper capitalization for some common words and prefixes.

    knowledgejunkie authored nriendeau committed
    Thank you Nick!
    Closes #10380
    Signed-off-by: Nicolas Riendeau <>
Commits on Feb 25, 2012
  1. @stuartm

    Populate the infomap from both RecordingRule and RecordingInfo if ava…

    stuartm authored
    …ilable. This fixes recording rule specific information not being available in the sub screens of the schedule editor.
  2. @stichnot

    Fix minor problems editing in-progress recordings.

    stichnot authored
    Revert 5d7c01d, and add logic that also updates the edit bar when
    the total number of frames in the recording changes.
    Modify 1ce3d21 to use a better test for an in-progress recording.
    If the recording finishes during the editing session, the original
    test would continue to indicate that the recording was in-progress.
    Binary version is updated.  Refs #10329.
  3. @stuartm

    Fix recording status strings being set to 'Not scheduled' when the re…

    stuartm authored
    …ason for that is they were in the past and should have been marked as Recorded/Aborted/Failed/Cancelled/etc instead
  4. @wagnerrp

    Correct handling of unrelated argument after arg-less keyword.

    wagnerrp authored
    This corrects an issue where an independent argument following a keyword
    that did not take an argument would produce a fault in the parser. This
    handles it properly and passes the argument back through for further
  5. @wagnerrp

    Delay String processing until after QApplication has been initialized.

    wagnerrp authored
    This stores strings, stringlists, and maps containing strings as
    QByteArrays instead, and waits until QApplication has been created and
    configured proper text behavior based off the system locale, before
    processing them into QString values.  This resolves an issue where
    >8-bit unicode text passed on the command line would result in invalid
    strings when used.
    This bumps the library version.
    Fixes #10371
  6. @gigem

    Fix LiveTV switching by videosource to honor livetvorder.

    gigem authored
    Change the GET_FREE_RECORDER_LIST protocol request to return the
    cardids in the desired order based on livetvorder.  Change TV to not
    re-sort the cardids when mapping sources to an inputid.
    N.B. There is still a corner case involving analog cards connected to
    multiple sources that could still result in the wrong input being
    chosen.  Fixing that would require a bigger change including a
    protocol change and can wait until after the 0.25 release.
  7. @stichnot

    Exit cutlist edit mode cleanly when executing a jump point.

    stichnot authored
    When the user executes a jump point in the middle of an editing
    session, make sure the edit session closes cleanly.  The work is not
    explicitly saved, but the auto-save state is retained in the database
    so that it will be auto-loaded next time the user edits that program.
    This makes a slight modification to the MythPlayer class and therefore
    the binary version is bumped, so be sure to do everything that
    Refs #7939.
  8. @knowledgejunkie @nriendeau

    Updated British English translation MythFrontend translation

    knowledgejunkie authored nriendeau committed
    from Nick Morrott.
    Thank you Nick!
    Closes #10379
    Signed-off-by: Nicolas Riendeau <>
Commits on Feb 24, 2012
  1. @stichnot

    Make jump points work during playback when the OSD is present.

    stichnot authored
    This works by sending an EXIT_TO_MENU event at the start of jump point
    execution.  Any existing TV playback instance will receive this message
    and manually shut down playback, including tearing down the OSD.
    This is not the fully general solution desired, but it is a big
    improvement until then.
    One issue is that an editing session is not shut down cleanly.  That
    will be addressed in a subsequent commit.
    Refs #7939
  2. @stuartm

    Change dynamic cast to static. This is troubling because we rely on d…

    stuartm authored
    …ynamic casts in so many places but for at least one person this cast was failing and causing popup menus in MythMusic to no longer have any effect. Maybe a buggy version of gcc?
  3. @stichnot

    Subtitles: Fix a logic error.

    stichnot authored
    At a minimum, this error resulted in only the first CEA-708 window
    being displayed when multiple windows were active.
  4. @wagnerrp
  5. @stichnot

    Cutlist editor: Update the edit bar for in-progress recordings.

    stichnot authored
    When editing an in-progress recording, the edit bar only updates when
    the cutlist is modified, but not when the recording length changes.
    However, the position indicator does get updated, resulting in the
    edit bar graphics getting out of sync.
    This is fixed by disabling the (seemingly premature) optimization that
    updates the edit bar only when the cutlist changes.  This fix will
    need to be reevaluated after the 0.25 release.
    Refs #10329
Something went wrong with that request. Please try again.