m_t has quit [Remote host closed the connection]
m_t has joined ##openfpga
<rqou> >have a CS degree
<rqou> >can't make printer print
<pie_> :I yes
<zino> Printers are are electronics and mechanical objects combined to cause the maximum possibility of failure and suffering for the person wanting to print.
<lain> ^
<azonenberg> And software
<azonenberg> dont forget the software
<azonenberg> things like "buy moar ink here" links popping up
* azonenberg looks at HP
digshadow has quit [Quit: Leaving.]
<lain> it's because printers are a sin against computing - computers replace dead trees, printers upset the balance.
digshadow has joined ##openfpga
<azonenberg> Nature abhors a... printer?
<azonenberg> :p
<lain> :D
<zino> It is known.
<rqou> network/wifi printers are the worst
<jn__> the firmware running on printers is also a little weird, as far a i've seen it
<rqou> why can't somebody just make a "stupid mode" print server that just maps USB EPs<->sockets?
<jn__> rqou: the USB/wifi/network stack in my dead epson printer is so weird
<qu1j0t3> Because printing is required to be very hard
<rqou> alternately why can't i just netcat a pdf/ps/pcl/etc. to a por?
<rqou> *port
<qu1j0t3> lain: I had to stop reading that once the real world overtook it
<lain> haha
<qu1j0t3> trutho:(
<jn__> rqou: some printers support that
<rqou> i guess the problem is i have a super-fancy multi-function consumer-grade wifi-enabled bullshit printer
<rqou> i actually have no idea what the internal representation of jobs for this printer is
<pie_> maybe if they wrote printer firmware in a functional language
<rqou> it's not a postscript printer for sure
* pie_ dumps a load of silver bullets on printers
<pie_> they all bounce off
<jn__> business hardware is the best, once again
* azonenberg wants to go Office Space on a printer at some point
<azonenberg> jn__: yes, commercial/industrial gear tends to be less full of BS than consumer
<jn__> pie_: does postscript qualify as a functional language?
<azonenberg> i was in the home depot today buying rock salt to melt ice on my driveway
<pie_> also is it just me or have ink cartridges gotten smaller
<pie_> jn__, idk
<azonenberg> i passed a $3500 refrigerator
<rqou> my roommate actually has a brother laser printer that's actually less crap
<pie_> jn__, does it work?
<rqou> it's a prosumer/small business printer
<qu1j0t3> jn__: hmmmmm i wrote a lot of postscript, that's a question i've never asked myself
<azonenberg> That had a full ~17" or so HD LCD display in one door
<azonenberg> i have no clue what that thing does
<azonenberg> and i dont want to know :p
<pie_> azonenberg, it hums
<pie_> when it gets bored
<qu1j0t3> jn__: i'd really have to think about that.
<pie_> and then plays with itself
<azonenberg> inb4 "Send 0.1 BTC to this address or all your food spoils" ransomware
<qu1j0t3> jn__: basically no, since current state is mutable. but it still has a bunch of interesting features
<azonenberg> or your food randomly not being chilled when it reboots to install a windows update
<azonenberg> :p
* pie_ shudders
<qu1j0t3> azonenberg: so you've seen the megapopular tweet about that
<pie_> azonenberg, actually industrial ransomware sounds pretty terrifying
<qu1j0t3> pie_: 3..2..1..
<pie_> qu1j0t3,
* pie_ glances around
<pie_> shhh the idea didnt come from me
<qu1j0t3> pie_: it's kind of infuriating how bad things are.
<rqou> but we added passwords!
<rqou> even for root!
<rqou> :P
<pie_> ransomware sounds like good business
<jn__> (in case someone was searching for the comic titled "the internet of ransomware things": http://www.geekculture.com/joyoftech/joyarchives/2340.html)
<pie_> kinda makes me sad im a bit of an altruist :P
<qu1j0t3> pie_: I mean you do know the story of Stuxnet right
<pie_> sure but it wasnt ransomware
<qu1j0t3> not too different really. worse
<pie_> and it was targeted
<qu1j0t3> Stuxnet is also the future
<qu1j0t3> yay
<rqou> nah, chinese webcams are :P
<qu1j0t3> well, we have a multifaceted future
<pie_> though im not sure if people want to start taking countries hostage zet
<pie_> *yet
<qu1j0t3> chinese webcams are a big part of it, sure
<qu1j0t3> a bit horrifying part
<qu1j0t3> big*
<pie_> hack all the nudes
<rqou> hey, iot vibrators exist
<qu1j0t3> welp
<pie_> sure but what, ...lmao
<pie_> youre gonna take a dildo hostage?
