| Commit message (Collapse) | Author | Age | Files | Lines |
| |
|
|
|
|
|
|
|
|
|
| |
Version 7 introduced slight API changes which are handled accordingly.
Legacy version 6 has equivalent support in ODR-PadEnc and there are no
plans to remove support for the next years.
See also #3.
|
|
|
|
|
| |
Move them right before the loop, as before a break (e.g. while waiting
for the audio decoder opening the pipe) didn't lead to termination.
|
| |
|
|
|
|
|
|
|
|
|
| |
When the opposite side of the PAD FIFO disconnects (e.g. due to crash of
the audio encoder), ODR-PadEnc crashed as well due to SIGPIPE.
This commit changes the behaviour. From now on instead error messages
are displayed and it is possible for a new instance of the audio encoder
to reconnect to the still running ODR-PadEnc instance.
|
|
|
|
|
|
| |
- fix compile warning w/o MagickWand
- limit variable scope
- fix variable types
|
|
|
|
|
|
|
| |
The variable survived a refactoring and is orphaned since. Only
compilation without MagickWand was affected.
Fixes #4.
|
|
|
|
| |
Affects #3.
|
|
|
|
|
| |
Using the C++ versions does not guarantee the availability of the
global namespace functions.
|
|
|
|
|
| |
Add signal handler for SIGINT and SIGTERM that closes the output socket
and deinits ImageMagick (if used).
|
| |
|
| |
|
|
|
|
|
|
|
| |
Ubuntu 16.04 currently ships an ImageMagick version that was affected by
a memleak and is fixed since v6.9.2-2: http://git.imagemagick.org/repos/ImageMagick/commit/6790815c75bdea0357df5564345847856e995d6b
Closes #2.
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
| |
A memleak occured when non-raw mode was used and a slide complied to the
following conditions:
- PNG/JPG file
- 320x240 resolution or less
- size above max slide size (~50 KB)
Closes #2.
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
|
| |
Up to now the delay between two loop runs was slightly higher than the specified
sleep delay as the sleeping process did not consider the time the PAD encoding
needs but rather sleeped for the set sleep delay.
This is now fixed by using a steady clock.
|
| |
|
|
|
|
|
|
|
| |
The alphabetic sorting within the slide dir was ignored so far. Now the sorting
is considered while adding slides for transmission.
Furthermore now also files starting with "." are added (except the ./.. dirs).
|
| |
|
| |
|
| |
|
| |
|
|
|
|
| |
- SLSManager: store PADPacketizer ref + fix spacing
|
| |
|
|
|
|
|
|
|
|
| |
- DLSManager: store PADPacketizer ref
- DABCharset: convert enum to enum class
- DL cmd prefix: output reserved instead of charset (does not make a difference)
- MOT header: get charset from enum class
- .gitignore: add Eclipse files
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
When a slide is transmitted, now every 50 PADs the current DLS is inserted
(and thereby reread from file). Hence in the best case (AAC-LC @ 48 kHz), DLS
is inserted every 1000ms; in the worst case (HE-AAC @ 32 kHz) it is inserted
every 3000ms. This way a listener will get DLS much earlier after switching to
a service, compared to the previous situation where the slide transmission was
not interrupted for DLS insertion.
Note that there still remains a delay without any PAD, between the end of a
slide transmission and the start of the next one's transmission.
Also note that adding a feedback channel from the audio encoder to ODR-PadEnc
will make this improvement obsolete.
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
|
| |
Parses an optional ".sls_params" file for each slide and hence allows to set:
* CategoryID/SlideID
* CategoryTitle
* ClickThroughURL
* AlternativeLocationURL
|
| |
|
|\
| |
| | |
converted the crc impl into c++ with namespace.
|
|/
|
|
|
| |
Since it conflicts with other crc implementation in imagemagic lib.
This resulted in png images that would not open due to "crc error".
|