| 00:00.51 | *** join/#edev Madsy (n=Madsy@fu/coder/madsy) [NETSPLIT VICTIM] |
| 00:01.34 | solar | might take a job up in san mateo |
| 00:08.37 | landley | Back from dinner! |
| 00:08.42 | landley | solar: um, congrats? |
| 00:08.48 | landley | How far away is that from wherever you are now? |
| 00:08.54 | landley | doesn't know his way around california. |
| 00:16.40 | calculus | I would estimate about 30-40 miles north of San Jose |
| 00:17.03 | calculus | it is closer to my place of work |
| 00:17.53 | landley | I know that Los Angeles is down south, and San Francisco's up north, and silicon valley's either somewhere between the two or south of Los Angeles. And that San Jose is nebulously attached to Silicon Valley via transporter beam or some such. |
| 00:18.02 | landley | And then there's northern california, where "Better Off Dead" was set. |
| 00:18.07 | calculus | heh |
| 00:18.45 | calculus | silicon valley would be just south of San Francisco |
| 00:18.54 | landley | It apparently has a ski slope, and John Cusack in a top hat. |
| 00:19.26 | calculus | silicon valley to SF is about an hour's drive (give or take with traffic) vs 5-7 hours south to LA |
| 00:19.29 | landley | And then there's Orange County, which is an evil conservative empire presided over by The Terminator. |
| 00:19.38 | landley | Or was it clint eastwood? |
| 00:20.05 | landley | In Texas it's 8 hours drive from El Paso to Austin. (Done it.) |
| 00:20.26 | landley | And 4 hours from Austin to the outskirts of Dallas. Plus about 3 years to get from one side of dallas to the other if traffic's against you. |
| 00:20.52 | landley | And of course crossing Houston along I-10 requires _provisions_, and innoculations against dysentary. |
| 00:34.11 | solar | calculus: I'd call the valley pretty much SFO to SJC. |
| 00:34.56 | solar | landley: 40min train ride or so north. But if all works out well. I should get a 40-60% pay bump |
| 00:38.20 | calculus | solar: oh, between the airports... I think that is fair |
| 00:38.59 | calculus | we can be train ride buddies! |
| 01:05.31 | *** join/#edev kergoth (n=kergoth@ip24-251-170-95.ph.ph.cox.net) |
| 01:19.24 | *** join/#edev JoeLlama (n=snork@206.169.63.10) |
| 01:24.45 | *** join/#edev JoeLlama (n=snork@206.169.63.10) |
| 01:26.53 | landley | solar: if you can use your laptop in the train, it might be worth it. |
| 01:31.30 | *** join/#edev JoeMoose (n=snork@206.169.63.10) |
| 01:44.49 | solar | landley: I'd move either way asap |
| 01:47.34 | *** join/#edev GPSFan (n=kenm@64.92.145.112) |
| 01:47.34 | *** mode/#edev [+o GPSFan] by ChanServ |
| 01:47.39 | *** join/#edev bsdfox (n=h36sa@c-67-161-174-21.hsd1.ca.comcast.net) |
| 03:10.02 | *** join/#edev chiques (n=working_@pool-71-189-111-152.lsanca.fios.verizon.net) |
| 03:58.25 | *** join/#edev chiques (n=working_@pool-71-189-111-152.lsanca.fios.verizon.net) |
| 03:59.36 | *** join/#edev SandGorgon (n=OmNomNom@122.173.249.152) |
| 04:25.35 | Mirell | Oh God 2012 was bad |
| 04:26.56 | Mirell | landley: Sacramento is nowhere near Orange County |
| 04:49.21 | *** join/#edev mbuf (n=mbuf@118.102.130.6) |
| 05:05.07 | calculus | oh, he was referring to the governor |
| 05:19.51 | Mirell | Indeed. |
| 05:20.11 | Mirell | Although the Irvine Company is definitely pretty bad. |
| 05:21.48 | landley | Mirell: that's what Jeph Jacques (the guy who draws questionable content) twittered about 2012. |
| 05:26.57 | Mirell | Mmm. |
