| 00:52.26 | *** join/#edev g1powermac_PB (n=g1powerm@CPE-76-177-156-202.natcky.res.rr.com) |
| 01:26.22 | *** join/#edev Crofton (n=balister@66-207-66-26.black.dmt.ntelos.net) |
| 03:11.43 | *** join/#edev linac (n=lin@216.80.76.222.broad.fz.fj.dynamic.163data.com.cn) |
| 03:39.18 | *** join/#edev DrZimmerman (n=theo@2001:5c0:8fff:fffe:0:0:0:6af1) |
| 03:50.08 | *** part/#edev pierrelux (n=pierre-l@144-125.sh.cgocable.ca) |
| 04:10.33 | *** join/#edev T0mW (n=Tom@24.238.86.158.res-cmts.sth.ptd.net) |
| 05:51.17 | steliosk | prpplague : For a product we want to make. I am looking for something we can get off the self, instead of designing it |
| 05:57.05 | steliosk | prpplague : but can't see to find any |
| 05:58.02 | T0mW | such is the way of the world |
| 05:58.28 | T0mW | nothing fits "just right" |
| 06:07.29 | Exiles | usb to sram? |
| 06:14.02 | steliosk | Exiles : yes |
| 06:23.55 | Exiles | well, use a micro with USB support, and your set |
| 06:26.00 | T0mW | LPC2148 if you only need a USB slave |
| 06:27.22 | Exiles | what do you need to do with the sram, and how much? |
| 06:29.17 | steliosk | Exiles : 1MB min, +battery backup would be ideal. Should conform to the usb ata (i.e look like as usb drive to the host OS in our case linux) |
| 06:32.28 | Exiles | why use sram? |
| 06:33.24 | Exiles | 1mb sram is very costly 1 mb as in 1 Mega byte, or 1 mega bit? |
| 06:33.44 | steliosk | Megabyte or 8 mbit :) |
| 06:34.23 | Exiles | costly |
| 06:35.16 | steliosk | well we need to write to it many times very often and if we take into account the flash specs, if we use a flash drive it looks like it will wear out in less than a year |
| 06:36.16 | Exiles | well, chips of that size do exist |
| 06:37.00 | steliosk | yeah, does not have to be a single chip either, no need to be in the usb stick "form factor" |
| 06:37.54 | steliosk | the idea was to look for something allready in the market as initial quantity would be like 200+ units |
| 06:38.09 | Exiles | dosn't exist |
| 06:38.48 | steliosk | mram is also nice, but even more expensive than sram, but you don't need a battery to backup data |
| 06:39.24 | Exiles | I have a usb board |
| 06:40.13 | Exiles | wonder what it would take to make it act as a mass storage device and stick sram outside it |
| 06:40.28 | Exiles | my sram chips atm have built in batter backup |
| 06:40.34 | Exiles | battery* |
| 06:41.54 | Exiles | but at 2M x 8 |
| 06:42.21 | steliosk | Exiles : Software :) |
| 06:43.44 | Exiles | basically 2MB |
| 06:43.58 | Exiles | this chips a little expensive |
| 06:45.11 | Exiles | http://www.maxim-ic.com/quick_view2.cfm/qv_pk/2677 |
| 06:45.13 | steliosk | which chipset your board uses ? |
| 06:45.28 | Exiles | pic18f4550 |
| 06:46.13 | Exiles | http://www.digikey.com/scripts/DkSearch/dksus.dll?Detail?name=DS1270Y-70%23-ND |
| 06:46.31 | Exiles | well that sram part is expensive |
| 06:47.27 | steliosk | is this a custom board ? |
| 06:47.34 | Exiles | yeah |
| 06:47.56 | steliosk | the pic has a usb2 or usb1.1 interface ? |
| 06:48.27 | Exiles | 2, full only has upto 11mbit speeds |
| 06:49.15 | Exiles | so if plugged into a usb2 hub it won't slow down the rest of the devices pugged in |
| 06:49.58 | Exiles | http://www.digikey.com/scripts/DkSearch/dksus.dll?Detail?name=DS1265AB-70-ND |
| 06:50.10 | Exiles | theres a 1mb sram chip |
| 06:50.17 | Exiles | http://www.maxim-ic.com/quick_view2.cfm/qv_pk/2675 |
| 06:52.03 | steliosk | mrams are cheaper >http://www.digikey.com/scripts/DkSearch/dksus.dll?KeywordSearch |
| 06:52.28 | Exiles | that url is useless |
| 06:53.13 | steliosk | do a search for mram in digikey |
| 06:53.49 | Exiles | but what about 8mbits worth? |
| 06:54.53 | Exiles | well, I don't want to solder those |
| 06:57.03 | steliosk | well you need 2 mram chips 4mbit each, which would be around 50$ |
| 06:59.26 | steliosk | as for the soldering, we have our inhouse smd production line ;) |
| 06:59.46 | Exiles | nice |
| 06:59.46 | Exiles | I have my crap soldering iron |
