<awygle>
Yeah probably. I have a couple features I want an mpi library to provide so let me check openmpi for them after work today
wpwrak has quit [Remote host closed the connection]
ZipCPU|Laptop has joined ##openfpga
wpwrak has joined ##openfpga
wpwrak has quit [Remote host closed the connection]
m_w has quit [Quit: Leaving]
wpwrak has joined ##openfpga
wpwrak has quit [Quit: Leaving]
wpwrak has joined ##openfpga
<cr1901_modern>
azonenberg: How stable are nightlies for Kicad?
<cr1901_modern>
in your experience*
<cr1901_modern>
I think awygle was telling me the something changed about making PCB parts recently
<cr1901_modern>
s/the/that/
Bike has joined ##openfpga
<rqou>
i was warned not to use nighties
<cr1901_modern>
Well I use pulseview nightlies and they've worked fine for me
<cr1901_modern>
otoh, pulseview is one of those rare pieces of software that has gotten light years better each time I update it
<awygle>
cr1901_modern: it seems to be not really a big deal. There are are
<awygle>
A few changes to the UI in terms of library management but the actual "put pins on a canvas" part is the same iiuc
<awygle>
Also you get my bug fix if you run a nightly :-D
<rqou>
hey, are fancy document feed scanners supposed to take many minutes to process and send you your scan?
<rqou>
is this just the usual "printers suck"?
<awygle>
That feel when you solve a seemingly intractable problem but the solution can't be justified at all :-/
<rqou>
that's called "software engineering" :P
ZipCPU|Laptop has quit [Ping timeout: 260 seconds]
wpwrak has quit [Remote host closed the connection]
<awygle>
It was an electrical problem and solution tho
wpwrak has joined ##openfpga
<rqou>
then it's called "electrical engineering" :P
<rqou>
offtopic: if i slice up all the pages of a pdf, rotate them, and then reassemble it, is this operation lossless?
<rqou>
(yes, the stupid scanner scanned all my pages wrong, how did you guess? :P )
Dolu has quit [Ping timeout: 248 seconds]
wpwrak has quit [Remote host closed the connection]
<awygle>
Hm. I wonder if what we talked about yesterday or the day before with the FBG484 Kintex supporting 10G also applied to the FBG484 Zynq...
m_w has joined ##openfpga
<azonenberg>
Good question
<awygle>
Do the slow GTX transceivers still have 64b/66b support and whatnot? Is that what makes them different from GTPs?
<azonenberg>
Hmm i cant remember if GTXes have 64/66
<azonenberg>
let me check
<azonenberg>
if GTPs do*
<awygle>
Pretty sure they don't
<azonenberg>
7 series GTP supports 8/10, 64/66, 64/67
<azonenberg>
however the 64/66 and 64/67 is just a gearbox
<azonenberg>
and requires the scrambling to be done in FPGA fabric
<azonenberg>
the 8/10 is a full coding block
<awygle>
Right
<azonenberg>
RX side, it has a gearbox with bitslip
<azonenberg>
but requires descrambling and block sync in FPGA logic
<awygle>
But the GTXs, the fast ones, have a whole coding block for the 64b/66b
<azonenberg>
Yeah
<azonenberg>
Because normally most protocols that use 64/66 are going to be >6 Gbps anyway
<azonenberg>
So they didnt bother implementing it fully in the GTP
<azonenberg>
Makes sense to me
<rqou>
why is the GTx naming so confusing?
<azonenberg>
so the full name is actually a bit longer
<azonenberg>
it's GT (gigabit transceiver), a letter indicating the type, then a suffix indicating the IP core version
<azonenberg>
so GTPE2 for example is version E2 of gigabit transceiver type P
<azonenberg>
As a general rule, for any chip family
<azonenberg>
GTP is the low-end transceiver, GTX is the high end, anything else is going to be ultra high end
<rqou>
iirc there's also a GTH?
<azonenberg>
(GTH/GTY/GTZ)
<azonenberg>
Those are all >10 Gbps high end transceivers, not sure of the specific differences
<azonenberg>
i've only ever actually used GTPs
<rqou>
i seem to recall 7-series had two mid-range transceivers?