| 05:27.51 | Mirell | Mmm, debate. Part of me wants to write this using environment variables to hold stuff, other part of me wants to add the ability to use arguments... |
| 06:19.48 | *** join/#edev mbuf (n=mbuf@118.102.130.6) |
| 06:44.17 | *** join/#edev chiques (n=working_@71.189.111.152) |
| 06:54.50 | landley | Ok, http://www.cbs.com/classics/star_trek/ is unfair. |
| 06:55.02 | landley | Putting every episode of star trek TOS online... I'm trying to _work_ here, people... |
| 06:56.08 | landley | Youtube's bad enough: http://www.youtube.com/watch?v=N4dhoDxEK5w |
| 06:57.57 | landley | Ah, the star trek video doesn't work on Linux. Back to productivity... |
| 07:07.32 | *** join/#edev roxfan (i=dunno@81.246.135.52) |
| 08:40.08 | mnemoc | "The video you have requested is not available for your geographic region" ... no star trek for me either :( |
| 08:47.20 | *** join/#edev wiml (n=wiml@underhill.hhhh.org) |
| 08:47.35 | *** part/#edev wiml (n=wiml@underhill.hhhh.org) |
| 09:24.25 | *** join/#edev florian_kc (n=fuchs@Maemo/community/contributor/florian) |
| 09:24.58 | *** join/#edev m4t (i=matt@theorize.org) |
| 09:47.24 | *** join/#edev mbuf (n=mbuf@118.102.130.6) |
| 09:55.50 | *** join/#edev JoeLlama (n=snork@206.169.63.10) |
| 11:21.21 | *** join/#edev pirho (i=pirho@gateway/gpg-tor/key-0x2CEEC9CB) |
| 11:43.04 | *** join/#edev Crofton (n=balister@pool-71-171-31-89.nwrknj.east.verizon.net) |
| 11:57.26 | *** join/#edev davidc__ (n=davidc__@user-387g3od.cable.mindspring.com) |
| 12:09.06 | *** join/#edev gustavoz (n=gustavoz@host108.201-253-243.telecom.net.ar) |
| 12:20.23 | *** join/#edev mbuf (n=mbuf@118.102.130.6) |
| 13:09.59 | *** join/#edev linac (n=lin@122.90.115.145) |
| 13:40.49 | *** join/#edev pirho (i=pirho@gateway/gpg-tor/key-0x2CEEC9CB) |
| 13:46.26 | *** join/#edev likewise (n=chatzill@82.171.51.231) |
| 14:33.21 | *** join/#edev kergoth (n=kergoth@nat/montavista/x-qodniqwzmvjervow) |
| 14:37.33 | *** join/#edev GPSFan (n=kenm@64.92.145.112) |
| 14:37.34 | *** mode/#edev [+o GPSFan] by ChanServ |
| 14:59.32 | *** join/#edev mbuf (n=mbuf@122.169.154.3) |
| 15:19.31 | *** join/#edev SandGorgon (n=OmNomNom@122.173.249.152) |
| 15:49.46 | *** join/#edev CosmicPenguin (n=nobody@66-162-99-237.static.twtelecom.net) |
| 17:05.44 | *** join/#edev VK4MSL (n=vk4msl@2001:388:f000:0:0:0:0:71) |
| 17:40.43 | Mirell | HateHateHateHate Tcl |
| 17:49.03 | calculus | play nice now |
| 17:49.07 | calculus | tcl loves you |
| 17:50.20 | CosmicPenguin | tcl is lonely - it loves anybody who pays attention to it |
| 17:58.53 | *** join/#edev florian (n=fuchs@Maemo/community/contributor/florian) |
| 18:34.05 | prpplague | anyone know what the difference between standard "windows xp professional" and "windows xp professional for system builders" is? |
| 19:06.50 | solar | likes tcl |
| 19:22.18 | prpplague | ho ho hum |
| 19:22.34 | prpplague | MonMotha: ping , did you get your copy of altium? |
| 19:30.33 | prpplague | can anyone recommend a bittorrent client? |
| 19:30.58 | kergoth | what OS? console or gui? |
| 19:31.10 | kergoth | is kind of partial to transmission, personally |
| 19:31.17 | kergoth | utorrent in windows |
| 19:31.38 | prpplague | kergoth: gui + linux |
| 19:31.46 | prpplague | gnome specifically |
| 19:36.10 | prpplague | installs transmission |
| 19:36.46 | SpeedEvil | mldonkey works for me |
| 19:38.07 | Mirell | solar: ....why?!?!?!?! |
