• 1
  • 43
  • 44
  • 45(current)
  • 46
  • 47
  • 111
OpenELEC Testbuilds for RaspberryPi (Kodi 15.0) Part 1
New OpenELEC I****** build #0219: RPi / RPi2
(Supercedes previous build)

Code:
# uname -a
Linux rpi512 3.19.0 #1 Fri Feb 20 03:02:17 GMT 2015 armv6l GNU/Linux

# vcgencmd version
Feb 18 2015 16:20:59
Copyright (c) 2012 Broadcom
version bcbf8eb39cfacb4b2fd648a080d19d22152c359b (clean) (release)

# lsb_release
OpenELEC (Milhouse) - Version: devel-20150220030127-#0219-g805bf0b [Build #0219]

# vcdbg log msg 2>&1 | grep DTOK
001538.692: Kernel trailer DTOK property says yes

# Kernel device tree status: Enabled

Based on tip of OpenELEC master (805bf0b8, changelog) and tip of XBMC master (d6f8a655, changelog) with the following modifications: Build Highlights:
  1. Improved MySQL transaction handling
  2. VideoDB queueing
  3. Revert SSE change that breaks non-x86 builds
  4. Add support for jumping to the first unwatched tv show season/episode
Build Details:
  1. XBMC:
    • [depends] enable C++11 for native builds (PR:6485, 1 commit, 1 file changed)
    • Add optional .index() to Control.GetLabel() allowing access to secondary... (PR:6327, 1 commit, 4 files changed)
    • [configure] enable SSE when using gcc-4.9 (PR:6487, 1 commit, 1 file changed)
    • [configure.in] use pkgconfig to detect curl and force the distros to use... (PR:6407, 1 commit, 1 file changed)
    • [gui] remove eventual context menu items for the "go to parent" item (PR:6290, 2 commits, 2 files changed)
    • Locale fixes (PR:6402, 9 commits, 13 files changed)
    • fixes for a few coverity issues (PR:6422, 7 commits, 9 files changed)
    • video library: queue database write processes (PR:6406, 13 commits, 33 files changed)
    • [mysql] fix transaction handling (PR:6467, 1 commit, 1 file changed)
    • Add support for jumping to the first unwatched tv show season/episode (PR:6489, 1 commit, 7 files changed)
    • revert xbmc.gui bw-compatibility change (PR:6494, 1 commit, 1 file changed)
    • cmake: fix include of check_target_platform.cmake (35b7ab3f)
    • [Confluence] api bump (d6f8a655)
  2. Additional commits/pull requests/changes not yet merged upstream:
    • Reverted: PR:6487: [configure] enable SSE when using gcc-4.9 (Reason: breaks non-x86 builds)
Texture Cache Maintenance Utility: Preload your texture cache for optimal UI performance. Remotely manage media libraries. Purge unused artwork to free up space. Find missing media. Configurable QA check to highlight metadata issues. Aid in diagnosis of library and cache related problems.
I cannot get any liveTV to work past release #213
#219,#218,#217 not working

I am using Argus TV

On selecting live TV Gui and schedule database work fine but selecting live TV busy notify briefly appears followed by black and causes kodi to crash and then reboot??

No changes made to any settings

Logs here
https://www.dropbox.com/sh/31t8upajt9zw1...5Pfja?dl=0
Location UK; Media server Windows 7 with ArgusTV 2.3 with TBS6981 DVB-S2 x4 and DVB-T x 2:All network connections cabled on 1Gb router Raspberry Pi2 1GB x 2; RPI 3 x1; PiB+512MB x 3; TV Samsung 55" C8000; AV Denon AVR X2200W
So #0213 is the last working build, or the first non-working build? Are you able to test on a Pi1, do you get the same crashes?
Texture Cache Maintenance Utility: Preload your texture cache for optimal UI performance. Remotely manage media libraries. Purge unused artwork to free up space. Find missing media. Configurable QA check to highlight metadata issues. Aid in diagnosis of library and cache related problems.
edit: brainfart
Well, it's not wrong - I stand by what I wrote. Yes, adding <seeksteps> to as.xml only changes the allowable values, and you then have to select/apply (ie. configure) the values in Settings in order to actually use them. I admit the final step might not have been made clear.

If you don't apply new settings, you'll still be using the original skip steps even though the allowable values in as.xml may now be completely different.
Texture Cache Maintenance Utility: Preload your texture cache for optimal UI performance. Remotely manage media libraries. Purge unused artwork to free up space. Find missing media. Configurable QA check to highlight metadata issues. Aid in diagnosis of library and cache related problems.
(2015-02-20, 13:39)Milhouse Wrote: So #0213 is the last working build, or the first non-working build? Are you able to test on a Pi1, do you get the same crashes?


Just checked Rpi 1 crashes just the same on builds #014 onwards... last working version #013
Location UK; Media server Windows 7 with ArgusTV 2.3 with TBS6981 DVB-S2 x4 and DVB-T x 2:All network connections cabled on 1Gb router Raspberry Pi2 1GB x 2; RPI 3 x1; PiB+512MB x 3; TV Samsung 55" C8000; AV Denon AVR X2200W
Now that I understand the new seekstep feature, it seems to me that the default 1st step of 7 seconds is too short. For one thing, it takes several seconds, on a pi1 at least, for the skip to happen. It feels broken. I can tell you, until we figured out what was going on, there was much wailing and gnashing of teeth in our family room.

