<sorear>
possible but weak explanation: it's hard to convince people that you didn't "have the factors all along"
<sorear>
publishing the prime factors of floor(pi * 2^1024) would have some value, but "random integers" have a different factor distribution than RSA moduli and I think ECM is better than GNFS for such numbers
<sorear>
if someone drops a pastebin with the prime factors of every RSA key in the mozilla CA bundle, that's your cue scalable quantum computing has been proven possible
<sorear>
google's not going to do that though, not responsible disclosure-y enough
* sorear
doesn't understand the algebraic number theory parts of GNFS very well
genii has joined ##openfpga
<OhGodAGuardian>
sorear: RSA moduli are random integers - just random ones which are the product of two random ones passing primality tests
<OhGodAGuardian>
As far as I'm concerned...
<Bob_Dole>
wot
<OhGodAGuardian>
... they were already proven scalable the moment we proved we could do the shit. It was less than an average human lifetime between our first powered flight and landing on our moon. Not taking that chance. :P
<sorear>
(a) when I say "random" without further qualification I mean "uniform". a uniform random integer between 1 and L has ln(ln (L)) prime factors
<sorear>
(b) I am talking to azonenberg_work, who has made clear that he believes QC research to be a scam/irrecoverably misguided, so I'm codeswitching a bit
<azonenberg_work>
i didnt say QC research is a scam
<azonenberg_work>
I'm just skeptical of some current implementations of it
<sorear>
it's fairly simple if you don't care about the error bounds
<OhGodAGuardian>
sorear: thing is, it has to be a hell of a hard target, or no one cares
<OhGodAGuardian>
i.e. it has to pass a shitton of iterations of probabilistic primality tests - both numbers
<OhGodAGuardian>
and then the product being one portion of the pubkey, the other being the public exponent
<OhGodAGuardian>
(usually 65537)
<sorear>
which is dumb, but such is the world we live in
<OhGodAGuardian>
actually, it's better than the original
<OhGodAGuardian>
3
<sorear>
2 or 3 is a much better choice with correct padding
<OhGodAGuardian>
Padding is easy to fuck up :P
<OhGodAGuardian>
btw
<OhGodAGuardian>
if anyone is itching to break a 768-bit key...
<OhGodAGuardian>
I got one for ya. It's the one AMD's using for Vega10 VBIOS signatures, and it's a royal pain in my ass.
<sorear>
they released a product *when* using 768-bit RSA signatures?
<OhGodAGuardian>
sorear: Hey, before that, it was technically not even a sig
<OhGodAGuardian>
but an obfuscated checksum
<OhGodAGuardian>
they're moving up. :P
<sorear>
there's no excuse in 2k18 to use anything other than a hash-based algorithm for firmware signing
<sorear>
assuming mb-scale firmware
<OhGodAGuardian>
Well, this isn't the FW itself, but quite important, and useful. I got around it at first
<OhGodAGuardian>
by making the driver load an image off disk via kernel modification
<OhGodAGuardian>
now they have the FW check the image loaded >.>
<sorear>
image of … what?
<OhGodAGuardian>
The VBIOS. Has a lot of settings, information, etc
<pie__>
(listening to you guys makes me forget that "professionals" as a notion actually exists in reality, you are all freaks of nature)
<OhGodAGuardian>
Which, fine, I CAN also write directly to the GPU at runtime
<sorear>
*somebody* is confused about trust relationships here
<OhGodAGuardian>
but being able to commit some of the good changes...
<OhGodAGuardian>
that was nice.
<OhGodAGuardian>
They're signing it.
<pie__>
(to be clear that was meant as a compliment i guess)
<sorear>
the card should not be trusting the vbios in any way, so there is no reason for the card to do any kind of attestation
<OhGodAGuardian>
It sort of has to.