| 19:41.17 | solar | Mirell: dunno. Just found it's set of built-in functions to be quite useful over many other scripting langs. syntax takes a bit of getting used to. But as long as one knows c and bash. He/she will do fine with tcl. |
| 19:41.34 | solar | puts "[lsort [info commands]]" |
| 19:41.37 | solar | helps a lot.. |
| 19:42.36 | Mirell | Everything's a string! |
| 19:42.39 | Mirell | Except when it's not! |
| 19:44.30 | prpplague | hmm, i must be doing something wrong |
| 19:47.17 | *** join/#edev SandGorgon (n=OmNomNom@122.173.249.152) |
| 19:47.51 | prpplague | hmm no connected peers |
| 19:54.24 | Mirell | Poor prpplague |
| 19:54.30 | Mirell | No one wants to seed him. |
| 20:05.36 | *** join/#edev }mat (n=marc@113-174.78-83.cust.bluewin.ch) |
| 20:06.30 | *** join/#edev prpplague^2 (n=Dave@adsl-76-255-85-233.dsl.rcsntx.sbcglobal.net) |
| 20:17.53 | *** join/#edev JoeLlama (n=snork@206.169.63.10) |
| 20:18.04 | *** mode/#edev [+o prpplague] by ChanServ |
| 20:22.25 | *** join/#edev JoeLlama (n=snork@206.169.63.10) |
| 21:13.59 | *** join/#edev }mat_ (n=marc@237-176.78-83.cust.bluewin.ch) |
| 21:45.44 | MonMotha | prpplague: hi, and not yet |
| 21:45.55 | MonMotha | should happen soon. why? |
| 21:46.05 | MonMotha | there have been some delays in getting things started at work |
| 21:46.48 | roxfan | prpplague: for system builders = oem methinks |
| 21:47.42 | MonMotha | prpplague: yeah, the former is probably full retail while the latter is OEM license terms |
| 21:47.54 | MonMotha | if you plan to just buy it an install it on a single machine and be done, the OEM is fine |
| 21:50.19 | prpplague | MonMotha: just wanting to know how well your purchase goes |
| 21:50.27 | MonMotha | prpplague: ah. I'll let you know |
| 21:50.32 | MonMotha | hopefully things will happen within a week or so |
| 21:50.36 | prpplague | MonMotha / roxfan yea that's what i figured, i was just curious |
| 21:51.01 | prpplague | i got a new box i need to run some apps on and i need window xp pro rather than vista |
| 21:51.15 | *** join/#edev Ian_Daniher (n=it@nat-pool-128-88.olin.edu) |
| 21:51.58 | prpplague | tries to decide if he needs 1 or 2 avr's to do the job |
| 21:52.00 | MonMotha | 7 isn't awful, but XP is still faster, of course |
| 21:52.09 | MonMotha | issue is that XP64 is well nigh unusable due to lack of drivers |
| 21:52.21 | prpplague | i have an older copy of pads2005 that only runs on xp |
| 21:52.57 | prpplague | i wish i could find a second hand copy of pads2007 |
| 21:53.23 | MonMotha | I was under the impression they were illegal to resell |
| 21:53.27 | MonMotha | at least so says the EULA |
| 21:54.34 | roxfan | well, if you break the eula you just lose company support... they can't make the resale illegal :) |
| 21:54.47 | MonMotha | that's still up for debate |
| 21:54.55 | MonMotha | it's clear that if you never accept the EULA, you can resell it |
| 21:55.15 | MonMotha | what's not clear is whether one can restrict resale in a EULA at all |
| 21:55.31 | MonMotha | in general, I wouldn't want to have Mentor Graphics mad at me |
| 21:55.32 | prpplague | MonMotha: they are |
| 21:55.58 | prpplague | MonMotha: they are illegal to resell |
| 21:56.10 | prpplague | MonMotha: or atleast that is the attitude that mentor graphics takes |
| 21:56.59 | MonMotha | well, that isn't surprising |
| 21:57.10 | MonMotha | the question is is it actually illegal to resell |
| 21:57.15 | prpplague | MonMotha: mainly i'm just frustrated with the whole pcb layout issue |
| 21:57.23 | MonMotha | I don't think PADS is very good, anyway |