<azonenberg>
GTP is the artix transceiver
<rqou>
but the max rate only differed by 1gbps or so
<azonenberg>
kintex has GTH
<azonenberg>
sorry, GTX
<azonenberg>
Then virtex has a mix of GTX and higher end stuff like GTH/GTY/GTZ
wpwrak has joined ##openfpga
<awygle>
rqou: GTP and crappy!GTX only different by like 350Mbps
<awygle>
Afaict
<rqou>
yeah, that
<rqou>
that always seemed kinda dumb
<azonenberg>
rqou: it makes sense
<awygle>
Very irritating that there's two (three?) GTXd
<azonenberg>
my understanding is, crappy GTX is the same silicon as nice GTX
<awygle>
*GTXs
<azonenberg>
but with a cheaper package
<azonenberg>
so worse power and signal intergity, and thus lower Fmax
<azonenberg>
But crappy GTX gives you all of the GTX frills like hardware 64/66 encoding etc
<awygle>
azonenberg: re: my earlier Zynq question, the datasheet doesn't say anything about the FBG484 doing 10.3G
<awygle>
But it's ~the same silicon and actually the same package
<awygle>
So it seems to me like it ought to work
<awygle>
Is that something vivado would tell me if I asked it?
<rqou>
afaik no
<azonenberg>
i wouldnt rely on it working
<azonenberg>
if it wasnt documented as such
<rqou>
when i was trying to figure out which GTXs did only 6gbps and which did 10gbps, i found a random forum post where some guy said "our factory screwed up and fitted the wrong part but everything still seems to work, is this ok for testing?"
<azonenberg>
alternatively, if you only needed it to work with one particular experimental setup and one particular chip
<azonenberg>
measure and see :p
<rqou>
iirc even the IDCODE doesn't change
<azonenberg>
of course not
<azonenberg>
they're the same die
<azonenberg>
the difference is the flip chip fanout substrate
<awygle>
rqou: sure. But I would expect that if I turn on vivado and say "I have an XC7Z030-2FBG484E, please set the GTX to 10.3G" it should either say "K" or "nah son"
<azonenberg>
for that matter, FTG256 and FGG1156 have the same die and IDCODE
<azonenberg>
And are indistinguishable over jtag
<rqou>
what
<awygle>
azonenberg: lulz
<azonenberg>
AFAIK coolrunner is the only xilinx part that has package info burned into the jtag ID
<rqou>
so telling them apart requires a Mark I Eyeball? :P
<azonenberg>
all others have the *die* id
<azonenberg>
but there is no electronic way to distinguish package/speed grade
<azonenberg>
i.e. its not in efuse anywhere
<azonenberg>
Since about 6 months ago, it's machine readable
<azonenberg>
but not over jtag
<azonenberg>
there's a datamatrix 2D barcode laser marked on the top of the chip
<cr1901_modern>
Mark I Eyeball <-- TIL. Who the hell says that ._.?
<azonenberg>
cr1901_modern: military
<awygle>
It's a mili-damn
<rqou>
offtopic: evince is slooow
* azonenberg
uses qpdfview now
ZipCPU|Laptop has joined ##openfpga
ZipCPU|Laptop has quit [Ping timeout: 265 seconds]
wpwrak has quit [Quit: Leaving]
wpwrak has joined ##openfpga
digshadow has quit [Ping timeout: 248 seconds]
wpwrak has quit [Remote host closed the connection]
digshadow has joined ##openfpga
wpwrak has joined ##openfpga
Bike has quit [Quit: Lost terminal]
<rqou>
azonenberg: i just remembered: whatever happened to the xc2 chips i mailed you? :P
<azonenberg>
They're still on my bench waiting for me to dump them
<azonenberg>
i've been traveling all month
<rqou>
ah ok
<azonenberg>
havent had time
<rqou>
(potentially-sensitive question, feel free not to answer) how many frequent flier miles have you accumulated? :P
brizzo has quit [Ping timeout: 255 seconds]
<azonenberg>
rqou: total? this year?
<rqou>
this year, since it seems you've traveled an insane amount :P
<azonenberg>
53,456
<azonenberg>
on Delta
<azonenberg>
plus smaller but nonzero amounts on a few other airlines
<azonenberg>
one leg on ryanair i know off the top of my head
<azonenberg>
31 flight segments
<azonenberg>
And i still have christmas travel coming up
<rqou>
lol
<rqou>
i hope IOA thinks they're getting a good ROI :P
<azonenberg>
They must, since they keep sending me places
<azonenberg>
rqou: and fwiw, i dont think i've ever redeemed any miles on delta yet
<azonenberg>
And i have 95,587 total
<azonenberg>
so over half of my lifetime total this year
<azonenberg>
lol
<rqou>
nice
<rqou>
unfortunately i usually fly on the evil airline that beats up passengers :P
<azonenberg>
lol
<azonenberg>
seattle is a big hub for delta and alaska so i do most of my travel on one or the other
<azonenberg>
i usually request delta as they fly to more destinations and it lets me keep all my points in one place
<azonenberg>
100K FF miles scattered among ten airlines arent all that useful
<rqou>
right, but the evil airline that beats up passengers is (was) more friendly towards corporate-types
<rqou>
so my father flew on them a lot
<azonenberg>
lol
<rqou>
although we've (mostly my sister :P ) been discovering that apparently no airline can sanely run any routes out of Ithaca NY
<rqou>
apparently this is due to a combination of this thing called "snow" combined with EWR telling small flights to get lost :P
<rqou>
azonenberg: was Troy any better? :P
<azonenberg>
Albany airport? i've had no problems with delays there
<azonenberg>
I've had problems en route to Albany due to my luggage getting lost in Minneapolis...
