[01:35] <MP0> Hmm, is the TPV5150 INTREQ pin tied to the DM320? I don't see such a connection on the schematic block diagram on open.neurostechnology.com, but I'm not sure that's a complete representation.
[01:37] <nerochiaro> MP0: i don't know, but if you ask in ML someone should be able to answer you
[01:38] <MP0> k
[01:38] <MP0> I got CC data out of the chip btw... I still need a way to get it into the output video stream.
[01:39] <nerochiaro> i saw that, was pretty impressive
[01:40] <nerochiaro> i don't think you need to put it _into_ the video stream. you actually need to make it available with a system that notify people when it change. then people can use the GUI layer to show that information on top of video
[01:40] <nerochiaro> i would do that, at least
[01:41] <nerochiaro> make a character device that people can select() on, and read() the CC text, maybe
[01:41] <MP0> it already is...
[01:41] <nerochiaro> so what's the issue ?
[01:43] <MP0> While that approach works, there's two issues. The minor one is that now it's open captioned rather than close captioned if you ever wanted to record it back onto tape.
[01:44] <MP0> The other reason is it makes the system modal: captioning then has two different behaviors depending on the video source.
[01:45] <MP0> if it all possible, I'd like to avoid that.
[01:46] <MP0> but okay, assuming I did build the complete CC decoder into the OSD, then the next question is how to store the data in an MP4 file. I've been searching the net and I find all sorts of references to embedding CEA-608 captioning into MP4, but no examples or specifications.
[01:46] <nerochiaro> i very much doubt you can render text into the video itself, you would have the DSP do that, and even we don't have souce for the DSP codecs (let alone how the DSP codec gets the CC data)
[01:46] <MP0> I was thinking about using the overlay
[01:47] <nerochiaro> MP0: what you mean with overlay ?
[01:47] <nerochiaro> MP0: overlay to render the text, or overlay to store CC in MP4
[01:47] <nerochiaro> ?
[01:47] <nerochiaro> i kinda mixed topics, sorry
[01:47] <MP0> okay, yeah, we're talking about a few different things
[01:48] <MP0> I was thinking it might be possible to use the overlay to get the analog TV out to contain caption data.
[01:48] <MP0> As for getting the CC data into the output of the codecs...
[01:49] <nerochiaro> MP0: to get into the analog TV output the CC data, rendering text to the GUI layer (overlay) is the way to go
[01:50] <nerochiaro> MP0: for the codecs, it seems really not feasible with the current system
[01:50] <MP0> it's an approach, although I'd prefer to find a way to get the raw caption data out instead.
[01:50] <nerochiaro> MP0: save it to text files (subtitle files) ?
[01:51] <MP0> you still need timing information.
[01:51] <MP0> as for the codecs, I don't necessarily need to modify the codecs, I'll probably just patch the output. I'm not really worried about the closed nature of the codec.
[01:51] <nerochiaro> subtitle formats (.srt, .sub etc...) allow that timing info
[01:52] <nerochiaro> MP0: "patch" the output ?
[01:52] <MP0> yes, but you need to get the timing from somewhere
[01:52] <MP0> which means having an idea how far the encoder has gotten.
[01:53] <nerochiaro> MP0: or you can use the OS time functions to just timestamp the start of the recording, and then the difference from that at each CC change
[01:54] <MP0> won't you get out of sync after a while?
[01:54] <nerochiaro> good point. i guess it depends if you drop frames or not.
[01:55] <MP0> I was going more for:
[01:55] <MP0> proprietary codec -> fwrite -> me combining with CC data -> fwrite -> file
[01:55] <nerochiaro> you do the combining on the file on disk as it's being written ?
[01:57] <MP0> I haven't investigate it enough at this point. I will probably try to at least get timing information from the imanage api.
[01:58] <nerochiaro> that is ok, it has that information. i think you can get it from the iencode api actually
[01:58] <nerochiaro> but more than that, i'm not at liberty to help you :(
[01:59] <MP0> I wouldn't worry about me :-) The proprietary stuff hasn't given me a problem yet.