| 21:57.46 | prpplague | yea, problem is that everyone i'm working with wants to use orcad+pads |
| 21:57.47 | MonMotha | there is a 30 day trial of Designer if you want to play with it |
| 21:57.52 | *** join/#edev bsdfox (n=h36sa@c-67-161-174-21.hsd1.ca.comcast.net) |
| 21:57.56 | Mirell | GoGo First Sale Doctrine |
| 21:58.10 | MonMotha | yeah, somehow OrCAD+PADS became the "industry standard" despite how much it sucks and how expensive it is |
| 21:58.14 | prpplague | MonMotha: yea i plan to download a copy of the trial in a few week after i finish this project |
| 21:58.20 | MonMotha | PADS has a schematic editor, but it's terrible, so everybody keeps using OrCAD |
| 21:58.27 | MonMotha | so you end up buying two software suites |
| 21:58.40 | prpplague | MonMotha: and the difficulty in back annotation when using the two |
| 21:58.42 | MonMotha | right |
| 21:58.43 | MonMotha | they don't get along very well |
| 21:58.48 | MonMotha | Designer is very integrated |
| 21:59.00 | MonMotha | it'll highlight schematic stuff in real time as you plop things down in layout |
| 21:59.02 | prpplague | MonMotha: eagle is reasonable priced, but i just hate how eagle works |
| 21:59.09 | MonMotha | and it does import PADS/OrCAD stuff |
| 21:59.13 | MonMotha | I just don't know that you can export back |
| 21:59.26 | prpplague | MonMotha: OH it does? |
| 21:59.29 | MonMotha | well, Eagle is pretty minimal |
| 21:59.44 | prpplague | MonMotha: i can take a PADS layout and import it into designer? |
| 21:59.47 | MonMotha | yeah, Designer will import PADS, OrCAD, Allegro, P-CAD, Protel, and some other stuff |
| 21:59.48 | MonMotha | yup |
| 21:59.54 | prpplague | hmm |
| 21:59.59 | MonMotha | you can also import the OrCAD schematics |
| 22:00.08 | prpplague | wonders if he should go ahead and download designer to have a look |
| 22:00.20 | MonMotha | I know the schematic importer works alright, but I've not used the PADS layout importer |
| 22:00.23 | prpplague | iirc you can also pay a monthly usage for designer |
| 22:00.25 | MonMotha | they claim it's good, of course, but this is why there ar etrials |
| 22:00.33 | MonMotha | you can, but you have to buy at least a year up front |
| 22:00.37 | prpplague | ahh |
| 22:00.47 | MonMotha | and then I think you buy it a year at a time |
| 22:01.00 | prpplague | switches to his winblows box to download designer |
| 22:04.57 | davidc__ | I'd use Altium, but windows only is a big pain in the ass for me... |
| 22:06.22 | prpplague | davidc__: yea that is why i use windows XP pro is so i can do rdesktop |
| 22:08.56 | MonMotha | you want want to rdesktop Designer |
| 22:09.02 | MonMotha | it's very hardware accelerated |
| 22:09.21 | MonMotha | unfortunately, none of the good ECAD software runs on Linux |
| 22:09.46 | MonMotha | it used to all run on UNIX (Solaris, HP-UX, etc.), then everybody demanded that they port it to Windows, and nobody's looked back |
| 22:11.26 | prpplague | MonMotha: hmm, i would expect it to run ok over rdesktop |
| 22:11.32 | prpplague | MonMotha: oh well, no biggy |
| 22:11.36 | prpplague | MonMotha: downloading now |
| 22:11.45 | MonMotha | prpplague: it might, but I somehow doubt it |
| 22:11.51 | MonMotha | it needs a pretty beefy machine to run |
| 22:11.54 | prpplague | MonMotha: the PCB layout is my big sticking point right now |
| 22:12.06 | prpplague | MonMotha: hmm |
| 22:12.10 | prpplague | MonMotha: how beefy? |
