Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.


Messages - phenigma

Pages: 1 ... 60 61 [62] 63 64 ... 77
916
Users / Re: 810 beta 2 - UI2 Audio Navigation with keyboard and mouse
« on: April 06, 2010, 06:22:28 pm »
That's really odd, what's your setup?

Zotac ION 330 miniITX (VGA 720P) with vdpau enabled.  I have >500 <1000 videos and around 44000 audio tracks, most have attributes and cover art applied. 

The only problem I experience is slow queries in mysql, some of which I've sped up others not.  I've no idea what is causing your problem.

J.

917
Users / Re: 810 beta 2 - UI2 Audio Navigation with keyboard and mouse
« on: April 03, 2010, 12:23:58 am »
Works fine here.

J.

918
Users / Re: Edit sources.list
« on: March 29, 2010, 06:31:05 pm »
I have a few questions about the repository's. I've tried several times to install LinuxMCE on top of Hardy(8.04) and well I get nothing but pure frustration

LMCE is not compatible with Hardy (8.04).  There are two flavours of LMCE you may install LMCE 0710 (for Kubuntu 7.10) or LMCE 0810 (for Kubuntu Intrepid, 8.10). 

You should perform an 0810 install if you are using newer hardware, follow these instructions: http://wiki.linuxmce.org/index.php/Installing_0810

J.

919
Users / Re: Multiple TV Tuners Question
« on: March 16, 2010, 02:50:40 pm »
MythTV will automatically select the next available un-used tuner based on priorities set for the tuners in MythTV.  If you have only one analog tuner and it is in use then you will not be able to use that tuner on another MD until it is no longer in use.  All channels will show in one big list.  With multiple tuners if the backend is recording and you choose liveTV the second tuner will be automatically selected.

J.

920
Users / Re: MPlayer fails to play on Media Director
« on: March 15, 2010, 10:04:24 pm »
My ION MD has the latest version installed from the avendard repository 195.36.08 for both nvidia-195-libvdpau and nvidia-glx-195.  Mplayer is working for me with, and without, vdpau acceleration.

dcerouter_1006499:~# apt-cache policy nvidia-195-libvdpau
nvidia-195-libvdpau:
  Installed: 195.36.08-0ubuntu1
  Candidate: 195.36.08-0ubuntu1
  Version table:
 *** 195.36.08-0ubuntu1 0
        500 http://www.avenard.org intrepid/release Packages
        100 /var/lib/dpkg/status
     195.36.03-0ubuntu2 0
        500 http://deb.linuxmce.org intrepid/beta2 Packages


For testing HD material I use trailers from here: http://www.h264info.com/?page_id=20 (rename .mp4 to .m2ts)

Almost every media format is handled by Xine, which is the default player for LMCE.  Renaming files to .m2ts causes them to be played by mplayer, which is not well integrated with LMCE...  Is there a reason you have renamed this file to .m2ts?

J.

921
Users / Re: MPlayer fails to play on Media Director
« on: March 15, 2010, 05:59:32 pm »
The GeForce 6800 is not VDPAU compatible, try setting your MD for opengl.  See: http://ru.download.nvidia.com/XFree86/Linux-x86/190.32/README/appendix-a.html

Turn the de-interlacing off (or lowest) until you get things working.

J.

922
Users / Re: apt-get tries to downgrade pacakges on MD
« on: March 10, 2010, 09:40:11 pm »
The script /usr/pluto/Diskless...TBZ.sh will re-create the MD images completely from scratch, including the /usr/pluto/deb-cache folder.

With the 'deb...' line removed from /etc/apt/sources.list the MD will not try to use files from that directory, you can use and update your MD as if nothing was different.  Apt's cache of downloaded .debs is still active and not affected by this change.

What remains is to determine why/when/how the filenames are being corrupted and by what.

J.

923
Users / Re: apt-get tries to downgrade pacakges on MD
« on: March 10, 2010, 08:17:27 pm »
I believe the /usr/pluto/deb-cache is a repository of .debs, used to build the diskless images, at the time you ran /usr/pluto/bin/Diskless_CreateTBZ.sh on your core. 

What is puzzling me is why apt isn't using the newer .debs from the online repos.

J.

924
Users / Re: apt-get tries to downgrade pacakges on MD
« on: March 09, 2010, 08:32:02 pm »
I see two possibilities:

1. Update your core and re-run /usr/pluto/bin/Diskless_CreateTBZ.sh then rebuild your MDs.

or

2. Remove the entry "deb file:/usr/pluto/deb-cache/ ./" from /etc/apt/sources.list on the affected MDs and then update/upgrade them.

I'm not sure why this is happening.

J.

925
Gotcha. 

J.

926
Hi posde, yes I realize that, the link is always by PK and FK.  So let me explain what I mean.  The PK for a related table row (eg. Manufacturer) could be different on the system it's being imported to than it was on the exporting system because of sqlCVS updates.  To get around this the psc_id could be used to find the proper PK on the import machine (if it's different).

