Scheduling a Compression.

[Slugger]

Originally Posted by nyplayer

# Only submit transcode jobs between midnight and 6am
* TimeOfDay: 06:00:00 - 20:00:00
# Only transcode tv recordings
= IsTV: true
# Don't try to transcode a recording that's in progress
= IsActivelyRecording: false
# Don't transcode recordings that have already been watched
= IsWatched: false
# Only transcode recordings of Planet of the Apes
= Title: Beneath the Planet of the Apes
# TRANS is an job id used to prevent the same file from being transcoded more than once; use whatever you want here
# __TRANSCODE is the special exe name used to queue up transcoding jobs in the SageTV core
# "MPEG4..." is the name of a valid transcode format as defined in the SageTV core
# The true/false on the end is a flag used to determine if the core will delete the source file after
#     transcoding or leave it; true means remove the source and false means leave the source
# NOTE: The action line below MUST start with a tab character (not 5
# spaces as pasted in here)
	TRANS__TRANSCODE "MPEG4-High Quality AVI" true

	TRANS __TRANSCODE "MPEG4-High Quality AVI" true

The format of the action line for transcoding is:

<tab><id> __TRANSCODE <format> <bool>

Notice the space between "TRANS" and "__TRANSCODE". Might be worth noting that the id (TRANS in this case) can be anything you want, just ensure it's unique.

[nyplayer]

Got a lil further but still errored.


SJQ v1.0.0b2
Started: Fri Aug 24 15:17:39 PDT 2007

Reading ruleset from 'C:\Program Files\SageTV\SageTV\rules.sjq'
Transcode action lines must contain 2 parameters on line 18
java.io.IOException: Transcode action lines must contain 2 parameters on line 18
at slugger.sagetv.sjq.RulesParser.<init>(Unknown Source)
at slugger.sagetv.sjq.SageJobQueue.run(Unknown Source)
at java.lang.Thread.run(Unknown Source)
*** Sleeping until Fri Aug 24 15:32:00 PDT 2007 ***

[Slugger]

Originally Posted by nyplayer

Got a lil further but still errored.


SJQ v1.0.0b2
Started: Fri Aug 24 15:17:39 PDT 2007

Reading ruleset from 'C:\Program Files\SageTV\SageTV\rules.sjq'
Transcode action lines must contain 2 parameters on line 18
java.io.IOException: Transcode action lines must contain 2 parameters on line 18
at slugger.sagetv.sjq.RulesParser.<init>(Unknown Source)
at slugger.sagetv.sjq.SageJobQueue.run(Unknown Source)
at java.lang.Thread.run(Unknown Source)
*** Sleeping until Fri Aug 24 15:32:00 PDT 2007 ***

Yup, saw this error while I was investigating your first post. Compiling beta 3 build now... should be about 15 mins and then I'll post it on my web server.

This is what happens when I can't actually fully test the feature locally.

[Slugger]

Beta 3 is ready. If this build doesn't go through properly then I'll record something here and try to transcode it so beta 4 will definitely be good.

[nyplayer]

With Beta 3 the SageTVClient starts shuts down right away and nothing happens. Your log file is empty.

#
# An unexpected error has been detected by Java Runtime Environment:
#
#  EXCEPTION_ACCESS_VIOLATION (0xc0000005) at pc=0x6bafe9a0, pid=3508, tid=540
#
# Java VM: Java HotSpot(TM) Client VM (1.6.0_02-b06 mixed mode, sharing)
# Problematic frame:
# C  [sqlitejdbc.dll+0x3e9a0]
#
# If you would like to submit a bug report, please visit:
#   http://java.sun.com/webapps/bugreport/crash.jsp
#

---------------  T H R E A D  ---------------

Current thread (0x04191400):  JavaThread "Startup-slugger.sagetv.sjq.SageJobQueue" daemon [_thread_in_native, id=540]

siginfo: ExceptionCode=0xc0000005, reading address 0x00000030

Registers:
EAX=0x00000000, EBX=0x04191a98, ECX=0x05d9fd64, EDX=0x00000004
ESP=0x05d9f908, EBP=0x05d9f908, ESI=0x00000000, EDI=0x00000000
EIP=0x6bafe9a0, EFLAGS=0x00010202