| 22:12.16 | MonMotha | unlike the old crap like PADS, it actually has decent graphics |
| 22:12.32 | davidc__ | I think I'm gonna just take 2 weeks off and hack on KICAD / geda, and build some bastardized thing which has weird-ass keyboard shortcuts that nobody but me would love. |
| 22:12.35 | MonMotha | ~2GHz Core2 + 8800GT-ish or better to get good performance |
| 22:12.48 | MonMotha | make sure it can have at least a gig of RAM to itself, better 2 |
| 22:13.26 | MonMotha | all the layout stuff is hardware accelerated since it has 3D features |
| 22:13.48 | prpplague | davidc__: hehe |
| 22:13.54 | davidc__ | MonMotha: 3D rendering, or do they use the GPU for autorouting? |
| 22:14.03 | davidc__ | MonMotha: also, how good is its autorouter? |
| 22:14.58 | prpplague | Pentium(R) Dual-Core CPU E5300 @ 2.60GHz MemTotal: 6162052 kB |
| 22:15.37 | MonMotha | the autorouter is about as good as what I've seen from anything else, but it has some very nice interactive routing features (e.g. route a whole bus at a time with push/shove) |
| 22:15.46 | MonMotha | the GPU is only used for graphics, not autorouting |
| 22:15.56 | MonMotha | I don't generally use autorouters |
| 22:16.08 | MonMotha | they seem to not get along well with modern packages |
| 22:16.48 | davidc__ | MonMotha: the only autorouter I've ever used is eagle's - which is shite. I never use autorouters either for that reason. |
| 22:16.58 | davidc__ | I was merely wondering if there was any better out there. |
| 22:17.53 | MonMotha | well, it's probably better than Eagle's. It sure beat the pants off the old OrCAD layout autorouter. |
| 22:18.03 | MonMotha | but I always found the "interactive" features better |
| 22:18.20 | MonMotha | takes a lot of the repetitive work out of a layout while still letting you make the high-level decisions so you end up with a "good" layout |
| 22:18.27 | davidc__ | yeah - kicad apparently has a nice bridge to this free-as-in-beer push+shove router |
| 23:01.14 | *** join/#edev poakinx (n=me@pool-72-66-58-55.washdc.east.verizon.net) |
| 23:06.06 | poakinx | do DSPs help alot in general multiplication? or are they only specialized for matrix calculations? |
| 23:07.25 | SpeedEvil | It depends. |
| 23:07.57 | poakinx | SpeedEvil: could you elaborate? |
| 23:07.59 | SpeedEvil | 'proper' processors - desktop ones - tend to have integer units or FP units that can do multiplication in one cycle. |
| 23:08.03 | SpeedEvil | So maybe not so much. |
| 23:08.11 | poakinx | i see |
| 23:08.31 | poakinx | then whats so different about DSPs? |
| 23:08.39 | SpeedEvil | Smaller processors tend to make the tradeoff for size over speed - so a wide multiply may take many cycles. |
| 23:08.57 | SpeedEvil | DSP is generally optimised for it, where your average microcontroller costing the same won't be |
| 23:09.20 | poakinx | oh, i am choosing between an ARM / DSP / or ARM+DSP solution |
| 23:09.52 | SpeedEvil | Look at the instruction set docs. |
| 23:10.10 | SpeedEvil | See how many cycles each takes for the operation you need. |
| 23:10.10 | poakinx | ok |
| 23:10.15 | davidc__ | DSP usually can also do Multiply + ADD faster than a processor |
| 23:10.23 | SpeedEvil | See if you can pervert other faster instructions to do what you want. |
| 23:10.25 | davidc__ | and they sometimes have special looping constructs [zero overhead loops] |
| 23:10.35 | davidc__ | poakinx: what exactly are you trying to do? |
| 23:10.36 | SpeedEvil | And wierd addressing modes. |