<sorear>
the vbios *could* compromise secure boot, so it may need to be signed for that reason, but it'd be signed by microsoft/intel in that case, not amd
<OhGodAGuardian>
There's no possible standardized way to do a lot without it
<OhGodAGuardian>
also, it's not needed to run at that time
<OhGodAGuardian>
it can be left alone - the device behaves as a dumb VGA device
<OhGodAGuardian>
until driver load - which can be in the OS
<OhGodAGuardian>
it's not attestation EXACTLY, more tamper-evidence
<OhGodAGuardian>
And the more I learn about EFI, the more I dislike it anyways
<OhGodAGuardian>
I did full-custom OS dev before it, so believe me, I KNOW that hell
<OhGodAGuardian>
and why a replacement is needed...
<OhGodAGuardian>
But EFI is a little "kitchen sink included" for me.
<OhGodAGuardian>
Besides, secure boot isn't a lot of good when Intel does retarded shit like even THINKING of putting Intel ME on any-goddamned-thing.
<OhGodAGuardian>
But they didn't even restrict it to actual servers and shit
Bob_Dole has quit [Ping timeout: 260 seconds]
<OhGodAGuardian>
no, they stuck the bitch on EVERYTHING
<whitequark>
laptops with antitheft
<whitequark>
are a huge thing in corp environments
<OhGodAGuardian>
This is a legit usage.
<OhGodAGuardian>
Kineta's board is a fuckin' EATX.
<OhGodAGuardian>
She's going nowhere.
<OhGodAGuardian>
Not saying it has NO legit usage.
<OhGodAGuardian>
I am saying they put this on a lot of shit shipped out to the consumers who have no idea they're carrying a fucking exploit waiting to happen.
<OhGodAGuardian>
ME has had more bugs than a rain forest.
<Zorix>
they do it because they know people don't really have much of a choice
lovepon has quit [Ping timeout: 264 seconds]
<OhGodAGuardian>
This is the issue. Even considering AMD - people really don't have a choice. When there's two diff parties who might just do the same idea cause they feel like it, you're fucked.
<sorear>
they're not doing the same thing because they hate you and are colluding, they're doing the same thing because they have no economic incentive to care what you think
<OhGodAGuardian>
Thing is, while people don't have a choice, I wish it got more press. They act downright fucking shady about the whole thing - which, to an extent, would be understandable were it limited to corporate envs.
<OhGodAGuardian>
sorear: No shit.
<Zorix>
go ask random people you know, or even IT people, 99% of them will say.. they don't have anything to hide or don't care one way or another
<sorear>
intel answers to disney here
<OhGodAGuardian>
sorear: Honestly, I don't care if they like me or personally hate my guts, end result is the same.
<OhGodAGuardian>
Zorix: I wouldn't say 99%, but close, yeah
<Zorix>
everyone i know in person, i have asked this question to, 100% of them don't care
<OhGodAGuardian>
The counter for that is noting that ME has bugs out the ass, meaning it's not just the govt getting access, it might be a dude wanting your bank info.
<zkms>
well NSA sure cares
<zkms>
hence the HAP bit :^)
<OhGodAGuardian>
Right.
<OhGodAGuardian>
I'm considering flicking it on, but this board I do cherish. Might give it a go on one I care less about first.
<OhGodAGuardian>
4+ years ago now, I built this system. Still feel no need to upgrade.
<Zorix>
i do a cpu/motherboard upgrade every 8 years
<OhGodAGuardian>
No, I mean, not a goddamned thing. MAYBE better monitors.
<OhGodAGuardian>
That's about it.
<Zorix>
im usually starving for ram at some point thats about it
<OhGodAGuardian>
also, wow. 8y.
<Zorix>
yea the motherboards start getting pretty unreliable at about 7 years heh
<OhGodAGuardian>
yeah, I build shit on my machine, I don't have time to waste xD
<Zorix>
guarantee if i go try to power on my old motherboard it wont
<sorear>
I jumped from an early-stepping P4 to a 2012 macbook
<OhGodAGuardian>
It's like Kineta's board, the ASRock X99 WS, was built for me. I have NO clue what market they were going for
<OhGodAGuardian>
besides "me"
<sorear>
CPU improvements in that timeframe are not really noticable but getting a SSD for the first time is w h o a
<SolraBizna>
I only upgraded from my PowerBook G3 because its backlight was failing
<OhGodAGuardian>
Dual LAN with teaming, 5x PCIe Gen3 x16 (all run at x8, even full up), 1x PCIe Gen2. 10x SATA 3, 2 4-pin Molex (optional) to provide more power to PCIe area... even a POST code display.
