08:40 <wpwrak> roh: btw, what cad do you use for 3d these days ? i did my "best" 3d work with fped plus a few dirty hacks, using fped to draw vertical slices.

08:44 <wpwrak> this approach has a few drawbacks, though: a) it's of course rather exotic, b) while i can make toolpaths, i don't have a way (so far) to go from that to a mesh, c) this limit me to all things in the Z direction to be strictly vertical - not nice for molds (which need slightly inclined walls) or fillets (both for mechanical stress and to avoid corners resins may not be able to reach)

09:40 <larsc> pcercuei: Can I add your SoB to "MIPS: JZ4740: LCD: Fixed LCD_CMD consts and completed LCD_STATE consts"

09:40 <larsc> ?

09:41 <pcercuei> hi

09:41 <pcercuei> I did that?

09:43 <pcercuei> ah no, signed-off-by doesn't mean I'm the author

09:44 <larsc> according to the log you are the author

09:44 <larsc> but the SoB is missing

09:45 <pcercuei> looks like it's mine yes

09:45 <pcercuei> sure, you can add my SoB

09:46 <pcercuei> Date: Wed May 2 23:45:35 2012

09:46 <pcercuei> that was a long time ago :)

09:46 <pcercuei> you're preparing a push to mainline?

09:47 <larsc> yes, it's a rather trivial change

09:49 <pcercuei> I still have commits to fbcon I need to re-send mainline

09:50 <pcercuei> I sent 5 patches to the maintainer, but he only added one, lol

09:51 <larsc> the binding code changed in upstream

09:53 <pcercuei> well we still have to apply 4 patches

09:53 <pcercuei> it changed since 3.11?

09:54 <larsc> it was changed in 3.11

09:54 <larsc> I did my best to forward port the patch

09:57 <pcercuei> I mean those patches:

09:57 <pcercuei> fbcon: use kstrtouint instead of deprecated function simple_strtoul.

09:57 <pcercuei> fbcon: continue parsing parameters after an error

09:57 <pcercuei> fbcon: optimize parameters parsing loop

09:57 <pcercuei> fbcon: add parameter to toggle bind on startup

09:57 <pcercuei> the first three are here to fix or clean the code

09:58 <pcercuei> so at least those should be accepted, I guess

09:58 <pcercuei> the last one is a functional change

09:58 <larsc> and I'm not sure how well it works with 3.11

10:15 <pcercuei> larsc: it works fine, just tested it

10:16 <pcercuei> the one we have on our github repo

13:30 <mth> larsc: this is how I modified the bind on startup patch: https://github.com/gcwnow/linux/commit/ef73228a

13:30 <mth> (for 3.11)

19:13 <whitequark> http://electronics.stackexchange.com/questions/67598/how-are-integrated-circuits-fabricated/67604#67604

19:40 <ysionneau> :)

20:08 <larsc> all you ever wanted to know about C http://www.seebs.net/faqs/c-iaq.html

20:09 <larsc> Q: 'What's the auto keyword good for?' A: 'Declaring vehicles."

21:10 <whitequark> excellent

21:11 <viric> Maybe 'auto' can help in some fancy macros

21:12 <viric> I can't think anything much useful though

21:13 <whitequark> viric: do you know what auto does in C?

21:14 <whitequark> well. nothing

21:14 <whitequark> it used to say to the compiler that the variable should have *auto*matic storage, i.e. (I guess) placed on stack

21:15 <viric> I thought of a macro that used the 'auto' keywords, to avoid macro users to set it static

21:15 <whitequark> but it's a no-op in every single compiler I know, perhaps except original K&R C

21:15 <viric> keyword

21:15 <viric> ah ok

21:15 <viric> I didn't think about the implementations. :)

21:16 <whitequark> I'm horrified to even think about how much incompatible ways you could invent for handling a combination of auto and static

21:16 <viric> I didn't reach anything :)

22:16 <whitequark> roh: so I'm looking at STM32L1's GPIOs right now

22:16 <whitequark> it's sitting directly on AHB, so zero wait states

22:17 <whitequark> >Fast toggle capable of changing every two clock cycles

22:17 <whitequark> ^ states the manual

22:18 <whitequark> (I guess that's because you need one cycle for ICode read and another for AHB write, per toggle)

22:18 <whitequark> it's definitely more work than on avrs though.

22:20 <whitequark> stm32l1 has some insane i/o... complete with a huge interconnect matrix which allows to route peripherals almost everywhere you may want

22:21 <whitequark> and arbitrary combinations of gpio/afio, input/output, push-up/down/floating, open-drain/push-pull