Top of Stack: (sp=0x05d9f908)
0x05d9f908:   05d9fa28 6baee27b 00000000 00000000
0x05d9f918:   0000000c 0000000c 00000000 7c9106eb
0x05d9f928:   03d7d940 04120c30 04087720 04087728
0x05d9f938:   00000000 00000000 041b0400 04120c30
0x05d9f948:   00000007 0000000c 00000000 6d8344a3
0x05d9f958:   00000002 6d833953 26f4ae58 26f4aec0
0x05d9f968:   0000002d 010dad30 0000001a 6d833840
0x05d9f978:   00000000 04191a98 010dad30 6d917cf3 

Instructions: (pc=0x6bafe9a0)
0x6bafe990:   24 04 8b 7c 24 08 89 ec 5d c3 55 89 e5 8b 45 08
0x6bafe9a0:   81 78 30 97 a6 29 a0 75 0e c7 40 30 06 79 3b f0

[Slugger]

Originally Posted by nyplayer

With Beta 3 the SageTVClient starts shuts down right away and nothing happens. Your log file is empty.

Hmm... that's weird... I'm recording something now so I can transcode it. If I don't post anything by 8:30pm EDT then check back around lunch tomorrow. Heading out for the night then and if I tell them I'm late because I'm compiling plugins for my PVR then I'll never hear the end of it all night.

[Slugger]

Beta 3 seems to be working for me...

SJQ v1.0.0b3
Started: Fri Aug 24 18:54:02 EDT 2007

Reading ruleset from 'C:\Program Files\SageTV\SageTV\rules.sjq'
Reading data store from 'C:\Program Files\SageTV\SageTV\sjq.sqlite'

<snip>

Processing media object: 'Late Show With David Letterman'
  Checking if media object '2553346' has been processed by action type 'COMM'
  Skipping: This object has already been processed by this action
  Checking if media object '2553346' has been processed by action type 'TRANS'
  Executing action line as all tests for rule 1 returned true
  Transcode job submitted to core; job #1 created in core queue
  Marking media object '2553346' as processed by action type 'TRANS'
*** Sleeping until Fri Aug 24 22:02:00 EDT 2007 ***

And I've confirmed that there is a SageTVTranscode process bogging down my server at the moment.

Can you constantly reproduce the crash? I notice from the output you pasted that you're running Java 6u2. I'm running on Win XP SP2, Java 6u1. I find it hard to believe, though not impossible, that my jar is causing the jvm to crash. If you can constantly reproduce it then I'd be more inclined to believe it, though not sure what I can do since here it's not crashing.

I'm going to leave my beta 3 jar running on my server overnight and I'll check to see if I'm still up and running when I return home tomorrow morning. In the meantime, if you can try to reproduce the crash by getting sage to start successfully without loading my jar and then getting it to crash constantly by loading my jar I'd like to know. Perhaps the zip download was corrupt? The file size on my web server is 389,830 bytes.

[Slugger]

Another thought... are you running 64bit Windows by any chance? I never really thought about it until now, but I believe the SQLite DLL I package in the zip is a 32bit intel build. I don't have a 64bit Windows install to test if the DLL is good or not on 64bit, but if it wasn't then the jvm's attempt to load the dll for the jdbc driver might just be something that could cause the jvm to crash.

Just throwing it out there...

[Slugger]

And finally... if you're still having problems with a crash can you please try to upgrade the SQLite JDBC driver?

http://www.zentus.com/sqlitejdbc/

You have two options:

* Upgrade the Windows native driver by downloading the latest version and replacing the files in the SageTV folder.

* Delete the sqlitejdbc.dll and jar that are currently in your SageTV folder (dll is in SageTV and jar is in JARs) and switch to the pure java driver. If you decide to go this route then delete the old drivers first and you can just install beta 4. There are no code changes in beta 4, I've just repackaged the JDBC driver. You must remember to delete the old dll and jar before installing the new pure java driver.

I've tested the pure java driver and it works, but there is a performance hit. However, I think I'm going to move to this driver since with the pure java driver I automatically support Linux installs. And since this is hardly a high usage database, the performance hit isn't really much of an issue.

Please report back on your progress/issues.

Thanks.

[nyplayer]

Slugger,

Beta4 is working... just fine. I am using 32 bit not 64. I deleted the dll etc.. and just installed your package.