| 07:00.31 | Exiles | so smd work is ually out of the question |
| 07:00.44 | Exiles | how fast do you need to read/write to them? |
| 07:01.52 | steliosk | http://www.autotech.gr/main/index.php?option=com_content&task=view&id=23&Itemid=57 |
| 07:03.57 | Exiles | sounds like you could handle my FPGA's |
| 07:04.34 | Exiles | XILINX XCV1000 |
| 07:04.53 | Exiles | 560 BGA |
| 07:05.15 | steliosk | well we could but we don't have bga inspection station :( |
| 07:05.38 | Exiles | okay, who's got the XRAY Machine |
| 07:06.09 | Exiles | I have 4 of them |
| 07:06.48 | Exiles | as far as I know they work |
| 07:06.55 | Exiles | million gate fpga's |
| 07:07.38 | Exiles | my standing offer is, get me a working board with one, and the rest are up for grabs |
| 07:09.23 | steliosk | what's the cards form factor ? pci/custom/eurocad ? |
| 07:10.05 | Exiles | cards? |
| 07:11.07 | steliosk | the board |
| 07:11.30 | Exiles | custom, development board |
| 07:11.58 | Exiles | no one has designed one yet |
| 07:12.55 | steliosk | i am working on a project with virtex 4 now, adding also support for the xilinx ppc cores in openembedded |
| 07:13.11 | Exiles | cool |
| 07:13.22 | Exiles | I was given these 4 chips |
| 07:13.39 | Exiles | thats why I won't sell them |
| 07:13.52 | Exiles | but I'll trade with them |
| 08:49.13 | *** join/#edev hw__ (n=hw@p578b3905.dip0.t-ipconnect.de) |
| 12:01.26 | *** join/#edev linac (n=lin@121.71.52.28) |
| 12:13.45 | *** join/#edev g1powermac_PB (n=g1powerm@CPE-76-177-156-202.natcky.res.rr.com) |
| 12:15.10 | *** join/#edev leoncamel (n=leoncame@p7146-ipbfp502fukuokachu.fukuoka.ocn.ne.jp) |
| 13:24.20 | *** join/#edev brad_mssw (n=brad@shop.monetra.com) |
| 13:39.44 | *** join/#edev hw_ (n=hw@p578b3905.dip0.t-ipconnect.de) |
| 13:41.49 | *** join/#edev prpplague (n=Dave@mail.americanmicrosystems.com) |
| 13:48.35 | *** join/#edev prpplague (n=DaveAnde@mail.americanmicrosystems.com) |
| 13:56.14 | *** join/#edev leoncamel (n=leoncame@p7146-ipbfp502fukuokachu.fukuoka.ocn.ne.jp) |
| 14:05.39 | prpplague | CP|too: you alive? |
| 14:06.07 | prpplague | ho ho hum |
| 14:11.32 | *** join/#edev CosmicPenguin (n=nobody@crank.laptop.org) |
| 14:11.42 | prpplague | CosmicPenguin: morning |
| 14:13.32 | CosmicPenguin | morning |
| 14:28.10 | prpplague | CosmicPenguin: you got a second? |
| 14:28.15 | chouimat | morning |
| 14:28.24 | prpplague | chouimat: greetings |
| 14:29.56 | chouimat | if a device battery is charging for 45 minutes to 90 minutes and then shutdown the device and you can't boot after it ... does this mean I got a bad battery? |
| 14:30.36 | prpplague | sounds so, more than likely the internal "fuel gauge" chip is reporting a problem |
| 14:31.48 | chouimat | sound like this ... I was able to get it working after that and the battery was critically low even if I left the device plugged for 8 hours ... and even if I plugged it all night I can't get the phone working at all |
| 14:32.15 | CosmicPenguin | prpplague: fire wawya |
| 14:32.21 | CosmicPenguin | away even |
| 14:32.35 | prpplague | http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=blob;f=arch/arm/mm/ioremap.c;hb=6d78b5f9c6cf59c98d3833e09d0ed6aebd6a33d3 |
| 14:32.41 | prpplague | function starting at line 44 |
| 14:33.04 | prpplague | shouldn't the section "bad:" return -EFAULT ? |
| 14:33.45 | CosmicPenguin | its not going to make it back from a BUG() |
| 14:34.10 | prpplague | ahh |
| 14:34.24 | prpplague | problem is the compiler gives a warning |
| 14:34.34 | prpplague | so i guess what it returns isn't really important |
| 14:34.34 | CosmicPenguin | yeah - probably |
| 14:34.50 | CosmicPenguin | thats one of those false positives that people who don't like compiler warnings glom onto |
| 14:35.39 | CosmicPenguin | you don't like them in that you want to fix them, or you don't like them because you think they're not useful? |
| 14:36.20 | prpplague | i don't like them and want to fix them |
| 14:36.47 | CosmicPenguin | I was talking about the other people |
