Friday, 2014-07-04

*** tpb has joined #timvideos00:00
*** rohitksingh has quit IRC01:10
mithroayush3504: ping?03:18
mithrotariq786: you about?03:29
tariq786mithro: yes03:39
tariq786mithro: i am talking about image buffer that stores video pixels03:40
tariq786mithro: i want to know how HDMI2USB is handling hsync and vsync?03:40
tariq786mithro: does image buffer store pixels keeping in mind pixel's relationship with hsync and vsync or it stores blindly?03:42
tariq786mithro: does my question make the point clear?03:42
mithroI'll be back in an hour03:53
*** techdragon has quit IRC04:15
*** techdragon has joined #timvideos04:18
mithrotariq786: I'm back now05:37
mithroyou still here?05:37
mithrotariq786: I'm not particularly interested in the hsync/vsync stuff at the moment, I'd really like to see the rtp packets coming out of the ethernet port05:45
tariq786mithro: that is great. but i need to interface with HDMI2USB05:46
mithrotariq786: I don't care if the image produced from the ethernet port is a static image05:46
*** techdragon has quit IRC05:47
tariq786mithro: for the image to appear correctly, hsync and vsync information is needed05:47
tariq786mithro: or you don't care about the image as well?05:47
mithrotariq786: I want to connect my laptop to the Atlys board, fire up wireshark on my laptop and then see correctly formed rtp packets coming out the Atlys board05:48
tariq786mithro: that is understood05:48
JoelwJust a single colour or a test pattern should be fine05:49
mithrotariq786: the actual image contents can be taken from the HDMI port, but I would be just as happy with a test pattern or similar05:49
tariq786joelw: sure. I just want to be careful because there are a lot pixels that should be ignored05:50
tariq786joelw: these are the ones in the front porch, during hsync or vsync = 1 and in the backporch05:50
tariq786mithro: all i want to know is does image buffer distinguish between these or not?05:51
JoelwIs there any significance to them? Do they actually contain any data? It could all be zeros or teletext data or something.05:51
mithrotariq786: just assume you have a buffer which has a single frame of image data you can read from05:51
JoelwIf it's not actually part of the image, and it's not any other data you're interested in, you probably don't need to worry about it.05:51
tariq786joelw: i am curious to know does image buffer in HDMI2USB has this data or not?05:52
JoelwDunno. Have you had a look at the code?05:52
tariq786joelw: if it has this data, i should ignore these while making rtp packet05:53
JoelwDoes the HDMI spec say what the nature of the data is?05:53
JoelwIf it's not part of the image and it doesn't encode any other useful information about the image, it's probably not too important.05:53
tariq786joelw: yes i did. Due to unfamiliarity with vhdl, i am struggling with the code base05:53
tariq786joelw: vtc_demo clearly demarcates good pixels from the ones that need to be ignored by hcount and vcount05:54
mithrotariq786: ignoring all the HDMI2USB code, if you just have a buffer with has valid pixel data in it05:55
mithrotariq786: take that data, put it in the correct rtp data and push it out the ethernet port05:56
tariq786mithro:that is what i have at the moment but chris wants to have integration with hdmi2usb05:56
tariq786actually that is why i did vtc_demo from Xilinx App 495,so as to generate valid pixels05:57
mithrotariq786: so, I can download your code to my Atlys and see the UDP packets coming out the ethernet port?05:57
tariq786mithro: yes05:57
tariq786mithro: this is what i am doing at the moment05:57
tariq786integrating MAC core05:57
mithrotariq786: can you please write up some instructions for people to reproduce this?05:58
tariq786mithro: reproduce what?05:58
mithro--> so, I can download your code to my Atlys and see the UDP packets coming out the ---ethernet port?05:58
*** tija has joined #timvideos05:59
tariq786mithro: once i am done with testing, i will document it fully. We also have a spec document that is updated to reflect what is the overall picture05:59
tpbTitle: HDMI2Ethernet Specification - Google Docs (at
tijashenki: 30 fps!!! @ 1024x76806:00
mithrotariq786: software development is an iterative process, there is no "once I'm done with testing" there will always be more testing to do06:00
tariq786mithro: true.06:00
tariq786mithro: but i have n't started testing yet on real hardware. This is what is going on06:01
mithrookay, looking at your first diagram in that document06:01
mithroFigure 1: HDMI2Ethernet Architecture Pipeline06:01
tariq786mithro: ok06:01
mithrotariq786: if the "memory-buffer" can be filled with contents from a test image generator rather then the HDMI subsystem06:03
mithrotariq786: then you can test the system without needing to integrate with the HDMI subsystem06:03
tariq786mithro: this is what i was heading but chris suggested to integrate with hdmi2usb06:04
tijatariq786: image buffer stores only pixels. It uses hsync and vsync to calculate the resolution.06:04
tijatariq786: resolution is calculated using a module called calc_res06:04
tariq786tija: hi, yes i saw that06:05
mithrotariq786: I really want to get something from you that we can reproduce here06:05
shenkitija: that's awesome06:06
tariq786mithro: you bet06:06