One more thing how do I make sure that an EDL exists before compression is scheduled?

Thank You,

If you need any more testing let me know

[Slugger]

Originally Posted by nyplayer

Slugger,

Beta4 is working... just fine. I am using 32 bit not 64. I deleted the dll etc.. and just installed your package.

One more thing how do I make sure that an EDL exists before compression is scheduled?

Thank You,

If you need any more testing let me know

Glad to here things are working smoothly. As to the edl problem, there currently is no way to check an external condition in SJQ. What I could do, is add a new comparison that allows you check the age of a recording. With this you could then add a test that says, "only transcode recordings that are more than x hours old". And if you set x to be big enough that you know your edl will have been generated for that recording then you're good to go. Would this solution address it sufficiently? I'm heading out for the day now, but I could probably whip this up sometime tomorrow. Let me know if this will work or throw any other ideas up in the air.

[Slugger]

Originally Posted by Slugger

Glad to here things are working smoothly. As to the edl problem, there currently is no way to check an external condition in SJQ. What I could do, is add a new comparison that allows you check the age of a recording. With this you could then add a test that says, "only transcode recordings that are more than x hours old". And if you set x to be big enough that you know your edl will have been generated for that recording then you're good to go. Would this solution address it sufficiently? I'm heading out for the day now, but I could probably whip this up sometime tomorrow. Let me know if this will work or throw any other ideas up in the air.

Download beta 5. It includes the ability to check for edl files (or any type of file for that matter). I've updated the docs for beta 5. Have a look at the new example rules file in the doc (it shows how you can check for edl files before transcoding). Pay special attention to the fact that the file is no longer automatically appended to action lines (see the media file variables section for more details).

I've done enough testing of b5 myself to feel fairly confident in calling this the official v1.0.0 release, but I'd really appreciate some feedback from anyone else who tries this over the next few days to help confirm my testing results.

[TiltRod]

Slugger,

Love the plugin. Logics work great for setting up different transcode formats based on content. However, I cannot get the boolean arg to work for __Transcode to replace the original MediaFile.

[Slugger]

Originally Posted by TiltRod

Slugger,

Love the plugin. Logics work great for setting up different transcode formats based on content. However, I cannot get the boolean arg to work for __Transcode to replace the original MediaFile.

I won't be able to experiment until later in the week, but I did take a quick look at the code. My docs might be backwards. From what I'm seeing in the API along with my code, if the boolean is true then the source file will be deleted and if it's false then the source file will be kept. Did you try to flip the boolean and see what happens? What I've described here is how it's suppose to work, regardless of what my docs might say.

Let me know and if both values keep the source around then I'll have to dig a little deeper when I return home later in the week.

[TiltRod]

Yes, I tried both true and false. Neither resulted in a replaced file.

[davephan]

Originally Posted by Folkboat

Indeed, compressing a recording during "prime time" utilizes system resources.

Transcoding a 2hr recorded TV movie to a DVD format can take a looooong
time on a slow system.

Automating the process just makes sense to me. OTOH its not an earth shattering issue!

Thanks

You might want to consider compressing video with the Roxy99 programs/scripts. You could then offload the video compression job to other computer(s) if you have other computers that are not being used. However, if you use really old computers to compress video, it will take many hours to compress each hour of MPEG into Xvid AVI.

Dave

[Slugger]

Originally Posted by Slugger

I won't be able to experiment until later in the week, but I did take a quick look at the code. My docs might be backwards. From what I'm seeing in the API along with my code, if the boolean is true then the source file will be deleted and if it's false then the source file will be kept. Did you try to flip the boolean and see what happens? What I've described here is how it's suppose to work, regardless of what my docs might say.

Let me know and if both values keep the source around then I'll have to dig a little deeper when I return home later in the week.

v1.0.0 released. This release includes all the transcode job features from the betas discussed in this thread. Note: The various beta builds linked to in this thread have been removed from my server. All code in those betas is available in the v1.0.0 build. Unfortunately, I'm unable to fix the boolean flag issue for the transcode action line. I'll try again when I have some spare time, but I decided to go ahead and release what's there since there are a lot of improvements over the last release. The effect of the bug is that the source files of anything that is transcoded via SJQ is not deleted by SageTV after transcoding completes regardless of the value used on the command line.