<qu1j0t3> pie_: You're young enough to enjoy all this Your Whole life
<rqou> iirc one of them had a temperature sensor in it
<pie_> i mean why not i guess
<qu1j0t3> pie_: while i'll spend the rest of mine just trying to have a quiet safe existence kthxbai
<rqou> ostensibly to monitor the motor
<qu1j0t3> OSTENSIBLY
<pie_> uterably
<qu1j0t3> unutterably
<qu1j0t3> exploitably
<pie_> dildo thermostat
<pie_> if your insides get too cold it takes the temperature up a bit
<pie_> though i suppose at that point youre probably dead
<pie_> oh have severe hypothermia
<pie_> *or
<pie_> homeostasis too boring for you? use a dildo thermometer!
<pie_> lets call it a tildo
<pie_> or a thildo?
<pie_> anyway
* pie_ goes back to trying to learn idris
<qu1j0t3> Idris++
<qu1j0t3> pie_++
<pie_> pie_ is not an lvalue
<jn__> pie__ := pie_ + 1
<pie_> i still dont think yu can do that?
<qu1j0t3> (let ((pie (+ pie 1))) ...)
<pie_> unless that like defines a new thing butunder tha same nae
<pie_> *name
<qu1j0t3> lexical scoping, shadowing
<pie_> different namespace or whatever
<pie_> yeah that. i think
<jrernst> Marex: are you still there?
<Marex> jrernst: kind of
m_t has quit [Quit: Leaving]
<pie_> jrernst, you guys are reversing bitstreams right?
jrernst has quit [Remote host closed the connection]
DocScrutinizer05 has quit [Excess Flood]
DocScrutinizer05 has joined ##openfpga
jrernst has joined ##openfpga
<Marex> jrernst: 'sup ?
digshadow has quit [Quit: Leaving.]
<jrernst> ah, hey!
<jrernst> I investigated the LE_BUFFER connections and found a weird thing
<jrernst> I scan QSF for the names and the positions and the RCF for the connections.
<jrernst> In the RCF I search for keyword LE_BUFFER:...
<jrernst> From the signal = <var> { I get the position of <var> from QSF.
<jrernst> So then I listed them and found that the I-numbers don't much vary
<jrernst> let me give 4 examples
<jrernst> X17_Y14_N20 <-- X17Y14S0I20 <-- VIC20:box1|T65:cpu|PC[12]~16
<jrernst> X30_Y18_N16 <-- X30Y18S0I17 <-- VIC20:box1|VIC20_VIC:vic|Equal21~2
<jrernst> X26_Y22_N17 <-- X26Y22S0I16 <-- VIC20:box1|VIC20_PS2_INTERFACE:keybd|reset_cnt[1]
<jrernst> X20_Y22_N25 <-- X20Y22S0I25 <-- VIC20:box1|VIC20_VIC:vic|last_matrix_cnt[4]
<jrernst> okay.
<jrernst> the first column are the positions found in QSF by variable name
<jrernst> second are the positions found by LE_BUFFER in RCF file
<jrernst> and third is the variable or signal name by which QSF and RCF can be referenced
<jrernst> As you can see the N-numbers and the I-numbers are quite similar
<jrernst> there are 4 cases: even numbers both, even/odd, odd/even, and both odd
<jrernst> actually there are 8 cases. 4 for the upper LEs and 4 for the lower LEs.
<jrernst> The numbers go from 0 to 31. The N-numbers and the I-numbers as well.
<jrernst> I correlated them and found that only the even/even and even/odd cases yields bits in the stream.
<jrernst> You found them too.