[01:59] <nerochiaro> that's great :)
[02:00] <nerochiaro> regarding the thing you said before about "combining" the CC data with the video, i'm not sure how you plan to do that
[02:00] <MP0> well, I know MP4 supports embedding CEA-608 caption data into a stream.
[02:00] <MP0> lots of products mention having the capability to work with that
[02:01] <MP0> like Apple's QuickTime
[02:01] <MP0> so I think MP4 supports having the data
[02:01] <nerochiaro> the MP4 encoder is closed as well, though
[02:01] <MP0> I don't need to touch the encoder.
[02:01] <nerochiaro> you can probably post process the stream and inject the data into the stream later
[02:01] <MP0> exactly
[02:02] <MP0> and like you said, there's also the option of generating separate subtitle files
[02:02] <MP0> however I'd like to get the caption information combined into the MP4...
[02:03] <MP0> but so far I haven't found an example MP4 file with close caption data in it.
[02:03] <nerochiaro> me neither
[02:04] <MP0> lots of products mention supporting it, but google doesn't seem to reveal what the specification is.
[02:04] <nerochiaro> might be one of these awful specs that you have to pay for just to read them
[02:06] <MP0> Do you guys have the MP4 specs?
[02:06] <nerochiaro> not that i know of. we didn't write the codec and the parser, we just tweaked it a bit
[02:07] <MP0> k
[02:07] <nerochiaro> the parser, that is. the codec, totally untouchable for us
[02:07] <MP0> okay. Do you know anyone who might be familiar with the spec?
[02:08] <nerochiaro> no, sorry
[02:08] <MP0> alright, thanks
[02:08] <nerochiaro> the people at video-specific forums, maybe, like doom9 or such places ?
[02:08] <nerochiaro> or the ffmpeg people ?
[02:09] <nerochiaro> did you check if ffmpeg has code related to that stuff ?
[02:09] <MP0> I'm investigating that... good idea about doom9
[02:22] <MP0> got it! tecnocato.podbean.com
[02:22] <MP0> a previous post gives a sample captioned video
[02:27] <nerochiaro> good luck then !
[02:27] <MP0> Give it a few more weeks :-)
[02:29] <nerochiaro> MP0: is the CC char device already available somewhere ?
[02:30] <MP0> paul.bluereboot.net
[02:30] <MP0> right now it only does NTSC captioning so if you're out of the US... oops
[02:30] <nerochiaro> MP0: i mean the source, sorry
[02:30] <MP0> ah, that's a piece of crap but if you really want to see it
[02:31] <crweb> nerochiaro: i'm having the hardest time making a upk that will flash
[02:31] <crweb> its blocking everything
[02:32] <nerochiaro> blocking as in freezing ?
[02:32] <crweb> no, blocking progress
[02:32] <crweb> i can't test anything I've written
[02:32] <gremlin[it]> morning nerochiaro & crweb
[02:32] <crweb> its crunch time and i can't get the blepity bleep... bleep..
[02:32] <MP0> paul.bluereboot.net
[02:33] <nerochiaro> crweb: can you please talk with Peter and confirm that you two are following the same process to create the UPKs ?
[02:33] <crweb> nerochiaro: ./build-helper.sh all .... worked friday
[02:33] <nerochiaro> MP0: thanks, i'll just give it a look out of curiosity later
[02:33] <nerochiaro> crweb: what did you change since then ?
[02:34] <crweb> i changed some bootup scripts in a fresh checkout
[02:34] <nerochiaro> if you rollback to friday's code, it works again ?
[02:34] <crweb> says bad checksum when I try to flash..
[02:34] <crweb> hmm..
[02:35] <crweb> do you have time to help me with svn, or should I just go read?
[02:35] <nerochiaro> crweb: -r option, in a nutshell
[02:35] <crweb> i'd need to do all trees though
[02:35] <crweb> ok
[02:36] <nerochiaro> for d in * ; do svn up -r whatever $d ; done
[02:36] <nerochiaro> or if you can stand a full checkout, svn co has the same option, IIRC
[02:37] <nerochiaro> and you will probably get tricked by timezones and such, so beware
[02:37] <crweb> well, any time prefriday should work
[02:37] <nerochiaro> indeed
[02:37] <nerochiaro> another more quick thing you can try
[02:38] <nerochiaro> is to try making upks that have only one component in them (uboot, rootfs, etc) and see if they flash, and what is the componenet that cause issues
[02:39] <crweb> well it goes:
[02:39] <crweb> Loading your update package: ....
