Flyspray:: Flyspray:: Rockbox: Recently opened tasks 2014-09-29T09:26:16Z FS#13007: FM Region "Other": Enable full frequency range. 2014-09-29T09:26:16Z 2014-09-29T09:26:16Z
This patch changes the frequency range when selecting "Region: Other" for the FM radio:

With this patch, the "full range" from 76 MHz to 108 MHz, in 0.05 MHz steps, is available when selecting the "Other" region.
Oid Maps
FS#13006: Gapless playback of OGG Vorbis is not working. 2014-09-27T05:58:58Z 2014-09-27T05:58:58Z
Gapless playback of OGG Vorbis is not working. High pitched background noise for a while before next track begins to play.
Official firmware plays same files gaplessly without the background noise.
Osmo Hyttinen
FS#13005: High-pitched background noise NOT PRESENT on official firmware 2014-09-27T20:49:00Z 2014-09-27T05:54:39Z
High-pitched background noise when starting playback, file changes during playback or during heavy I/O operations.
NOT PRESENT on official firmware
Osmo Hyttinen
FS#13004: ATA Error 11 with mSata SSD installed using mSata-ZIF adapter 2014-09-26T09:13:35Z 2014-09-26T09:13:35Z
I have an upgraded iPod 5.5 with CF 32Gb installed in CF-ZIF adapter.
Both iTunes & Rockbox are working fine with this setup.

Then I tried to increase the storage space by using mSata SSD 64 GB + mSata-ZIF adapter. iTunes recovered the iPod successfully and it is working just fine. I installed Rockbox latest stable (non-stable latest = same thing) version & after the Rockbox boot-up it reporting an ATA error -11.

I wonder if this can be solved.

FS#13003: Scrolling line broken if put directly after sublines. 2014-09-24T08:38:27Z 2014-09-24T08:38:27Z
## This was tried out with Sansa Fuzev2 simulator version d392da8-140922, built from git on a 64 bit Linux system ##

I discovered something which seems to be a bug somewhere in the place where WPS get interpreted or rendered.

When I use a scrolling line just after a line containing sublines, the scrolling line stops scrolling as soon as the sublines change the first time.

Below are steps to reproduce it on the UI simulator for the Sansa Fuzev2. I think it also works without using a clear configuration. I haven't tested with a real device, only with the simulator.

Attached are a .wps-file showing the buggy behaviour ('scrollline_after_sublines_bug.buggy.wps'), and a .wps-file working correctly ('scrollline_after_sublines_bug.workaround.wps'), which are the same files as quoted below.


(0) Use Sansa Fuzev2 simulator version d392da8-140922, build from git on a 64 bit Linux system.

(1) Use a clear configuration, so erase
in your 'simdisk' directory.

(2) Start the rockbox UI simulator

(3) Change the theme to 'rockbox_failsafe', i.e. in the menu, do the following:
* 'Settings' -> 'Theme Settings' -> 'Browse Theme Files' -> 'rockbox_failsafe.cfg'
* 'Settings' -> 'Theme Settings' -> 'While Playing Screen' -> 'rockbox_failsafe.wps'
* 'Settings' -> 'Theme Settings' -> 'Radio Screen' -> 'rockbox_failsafe.fms'
* 'Settings' -> 'Theme Settings' -> 'Base Skin' -> ''

(4) Use the following .wps-file as the "While Playing Screen" skin:

# Two sublines:
# A long line, scrolling:
%sA very long text which is meant to scroll because it is longer than the screen is wide.
# Another line of just text (Not needed to reproduce the bug, but if it is swapped with the long line above, the long line scrolles correctly):
#### Here the following Bug appears:
# When the Line 'Text1;Text2' changes the sublines the first time, the subsequent line stops scrolling. It happens when the sublines change; as when you use '%t(5)Text1;Text2', the scrolling stops only after 5 seconds, when the sublines actually change.

(4a) Put the scrolling line not directly after a line with sublines. The following .wps as the "While Playing Screen" skin works:

# Two sublines:
# Another line of just text. Needed to make the long line scrolling correctly:
# A long line, scrolling:
%sA very long text which is meant to scroll because it is longer than the screen is wide.
#### Here the bug does not appear, the long line scrolls corretly. But when the long scrolling line is put just below the line with the sublines, it stops scrolling as soon as the sublines change for the first time.
Oid Maps
FS#13002: Fails to build from source under long build path with "execvp: /bin/sh: Argument list too long". 2014-09-22T15:21:44Z 2014-09-22T15:21:44Z
When I try to build rockbox, I get the following error:

make: execvp: /bin/sh: Argument list too long

My build path is a fairly long path. Building it at a location with shorter path it works. Details:

Steps to reproduce:

* Create a location where you want to have the sources; use a fairly long path name, e.g.

mkdir -p /home/myuser/download/mediaplayer_hardware/SanDisk/Sansa_Fuze_v2/rockbox/firmware/source/

* Change to that directory:

cd /home/myuser/download/mediaplayer_hardware/SanDisk/Sansa_Fuze_v2/rockbox/firmware/source/

* Download git sources (Tried with version from 2014-09-22, 10:55:11 GMT+1, Commit by "Marcin Bukat <>", Change-Id: "Ie3aa9b208e3f4f17d4d02f11f69839e9b381217d") to a long path:

git clone git://

* Make a build directory and change into it:

mkdir -p /home/myuser/download/mediaplayer_hardware/SanDisk/Sansa_Fuze_v2/rockbox/firmware/source/build_FuzeV2
cd /home/myuser/download/mediaplayer_hardware/SanDisk/Sansa_Fuze_v2/rockbox/firmware/source/build_FuzeV2

* Configure the build:

../rockbox/tools/configure --target=63 --type=N

prodces the output

Using temporary directory /tmp
Platform set to sansafuzev2
Normal build selected
Using source code root directory: /home/myuser/download/mediaplayer_hardware/SanDisk/Sansa_Fuze_v2/rockbox/firmware/source/rockbox
Using arm-elf-eabi-gcc 4.4.4 (404)
Using arm-elf-eabi-ld
Found and uses ccache (/sbin/ccache)
Automatically selected arch: arm (ver 5)
Created Makefile

* Running "make":


produces the output

Generating dependencies
make: execvp: /bin/sh: Argument list too long
make: *** No rule to make target '/home/myuser/download/mediaplayer_hardware/SanDisk/Sansa_Fuze_v2/rockbox/firmware/source/build_FuzeV2/make.dep', needed by 'all'. Stop.

and returns with exit code 2.


Building it in a directory with shorter path, e.g. doing everything in "/tmp", works.

But that is only a workaround; it should not fail when building in arbitrary locations, and especially not with a message not making it clear _what_ the underlying problem ist. (I had to do a long internet search to get to the idea that the build place can be the problem).
Oid Maps
FS#13001: Portugal Portuguese Update 2014-09-21T22:37:33Z 2014-09-21T22:37:33Z
Attached you'll find a DIFF for Portugal Portuguese.

Tiago Medeiros
FS#13000: Audio drop-out with hi-res file playback 2014-09-19T20:47:11Z 2014-09-19T20:47:11Z
I have a couple of hi-res files (24/44) that seems to have audio volume drop-outs at certain points, as if it is an over-reacting dynamic clip prevention. If I try the same file in Foobar 2000, I do not get the sound drop-outs. Dithering is turned on, ReplayGain is on "album", and I tried it both with and without clipping prevention, but both times I got the drop-outs.

The file is too large to attach here, but I can provide it via other means. They are tracks from Steven Wilson - The Raven That Refused to Sing. I resampled and dithered the official 24/96 release to 24/44 using iZotope RX 3 Advanced. I can provide both resampler and dither settings if desired, or I can just provide the effected tracks (title track, "The Watchmaker", both near the end), or the files themselves.
Aaron Tate
FS#12999: Manual says there are 2 different builds of iPod 5G 2014-09-16T19:29:39Z 2014-09-16T18:41:12Z
Getting last iPod 5G manual (found here: page 17), it says there are two different versions of Rockbox for this player. It was true some time ago, but now (even in release page), there are only one version that works with any model of iPod 5G.
Arthur Ferrai
FS#12998: Vertical progress bar (%pb) glitches at the start of a song 2014-09-16T17:12:26Z 2014-09-16T17:12:26Z
Consider the following theme markup:


That draws a vertical progress bar that will get filled from bottom to top. When a song starts to play, the progress bar is one pixel high, at the correct position (at the bottom of the viewport). That 1px-high image quickly jumps up one pixel (moves up by one pixel), then comes back down one pixel (at the correct position) whenever the progress percentage is high enough. The progress bar then displays normally and grows as the song is playing. See attached screenshots.

Expected behavior: the position of the progress bar (relative to the viewport) should not change. The bug has been observed in the iPod Classic and Clip Zip sims (builds 1bc89bd and 77bfff5) as well on the actual devices. Inverting the progress bar (with the "invert" parameter) shows the correct behavior (inverted bar with no such glitch).

Affected themes: Ash (iPod Classic), SugarBrown (Clip Zip).
Guillaume Cocatre-Zilgien