<SolraBizna>
I would still be using that thing today if not for that
<Zorix>
not participating in the ssd craze until they do something about endurance.. i dont like renting my storage
<sorear>
then replaced that in 2015 with another macbook with a bigger disk, then replaced that with an xps13 when I quit and had to return the work laptop a year ago
<OhGodAGuardian>
Zorix: uh, you're late.
<OhGodAGuardian>
What rock have you been under?
<sorear>
probably getting something new so I have a backup computing device ~soon
<sorear>
Zorix: what kind of endurance
<Zorix>
wear leveling isnt a solution, it's a workaround, nand has been losing endurance every new generation
<SolraBizna>
zram: my anti-random-failure-of-my-SSD-causing-me-to-lose-all-my-data-at-once-with-no-warning
<sorear>
modern SLC is worse than old SLC?
<Zorix>
you should see the pile of SSDs that hit the recycle boxes at work because of failures
<sorear>
QLC is going to be terrible, that's a given
<zkms>
that's why they invented backups and mirroring.
<Zorix>
those are also workarounds
<OhGodAGuardian>
I'd probably trust some of my Samsung Evo drives more than some WDs.
<Zorix>
i can use a magnetic hard drive and have very reliable storage, longevity, better price, and at the cost of a few seconds of extra load time.. the pros outweigh the cons
<OhGodAGuardian>
Although, I still have magnetic storage
<sorear>
i've lost far more data on magnetic hard drives than ssds
<OhGodAGuardian>
but no, it's not worth it when you need access all the time.
<OhGodAGuardian>
sorear: I heard that.
<zkms>
sure, until the centimetre-scale actuators that need to achieve nanometer precision and fly single-digit nanometers above a fast-moving surface decide to give out
<zkms>
no storage is reliable.
<sorear>
nm-precision cm-scale actuators, it's a miracle they work at all and they tend to stop at a moment's notice
<OhGodAGuardian>
zkms: reliability is a spectrum :P
<zkms>
in the end everything is shockingly fragile and small bit cells backed up by soft-decode LDPC :3
<sorear>
I regard "PV/SSDs/etc are non-renewable because they have limited lifetime" arguments as quite specious, the only thing stopping us from grinding up 'dead' silicon and sending it back through the fab is how cheap sand is
<OhGodAGuardian>
The people who know the reliability best is likely the makers... Samsung offers a warranty of 5 years or 2,400 TB written for a 4TB SSD.
<sorear>
the only human technology that permanently consumes atoms is nuclear power
unixb0y has quit [Ping timeout: 246 seconds]
unixb0y has joined ##openfpga
<whitequark>
Zorix: i like being able to drop my laptop and not worry about it
<whitequark>
now *that* is reliability
<whitequark>
just fuckin smash it into a wall
<whitequark>
who cares
<Zorix>
you think that wont pop a chip off a pcb when it's running at high temp? heh
<Zorix>
at the very least crack a bga joint
<Zorix>
ive seen flash drives damaged by dropping on the ground that i could recover by pressing on the bga chip while copying the data off
<whitequark>
Zorix: yes
<whitequark>
i am very sure
<whitequark>
because i have both literally soldered on my laptop with a hot air gun *and* threw it into a wall repeatedly
<Zorix>
heh what did it do to you to deserve that
noobineer has joined ##openfpga
Bob_Dole has joined ##openfpga
<sorear>
wtf
<sorear>
here and I've not dropped my laptop more than two inches
<sorear>
i assumed that the solder joints were fragile af
<whitequark>
Zorix: i used to have severe anger issues
<Bob_Dole>
OhGodAGuardian, do you know if there's any plans to add the new GDDR5X 1060 to the ohgodanethlargementpill?