<rqou>
probably because it's bigger than ithaca :P
<rqou>
i wonder if i'm going to get flagged by german airport security again this year? :P
<rqou>
every time i've flown out of hamburg i get flagged for "there's way too much crap in this bag blocking x-rays and we can't see anything"
<rqou>
interestingly american security theater never flags this as a problem
<azonenberg>
Probably because most american airport x-ray systems are one of a small handful of models
<azonenberg>
that are dual energy
<azonenberg>
If you've ever looked at the displays on the x-rays you'll notice they have orange-yellow and blue-green images
<rqou>
oh interesting
<azonenberg>
It's essentially a false color image made by stacking images at two energy levels
<rqou>
hmm, when i saw the screen in hamburg, the image was 100% blue :P
<azonenberg>
This allows good penetration of dense objects while not washing out less dense items
<azonenberg>
It's possible their NRC equivalent has a lower cap on x-ray security screening energy levels
<rqou>
other normal people had mostly orange with small blue objects
<azonenberg>
Yeah they probably used a similar scheme then
<azonenberg>
Weird that nobody else complained
<azonenberg>
What are you flying with, lead bricks?
<rqou>
a bag full of random cables and devboards
<rqou>
cables are pretty opaque to x-rays :P
<rqou>
especially when it's _after_ CCC so none of them are neatly organized anymore :P :P
<azonenberg>
lol
<rqou>
there was some youtuber who actually tried to fly with lead (or some other very dense material) in checked luggage
<rqou>
because it was surrounding some uranium ore
<rqou>
apparently it got flagged and inspected and the inspector just shrugged and said "hmm, rocks?"
<rqou>
or something like that
<rqou>
because it did arrive at the other end :P
<rqou>
offtopic: i'm watching the latest NileRed video on Dulcin (an artificial sweetener)
<rqou>
a very interesting comment: because it's not very soluble in water, it tastes weird because the texture is all wrong
pie_ has quit [Ping timeout: 256 seconds]
<azonenberg>
lol i just got through the one where he made urea
<azonenberg>
i didnt get to see him use it for anything
<azonenberg>
s/made/extracted/
<rqou>
yeah, he uses it for synthesizing dulcin afterwards
<rqou>
after watching too much nurdrage/nilered i really want to learn o-chem "for real"
<rqou>
:P
<rqou>
offtopic: anybody got a set of good books/websites/lectures/etc. for learning o-chem? (whitequark?)
<rqou>
my formal chem background goes up do doing ap chemistry in high school
<rqou>
which was... *calculates* 5+ years ago wtf
<rqou>
so a) i'm old wtf b) i probably don't remember all the chem i learned
<whitequark>
practical orgchem: the entire room is smelling like bromine, the fucking UV light does jack shit instead of creating free radicals, and roommate tried to nitrate something that blew so badly that the table has a ~1cm deep impression of a hot plate's rubber legs in it
Dolu has quit [Ping timeout: 240 seconds]
FabM has joined ##openfpga
Dolu has joined ##openfpga
wpwrak has quit [Remote host closed the connection]
wpwrak has joined ##openfpga
wpwrak has quit [Remote host closed the connection]
<cr1901_modern>
I wish there was a program to trade in my physical textbooks (which are 5-10 years out of date) for ebook versions of current at a discount.
wpwrak has quit [Remote host closed the connection]
wpwrak has joined ##openfpga
wpwrak has quit [Remote host closed the connection]
wpwrak has joined ##openfpga
Dolu has quit [Ping timeout: 248 seconds]
<qu1j0t3>
yeah. it's such a clusterf
<awygle>
Speaking of clusters, openmpi seems fine
<qu1j0t3>
fpeaking of clusterf, openmpi feemf sine