<jrernst> You called them
<jrernst> M/m - C input mux configuration (combinatorial mux #1)
<jrernst> wait I made a mistake. my bits are mirrored.
<jrernst> Again... you called them
<jrernst> Uncertain: Q/q - enable COMBOUT routing outside of the LE ?
<jrernst> That bits for the the even/even case
<jrernst> You didn't find the bits for the even/odd case. They sit between them.
<jrernst> it's strange.
<jrernst> according to the numbering even means the LCCOMB and odd means FF
<jrernst> so even/even should be LCCOMB and LE_BUFFER
<Marex> keep in mind there are multiple buffers per LE output
<Marex> it could very well be that you can enable each of those
<jrernst> don't think that's the explanation for this behaviour
<jrernst> because it's so selective only even/even and even/odd is found
<jrernst> or maxbe you are right but only for the odd/even odd/odd cases
<jrernst> maybe even/even and even/odd only have on wire path
<jrernst> but the reverse odd/even and odd/odd cases have multiple wire paths. Then it's possible that the correlation fails.
kuldeep has quit [Ping timeout: 272 seconds]
<jrernst> I used a design with 2699 LEs so statistically it gives a good distribution
<jrernst> okay let's assume M/m is right and has only a single routing
<jrernst> you wrote
<jrernst> M/m - C input mux configuration (combinatorial mux #1)
<jrernst> 0 ... input from external port
<jrernst> 1 ... input from REGOUT signal
<jrernst> maybe it's the LCCOMB input, it can come from extern or from FF REGOUT so there's not many paths
<jrernst> and from symmetry I think the missing cases are the N/n bits
<jrernst> you wrote
<jrernst> N/n - C input mux configuration (combinatorial mux #2)
<jrernst> 0 ... input from CIN signal
<jrernst> 1 ... input from combinatorial mux #1
<jrernst> I don't understand what this means
<jrernst> what's CIN?
<Marex> jrernst: see lab.txt
<jrernst> ok I found it
<jrernst> and comb mux #1? where is it?
<jrernst> is it DATAC?
<Marex> they are all numbered ;-)
<Marex> line 24 therein
<jrernst> ok, yes it's the 2-input mux for DATAC
<jrernst> hmmm... with this N/n makes no sense for the missing cases
<jrernst> okay one thing is matching
<jrernst> the cases even/ven and even/odd have signal names with N-numbers even.
<Marex> ummmm, I don't think I understand
<Marex> do you refer to the buffer numbering ?
<Marex> the buffer numbering is a mystery to me thus far
digshadow has joined ##openfpga
<jrernst> and even means the LCCOMB. and it has to be the source. RCF lists LE_BUFFER as a intermediary destination on a route
digshadow has quit [Client Quit]
<Marex> I have an idea, but not a clear one, yet
<jrernst> yes the numbering. it's quite simple. even N is LCCOMB and odd N is FF
<jrernst> it's what the chip planner is displaying when you click the left or the right box of the LE
<Marex> ah that
<Marex> there is imo more to it, there are physical buffers to which the global interconnect connects
<jrernst> how did they show up in the RCF?
<jrernst> Aren't they the C4, C16, R4 and R16?
<jrernst> in the RCF I saw also LOCAL_INTERCONNECT and LOCAL_WIRE but it's local as it says
<jrernst> but before I was close to the LCOMB and the FF
<jrernst> the nearest stop is the LE_BUFFER on the route
<Marex> I dont know where C16 is configured, although I have a hypothesis
<Marex> also I know where R4 is, but that's not something I looked into in detail yet
<Marex> I suspect what they call LE_BUFFER is really the blob of R4/C4 confiuration bits
<jrernst> I don't think so because there are lines in the RCF. a) signal with LE_BUFFER and with C4 and b) only with LE_BUFFER and no C4
<Marex> jrernst: which is why I suspect there are multiple buffers and they can drive various things ...
<Marex> look into cyclone II handbook, there is a nice schematic there of the LE
<Marex> it shows one buffer which drives local interconnect and two buffers which drive R4/C4
<jrernst> is this right. cyclone II (II like 2) ?