| 23:12.05 | poakinx | davidc__: i havent made the algorithm yet, but basically calculating ~100 pwms that coordinate with eachother from 2 input adc |
| 23:12.28 | poakinx | davidc__: kind of like controlling a linear motor with inputs for speed and direction |
| 23:12.54 | SpeedEvil | what do the pwms do? |
| 23:13.20 | davidc__ | poakinx: and how much calculation is needed for each? |
| 23:13.29 | SpeedEvil | drive |
| 23:13.36 | davidc__ | poakinx: and at what frequency do you need to update them? |
| 23:14.08 | poakinx | davidc__: to be honest, i have not worked out the details yet. i am still assessing the situation/components needed |
| 23:14.20 | SpeedEvil | coilgun? |
| 23:14.25 | davidc__ | poakinx: figure out the algorithm first - you're going at this the wrong way around. |
| 23:15.00 | roxfan | some dsp (like c55x) can do two multiply+add in parallel |
| 23:15.22 | SpeedEvil | Yeah. |
| 23:16.01 | poakinx | davidc__: i dont think i can at this moment because the system seems fairly complex. also i have no idea if it will actually work if i dont make it |
| 23:16.17 | davidc__ | poakinx: erm, are you building just one? or a whole bunch? |
| 23:16.27 | SpeedEvil | Work out the algorithm. Read over the instruction set. See if inspiration strikes and inspires you to ways to do it much more efficiently on a processor. |
| 23:16.27 | SpeedEvil | Can you bit-slice it for example. |
| 23:17.06 | davidc__ | poakinx: also; how are you planning on generating 100PWM channels? |
| 23:17.13 | poakinx | looks up bit slicing |
| 23:17.25 | poakinx | davidc__: via fpga/cpld |
| 23:17.36 | SpeedEvil | http://www.darkside.com.au/bitslice/ |
| 23:17.47 | poakinx | thanks |
| 23:17.56 | SpeedEvil | Basically using conventional AND and OR/... instructions to make a virtual logic circuit |
| 23:18.17 | SpeedEvil | And using the parallelism of the data width of the processor to operate 32 or whatever of these at once |
| 23:18.31 | poakinx | nice |
| 23:22.17 | poakinx | by the way, on ARM+DSP solutions, from what I have read, the DSP stuff is transparent from the developer? |
| 23:22.49 | roxfan | not sure what you mean by "transparent" |
| 23:24.08 | poakinx | that the developer can just program it like a regular arm chip |
| 23:24.22 | roxfan | well it IS an arm chip |
| 23:24.41 | roxfan | usually dsp connection is done via some memory area |
| 23:24.52 | poakinx | mhm |
| 23:27.58 | roxfan | btw does your algo involve fft? some dsps have hw acceleration for that.. |
| 23:30.15 | poakinx | fft? isnt that signal filtering? then no. the algo is just varing those 100 pwm freq by 2 adc inputs |
| 23:32.37 | SpeedEvil | Do they all vary differently? |
| 23:34.27 | poakinx | they manipulate the same and only one object |
| 23:41.21 | davidc__ | poakinx: why not just build your PWM control board, then use a computer to do all the calculations to start with |
| 23:41.36 | davidc__ | poakinx: or make software model of the object you're controlling |
| 23:41.43 | davidc__ | and then develop the algorithm against that |
| 23:42.41 | poakinx | hmm |
| 23:43.09 | poakinx | well the end product has to be standalone / cant interface to a computer |
| 23:43.17 | poakinx | but what do you mean by a software modell |
| 23:43.34 | poakinx | sounds like a good idea |
| 23:44.33 | davidc__ | aka, make a piece of software that takes the PWM inputs, and emulates what would happen in the real world |
| 23:45.36 | poakinx | so i would need to use a simulation software |