I suggest that the 7 second interval be disabled by default.

Is this in master or newclock4?
(2015-02-20, 16:22)allan87 Wrote: Now that I understand the new seekstep feature, it seems to me that the default 1st step of 7 seconds is too short. For one thing, it takes several seconds, on a pi1 at least, for the skip to happen. It feels broken. I can tell you, until we figured out what was going on, there was much wailing and gnashing of teeth in our family room.

I suggest that the 7 second interval be disabled by default.

Is this in master or newclock4?

let me remind you that not everything revolves about the r-pi and some/most people have a device that is fast enough to handle the 7seconds
Read/follow the forum rules.
For troubleshooting and bug reporting, read this first
Interested in seeing some YouTube videos about Kodi? Go here and subscribe
(2015-02-20, 16:22)allan87 Wrote: Is this in master or newclock4?

This is in master.
(2015-02-20, 16:22)allan87 Wrote: Now that I understand the new seekstep feature, it seems to me that the default 1st step of 7 seconds is too short. For one thing, it takes several seconds, on a pi1 at least, for the skip to happen. It feels broken. I can tell you, until we figured out what was going on, there was much wailing and gnashing of teeth in our family room.

I suggest that the 7 second interval be disabled by default.

Is this in master or newclock4?

I did comment on this before as I had issues with the skips.
after pressing the skip button it is waiting for you to press again (for a higher skip value).
This results in you getting a delay before playback is resumed (if you are only skipping once).
I found the best thing to get old behaviour back was to disable the skip variants that you don't wan't.
I.E. disable all but +/- 30 seconds (or which ever one you want to keep as default)
then pressing the skip button immediately performs that skip action

I left mine on 7 seconds only and as that is the only option it doesnt keep waiting for more.
so if i want a short skip i can just press the button once.
However if i want a longer skip i can hold the right button and it will keep jumping 7 seconds rapidly as there are no other presets
(2015-02-20, 16:30)Martijn Wrote:
(2015-02-20, 16:22)allan87 Wrote: Now that I understand the new seekstep feature, it seems to me that the default 1st step of 7 seconds is too short. For one thing, it takes several seconds, on a pi1 at least, for the skip to happen. It feels broken. I can tell you, until we figured out what was going on, there was much wailing and gnashing of teeth in our family room.

I suggest that the 7 second interval be disabled by default.

Is this in master or newclock4?

let me remind you that not everything revolves about the r-pi and some/most people have a device that is fast enough to handle the 7seconds
i understand that, but,
1. A lot of people do have pis or other lower power devices, and to them, skipping will seem broken. I appreciate that the problem can be alleviated by tweaking a setting, but watching tv should not require a learning curve.
2. In my view, 7 seconds is of minimal utility, even if you have a faster device. In any event, people with faster devices will be minimally inconvenienced (if at all) by a 15 second skip as default, compared to the situation with the pi.
3. If you were a mechanical engineer, would you design overpasses with 2 meter clearance because not everyone drives a truck?
And anyway, does "not everything revolves around the pi" mean what it sounds like? Does it mean that everything revolves around expensive high powered devices?Is there a development policy that says low power devices are disregarded in design decisions? That would explain the (adjective deleted) rejection of the PR to replace the current, CPU wasting, busy spinner with an efficient, attractive alternative. No, it made sense - to somebody - to exhaust the CPU unnecessarily, exactly when its capacity is needed elsewhere, because not everything revolves around the pi.

Think about it. The entire set top box firmament is composed of low power devices. Apple TV, Fire TV, Chromecast, Roku, smart TVs. Low power devices are where it's at. No design decision should be made without regard for that.
(2015-02-20, 16:22)allan87 Wrote: I suggest that the 7 second interval be disabled by default.

+7 seconds (seeking forwards) is already disabled by default.

However -7 seconds (seeking backwards) is enabled by default, which seems a bit unnecessary when there's already <smallstepbackseconds> set to 7 seconds and so does the same thing.
Texture Cache Maintenance Utility: Preload your texture cache for optimal UI performance. Remotely manage media libraries. Purge unused artwork to free up space. Find missing media. Configurable QA check to highlight metadata issues. Aid in diagnosis of library and cache related problems.
(2015-02-19, 00:07)Milhouse Wrote: Are people that reported problems with #0216 hanging on startup still having problems with #0217 or #0218?

Hi Milhouse
I had problems with the hanging with build 0216. Installed now the build 0219 and it is solved
Using a raspberry pi2

Thanks for all the work
Live tv from tvheadend on linux server plays in slowmotion on rpi 2 in todays build , know issue or do you need logs? Worked fine on last build i used (from 2015.02.08)
  • 1
  • 43
  • 44
  • 45(current)
  • 46
  • 47
  • 111

Logout Mark Read Team Forum Stats Members Help
OpenELEC Testbuilds for RaspberryPi (Kodi 15.0) Part 112