| *** tpb has joined #timvideos | 00:00 | |
| mithro | MaZderMind: ping? I feel really bad about not having a really close look at the sync issue | 00:16 |
|---|---|---|
| *** Niharika has quit IRC | 00:22 | |
| *** Niharika has joined #timvideos | 00:22 | |
| CarlFK | mithro: how long will you be in Germany? | 00:41 |
| *** 64MAC5YBM has quit IRC | 00:41 | |
| CarlFK | if there is some TV sprinty stuff I might see about joining you | 00:41 |
| *** 64MAC50LY has joined #timvideos | 00:43 | |
| *** 64MAC50LY has quit IRC | 01:04 | |
| *** 16WABBNFO has joined #timvideos | 01:06 | |
| *** CarlFK has quit IRC | 01:49 | |
| *** CarlFK has joined #timvideos | 02:10 | |
| *** ChanServ sets mode: +v CarlFK | 02:10 | |
| mithro | CarlFK: the plan is definitely to be doing TimVideos hacking and meet both the Chaos Club VOC people and DebConf video team in person | 02:43 |
| CarlFK | mithro: I think it would be a little odd to hack on TV stuff while they are clicking away on DVswitch | 02:47 |
| CarlFK | and me flying to Germany to click DVswitch is silly | 02:47 |
| CarlFK | click DVswitch and hack on TV for a few hours each night is only a little better | 02:48 |
| mithro | CarlFK: my plan is to hack on TV stuff for all of Chaos Comms Camp and then figure out DebConf when we get there. | 02:59 |
| CarlFK | yeah, Ill be in NYC during CCC. | 03:00 |
| mithro | CarlFK: I'm not planning on participating in the actual A/V recording except if we have a non-DVSwitch system running | 03:00 |
| CarlFK | I hope you can keep from getting sucked in | 03:01 |
| CarlFK | I guess they are only doing 2 rooms, they should have plenty of help, so maybe. | 03:01 |
| mithro | CarlFK: we'll see | 03:01 |
| *** 16WABBNFO has quit IRC | 03:28 | |
| *** 7GHAAI47W has joined #timvideos | 03:30 | |
| *** hyades has joined #timvideos | 03:31 | |
| MaZderMind | mithro: no worries, i have put it aside to finish the remaining voctomix things, like the gui, documentation etc. | 05:36 |
| MaZderMind | if it dowsn't happen earlier, the camp is a good place to work on that one remaining issue | 05:37 |
| MaZderMind | CarlFK: we can implement the record-cut-button there ;) | 05:39 |
| MaZderMind | ah - now i understand - we will habe blackmagic hardware mixers for the lecture rooms snd angles (helpers) to operate them | 05:40 |
| MaZderMind | the voc-village will focus in postprocessing (not much to do), hacking snd having a party | 05:40 |
| MaZderMind | mithro: no dvswitch, all hardware mixers ;) nevertheless we have staff to operate them | 05:41 |
| mithro | MaZderMind: I think CarlFK was talking about DebConf I think | 05:42 |
| MaZderMind | hm kk ^^ | 05:42 |
| CarlFK | mithro: I don't care where he implements the button :p | 05:42 |
| mithro | I think, I think :P | 05:42 |
| * mithro is waiting for a dental appointment, not looking forward to it :( | 05:42 | |
| CarlFK | I'll be in NYC for sure and flying to DC the next day seems a little much | 05:43 |
| mithro | MaZderMind: snd? s/snd/and/ ? | 05:43 |
| MaZderMind | yes | 05:46 |
| MaZderMind | typing on my smartphone into an ssh terminal | 05:46 |
| *** 7GHAAI47W has quit IRC | 07:16 | |
| *** 64MAC52VY has joined #timvideos | 07:18 | |
| mithro | MaZderMind: no worries, I wasn't sure if you where abbreviated sound to snd or something else | 07:18 |
| *** 64MAC52VY has quit IRC | 07:39 | |
| *** 17SACV2B9 has joined #timvideos | 07:41 | |
| *** 17SACV2B9 has quit IRC | 07:42 | |
| *** 5EXABIRTX has joined #timvideos | 07:44 | |
| *** 5EXABIRTX has quit IRC | 07:59 | |
| *** 7GHAAI6UR has joined #timvideos | 08:00 | |
| *** sb0 has quit IRC | 08:06 | |
| *** 7GHAAI6UR has quit IRC | 08:54 | |
| *** 32NAB78RK has joined #timvideos | 08:56 | |
| *** sb0 has joined #timvideos | 09:04 | |
| *** sb0 has quit IRC | 09:12 | |
| *** 32NAB78RK has quit IRC | 10:00 | |
| *** scrollback has joined #timvideos | 10:02 | |
| *** tija has joined #timvideos | 11:30 | |
| *** sb0 has joined #timvideos | 11:37 | |
| tija | mithro: migen can covert any module into synthesizable verilog. So why not convert everything to verilog and then add the jpeg encoder. | 11:38 |