| 14:38.06 | prpplague | oh |
| 14:38.25 | chouimat | prpplague: does the fact that I got the device yesterday and it's brand new help to reach the conclusion that the battery is bad? |
| 14:38.30 | prpplague | well i'm going to add a return anyway to kill the warning |
| 14:38.52 | prpplague | chouimat: if it were one of our products you'd definetly get an rma |
| 14:39.15 | *** join/#edev g1powermac_ (n=g1powerm@adsl-147-182-120.sdf.bellsouth.net) |
| 14:39.42 | chouimat | prpplague: thanks |
| 14:39.45 | prpplague | np |
| 14:40.18 | prpplague | CosmicPenguin: so if you were looking at a patch that included a return -EFAULT in that function, would you accept it? |
| 14:40.19 | chouimat | prpplague: so far my experience with the greenphone is not good |
| 14:40.53 | *** join/#edev howtoo (n=howtoo@oslo-static64.bluecom.no) |
| 14:41.25 | prpplague | chouimat: doh |
| 14:41.41 | chouimat | but i like the size |
| 14:44.21 | CosmicPenguin | prpplague: would I accept it? Yes |
| 14:44.26 | CosmicPenguin | would Russell or Linus accept it? |
| 14:44.28 | CosmicPenguin | unclear |
| 14:44.55 | prpplague | lovely |
| 14:45.14 | chouimat | prpplague: this weird the device just reste and booted ... it's now charging |
| 14:45.47 | CosmicPenguin | sometimes the EC that is used to handle charging needs a kick in the head |
| 14:46.26 | chouimat | I just hope it will not stop after 45 minuies this time .... |
| 14:55.41 | *** join/#edev Crofton (n=balister@hc6521735.dhcp.vt.edu) |
| 15:02.47 | chouimat | ok went dark after 15 minutes |
| 15:03.20 | prpplague | chouimat: doh |
| 15:04.02 | chouimat | prpplague: I huess it's hardware defect |
| 15:04.45 | *** join/#edev g1powermac__ (n=g1powerm@adsl-147-182-37.sdf.bellsouth.net) |
| 15:09.52 | chouimat | prpplague: btw 4800x1200 is really nice for work |
| 15:11.20 | prpplague | hehe |
| 15:12.26 | chouimat | I only 3 19" CRT but this impress some of the customers ... |
| 15:12.43 | brad_mssw | :/ only got 3840 x 1024 (3 x 17" LCD) |
| 15:14.10 | brad_mssw | anyone have experience designing gumstix modules? |
| 15:14.23 | prpplague | ~lart brad_mssw |
| 15:14.23 | purl | readies the nuke launcher and fires some rounds at brad_mssw |
| 15:14.32 | prpplague | we don't do gumstix here, hehe |
| 15:15.04 | prpplague | brad_mssw: seriously though, i don't think anyone but tomw might have done a module for gumstix |
| 15:15.09 | chouimat | prpplague: what piss me more is to have to wait 10 hours to get it charging for 15 minutes |
| 15:15.20 | prpplague | chouimat: hehe |
| 15:15.31 | brad_mssw | may I ask why? I'm not tied to gumstix, just thought it might be the fastest way to get to market with a product |
| 15:15.46 | chouimat | definitly sound like a bad hardware ... now is this the battery of the phone itself |
| 15:16.13 | prpplague | brad_mssw: hehe, mainly because my employeer has been working on introducing a similiar product |
| 15:16.23 | prpplague | brad_mssw: http://www.elinux.org/Hammer_Board |
| 15:16.59 | brad_mssw | prpplague: ahh, cool |
| 15:17.35 | prpplague | brad_mssw: .1" pins so no problems doing protoboards |
| 15:17.59 | brad_mssw | yeah, that's actually my main problem with gumstix, the damn hirose connectors are a pain to solder ;) |
| 15:18.21 | brad_mssw | (at least for me, with nearly no soldering experience ... I'm a software guy) |
| 15:18.35 | prpplague | brad_mssw: yep |
| 15:21.58 | brad_mssw | I actually like that design a lot ... hmm |
| 15:23.24 | brad_mssw | does the Samsung chip on there provide a way to supply a V_BAT_BACKUP for the rtc? |
| 15:24.18 | chouimat | prpplague: the board sound cool but since me and soldering iron == need a new apartment I will pass :) |
| 15:24.27 | agaffney | brad_mssw: are you the same brad_mssw that did work on gentoo's genkernel? |
| 15:24.38 | prpplague | brad_mssw: yea the chip does, but the first Hammer board doesn't support it |
| 15:24.45 | brad_mssw | agaffney: hah, yeah, years and years ago |