[02:39] <crweb> Beginning Flash (Pong game)
[02:39] <crweb> red letters Error Error Error, then NSD
[02:40] <crweb> (neuros screen of death ;) Circle logo with one side orange ;)
[02:40] <nerochiaro> what does uboot console say ?
[02:41] <crweb> getting it one last time cleanly
[02:41] <crweb> i guess i have 1x sd cards.. cause.. damn
[02:42] <crweb> pastebin.com
[02:42] <crweb> heh, bad checksum
[02:44] <nerochiaro> crweb: can you try using another card ?
[02:44] <nerochiaro> maybe it's broken ? or updating from GUI
[02:44] <crweb> i tried MS, and SD. will try my phone SD
[02:44] <crweb> emergency flash
[02:44] <nerochiaro> ah, you need to test emergency, or you just need to test if your upk is ok ?
[02:45] <nerochiaro> in the latter case, please put it on an USB disk and try GUI upgrade
[02:45] <crweb> i need it to flash, so i can test my work
[02:45] <nerochiaro> see if you get bad checksum when doing the GUI upgrade then
[02:45] <nerochiaro> from a non-card device
[02:52] <nerochiaro> crweb: regardin that COP thing to close settings windows when recording start (from scheduler), i think we can keep the COP thing, but actually make it a notication like recorder notifyig people that it is starting (as opposed to a specific message for main-menu to close top windows, which is silly)
[02:52] <nerochiaro> crweb: what do you think ?
[02:55] <crweb> i think either you or I misunderstand
[02:55] <crweb> thats main-menu sending itself a cop message to close all its own windows
[02:56] <nerochiaro> no, i already shot down that case. main-menu receive the RECORD key, and thus can close the settings windows without any cop at all
[02:56] <nerochiaro> see mail
[02:56] <crweb> yeah, what other case is there?
[02:56] <nerochiaro> however when there's a scheduled recording, main-menu doesn't know
[02:57] <nerochiaro> so it can have settings open
[02:57] <nerochiaro> and they will stay open
[02:57] <crweb> ah got it.
[02:57] <crweb> yes, add notifications to app starting
[02:58] <crweb> but, nwm can do this
[02:58] <crweb> on app start, send app is starting
[02:58] <crweb> this is common
[02:59] <crweb> nerochiaro: gui says bad checksum
[02:59] <nerochiaro> hmm, i like the idea. however we may not need notification for all apps, just for some
[02:59] <nerochiaro> in this case recording
[03:00] <crweb> then it just doesn't listen
[03:00] <nerochiaro> you still waste time sending out notification that no one care about
[03:00] <crweb> but app start, main-menu should close its existing windows. we don't have room for them
[03:00] <crweb> no, it only goes to those who listen
[03:01] <nerochiaro> the act of "emit" doesn't just waste resources ?
[03:01] <nerochiaro> even if no one is listening
[03:01] <crweb> whos it going to send to if no one is connected?
[03:01] <crweb> no, its fine
[03:01] <nerochiaro> ok then
[03:01] <nerochiaro> i'll update the thread myself
[03:01] <crweb> msgs happen if no one is listening or not
[03:01] <crweb> err..
[03:01] <crweb> you get it
[03:01] <nerochiaro> i get the idea, yes
[03:02] <crweb> /media/SD-card/newpackage/r3.upk: line 1: Neuros: not found
[03:02] <crweb> /media/SD-card/newpackage/r3.upk: line 70: cannot create ???P?: Read-only file system
[03:02] <crweb> /media/SD-card/newpackage/r3.upk: line 70: I?D???programs_0.tar.gz2.00?Developer: not found
[03:02] <crweb> /media/SD-card/newpackage/r3.upk: line 71: syntax error: word unexpected (expecting ")")
[03:03] <crweb> this upk is 17.8mb but stuff can't be found?