<Marex> yes
<Marex> they removed that detailed schematic from III handbook and on
<jrernst> volume 1 ? (470 pages)
<Marex> I think so
<Marex> also read this : https://www.google.com/patents/US6262595?dq=interconnect+inassignee:%22Altera+Corporation%22&hl=en&sa=X&ved=0ahUKEwjA8Nm_kafOAhVFaRQKHQmLAZUQ6AEIWzAJ
<Marex> but be very careful as some information in there do NOT match C/IV architecture anymore
<Marex> but it does give you an idea where the altera interconnect came from
<Marex> in the PDF, see FIG 4 , FIG 5, FIG 6
<jrernst> there is no fig 4. there is figure 4-1
<jrernst> this does not match. wrong book?
<Marex> jrernst: I was referring to the patent I posted above ...
<jrernst> okay
<Marex> in the handbook, it's page 29
<jrernst> fig 5 reminds me to C4
<jrernst> Figure 2–2. Cyclone II LE ?
<Marex> yeah, see the buffers on the right
<jrernst> there are two new muxes in the CycIII LEs. And the CycIII LE diagram is identical to the CycIV LE diagram.
<jrernst> yes, always 3 buffers
<jrernst> I got an idea
<jrernst> M/n is the left MUX to DATAC of LCCOMB
<jrernst> case even/even means destination LCCOMB (DATAC) and source LCCOMB region (data 3)
<jrernst> case even/odd means destination LCCOMB (DATAC) and source FF region (REGOUT)
<jrernst> there is a mux between LCCOMB and the sync inputs of the FF
* Marex is lost :-)
<jrernst> case odd/even means dest FF (sync input) and src LCCOMB region (LUT output)
<jrernst> case odd/odd means dest FF (sync input) and src FF region (mux to register chain)
<jrernst> hypthesis check:
<jrernst> before I posted 4 cases with positions and signal names.
<jrernst> X17_Y14_N20 <-- X17Y14S0I20 <-- VIC20:box1|T65:cpu|PC[12]~16
<jrernst> X30_Y18_N16 <-- X30Y18S0I17 <-- VIC20:box1|VIC20_VIC:vic|Equal21~2
<jrernst> X26_Y22_N17 <-- X26Y22S0I16 <-- VIC20:box1|VIC20_PS2_INTERFACE:keybd|reset_cnt[1]
<jrernst> X20_Y22_N25 <-- X20Y22S0I25 <-- VIC20:box1|VIC20_VIC:vic|last_matrix_cnt[4]
<jrernst> case 3 and 4 are missing but as you can see the names are ..._cnt[?]
<jrernst> these are counters and the use the carry chain for sure.
<jrernst> maybe the switching of the two muxes are not simply two bits. this could explain why the correlation fails.
jrernst_ has joined ##openfpga
jrernst has quit [Ping timeout: 240 seconds]
jrernst_ has quit [Client Quit]
jrernst_ has joined ##openfpga
pie__ has joined ##openfpga
digshadow has joined ##openfpga
pie_ has quit [Ping timeout: 248 seconds]
pie_ has joined ##openfpga
kuldeep has joined ##openfpga
jrernst_ has quit [Quit: Leaving]
pie__ has quit [Changing host]
pie__ has joined ##openfpga
amclain has quit [Quit: Leaving]
DocScrutinizer05 has quit [Disconnected by services]
DocScrutinizer05 has joined ##openfpga
<rqou> why are smartcards so complicated?
<pie__> why not?
<lain> because complexity is the enemy of security
kuldeep has quit [Read error: Connection reset by peer]
kuldeep has joined ##openfpga
<pie__> cant get loot from secure ones though :C
<pie__> by the way, how long till internet of smartcards?