| 15:24.55 | prpplague | brad_mssw: the next release will be a low power version designed to run on two AA batteries |
| 15:24.57 | brad_mssw | agaffney: its ugly |
| 15:24.58 | agaffney | brad_mssw: I only know your name from the directory in CVS :P |
| 15:25.03 | agaffney | brad_mssw: it's a lot less ugly these days |
| 15:25.07 | agaffney | I'm one of the maintainers |
| 15:25.13 | agaffney | not by choice :P |
| 15:25.17 | prpplague | chouimat: you can get a carrier board to work with it |
| 15:25.17 | agaffney | I kinda fell into the roll |
| 15:25.31 | brad_mssw | agaffney: heh, good deal ... I had to write it because the predecessor didn't support x86_64, and I was the one starting the x86_64 port |
| 15:25.38 | agaffney | ah |
| 15:25.43 | brad_mssw | agaffney: and it was needed to build livecds |
| 15:25.56 | agaffney | gk 3.4.8 is pretty good about that kind of thing these days |
| 15:26.03 | agaffney | relatively flexible and cross-arch |
| 15:27.20 | prpplague | brad_mssw: there will also be a large 80-pin board called the "Sledge Hammer" which will have more i/o and power management options |
| 15:27.44 | brad_mssw | prpplague: do you all do custom embedded projects? |
| 15:28.11 | prpplague | brad_mssw: you mean the company i work for? or people here in this channel? |
| 15:28.20 | brad_mssw | prpplague: the people you work for I suppose |
| 15:29.24 | brad_mssw | agaffney: my main goal back then with genkernel was to remove all the bash-isms (as it's just plain ugly, and not portable), and get rid of as many external dependencies as possible (like python) |
| 15:29.34 | prpplague | brad_mssw: yea, we specialize in industrial handhelds, but we are jsut about to start retailing alot of the project designs that don't get used/purchased |
| 15:33.00 | brad_mssw | prpplague: we're looking at needing a board and case developed ... arm-based preferably, with CF or MMC/SD support, 10/100 ethernet (1 port needed, 2 ports would be nicer), 2 RS232 (1 for dedicated console), USB Host port, RTC with battery backup is a must, 32-128MB RAM (depending on price), and a modem of some sort, whether soft or hardware (like Multitech's RJModem) ... |
| 15:33.46 | prpplague | brad_mssw: no lcd? |
| 15:34.06 | brad_mssw | no lcd |
| 15:35.03 | brad_mssw | our current development systems are soekris net4801, mainly because it provided us with a PCI slot which we could put a modem in, we've shipped a couple out as demos to clients |
| 15:35.56 | prpplague | thats pretty close to the m5900 design minus the keypad and lcd |
| 15:36.10 | brad_mssw | not really a huge fan of soekris, especially since they won't answer e-mails or their phone, for 2-5k qty (plus, not really a fan of geode) |
| 15:36.18 | prpplague | http://www.elinux.org/M5900 |
| 15:36.29 | prpplague | that design could be relaid out |
| 15:36.38 | CosmicPenguin | ... |
| 15:37.10 | prpplague | hehe |
| 15:37.23 | CosmicPenguin | Maybe if AMD would hire some real engineers to work on Geode, rather then those bastards they have now |
| 15:37.27 | CosmicPenguin | bunch of lazy, no nothing hacks |
| 15:37.43 | prpplague | hehe |
| 15:37.57 | brad_mssw | prpplague: what's with the board in the background image? prototype or something? |
| 15:38.56 | brad_mssw | CosmicPenguin: heh, I just don't think x86 makes for a good embedded architecture ... no clue about the amd engineers working on geode |
| 15:39.02 | prpplague | brad_mssw: yea, we always do a large development board before going to actual size pcb |
| 15:39.20 | CosmicPenguin | brad_mssw: yeah, they're idiots |
| 15:39.45 | CosmicPenguin | heh - thats probably worse |
| 15:40.00 | prpplague | brad_mssw: CosmicPenguinis the lead linux guy at AMD for the geode stuff |
| 15:40.08 | brad_mssw | Jordan Crouse |
| 15:40.08 | brad_mssw | Senior Linux Engineer |
| 15:40.08 | brad_mssw | Advanced Micro Devices, Inc. |
| 15:40.12 | brad_mssw | nice |
| 15:40.57 | CosmicPenguin | And yes, I am an idiot |
| 15:41.05 | CosmicPenguin | prpplague can vouch for me |