[03:03] <crweb> I'm going to continue on with terry. thanks
[03:04] <nerochiaro> unexpected (expecting ")") << this means usually that you're running an executable that it's really not an executable
[03:04] <nerochiaro> it happens for example when i run x86 stuff on ARM or the other way around
[03:06] <crweb> gah..
[03:07] <nerochiaro> crweb: doesn't app channel already have notifications for app starting ?
[03:07] <crweb> no, it has message sent to start app
[03:07] <crweb> "App starting" message would go out if app actual begins
[03:08] <nerochiaro> so i'll tell them to add the app started in the app channel. seems the better place for it
[03:08] <crweb> thats what i said
[03:08] <crweb> yeah
[03:10] <nerochiaro> ok, done
[03:15] <crweb> i just open sourced all my private projects
[03:15] <crweb> free hosting at code.google.com
[03:15] <nerochiaro> it's been there since a long time
[03:15] <crweb> never knew..
[03:15] <crweb> uses svn too. pretty cool
[03:16] <nerochiaro> shoudld use git :P
[03:16] <crweb> yes it should
[03:16] <nerochiaro> and you should've known since you were a SoC mentor
[03:16] <crweb> i'm sure it will eventually
[03:16] <crweb> i didn't know they did project hosting
[03:16] <crweb> practicevision.googlecode.com
[03:17] <crweb> it is _not_ a good Qt code reference
[03:17] <crweb> i've learned much since 2005
[03:17] <nerochiaro> guess so
[03:17] <nerochiaro> weren't you seeling that app ?
[03:18] <crweb> no, it was privately funded
[03:18] <crweb> going to, soon as i rewrite the db layer
[03:18] <crweb> also, will be releasing topiabox
[03:18] <crweb> gui audio player for osd
[03:20] <nerochiaro> crweb: that was pretty ugly looking ;)
[03:21] <crweb> topiabox?
[03:22] <crweb> might be ugly looking. does things xmms dreams of though
[03:22] <crweb> like.. multi-db sources
[03:22] <nerochiaro> such as ?
[03:23] <nerochiaro> i'm sure xmms2 does things that your app dreams of too
[03:23] <crweb> storage of metadata on removable storage
[03:23] <crweb> nothing the osd needs ;)
[03:23] <crweb> its an OSS world. of couse mine is better
[03:23] <crweb> ;)
[03:24] * nerochiaro  pats crweb's ego
[03:24] <nerochiaro> now go back to that upk issue ;)
[03:25] <crweb> terry and i are compiling
[03:25] <crweb> its not going so well though
[03:25] <crweb> freaking language support is still breaking everything
[03:26] <nerochiaro> i just approved Jimmy's patch to makefiles
[03:27] <nerochiaro> what else is wrong with these lang files ?
[03:27] <crweb> ah, i guess they weren't committed yet?
[03:28] <nerochiaro> check it, but i guess they weren't
[03:28] <nerochiaro> at least not the || exit 1 fixes
[03:28] <nerochiaro> jimmy is online, be sure to check with him and get a resolution while he's still around
[03:28] <crweb> i'm pretty sure this one is from a patch i already ok'd
[03:28] <crweb> he is checking it already
[03:29] <crweb> qm_install error on neux
[03:29] <nerochiaro> ok, tell him of my email of now, he may have missed it as i just sent it
[03:33] <crweb> if i say, wanted to go ahead and commit/submitted the parted work so you can look at it and help me. that going to be ok?
[03:33] <nerochiaro> i don't understand
[03:33] <crweb> i got the external-libs/parted all setup and stuff
[03:34] <crweb> only thing to overcome is the compile error
[03:34] <crweb> want me to zip it and send it, or fix it while its in tree
[03:35] <nerochiaro> if it doesn't break anything else except the specific thing you're working on, then feel free to commit it to tree. if it break the whole build for everyone, don't
[03:36] <crweb> it would have to be added to the build to do so
[03:37] <crweb> terry is now waiting on parted
[03:37] <crweb> its end of day though, so he's not waiting very badly
[03:38] <nerochiaro> all i'm saying is: don't break the tree for everyone. if you can avoid that, feel free to commit half-working things
[03:52] <crweb> nerochiaro: Neuros-Cooler/external-libs/parted
[03:52] <crweb> make rebuild
[03:53] <nerochiaro> i just ran an build-helper up + all
[03:53] <nerochiaro> one minute ago. should i repeat ?