Example (Follow me here):
1.User1 installs a fresh 0810 system and creates a DeviceTemplate and then exports it and gives it to User2.  This device template refers to a manufacturer that was recently added to sqlCVS.

2.User2's system is older than User1's and User2 has created a new manufacturer before doing an sqlCVS update.  The sqlCVS update adds the manufacturer that User1's exported template is referring to.  Both users now have the same Manufacturer in their databases but with different PK's for that Manufacturer but the same psc_id.  This is how sqlCVS works isn't it?

3.Importing the DT into a new system by original PKs will link the imported Device to the incorrect Manufacturer on User2's system. 

The script will need to check the psc_id of the manufacturer and use it to select the the import machine's PK for that Manufacturer to properly link rows on the import machine... 

J.

927
It has occurred to me that when inserting devicetemplates into a new system that records should be linked on psc_id, where it exists, not by PK_ as the PK_ could be different where the psc_id should be identical across systems.  The script is still maintaining links by PK_ which could create problems where a DT depends on a table row added by an sqlCVS update.  I'll change this functionality to work on psc_id where it exists.

I'm only testing on 0810 but can I assume that psc_id's are not reused and a psc_id for a device under 0710 will still describe the same device in 0810?

J.

928
Sure.  Here's a slightly updated version exporting and import viewing work well.  I need to scrub my virtual machine and test some more imports with the updates.  Simple IR based AV devices seem to work on native databases.

Changes:
 * 2010-03-06 - Changes by phenigma phenigma@hotmail.com
 * - fixed updates all entries' FK's for new PK entries after initial inserts
 * - fixed string escaping using mysql function (utf friendly?)
 * - fixed tables gets a unique fakeID when previewing rather than all related tables=-3
 * Known Issues:
 * - does not export/import any files/scripts - only DB entries
 * - does not import DTs marked 'old', use &forceExportType=new during export to test importing
 * - exports redundant data with &forceExportType=new
 * - may create duplicate rows when importing 'old' DTs exported with &forceExportType=new
 * - I'm sure there's lots more...
 
As I learn more about the database/sqlCVS and php/mysql I'm finding more to implement and better ways of doing things... 

J.

929
I think this is an important abiltity to have and a presents a great learning opportunity to become more familiar with LinuxMCE, PHP, & MySQL.  Over the last couple days I've got the script more or less working to export and import basic user generated device templates.  I have tested this with a few IR A/V devices and I have also tried the PLCBUS template, programmed IR codes and ruby commands are exporting and importing .  The following is a list of some updates and issues:

Changes:
 * - fixed - quotes around field names are gone
 * - fixed - stores all rows from related tables rather than last row only
 * - fixed - updates DeviceTemplate FK's for new related table PK entries after inserts are done
 * - fixed - strings are escaped before inserting to db (ruby looks okay to me)
 * - added - preview insert/update statements before importing to DB
 * Known Issues:
 * - when previewing these are placeholder values PK_DeviceTemplate=-3, PK_InfraredGroup=-2
 * - does not set FK_Manufacturer, FK_DeviceCategory properly in related tables if new
 *   Manufacturer or DeviceCategory entries are inserted by the script
 * - does not export/import DHCPDevice PnpDetectionScript scripts - only DB entries
 * - does not import DTs marked 'old', use &forceExportType=new during export to be able to import
 * - no validation of data can create duplicate rows in tables when importing 'old' as 'new'
 * - I'm sure there's more...

Thanks posde and merkur2k I'm learning lots.  It has a ways to go to deal with all possibilities, and could be a little prettier, but it's in a state where it is worth doing some testing with.  I'd like to rename it at some point (posde?) as it no longer generates a PHP script.  DeviceTemplateRevolver (DTR) or DeviceTemplateXchanger (DTX)... or something.  I expect to tackle the issues mentioned above, feel free to add to the list if you find specific issues.  Backup your data before using.

Use:
 Exporting - the script will display the serialized data and ask you to confirm the export after which it will present the file download dialog.
 Importing - View or Import.  Default is View.  Choosing view will parse the selected file and display the generated insert/update statements without executing any of them.  Choosing Import will immediately execute the insert/update statements and display them at the same time.  Choose view the first time, verify the insert/update statements, hit the back button and choose Import to import the device.

Grab the updated script from: http://filebin.ca/mftsyp/DeviceTemplate2php.tar.gz

J.

930
Thanks Posde.  I am aware of that issue and I do see that happening with 2 of my external devices (ps3 and wii templates), when it does I use the orbiter to switch the device off and then select it again from the orbiter and the inputs are set correctly.  This is the issue that davegravy was describing earlier in this thread and I have always seen this behaviour.

My issue seems to be happening with only one of my external devices regardless of whether LMCE thinks the display is on or off. 

Seperate from the inputs not being set properly all of my orbiters show my television is always on the MD's input (resend AV codes screen), even when LMCE has properly set it to another input...  I'm going to keep investigating.

J.

Pages: 1 ... 60 61 [62] 63 64 ... 77