| 15:41.24 | prpplague | hehe |
| 15:41.38 | brad_mssw | good to know ;) |
| 15:41.49 | prpplague | CosmicPenguin: one idiot calling another person an idiot generally doesn't carry much weight |
| 15:42.01 | CosmicPenguin | Who better to know an idiot then you? |
| 15:42.03 | CosmicPenguin | hoo! |
| 15:42.07 | prpplague | hehe |
| 15:42.15 | prpplague | indeed |
| 15:42.31 | CosmicPenguin | prpplague is most definately not an idiot |
| 15:42.48 | prpplague | CosmicPenguin: you've been drinking too much moose drool, hehe |
| 15:44.58 | brad_mssw | prpplague: I see you all seem to like this s3c2410 samsung chip ... |
| 15:45.46 | brad_mssw | we'd probably prefer a slightly higher-frequency processor though |
| 15:45.56 | brad_mssw | but I'm sure that's no big deal |
| 15:46.27 | prpplague | brad_mssw: s3c2440 can run at 300/400 and the s3c2460 runs at 500/600 |
| 15:46.42 | brad_mssw | 400 is fine |
| 15:46.51 | prpplague | brad_mssw: but you'll find that the s3c2410 running at 200mhz will be about the same performance as a 400mhz xscale |
| 15:46.58 | brad_mssw | really? |
| 15:47.01 | prpplague | yea |
| 15:47.13 | brad_mssw | so basically the same as our gumstix connex 400 board |
| 15:47.22 | prpplague | pretty much |
| 15:47.30 | brad_mssw | interesting |
| 15:47.46 | prpplague | brad_mssw: xscale has some internal pipe issues that limited some operations |
| 15:48.01 | prpplague | brad_mssw: the standard arm9tdmi core doesn't have those issues |
| 15:48.21 | prpplague | brad_mssw: so most arm9tdmi cores out perform xscale at the same core freq |
| 15:48.37 | brad_mssw | nice |
| 15:49.10 | brad_mssw | where's your company based out of ? |
| 15:49.21 | prpplague | dallas ft.worth area |
| 15:51.44 | brad_mssw | cool, I may get back to you when I find out how this latest demo goes |
| 15:51.57 | prpplague | okie dokie |
| 15:52.20 | brad_mssw | only problem is I know they're in a hurry because their current solution has failed ... so that'll be their next question :/ |
| 15:53.25 | *** join/#edev ico2 (n=ico2@modem-1858.cougar.dialup.pol.co.uk) |
| 15:55.47 | steliosk | prpplague : will you be at the power.org devcon ? |
| 15:58.16 | prpplague | steliosk: sorry no |
| 16:34.04 | *** join/#edev leoncamel (n=leoncame@p7146-ipbfp502fukuokachu.fukuoka.ocn.ne.jp) |
| 16:46.31 | *** join/#edev ico2 (n=ico2@modem-854.cheetah.dialup.pol.co.uk) |
| 17:06.29 | *** join/#edev lyakh (n=lyakh@dynamic-unidsl-85-197-26-24.westend.de) |
| 17:58.13 | ds2 | donno if the guy asking for a USB SRAM drive is still here, but |
| 17:58.21 | ds2 | a gumstix would give you everything you need to do that |
| 17:59.56 | prpplague | ds2: did he say why he wanted a usb sram drive? |
| 18:01.27 | ds2 | prpplague: something about rewrites cycles |
| 18:01.49 | prpplague | ds2: ahh |
| 18:02.07 | ds2 | odd request but *shrug* |
| 18:02.21 | prpplague | ds2: well you could use something like the Hammer or Gumstix with a usb mass storage gadget driver mapped to a ramdisk |
| 18:02.39 | ds2 | prpplague: exactly... the gumstix has a USB dongle power setup ready to go so... |
| 18:03.27 | ds2 | personally, I'd do it on a bigger flash device...he wanted like 1M and claimed it would have a life time of about 1 year... well, get a 512M device, wear level it and you have a 512 year device |
| 18:03.36 | ds2 | and if your product lasts 512 years.... |
| 18:04.21 | prpplague | yea |
| 18:06.17 | Exiles | some people are odd though |
| 18:07.18 | *** join/#edev ico2_ (n=ico2@modem-1066.cheetah.dialup.pol.co.uk) |
| 18:09.40 | ds2 | it just occurred to me... the PXA255 is not very common, it seems |
| 18:10.12 | Exiles | I though the xscale chips were more commonly used |
| 18:10.22 | ds2 | the PXA270 seems to be more common |
| 18:10.34 | Exiles | well, these days |
| 18:10.43 | Exiles | as it's the faster counterpart |
| 18:10.52 | ds2 | I was trying to think of what device I know of other then the old gumstix, uses the PXA255 |
| 18:11.05 | Exiles | dell axmin x5 |