[03:53] <nerochiaro> it seems to have picked up parted already
[03:55] <crweb> its not part of buildhelper
[03:55] <crweb> that would break build for everyone
[03:55] <nerochiaro> it's part of what, then ?
[03:55] <crweb> cd Neuros-Cooler/external-libs/parted
[03:55] <crweb> make rebuild
[03:55] <nerochiaro> ah, ok, just if someone forced rebuild
[03:55] <nerochiaro> good
[03:56] <crweb> i can't get it to build, so couldn't add the binaries yet
[03:56] <nerochiaro> it's ok then
[03:56] <crweb> ok.. well, please take a look when you have time
[03:57] <crweb> more eyes will have to fix this. cause i'm at a loss
[03:57] <nerochiaro> ok, i will take a look later today
[03:57] <crweb> i disabled readline support, and it still tries to use readline support...
[03:59] <nerochiaro> i'll try some cross compile mojo on it
[03:59] <nerochiaro> finishing checking some big email threads, then i'll have a look
[03:59] <crweb> no problem terry's headed home for today anyway
[03:59] <nerochiaro> ok
[04:25] <nerochiaro> crweb: wow, i just checked out the changes to main-menu merge and file browser merge. good stuff there, it's really quite fast now
[04:25] <crweb> excellent
[04:27] <crweb> hopefully mine will flash and I can see too
[04:27] <crweb> gosh dang it
[04:27] <crweb> corrupt
[04:28] <nerochiaro> should i try flashing stuff too ?
[04:29] <crweb> please
[04:29] <crweb> I seriously don't have time for this...
[04:29] <crweb> need to test *now*
[04:30] <nerochiaro> ok, but even assuming i can flash... what does it change for you ?
[04:30] <crweb> network script startup and exec
[04:30] <crweb> support for rcU.d working
[04:30] <crweb> nfs detection in update scripts, etc
[04:30] <nerochiaro> no i mean, i can flash, you can't. so what do you gain ?
[04:31] <crweb> the knowledge that I should format the drive and then toss the PC out the window?
[04:31] <nerochiaro> ok, let me try
[04:31] <nerochiaro> arizona firmware ?
[04:31] <nerochiaro> right ?
[04:31] <crweb> yeah
[04:32] <nerochiaro> should i turn off NFS when doing that ?
[04:32] <crweb> that dev.upk that you just created in images/
[04:32] <crweb> it will overwrite that
[04:32] <crweb> you would have to resetup your nfs settings
[04:32] <nerochiaro> no problme
[04:32] <nerochiaro> trying now
[04:35] <crweb> pastebin.com
[04:35] <nerochiaro> it's upgrading. warned me about not official package, but UPK checksum was ok. it's flashing it now
[04:37] <nerochiaro> pastebin.com
[04:37] <crweb> mine is smaller than yours...
[04:37] <crweb> whats that about
[04:38] <nerochiaro> also: pastebin.com
[04:38] <crweb> you flash from CF?
[04:38] <nerochiaro> no, from SD
[04:38] <nerochiaro> that CF message, i dunno what it has to complain
[04:39] <nerochiaro> it's plugged in when i boot
[04:39] <nerochiaro> regarding the size, how much smaller is yours ?
[04:40] <crweb> 200k
[04:40] <nerochiaro> i'm thinking i may have some lefover imedia app crap
[04:40] <crweb> ok.. going to rm -rf..
[04:41] <nerochiaro> well, do a fresch checkout somewhere else
[04:41] <nerochiaro> why removing the stuff
[04:41] <nerochiaro> ?
[04:41] <crweb> if it doesn't work what good is it..