tariq786see the picture to identify valid pixels from invalid ones06:06
mithrotariq786: that is already done for you06:07
tariq786mithro: ok that means image buffer has only valid pixels06:07
mithrotariq786: yes, I think that is safe assumption to start with06:07
tariq786mithro: ok06:08
tariq786mithro: thanks06:08
mithrotariq786: if that turns out to not be the case, we can look at that later06:08
mithrotariq786: but I'd really like to be able to see the ethernet/IP data coming out of the physical Atlys hardware06:09
tariq786mithro:ok. I am going to try my best not to disappoint you06:09
mithrotariq786: If I understand correctly, that was the mid-term goal?06:10
mithrotariq786: what was the mid-term goal?06:11
tariq786mithro: midterm goal was rtp packet creation, synthesis and sending it to MAC core06:11
mithrotariq786: okay - that implies to me that I should be able to see ethernet/IP data coming out of the Atlys board to me06:12
tariq786mithro: before midterm was simulation. Now is emulation06:13
mithrotariq786: as if creating RTP packets and sending them to the MAC core, they should appear out the device right?06:13
tariq786mithro: mac has two interfaces. Upstream that interfaces with RTP and downstream that interfaces with PHY.06:14
tariq786mithro: upstream is done, the downstream needs to be tested on atlys06:14
JoelwWe already know that the MAC works - you don't really need to do much work to prove that.06:15
JoelwAs long as you're generating valid Ethernet packets and your code is synthesisable, it should pretty much just work.06:15
tariq786joelw: true06:15
tariq786joelw: that is what is going on now. mac testing. Righnow having issues with cable driver06:16
tariq786on ubuntu06:16
tariq786as i have ubuntu on the pc that has 1G ethernet card06:16
mithroshenki: gah, ajit left after saying that stuff :P06:19
tijamithro: I am still here :D06:20
mithrotija: oh, cool! are you going to be around in 1 hour?06:20
tijamithro: yes06:20
mithrogreat lets chat more then06:20
mithrotija: I think I'd like you and rohit to synergies a bit :P06:21
tijamithro: what does it mean?06:21
mithrotija: opportunities to help each other and amplify each others work06:22
tijamithro: yeah sure!06:22
*** techdragon has joined #timvideos06:22
mithrotariq786: so can we put everything on hold that isn't working towards the goal of getting the actual Atlys board producing Ethernet frames and some in else reproducing that result06:35
mithroI'm happy to talk to Chris about it06:38
tariq786mithro: ok06:41
mithrotija / shenki: if you where both available in 20-30ish minutes we could try and have a VC then06:43
tijamithro: I am in06:45
*** slomo has joined #timvideos06:50
*** CarlFK has quit IRC07:23
shenkii can't sorry07:24
shenkiyou guys should feel free to chat07:24
*** CarlFK has joined #timvideos07:24
*** ChanServ sets mode: +v CarlFK07:24
mithrotija: what was the URL to your mid-term report again?07:37
tpbTitle: Mid Term Report - Google Docs (at
mithrotija: adding comments to that now07:46
tijamithro: okay07:46
mparuszewskihello guys, hello mithro :) I am ready to talk. Unfortunately I've got poor Internet connection and video hangout won't work, yesterday I tried to talk with friends and it was impossible. :/07:58
mithromparuszewski: okay, text chat is probably okay08:00
mithromparuszewski: so where are we at?08:01
mparuszewskiWe should talk about deployment getting-started website.08:04
mithromparuszewski: is the latest code deployed somewhere?08:04
mparuszewskilatest code is deployed on pinoss.github.io08:04
mithromparuszewski: so what do we need to do to get the website to that?08:05
mparuszewskifirst, I need to create pull request to timvideos/getting-started to merge my changes to main repository08:06
mithroand how should we deal with bugs / issues in your current output08:07
mithromparuszewski: and how should the deployment change with the prose editor verses the old wiki paged editor?08:07
mithromparuszewski: do we delete the wiki from getting-started?08:08
mparuszewskiyes, we should delete wiki from getting-started and use only or deploy own with own gatekeeper as timvideos repository08:08
mithromparuszewski: so how do we do the gatekeeper part?08:09
tijamithro: going for lunch. will be back in 20 mins.08:10
mithromparuszewski: IE what are the exact steps we need to do to get the website deployed in a state where people can edit things and start reporting issues / bugs / feature requests08:10
mparuszewskiCurrently I have gatekeeper on my own heroku server, but I created new account only for timvideos and I will deploy the gatekeeper there.08:11
mparuszewskiMaybe I will prepare step-by-step tutorial how to deploy that page?08:11
mparuszewskiwith, gatekeeper, configuration, etc?08:11
mithromparuszewski: I'd like a set of instructions that anyone could follow to do the migration08:12
mparuszewskigreat, I will do that.08:12
mparuszewskiI think we should move issues from getting-started page to projects related to selected issue, because currently there is a lot of issues that, in my opinion, should not be in timvideos/getting-started issues but in HDMI2USB project, etc.08:13
mithromparuszewski: getting-started should be a collection of useful issues from all timvideos projects which you could get started with08:14