<whitequark>
i have dented multiple walls with my previous laptop
<whitequark>
i changed the keyboard at least three times
<whitequark>
i think four
<whitequark>
modern laptops are remarkably resilient
<Bob_Dole>
or is that kind of abandoned for public-consumption
<whitequark>
i eventually replaced it because i wanted a better display fwiw
<whitequark>
Zorix: sorear: also if even solder joints crack i can just diagnose and reflow them.
<whitequark>
not a big deal.
<whitequark>
shattered HD platter though kind of is
<Bob_Dole>
I type consumption, and immediately: thoughts of vore.
kuldeep has joined ##openfpga
<pie__>
wait this is ##openfpga
<Bob_Dole>
also OhGodAGuardian like DDR3 MRAMs are hard to come by but they -have no refresh cycle- and I'm curious how that'd make them work for like cryptonight mining
<Bob_Dole>
if one were to have them for other reasons, I'm sure the faster types of the big 'ol SRAM chips available would be a better choice if going for just mining
<Zorix>
whitequark, with enough force you can also fracture the silicon
<whitequark>
Zorix: I tried that on some old PS3 (?) PCB
<whitequark>
I took off the heatsink and started hitting it with a screwdriver as hard as I could
<whitequark>
it took me multiple minutes to even make a dent
<Zorix>
hmm
genii has quit [Remote host closed the connection]
<SolraBizna>
talk me into and/or out of putting Ethernet on this thing
<Bob_Dole>
using an ISA 10baseT chip?
<Bob_Dole>
like what the MyCPU uses?
<Bob_Dole>
because I'll talk you out of putting it on solder-test-board but not main board.
<Bob_Dole>
do it
<Bob_Dole>
I'm now getting targeted ads for the 86duino on facebook.
<Bob_Dole>
and I never did shoot dm&p an email about how available the vortex86 chips actually are
<whitequark>
Bob_Dole: or just make a PHY withsome CPLD :P
GenTooMan has quit [Quit: Leaving]
Miyu has quit [Ping timeout: 245 seconds]
<SolraBizna>
it would be on the IO board
<Bob_Dole>
..could the fe310 handle the tcp ip stack?
<SolraBizna>
I guess, but I was only going to have it do checksum offloading, if that
<Bob_Dole>
SolraBizna, you should join ##whitequark too
<SolraBizna>
but then I'll see more messages
<pie__>
hilarity ensues
<Bob_Dole>
it's a little less fpga-centric and gets other projects in there too.
<sorear>
i'm highly surprised that the "refresh cycle" has more than percent-level impact
<Bob_Dole>
I observe a lot of channels on a lot of platforms to pick up as much information as I can even if I can't understand that information.
<sorear>
also *cough* if you're activating every row every 16-32ms you don't need a separate refresh
<pie__>
Bob_Dole, yep
<sorear>
can probably go longer than that if you don't care about BER
<Bob_Dole>
pie__, yep?
<sorear>
if 0.01% of POW calculations are wrong, you've effectively lost 0.01% of throughput; it's impossible for errors to propagate beyond that
<Bob_Dole>
SolraBizna, I'm sure you have figured out that I know a fuck ton about stuff I don't actually understand At All, haven
<Bob_Dole>
't you
lovepon has joined ##openfpga
<SolraBizna>
and yeah, I was going to make my own using the FPGA you will almost certainly force me to put on the IO board
<Bob_Dole>
whitequark, tell SolraBizna about SD Cards
<Bob_Dole>
plz
<Bob_Dole>
that's about the only reason for fpga on io board unless you want to get a DAC for VGA on it
<Bob_Dole>
and not do resistor ladder dac
<SolraBizna>
the FE310 doesn't have quite enough GPIO ports to do everything I want... an FPGA on there would let me fix that, and also greatly simplify the display interface
<Bob_Dole>
oh yeah the eInk thing
<Bob_Dole>
'spose there more than VGA
<sorear>
it's amazing how much packaging has driven the semiconductor industry
<Bob_Dole>
more than vga and sd-cards*
<SolraBizna>
it could also scan the keypad / keyboard and maybe even make USB possible
<Bob_Dole>
whitequark was recently talking a lot about problems with sdcards and how they sometimes have the SPI interface broken and such so it's best to actually use them as sdcards and not as MMC card etc.