Bike has quit [Quit: leaving]
kuldeep has quit [Ping timeout: 255 seconds]
kuldeep has joined ##openfpga
m_t has joined ##openfpga
<eduardo__> azonenberg: I did a youtube video about inkjet printers (mainly teared one apart). Got 300k views. Title: "inkjet printers are criminal". People seem to be upset :-)
kuldeep has quit [Read error: Connection reset by peer]
kuldeep has joined ##openfpga
kuldeep_ has joined ##openfpga
kuldeep has quit [Read error: Connection reset by peer]
kuldeep_ has quit [Read error: Connection reset by peer]
Ardeshir has joined ##openfpga
<qu1j0t3> eduardo__: you must be doing something right!
seu_ is now known as seu
seu has quit [Quit: http://quassel-irc.org - Chat comfortably. Anywhere.]
seu has joined ##openfpga
eduardo_ has joined ##openfpga
eduardo__ has quit [Ping timeout: 240 seconds]
Bike has joined ##openfpga
Ardeshir has quit [Quit: Leaving]
pie__ has quit [Ping timeout: 240 seconds]
Amun_Ra has quit [Ping timeout: 258 seconds]
m_t has quit [Quit: Leaving]
m_t has joined ##openfpga
ZipCpu has joined ##openfpga
Tenacious-Techhu has joined ##openfpga
<Tenacious-Techhu> Any Mac IceStorm users in the house?
ZipCPU|Laptop_ has joined ##openfpga
ZipCPU|Laptop_ has quit [Remote host closed the connection]
<rqou> somebody at 33c3 was using macos
<rqou> afaik it should work just fine
<rqou> other than potential issues with drivers
pie_ has quit [Ping timeout: 248 seconds]
<Tenacious-Techhu> 33c3?
<Tenacious-Techhu> rqou, I was wondering if the instructions for using “brew” found on the Apio page could be ignored if you just use the installer straight from FTDI.
<lain> ^ 33c3
<rqou> i know nothing about brew or mac ftdi driverws
<rqou> other than i hated using them :P
<Tenacious-Techhu> I understand and respect what “brew” is trying to do, but I’ve been down that road with macports and fink before, and not enjoyed the headaches required to unistall everything properly.
<Tenacious-Techhu> I’d rather not install third party software just for the sake of installing third party software. :P
<nats`> maybe use a VM with a linux
<Tenacious-Techhu> nats`, not really what I was going for; I suspect they’re asking us to use brew to install something that could be easily downloaded from the manufacturer’s website as an installer...
<Tenacious-Techhu> In short, the instructions are too “Linux”y to be any good to Mac users in general.
<rqou> my experience was that every time i chose one of macports/brew/fink/whatever i would find a package that wasn't available
<rqou> and then it would ask me to switch to one of the other ones
<rqou> and then i gave up trying to do dev on mac :P
DocScrutinizer05 has quit [Remote host closed the connection]
<rqou> idk what instructions you are using but iirc you only need the xcode command line tools if you don't care about programming a real board
<Tenacious-Techhu> Development on Mac is fine, so long as you stick to actual development Apps, and not piddling little projects that don’t know what an Application Bundle is.
<nats`> that don’t know what an Application Bundle is. <= ...
<nats`> you can package it if you want
<nats`> I guess people wil be glad to have help on that
<nats`> packaging for proprietary OS is always a pain in the ass
<rqou> yeah, nobody has made a bundle
<rqou> feel free to work on that :P
<Tenacious-Techhu> What about the Apio thing?
<rqou> i have no idea what that is unfortunately
<rqou> link?
<rqou> hmm iirc somebody from that project showed up at 33c3 but other than that i've never used it
<rqou> i would say "ask Edmund" but he unfortunately doesn't really use irc
<jn__> eduardo_: ^
<rqou> anyways, ftdi driver fuckery is a big reason i stopped using macos
<rqou> is libftdi the proprietary userspace or the foss userspace?
<rqou> ah it's the libusb one
<rqou> Tenacious-Techhu: you can probably build libftdi without brew (using only the xcode tools)
<rqou> offtopic: do i still need to pay apple to get a kext signing cert?
<Tenacious-Techhu> rqou, isn’t the whole point of that driver to use FTDI’s chips? Isn’t their driver better by definition in this regard?
<rqou> every ftdi driver has a different set of bugs :P
<rqou> there's also like 5 of them
<rqou> let's see...