[04:42] <nerochiaro> you never know. you have terabytes of space anyway
[04:42] <nerochiaro> :P
[04:52] <nerochiaro> crweb: upgrade seems successful here
[04:55] <crweb> just for the record
[04:55] <crweb> when you start hearing of this issue on ML, here, and Forums
[04:56] <crweb> have them upgrade to latest torfu, or "run update-uboot" from uboot/serial
[04:56] <nerochiaro> eh ? peole shouldn't need to do that
[04:56] <nerochiaro> people
[04:57] <crweb> less than torfu 6 can't read the upk
[04:57] <nerochiaro> only upgrade to latest TORFU
[04:57] <nerochiaro> that's a requirement
[04:57] <nerochiaro> to upgrade to AZ
[04:57] <crweb> yeah , i know
[04:57] <crweb> just remember... it'll come up
[04:57] <nerochiaro> and with the new changes in upgrade we're doing now, it will be even more of a requirement
[04:57] <nerochiaro> ok, i'll remember
[04:57] <crweb> someone will do it
[04:57] <crweb> i was running pre-torfu
[04:58] <nerochiaro> they deserve failure if they do
[04:58] <crweb> right, but person may not realize
[04:58] <nerochiaro> there are upgrade instructions...
[04:58] <crweb> we still get questions about the 3 step upgrade to torfu if they waited to long
[04:58] <nerochiaro> well, ok, i'll keep this in mind
[04:59] <crweb> and it'll happen to the best of'm
[04:59] <nerochiaro> it doesn't help that the version numbers are such a total mess
[04:59] <nerochiaro> i keep getting confused by them
[04:59] <crweb> or in my case. the most retarded also.
[04:59] <crweb> i didn't grab my normal osd when I left home.
[04:59] <crweb> i grabbed my "traveling" bag
[05:00] <nerochiaro> ah, i see
[05:00] <nerochiaro> very old stuff
[05:00] <nerochiaro> so now, can you flash ? was that the issue
[05:00] <nerochiaro> ?
[05:00] <crweb> yeah
[05:00] <nerochiaro> oh very good
[05:00] <nerochiaro> i'll get a sandwich and then look into that parted stuff
[05:02] <crweb> we should change the error to: Bad Checksum or Uboot to old to recognize file
[05:03] <nerochiaro> to change the error, one has to have a newer uboot, so he would not have the problem...
[05:04] <crweb> lol
[05:04] <crweb> greyback: the common answer will be: upgrade to latest torfu then upgrade to AZ
[05:05] <crweb> its ok if you don't know to what I refer. you will in a few weeks
[05:05] <greyback> crweb: heh, I don't doubt that :)
[05:05] <greyback> No, I get you.
[05:05] <crweb> 5 freaking hours
[05:05] <crweb> 5 of them
[05:05] <greyback> Good luck!
[05:05] <crweb> thats how long it took to diag that problem
[05:06] <greyback> Ouch
[05:06] <nerochiaro> it happens man, don't bash yourself over it
[05:06] <crweb> i grabbed an outdated osd on my way out
[05:07] <crweb> nerochiaro: you'd be proud though. the first time it didn't work. I automatically plugged it into a TV (last friday's issue)
[05:08] <nerochiaro> crweb: wow, that's a step forward
[05:08] <greyback> Quick question, I notice that you don't need to source neuros-env any more. neuros-env has been moved to scripts/project-env. But it has a variable TOOLCHAIN_DIR that's not defined, so if I source it to build my own stuff, PATH gets a lil screwed up.
[05:09] <greyback> Have I missed something?
[05:09] <crweb> greyback: you still need neuros-evnv
[05:09] <nerochiaro> greyback: why dont' you need that anymore ?
[05:09] <crweb> its generated by build-helper.sh
[05:10] <greyback> Hmm, ok, I must have missed something
[05:11] * crweb  sticks a postit note with "upgrade to latest torfu" to cousins wall.
[05:11] <greyback> Ah, ok, I see it now. Duhh, my bad.
[05:12] <greyback> crweb: latest torfu beta? i.e. 3.33-1.74-00.82j
[05:12] <crweb> i think it is anything newer than 7.0
[05:12] <crweb>