11:33 <larsc> the pshdl guy got at least one thing right, he has a board for 30 euros

11:48 <mumptai> his title page said it should be like arduino, what else is there to say?

11:49 <larsc> arduinos sold pretty well

11:52 <mumptai> but technology wise they are limited and the design was never up to date, ignores many best-practises, contains trivial errors and is now locked into is own (backward) compatibility trap

11:52 <mumptai> but educationally and commercially a huge success

11:53 <mumptai> and arguably "better" hat basicstampes and pic-axe products ... if its only for the avr and gcc

11:55 <mumptai> s/hat/than/

11:55 <mumptai> (sorry for my typing, still recovering from an arm injury)

12:07 <sb0> larsc, as long as the maker community doesn't move away from 30-euro hacks, there is no room for serious stuff

12:11 <sb0> and they have money. but they spend it on Apple products :-)

12:14 <mumptai> it is similar with hobbyist robotics

12:17 <mumptai> but we got a lot of good and rather cheap parts availiable now, but people are only adapting them slowly and are rather reluctant to spend more than 50€ or more than a weekend on anything new

12:20 <mumptai> those that wanted to do this 10 years ago had to spend a few 1k€ or crawl ebay for months and years, now it can be done with a few 100€ and COTS parts, but the numbers of people haven't really climbed that much

12:24 <larsc> the good thing about cheap stuff in my opinion is that it might introduce the subject to people who wouldn't have considered it otherwise

12:24 <larsc> and some of those persons than might go on an do serious stuff

12:29 <mumptai> yeah, arduino might have done more for future generations of engineers than a lot of industrial engineers and academics

16:03 <mumptai> guten rutsch

20:38 <barmstrong> hmm, I'm thinking of making a patch so that Memory's init accepts a dict or a list. if you use dicts, you can do a sparse initialization

20:38 <barmstrong> any objections?

20:39 <barmstrong> nevermind, i'll just unroll it into a list before calling :)

20:48 <barmstrong> how do i peek around the Memory's values at simulation time?

21:03 <barmstrong> actually, nevermind, it doesn't even seem to be sim-compatible since it has no get_fragment? i'm not sure how that part works

22:46 <barmstrong> i think i'm running into some timing weirdness. i swapped out a memory module i made for the migen one and am getting some really strange issues with signals seemingly not updating

22:46 <barmstrong> it's probably my own bug, but i'm curious if this is a known issue?

23:47 <ysionneau> happy new year :)