| 18:11.09 | Exiles | axmin* |
| 18:11.14 | ds2 | well, that and, I think, more importantly, it fixes certain errata |
| 18:12.50 | ds2 | interesting... I wonder how they worked around the erratas then |
| 18:12.51 | prpplague | when most of the xscale stuff was sold to marvell, alot of projects dumped xscale |
| 18:12.57 | prpplague | marvell isn't known for the cooperation |
| 18:13.08 | ds2 | this was intel era |
| 18:13.16 | prpplague | i.e. all the pxa datasheets are no longer available online |
| 18:13.34 | ds2 | the gumstix folks claim there are erratas for sleep and frequency change (PM related errata) |
| 18:13.47 | ds2 | yes, I am too aware of the Marvell plague |
| 18:16.40 | g1powermac | hey prpplague |
| 18:16.45 | g1powermac | I got a question for ya |
| 18:16.55 | g1powermac | well, more about the stuff your company sells |
| 18:17.38 | g1powermac | we use Quickbooks Point of Sale for the store stuff, and they sell a handheld inventory thing: http://quickbooks.intuit.com/commerce/catalog/product.jhtml?prodId=prod0000000000008022447 |
| 18:17.49 | g1powermac | you can see it on the bottom right |
| 18:18.08 | g1powermac | do you guys offer anything compatible with quickbooks? |
| 18:22.18 | prpplague | g1powermac: we mainly just sell the hardware, we don't normally do the software integration, we have ISV's that do that |
| 18:23.01 | Exiles | to replace quickbooks |
| 18:23.15 | prpplague | g1powermac: they don't seem to give much info on what the handheld actually does |
| 18:24.01 | prpplague | g1powermac: our M5900 can work as an HID keyboard, so you could do your inventory ,then go to your pc, plug it in, and start a transfer where it types it in as if it were coming from a keyboard |
| 18:24.28 | brad_mssw | heh, quickbooks provides APIs |
| 18:24.36 | brad_mssw | most of their stuff is SOAP based |
| 18:24.44 | brad_mssw | so you just post back home |
| 18:24.56 | brad_mssw | via wifi or similar |
| 18:25.22 | g1powermac | hmm, cool |
| 18:26.32 | prpplague | then it should work pretty easy from linux in general |
| 18:26.50 | brad_mssw | well, except for the fact that they don't document it is soap based |
| 18:27.05 | brad_mssw | they document the xml body at least |
| 18:27.26 | brad_mssw | their forums have good discussion on accessing it from PHP or other scripting languages |
| 18:27.50 | brad_mssw | ... at least the quickbooks accounting ... Point-of-Sale _should_ be similar |
| 18:28.04 | brad_mssw | you just have to have their RDS Server installed which is the soap server |
| 18:28.13 | g1powermac | hmm |
| 18:28.25 | brad_mssw | otherwise you're limited to windows messaging |
| 18:29.20 | g1powermac | personally, I'd love to get away from quickbooks and windows in general |
| 18:29.43 | brad_mssw | yeah, I'd agree |
| 18:30.31 | g1powermac | interesting. . . . |
| 18:31.32 | g1powermac | mac software would work too, since half the machine here are now macs |
| 18:32.35 | brad_mssw | which pos |
| 18:32.39 | g1powermac | http://www.xsilva.com/launch.php |
| 18:32.45 | g1powermac | its called LightSpeed |
| 18:33.10 | brad_mssw | ok, yeah, don't know them |
| 18:33.20 | brad_mssw | there's a few mac pos's I'd steer clear of though |
| 18:33.39 | g1powermac | which ones? |
| 18:33.51 | brad_mssw | especially the filemaker ones |
| 18:34.06 | brad_mssw | really rather not list them as my company does business with them ;) |
| 18:34.17 | g1powermac | heh |
| 18:34.20 | g1powermac | well |
| 18:34.23 | brad_mssw | basically, before purchasing, find out if they use filemaker |
| 18:34.28 | brad_mssw | if they do, look elsewhere |
| 18:35.24 | g1powermac | this LightSpeed one is very intriguing |
| 18:35.50 | g1powermac | not really all that much cheaper than QuickBooks PoS, but it has all the big features we need |
| 18:36.03 | g1powermac | I'd have to look further on some of the smaller features |
| 18:37.37 | brad_mssw | they use OpenBase for their backend ... odd |
| 18:37.50 | brad_mssw | originally developed for OpenStep |