<Tenacious-Techhu> rqou, it depends; do you want your kext to work with Darwin, or to work with Mac OS?
<rqou> for a real mac os kext
<Tenacious-Techhu> Probably yes, but that’s a computer security matter sort of thing; you can still give users instructions on how to install it without the cert.
<Tenacious-Techhu> Signed kexts are for legitimate companies releasing legitimate products, I think.
<Tenacious-Techhu> So the cost has more to do with Apple spending the time to recognize you as a legitimate source of kexts.
<Tenacious-Techhu> I have had no problems with the FTDI-supplied driver. I see no reason why that shouldn’t work, but if there’s a good reason it shouldn’t, I wouldn’t mind knowing.
<rqou> iirc on mac os there are 3 kernel drivers: the apple one, the ftdi one, and the not-really-a-driver raw usb mode (via iokit or whatever)
<rqou> and iirc the apple one was the buggiest
<rqou> the ftdi one works great in serial port mode
<Tenacious-Techhu> If the Arduino project can use the FTDI-supplied driver, I see no reason why this thing can’t.
DocScrutinizer05 has joined ##openfpga
<rqou> but on the userspace side there's a split between the proprietary userspace and various open-source ones that use raw usb
<Tenacious-Techhu> Any FTDI driver made by Apple has to be developed with proper documentation and support from FTDI, which they may have felt too lazy to do.
<rqou> afaik arduino works in serial port mode and doesn't try to do anything fancy
<Tenacious-Techhu> FTDI, I mean.
<rqou> so afaik arduino works with either the apple or ftdi proprietary driver
<Tenacious-Techhu> I see no reason why the FTDI supplied driver shouldn’t be complete; it’s their chip, after all.
<nats`> if you think manufacturer of think do the best driver for their own chip you're wrong :D
<nats`> their fucking driver are a mess
<nats`> and I have less trouble with libusb one
<rqou> iirc the ftdi driver and the foss driver were just buggy differently :P
<rqou> anyways, the various fpga programming thingies use various advanced bitbang/mpsse modes of the ftdi chip rather than serial port mode
<rqou> and afaik this can be done with either the ftdi proprietary userspace talking to the ftdi proprietary kernel driver
<rqou> or a foss libusb userspace talking to raw iokit or whatever
<rqou> i don't think apple's ftdi driver works with any of them
<rqou> although on windows there might have been a foss userspace talking to the proprietary driver
<Tenacious-Techhu> So it shouldn’t make a difference whether you’re using the Open-Source one or the FTDI-supplied one?
<rqou> this whole thing is a mess :P
<rqou> libftdi isn't provided by FTDI
<Tenacious-Techhu> Bah.
<rqou> afaik apio wants to use libftdi, which is the foss userspace talking to iokit
<rqou> so it asks you to unload both the apple and ftdi kexts
<rqou> ftdi's userspace is named something like libftd2xx
<Tenacious-Techhu> Would that mean that, while I’m working with Apio, I wouldn’t be able to use other stuff that needed the legitimate FTDI-supplied driver?
<rqou> that's what it looks like
<rqou> you can probably make a "codeless kext" or whatever to block the ftdi driver only for the fpga boards
<rqou> btw this is all much more seamless on linux
<rqou> the linux kernel usb stack has a "kick out kernel driver" command
<Tenacious-Techhu> Well, of course it would be; this was a project designed by Linux geeks primarily for Linux geeks.
<Tenacious-Techhu> Such Linux geeks need to learn proper cross-platform design.
<rqou> driver/kernel crap is inherently platform-dependant
<jn__> iceprog uses libftdi, which (on linux) uses libusb, which afaik also works on MacOSX
<Tenacious-Techhu> Yes, but proper cross-platform design relegates that to cross-platform libraries that handle that platform-specific stuff for the developer.
<rqou> except you have to unload the serial-port-mode kext on macos
<rqou> same on windows
<rqou> Tenacious-Techhu: that's what libftdi tries to be
<Tenacious-Techhu> No, it doesn’t; instead of being an interface to the existing libraries, it tries to stampede over the existing libraries.
