**** BEGIN LOGGING AT Tue Oct 11 02:59:58 2016 Oct 11 13:58:37 bencoh: well, imx6 looks like it should be good enough for most smartphone needs today; but by the time devices in development right now go into mass production, it might feel rather obsolete... Oct 11 13:59:04 especially since nerds tend to frown upon anything but the "latest and greatest" :-( Oct 11 21:33:38 * DocScrutinizer05 tries hard to not start elaborating about character evaluations, which usually are very much mutually biased and subjective Oct 11 21:34:32 * DocScrutinizer05 however wonders what happened to his ignore list Oct 11 21:37:01 anyway it's probably pretty arrogant to claim character deficiencies on somebody, based on that other person's refusal to accept teachering about "the only correct way to do as-open-as-possible" hardware for the community Oct 11 21:39:13 pabs3: ((Neo900)) we just finished total schematics "redo" thanks to switch eagle->kicad, and are negotiating with several layouters for proto_v2 since Nikolaus bailed out as layouter Oct 11 21:41:00 eagle->kicad switch perceived necessary since eagle seems not up to the task without Nikolaus Schaller's homegrown router that only works on Mac and has no manual whatsoever (and thus is not FOSS) Oct 11 21:41:20 we also thought kicad better serves the "open" aspect of the project Oct 11 21:43:49 pabs3: https://neo900.org/news/migrating-away-from-eagle-to-kicad Oct 11 21:51:40 pabs3: also did you see Werner's **awesome** incredible Kicad supplementary tool eeshow? It's ground breaking! Oct 11 21:52:54 we expect it to get embedded into kicad seamlessly eventually, incl an augemnted git filesystem support and kicad-generic git-merge Oct 11 21:55:28 so kicad not only could compare recent schematics to an older git version, but also store directly to git (add, checkin, push) and do genuine merge of old / other branch git version schematic details into the current schematics Oct 11 21:55:53 for now eeshow is a very powerful git browser and schematics diff Oct 11 21:56:25 and I think it's absolutely unique and one of a kind so far Oct 11 21:59:09 antrik: ((by the time devices in development right now go into mass production, it might feel rather obsolete...)) that's the problem with ALL except the developments of the top ten hugest companies who can afford e.g. a dozen TI field engineers at (Nokia's) location, helping them (Nokia) to design a circuit with a SoC that's not even existing yet neither in silicon nor in technical reference manuals Oct 11 22:01:08 substitute TI and Nokia by wahtever comes to mind: ST, Qualcom etc for SoC, and Samsung, RIM, LG, Huawei for device under development Oct 11 22:06:08 "Hi $chip_manuf, we want to build a new device, projected sales volume 30 million units, and we thought your SoCs are nice... could you help?" Answer:"Sure thing, we have plans for a great new SoC we want to annouce next year. We send you a team of engineers helping to kick off your development right away. Just please sign this contract for 10million chips first" ---vs--- "Hi $chip_manuf, we want to build a new device, Oct 11 22:06:10 projected sales volume 3000 units, and we thought your SoCs are nice... could you help?" Answer:> --------- shhhhhhhhh [tumbleweed rolling by] ----------------- shhhhhhhhh < Oct 11 22:11:31 pabs3: btw watch out with rah, eventually he might accuse you too to be elitist closed minded and completely fucked up since you do service of the OM servers instead of handing this task to the community where it belongs Oct 11 22:13:11 just to make absolutely vlear: that's what *he* said, not what I think Oct 11 22:13:52 and I think it predates the time where you took over Oct 11 22:27:22 DocScrutinizer05: I'm aware of that situation (only that it's chips *two* years in the future they would be looking at, not one year) Oct 11 22:28:08 DocScrutinizer05: doesn't change the problem though that starting a design with something that is getting dated *right now* will guarantee it's totally obsolete by the time it's out... Oct 11 22:28:25 ...want to **announce** next year ;-) Oct 11 22:28:41 oh, right. sorry Oct 11 22:28:48 that means it's 2 years at least till firsz proto chips Oct 11 22:29:38 actually Pyra started on OMAP5 before any such chip been available to mere mortals Oct 11 22:30:18 at times it even been unclear if it ever will be Oct 11 22:38:30 and for OMAP4 I witnessed several of the above listed companies running into a very tough place when they developed (phone) devices based on that OMAP4 and a ST-Ericsson Thorium phone modem, and all those ST-E customers opened tickets against the Thorium modem that it stalls or the HSI interface misbehaves, until it turned out it was a SiErr in OMAP4 HSI resp IRQ handling that would have rendered _all_ those devices defunct if Oct 11 22:38:31 we at ST-E couldn't have patched the HSI drivers in the modem to handle that fuckup TI built into the OMAP4 (by actually *wiggling* the HSI IRQ line as long as it needs until OMAP detects there'S activity) Oct 11 22:44:43 the fun of engineering in the real world ;-) Oct 11 22:44:56 go figure: ticket "your modem randomly stops answering to any command" -- weeks of evaluation, then "our modem tries to answer but your application doesn't listen" -- "our drivers' logs clearly indicate that your modem never announced any response communication" -- weeks of hardware tests, then "here's a scope print that shows how your application (OMAP4) sends a command, then at time X1 our modem tries to respond and on that Oct 11 22:44:58 purpose it asserts the HSI IRQ line which stays that way and your application never responds" Oct 11 22:47:35 good old blameshifting -- just as relevant in engineering as in administration ;-) Oct 11 23:27:50 nah, here it was really to find and kill a bug that rendered the whole system a paperweight Oct 11 23:57:16 http://lists.openmoko.org/pipermail/community/2012-April/066797.html Oct 11 23:59:56 http://lists.openmoko.org/pipermail/community/2013-August/068855.html **** ENDING LOGGING AT Wed Oct 12 02:59:58 2016