| 18:38.04 | g1powermac | holy moly |
| 18:38.15 | g1powermac | this PoS already has multi-store stuff setup |
| 18:38.23 | g1powermac | for the price, thats amazing |
| 18:38.47 | g1powermac | that makes it a heck of a lot cheaper than Quickbooks PoS Multi-store |
| 18:45.20 | MothaWork | prpplague: have you actually ever heard anybody have legit concerns about the erase cycle limitations of modern flash devices (i.e. done the math and determined that their product would exceed the maximum number of erases even with wear leveling)? |
| 18:47.54 | ds2 | that would make a lot more sense of PoS stood for Piece Of Sh.... ;) |
| 18:48.35 | MothaWork | also, there's no reason on real static ram to not mix up the address and data lines if it makes the layout easier, right? |
| 18:48.56 | MothaWork | since fully static ram has no page properties or commands |
| 18:49.24 | ds2 | unless you want to at one point swap in a EEPROM |
| 18:49.30 | ds2 | think they share pinouts |
| 18:50.29 | MothaWork | not on this board, never |
| 18:50.36 | MothaWork | the RAM is a framebuffer for the LCD |
| 18:50.41 | MothaWork | I'd use PSRAM, but I can't bring out the BGA |
| 18:50.50 | MothaWork | and this bus has no DRAM controller |
| 18:50.54 | MothaWork | at least not a usable one |
| 18:51.27 | ds2 | what about debugging, if you had to snoop the bus? |
| 18:52.02 | MothaWork | I can't foresee needing to since it is just a framebuffer, and if I were to hook up to testpoints, I can just take into account the reordering while hooking up |
| 18:52.28 | ds2 | it just one more document to look at |
| 18:52.31 | MothaWork | and of course from the processor's point of view (i.e. JTAG ICE) it's just SRAM no matter how I hook it up |
| 18:52.49 | ds2 | I thought of that back in school and couldn't come up with a good reason to not do it |
| 18:52.56 | MothaWork | nor can I |
| 18:53.28 | MothaWork | obviously on paged devices with page mode accesses or on devices with command sets, you want to avoid that |
| 18:53.35 | MothaWork | but on fully static RAM, it shouldn't matter a whole lot |
| 18:54.03 | MothaWork | I guess I can see if I can make the layout work |
| 18:54.09 | MothaWork | I have 8 layers, so it should be possible |
| 18:54.44 | ds2 | heh... I backed off to avoid having to argue with the TAs and profs on why that is not a bad idea |
| 18:55.01 | prpplague | MonMotha: only for very excessive write cycles using jffs2 |
| 18:55.32 | MothaWork | ds2: the only reason I can think of that it would be a bad idea is the "well, this doesn't make sense" factor of a 3rd party looking at it, or if you wanted to swap in a page mode/command set based device |
| 18:55.58 | MothaWork | prpplague: the erase cycle limitations of flash seem WAAAAY overblown. It's always the first thing people mention, and is usually the last problem they will encounter |
| 18:56.10 | MothaWork | I've had more problems with design in terms of data retention limits than erase cycles |
| 18:56.21 | ds2 | MothaWork, I suppose it could make it hard, if not impossible to piggy back a EEPROM on top of it if you ever run out of capacity or need to do a emergency upgrade |
| 18:56.22 | MothaWork | 10 years isn't THAT long |
| 18:56.37 | MothaWork | well, there are two busses on this thing |
| 18:56.49 | ds2 | people using CF flash for swap have manage to wear them out |
| 18:56.51 | MothaWork | one bus has some SRAM that serves as the LCD framebuffer, while the other bus has NAND flash and SDRAM |
| 18:57.12 | MothaWork | if I need more space, I'll just put a larger NAND device on there, or put in a large SD card |
| 18:57.35 | MothaWork | the SRAM is surface mount TSOP anyway, and I've never seen EEPROM in that package |
| 18:57.43 | prpplague | MonMotha: we did wear some flash early in the m5900 dev cycle |
| 18:57.51 | ds2 | ah... I'm still thinking of them as 28pin DIPs ;) |
| 18:58.05 | prpplague | MothaWork: we had an application that was erasing one block and writing back to it about every 2 seconds |