<rqou> ftdi is partially to blame with their buggy chip with a bajillion different modes
manderbrot has joined ##openfpga
<rqou> idk if the macos driver even has "d2xx" (non-serial-port mode) support
<rqou> the linux driver definitely doesn't
<nats`> Tenacious-Techhu if you want to write a version using real ftdi driver you can
<rqou> on windows iirc there's a foss userspace to talk to the ftdi proprietary kernel driver
<rqou> my personal opinion is to stop using ftdi :P
<nats`> if I could I would
<nats`> +10
<rqou> one of my projects is to program a cheap cortex-m0 firmware with all the ftdi features but fewer bugs and stupid driver problems
<rqou> e.g. by using usb-acm for both bitbang and serial channels
<rqou> no driver fuckery required
<Tenacious-Techhu> Maybe so, rqou, but I’m not about to use brew just to do that. :P
<jn__> rqou: that's where all the ftdi clones come from! ;)
<rqou> no, i won't be using the ftdi vid/pid
<rqou> :P
<rqou> also some cortex m0s are cheaper than the ftdi asic
<Tenacious-Techhu> rqou, that’s a better argument for FPGA dev-board makers to stop using FTDI than what driver to use. :P
<rqou> but in conclusion for Tenacious-Techhu: given the current state of software/effort, you're screwed :P
<rqou> sorry
<Tenacious-Techhu> How so?
<rqou> as far as i can tell, there is no convenient way to get the ice40 programming tools to work on macos side-by-side with other ftdi devices
<rqou> and none of the dependencies are packaged "properly" for macos
<rqou> lol according to FTDI's own AN134 the proprietary userspace on macos has the exact same problem
<rqou> it also requires unloading apple's kext
<Tenacious-Techhu> Ew.
<rqou> so only on windows do you not need to fuck around with kernel drivers
<rqou> also apparently on macos installing the ftdi proprietary userspace involves just dumping a dylib into /usr/local/lib
<rqou> i wonder how well that works with the weird rootless thing?
<rqou> AN134 also suggests that you can reprogram the usb pid
<rqou> so the serial port drivers won't claim the device
<rqou> but then idk if apio or the icestorm tools will detect the new pid
<nats`> that's what they used at some point to brick fake ftdi
<nats`> reprogramming them to 0:0
<rqou> yeah i know that
<rqou> hilariously enough linux doesn't give a shit if your vid/pid is 0/0
<rqou> :P
<nats`> :p
<rqou> but yeah, even ftdi seems to treat macos as "just a more inconvenient and annoying linux" :P
<jn__> rqou: well
<jn__> rqou: they added supprt after the windows FTDI driver bricked clones by setting vid/pid to 0
<jn__> (they = linux)
<rqou> afaik you could always use vid=pid=0
<rqou> just that no kernel driver would load by default
<rqou> you could always override ftdi_sio's vid/pid to whatever you want
<jn__> touché
<rqou> oh that's just pid=0
<rqou> afaik linux even accepts vid=pid=0
<rqou> i've definitely seen vid=pid=0xffff
<rqou> as in, on an actual device that i've owned at one point
<jn__> did it autoload?
<rqou> iirc it was a usb bluetooth thingy from china
<rqou> and yes it did
<lain> ffff is for testing iirc
<rqou> because it has the correct usb class
<lain> classy
<jn__> yay, class drivers
eduardo_ has quit [Quit: Ex-Chat]
eduardo_ has joined ##openfpga
<cr1901_modern> thought 6666 was the vendor id for testing
* cr1901_modern isn't joking. This is what I thought the id was
<cr1901_modern> The USB Consortium must've knew that they were summoning Satan into this time period by creating USB
m_t has quit [Quit: Leaving]
<eduardo_> Tenacious-Techhu: Micah Scot did a Mac installation. https://www.youtube.com/watch?v=HD3hADW57WM
<eduardo_> During the video she also did a live compile and install . There is not yet a MAC package.
<rqou> @scanlime uses yosys/icestorm? TIL
<eduardo_> rqou: yes, there is also an icoBoard heading her direction.
<eduardo_> her current setup uses an iceStick as a frequency divider. She programmed it with Yosys. Can all be seen in the video.