| mithro | tija: I don't understand your statement? | 11:46 |
| tija | mithro: We have a open JPEG encoder in VHDL. We have reset of the component written in MIGEN. Instead of writing a JPEG encoder in MIGEN, MIGEN gives synthesizable verilog. So convert rest of the component to verilog add jpeg encoder to it. Do I make sense? | 11:48 |
| mithro | migen can use vhdl or verilog components | 11:49 |
| mithro | See the misoc/extcores stuff | 11:50 |
| mithro | hrm, actually, it might only support verilog..... | 11:51 |
| mithro | tija: but cfelton's verilog version of the jpeg encoder should be pretty drop in | 11:52 |
| tija | mithro: hmm. The place where I am working, they code in C++ and then use Vivado HLS to convert everything to verilog. Other than dynamic memory allocation, almost all C++ constructs are valid including classes. | 11:54 |
| mithro | tija: that is a terrible idea :P | 11:55 |
| tija | mithro: There are open source substitute for Vivado HLS. http://legup.eecg.utoronto.ca/ | 11:56 |
| tpb | Title: High-Level Synthesis with LegUp (at legup.eecg.utoronto.ca) | 11:56 |
| mithro | tija: still a terrible idea :) | 11:56 |
| mithro | tija: why would you choose something as horrible as C++? | 11:57 |
| tija | mithro: C++ is horrible? It makes more stable system than python I heard :P | 11:58 |
| tija | And is still better than VHDL | 11:58 |
| mithro | tija: there is a *huge* impedance mismatch between C++ and doing things in hardware | 11:59 |
| mithro | I'd much prefer the problems that come with a Python codebase then a large C++ codebase | 12:00 |
| tija | mithro: That is the whole idea of High level synthesis. Let the compiler search for parallelism, you just write the algorithm. | 12:00 |
| mithro | tija: yes, it's great in theory :) | 12:01 |
| tija | https://github.com/stiggy87/ZynqBTC | 12:01 |
| tpb | Title: stiggy87/ZynqBTC ยท GitHub (at github.com) | 12:01 |
| tija | Bitcoin miner written in HLS. | 12:01 |
| mithro | Compilers have gotten to the stage where they can often beat a human coder when transforming C++ code into ASM instructions these days | 12:03 |
| mithro | (for things like multimedia encoders / decoders for example) | 12:04 |
| mithro | It's not unreasonable to think that C++ -> "gates" might reach a similar position 5->10 years from now | 12:04 |
| mithro | But interestingly, JIT systems can sometimes beat compiled languages these days too | 12:06 |
| mithro | because they have more information about the system at the time they JIT so can apply better optimisations | 12:06 |
| tija | What is JIT | 12:07 |
| mithro | Just-in-time compilation, what things like the JVM and PyPy do | 12:08 |
| tija | There is a similar paradigm in hardware called partial reconfiguration. A part of fpga can be programmed while running. | 12:10 |
| mithro | tija: anyway, back to something more practical | 12:12 |
| mithro | tija: I'd ask in the #m-labs channel how to integrate an external existing core into misoc | 12:13 |
| tija | mithro: hmm. But first let me get a feel of migen/misoc. | 12:15 |
| mithro | tija: okay, I thought you were talking about it because you had gotten the feel part done already | 12:16 |
| mithro | http://0pointer.net/blog/the-new-sd-bus-api-of-systemd.html | 12:28 |
| tpb | Title: The new sd-bus API of systemd (at 0pointer.net) | 12:28 |
| aps | Heyo people. Long time :) | 12:51 |
| *** scrollback has quit IRC | 13:09 | |
| *** scrollback6 has joined #timvideos | 13:12 | |
| mithro | hey aps | 13:34 |
| mithro | home time for me | 13:34 |
| mithro | be back tomorrow | 13:34 |
| aps | mithro: good night. cya | 13:35 |
| *** tija has quit IRC | 15:04 | |
| *** scrollback6 has quit IRC | 15:52 | |
| *** 1JTAAB60Y has joined #timvideos | 15:54 | |
| *** 1JTAAB60Y has quit IRC | 17:13 | |
| *** 6A4ACYVUU has joined #timvideos | 17:15 | |
| *** 6A4ACYVUU has quit IRC | 18:41 | |
| *** 5EXABIXMQ has joined #timvideos | 18:44 | |
| *** hyades has quit IRC | 22:00 | |
Generated by irclog2html.py 2.13.1 by Marius Gedminas - find it at mg.pov.lt!