| 18:58.31 | MothaWork | prpplague: the problem there is lack of wear leveling |
| 19:00.09 | brad_mssw | ... stupid question, but reading this discussion ... where is wear leveling implemented ... is the hardware responsible for doing this, or is the software level responsible (talking mainly CF here) |
| 19:01.04 | ds2 | the CF card itself is suppose to do it |
| 19:01.04 | MothaWork | depends on the media |
| 19:01.10 | MothaWork | CF and SD usually wear level on their own |
| 19:01.23 | ds2 | SM is the nasty sucker |
| 19:01.23 | MothaWork | in SD it's part of the spec, and in CF all high end devices do it for ATA devices |
| 19:01.29 | brad_mssw | ok, cool |
| 19:01.35 | MothaWork | for NOR CF, the software is responsible |
| 19:01.47 | MothaWork | for direct access to NAND or NOR devices on a system bus, the software is also responsible |
| 19:02.00 | MothaWork | jffs2, a filesystem designed to run directly on top of flash, does implement wear leveling |
| 19:02.11 | MothaWork | as does yaffs |
| 19:02.16 | ds2 | I suppose with NAND, you don't have to wear level except for the few critical blocks |
| 19:02.23 | ds2 | use them til they wear out and tag them as bad |
| 19:02.24 | CosmicPenguin | grrr - germinate |
| 19:02.37 | MothaWork | ds2: true, since you're assuming things will randomly go bad anyway |
| 19:02.43 | brad_mssw | we just run a sqlite3 database on an ext2 fs, so I don't think there is any wear-leveling logic in either of those, and we're fairly write-heavy |
| 19:02.52 | MothaWork | and you don't ever have to worry about XIP from NAND |
| 19:03.02 | CosmicPenguin | Why do I get the feeling that all of debian/ubuntu is held together with ducttape and bailing wire? |
| 19:03.03 | MothaWork | or at least normally don't |
| 19:03.09 | MothaWork | since most hardware doesn't do NAND FTL |
| 19:03.19 | ds2 | aren't there NAND's with a cache block to let you do XIP? |
| 19:03.26 | MothaWork | probably |
| 19:03.47 | MothaWork | I know I've seen stacked NAND/SRAM devices, but I think they were separate |
| 19:03.51 | MothaWork | separate logically that is |
| 19:38.46 | *** join/#edev agaffney (n=agaffney@gentoo/developer/pdpc.active.agaffney) |
| 19:40.34 | *** join/#edev ico2__ (n=ico2@modem-3848.chimpanzee.dialup.pol.co.uk) |
| 19:50.10 | MothaWork | why oh why must autorouters suck so much? |
| 19:55.57 | prpplague | hehe |
| 19:56.28 | prpplague | okie dokie, just about got everything ready for the Hammer shipments |
| 20:18.46 | *** join/#edev SpaceMonkey (n=andrew@client-81-105-49-170.popl.adsl.virgin.net) |
| 20:40.26 | *** join/#edev Crofton|work__ (n=balister@ossie.wireless.vt.edu) |
| 21:04.59 | *** join/#edev ico2 (n=ico2@modem-845.cougar.dialup.pol.co.uk) |
| 21:38.00 | *** join/#edev lyakh (n=lyakh@dynamic-unidsl-85-197-26-24.westend.de) |
| 21:47.45 | *** join/#edev Crofton (n=balister@66-207-66-26.black.dmt.ntelos.net) |
| 21:51.54 | *** join/#edev pierrelux (n=pierre-l@144-125.sh.cgocable.ca) |
| 22:02.11 | *** join/#edev lyakh_ (n=lyakh@dynamic-unidsl-85-197-26-24.westend.de) |
| 22:04.35 | *** join/#edev g1powermac (n=g1powerm@adsl-147-182-37.sdf.bellsouth.net) |
| 22:22.07 | *** join/#edev chouimat (n=dieu@kde/developer/chouinard) |
| 22:28.31 | chouimat | back |
| 22:28.58 | CP|too | front |
| 22:29.07 | MothaWork | left |
| 22:29.32 | MothaWork | (now somebody is supposed to go "right...") |
| 22:34.56 | chouimat | ok a 74"x38" table with a 1.5" MDF top is heavy |
| 22:38.22 | MothaWork | yes |
| 22:41.43 | ds2 | okay... is there a way in Kconfig to make an option either module or nothing... i.e. never linked in |
| 22:42.14 | *** join/#edev joe_bleau (n=joe_blea@adsl-69-152-222-211.dsl.fyvlar.swbell.net) |
| 22:52.45 | *** join/#edev GPSFan (n=kenm@12.10.255.248) |
| 22:52.45 | *** mode/#edev [+o GPSFan] by ChanServ |
| 23:24.01 | *** join/#edev ico2 (n=ico2@modem-845.cougar.dialup.pol.co.uk) |
| 23:28.29 | *** join/#edev g1powermac_PB (n=g1powerm@CPE-76-177-156-202.natcky.res.rr.com) |