<rqou> hmm i don't actually follow @scanlime's work that much
<Tenacious-Techhu> eduardo_, so far, I’m seeing a nice video on command-line based place & route and synthesis... my questions were more about installation.
<qu1j0t3> maybe Homebrew formula can be used for Mac packaging
<Tenacious-Techhu> To me, Homebrew is the problem, not the solution.
<Tenacious-Techhu> Homebrew doesn’t follow Mac installation conventions.
<Tenacious-Techhu> Things should be packaged in an Application bundle that can be dragged to the Applications folder for installation, and removed from the Applications folder for uninstallation; things should not follow Unix/Linux expectations on where this or that esoteric file should be installed.
<Tenacious-Techhu> If additional libraries need to be installed, then a full installer may be necessary.
<eduardo_> Tenacious-Techhu: I did see the video live two or three weeks ago and made a video out of it. I think she did not specifically explain the installation of Yosys under Mac.
<eduardo_> Yes, as your describe this is the way it should be done. But noone of us own a Mac or knows how to use a mac or how to do a package.
<eduardo_> as long as noone steps up with this assets and the will to do it, we are out of luck.
<Tenacious-Techhu> Bundles are trivially easy; they’re really just fancy folders.
<rqou> i have a mac
<rqou> but it's running debian
<rqou> :P
<Tenacious-Techhu> You can probably stuff things in a folder and then convert it into a bundle with a handy App; it’s really trivial.
<nats`> so do it ?
<eduardo_> ok. I mean macOS
<rqou> at this point i don't care that much about building a bundle
<rqou> a bundle doesn't fix the kext stuff
<Tenacious-Techhu> nats`, in order for me to do it, I would have to install all the nasty Homebrew stuff first, compile it all, and then be at a complete loss as to how to uninstall all the garbage I was required to install along the way to get to that point.
<Tenacious-Techhu> rqou, no, it doesn’t.
<Tenacious-Techhu> That is an additional issue that you guys raised successfully that I was not aware of.
<Tenacious-Techhu> Regardless, Homebrew is not a solution for anything.
<rqou> in general cli tools don't make too much sense as bundles anyways
<Tenacious-Techhu> If it’s exclusively command-line stuff that will never see a user interface, it’s fine, because the Unix and Linux devs know what to expect on that front...
<Tenacious-Techhu> But when it’s something destined for an IDE at some point, a completely different approach is required.
<rqou> you can probably make apio or the ide thing that uses apio into a bundle
<Tenacious-Techhu> Right.
<Tenacious-Techhu> That’s where I was going with that.
<Tenacious-Techhu> Or maybe an Eclipse thing, or a Netbeans thing, or whatever.
<rqou> ugh
<rqou> those are all awful ides
<Tenacious-Techhu> I don’t assume they are good ideas.
<rqou> the only reason i use eclipse ever is that building java is even more awful
<Tenacious-Techhu> Just some sort of cozy IDE App of some kind.
<rqou> so i use the ide solely to hit the "build" button :P
<Tenacious-Techhu> Well yes, but that’s why you use your Mac to run Debian.
<pointfree> cyrozap, eric_j, carl0s: I once attempted to get Cypress PSoC Creator working on Linux under WINE: https://www.reddit.com/r/PSoC/comments/2uhm5h/i_did_a_full_install_of_psoc_creator_under_linux/
<pointfree> Anyway, there's a lot of little command line utilities under ~/.wine/drive_c/Program Files/Cypress/PSoC Creator/3.0/PSoC Creator/bin/
<pointfree> They run just fine under WINE on Linux.
<pointfree> It started up with splash screen and IDE ui, but it didn't do much more than that.
<pointfree> Of particular interest are tools like sjrouter.exe and sjplacer.exe
<pointfree> Just run some of them with WINE. They will show you the usage.
<pointfree> I'd like to get these running out of scripts on Linux so we can write/generate Verilog on Linux and do automated RE'ing from Linux with the newfound constraints + Mill's Methods.
<pointfree> Maybe we can find a MS Windows batch file that ties all these tools together?