<GitHub190>
[milkymist-ng/master] asmicon/multiplexer: correct read-to-write delay to prevent conflicts on the tag bus - Sebastien Bourdeauducq
Gurty` has joined #milkymist
Gurty has quit [Ping timeout: 265 seconds]
rejon has joined #milkymist
<lekernel>
phew, last framebuffer bug is gone :) now clean things up a bit and commit...
<lekernel>
or rather... sdram controller bug tickled by concurrent fb+cpu accesses
kyak has quit [Ping timeout: 246 seconds]
voidcoder has quit [Remote host closed the connection]
voidcoder has joined #milkymist
kyak has joined #milkymist
rejon has quit [Ping timeout: 264 seconds]
<wpwrak>
nice ! let's hope it keeps the distinguished title of being the last ;-)
<lekernel>
I'm thinking about integrating some of http://c-to-verilog.com/ into the dataflow system... could be nice e.g. for the next TMU/graphics pipeline
<lekernel>
(it's open source now)
<wpwrak>
oh, that sounds very nice ! i've been dreaming of such a thing for a very long time
kyak_ has joined #milkymist
kyak has quit [Read error: Operation timed out]
azonenberg has quit [Ping timeout: 245 seconds]
<lekernel>
one potential problem with this thing is everything is statically scheduled, so it may not play nice when connected to nasty systems with unpredictable latency like shared SDRAM
azonenberg has joined #milkymist
<lekernel>
this is why VLIW processors didn't work too well and got beaten up by superscalar OoO designs
<larsc>
lekernel: I heard you are maybe going to move the ehsm location?
<wpwrak>
chase CCC ?
<lekernel>
some people proposed it, but I'm not really sure I want it
<wpwrak>
having the whole city all for yourself is tempting, eh ? :)
<lekernel>
well there are more mundane considerations
<lekernel>
first, we have an awesome room for free in Berlin, and that was/will be hard to obtain
<wpwrak>
someone who wants to visit both events should be able to move between them in something like 3 hours anyway. in the same city, it may be one hour. not such a HUGE deal.
<larsc>
wpwrak: hamburg<->berlin are 1.5h by train
<larsc>
and the C3 venue is 500m from the train station
<lekernel>
the current location is about 30min away from the usual CCC location, including the walk and an average wait for the S-Bahn
<mumptai>
wpwrak, despite the higher cost
<wpwrak>
larsc: okay, so maybe 2.5 hours in total ;)
<lekernel>
I'm not very happy they moved it to Hamburg actually... didn't want to cause conflicts, but rather bring a much-needed advanced hardware complement
<lekernel>
c3 is too much about led blinkers etc.
<wpwrak>
lekernel: i don't think they moved because of you ;-))
<wpwrak>
you would have liked the sheer amount of arduino talks at FISL this year ;-) and then some pig^Hc, too :)
<lekernel>
yeah, sure, but in the end there are two conflicting events with somehow related topics in different cities
<mumptai>
lekernel, you didn't like blinken-area?
<lekernel>
(and yes, we did try to get in touch with the c3 organizers before, but...)
<larsc>
but nobody responded?
<lekernel>
mumptai: let's say I rather disliked the absence of anything else that would involve FPGAs, ultra high vacuum, molten metal, multilayer PCBs, plasma, etc. (not necessarily all in the same device ;)
<mumptai>
i see, i have a similar concern about arduino in web right now
<mumptai>
and 95% of it are trivial projects
<mumptai>
(but it makes people happy, resumably, so its kind of okay)
<mumptai>
s/resumably/presumably/
<mumptai>
i saw the schematics for the upcoming mm in kicad, are you really moving a way from altium?
<lekernel>
yes
<lekernel>
hopefully the m3 will be an all-kicad design even
<mumptai>
wow! so kicad is really ready for big stuf now?
<mumptai>
+f
<lekernel>
we'll find out ;)
<mumptai>
okay ;)
<mumptai>
that would be a major step for opensource in the h/w realm
<lekernel>
wpwrak: btw regarding your email I serendipitously met someone from https://www.changers.com/en this Wednesday... interesting for the mechanical design (and 2km from where I live)
<wpwrak>
lekernel: ah, they do mechanical work ? very good !
<lekernel>
not sure yet what amount they do internally, but they got that device done in the end