mparuszewskiif we do that, then we can have on getting-started repository issues/bugs/feature request related only to that repository08:14
mithromparuszewski: the reason the website ended up in that repo is that the wiki started tracking information on getting started too (IE descriptions of each project and such)08:15
mithromparuszewski: so maybe we need to restructure that08:15
mithromparuszewski: but we still very much want a place to track/include/hint at how to "get started" on timvideos08:16
mithrothere are a lot of moving parts and a lot of potential things to work on08:16
mithroso we want to make it easier on new people08:16
mparuszewskiyes, I think it's a good place to have all issues for those who want to start working on timvideos08:17
mithromparuszewski: so, how do we do that while giving you a place to record the website stuff?08:20
mparuszewskiWe can delete wiki part, leave issues in getting-started and if someone find bug or want to add some features to timvideos, should write an issue in getting-started repository with proper label. Currently we have many labels and they are self-explanatory, so it works in my opinion.08:24
mithrobut getting-started is for good issues to get started with, not for all website issues08:24
mparuszewskihmm, you are right. Maybe we should create new repository with content and issues related only to TimVideos developers page08:27
mparuszewskibut what we will have in getting-started repository except issues08:27
mithromparuszewski: that is a good question08:28
mithromparuszewski: because there is a huge overlap between the website itself and "getting-started documentation" right?08:28
mithromparuszewski: I have no idea how to deal with that :P08:29
mithromparuszewski: but maybe we just ignore that for now08:30
mithromparuszewski: I also think we want a good separation of the "content" from tools for using the content right?08:30
mparuszewskiMe neither, we can ignore that for now and maybe we will figure out something later.08:32
mparuszewskiRight, content and tools should be separated.08:32
mithromparuszewski: so what is our next step?08:33
mparuszewskiIn prose I can hide all files except pages with content, so someone who wants to edit content will not know that he is editing jekyll repository.08:34
mithromparuszewski: do you want to dump some steps into a Google doc somewhere that we can edit it together?08:34
mparuszewskiyes, I will prepare document right now08:34
mithromparuszewski: I'm happy to help work on it with you, link?08:35
tpbTitle: TimVideos Getting Started Deployment - Google Docs (at
mithrotija: BTW Did you see this update -> ?09:21
tpbTitle: Digilent Atlys Board: Loading Firmware · timvideos/HDMI2USB Wiki · GitHub (at
mparuszewskimithro: how about Supybot, how to make sure it always is running?09:43
mithromparuszewski: good question09:43
mithrotpb is the supybot and I original set it up for a different project09:44
mparuszewskiI created working configuration, but I don't know how to run it on generate.timvideos.us09:44
mithroit runs on roger.mithis.com09:44
tpbTitle: pinoss/supybot-configuration · GitHub (at
mithroWe really should migrate him somewhere better09:45
mparuszewskiCan I create crontab on to run planet supybot?09:47
mithromparuszewski: we don't want multiple bots09:47
tijamithro: I answered most of you comments.09:49
mithromparuszewski: we also have the IRC logs at ->
tpbTitle: Friday, 2014-07-04 (at
mparuszewskiI know that, these are really helpful for me.09:50
mparuszewskiMy configuration is in repository so it won't be lost, we can marge it later. :)09:55
mparuszewskimithro: Do you have more questions? :)09:56
mithrothe tpb bot config isn't commited anywhere :P09:57
mithrowe use a totally different supybot during events and it makes sense to keep them separate I think?09:58
mparuszewskiYes, of course.10:00
*** FeltonChris has joined #timvideos10:08
mithrohey FeltonChris10:13
*** tija has quit IRC10:19
*** tija has joined #timvideos11:16
mithromparuszewski: so where are we at?11:18
mithrotija: can you give me edit permission to ->
tpbTitle: Debug Design - Google Docs (at
mithro tija: BTW Did you see this update -> ?11:27
tpbTitle: Digilent Atlys Board: Loading Firmware · timvideos/HDMI2USB Wiki · GitHub (at
mithroHeading home now11:37
*** slomo_ has joined #timvideos12:05
*** slomo has quit IRC12:08
*** FeltonChris has quit IRC12:33
*** FeltonChris has joined #timvideos12:50
*** mparuszewski has quit IRC13:02
*** tija has quit IRC13:23
*** mparuszewski has joined #timvideos13:28
*** mparuszewski has quit IRC15:32
*** FeltonChris has quit IRC15:37
*** FeltonChris has joined #timvideos15:55
*** rohitksingh has joined #timvideos17:40
*** mparuszewski has joined #timvideos18:05
mparuszewskimithro: I am back, had some problems with Internet connection, now should be good, but I think I will have some problems for about 1-2 week to have good internet connection.18:07
mparuszewskiWhat should I do now?18:07
*** FeltonChris has quit IRC18:09
*** rohitksingh has quit IRC19:29
*** rohitksingh has joined #timvideos19:41
*** slomo_ has quit IRC21:41
*** rohitksingh1 has joined #timvideos22:58
*** rohitksingh has quit IRC22:58
*** rohitksingh1 is now known as rohitksingh22:58
*** mparuszewski has quit IRC23:52

Generated by 2.12.1 by Marius Gedminas - find it at!