<SolraBizna>
an FPGA would let me do that too
<Bob_Dole>
yeah
noobineer has quit [Ping timeout: 252 seconds]
<Bob_Dole>
are VGA-centric DACs even available anymore?
<Bob_Dole>
since the trend of moving the VGA DAC on-die with the gpu started back in the 90's
<sorear>
it's always been sort of curious to me that "RAMDAC" is discussed as one component
<Bob_Dole>
yeah but you need 3 8bit DAC channels for VGA
<Bob_Dole>
well for good VGA
<sorear>
VGA color components only have 6 bits of precision
<Bob_Dole>
really? every VGA-centric dac I had seen was 8bit
<sorear>
well, I'm talking about the capabilities of the original VGA
<SolraBizna>
Bob_Dole is talking about the signal, not literal VGA cards
<Bob_Dole>
but anyways: I found one that makes needing a resistor ladder.. unneeded, didn't find that part last time I looked probably a couple years ago but that one was damn easy to find
<Bob_Dole>
SolraBizna, you talked about making an eink vga display didn't you? maybe I should look for an ADC..
<SolraBizna>
as a way of testing the signaling with a variety of source images easily, and as a good source of rapidly changing images to test refresh rates
<Bob_Dole>
so ADC and DACs for dicking about with VGA exist, makes some Dumb Things I hope to commission to creation of seem more plausible.
<SolraBizna>
I've deliberately designed the "oh god soldering BGA by hand" test board to be useful for that test and others
<Bob_Dole>
OhGodASolderTest >.>
<Bob_Dole>
OhGodAGuardian, would we get sued by OhGodACompany if we silkscreen that onto a pcb
<sorear>
what exactly is the meme here
<Bob_Dole>
OhGodACompany, run by OhGodAGirl, released OhGodAnEthlargementPill and now there's a lot of OhGodA's. like OhGodAPaladin and OhGodAPizza etc.
<Bob_Dole>
or was run by OhGodAGirl I'm not sure how OhGodACompany is currently organized.
<sorear>
so OhGodAGuardian is neither OhGodAGirl nor OhGodAGeek?
<Bob_Dole>
OhGodAGuardian is Wolf9499 (that number string I think?) aka OhGodAPet.
<sorear>
and this is like, crypto internal politics stuff?
<Bob_Dole>
the OhGodA stuff is a crypto community stuff, wouldn't call it politics
<Bob_Dole>
I don't have any expectations of viability of competing with anything the people involved with OhGodACompany are making. too high end of parts, too much pre-existing experience.
<Bob_Dole>
for crypto* but I hope for things to exist outside of the crypto space.
<Bob_Dole>
SolraBizna's line about Oh God Soldering BGA By Hand made me think of how it's kinda turned into a meme, and I fuckin love memes.
<Bob_Dole>
a lot
Guest91359 has joined ##openfpga
Guest91359 has quit [Client Quit]
wpwrak_ has joined ##openfpga
wpwrak_ has quit [Remote host closed the connection]
wpwrak_ has joined ##openfpga
wpwrak_ has quit [Remote host closed the connection]
wpwrak_ has joined ##openfpga
Bob_Dole has quit [Read error: No route to host]
Bob_Dole has joined ##openfpga
wpwrak_ is now known as wpwrak
Bike has quit [Quit: Lost terminal]
_whitelogger has joined ##openfpga
_whitelogger has joined ##openfpga
zkms has quit [Quit: WeeChat 2.1]
zkms has joined ##openfpga
zkms has quit [Client Quit]
zkms has joined ##openfpga
zkms has quit [Quit: WeeChat 2.2]
zkms has joined ##openfpga
xdeller_ has quit [Remote host closed the connection]
<OhGodAGuardian>
Bob_Dole: lol, I doubt it.
<OhGodAGuardian>
Bob_Dole: Wolf9466.
<OhGodAGuardian>
AKA OhGodAPet
<Bob_Dole>
well then, OhGodASolderTest
rohitksingh has joined ##openfpga
<Bob_Dole>
and then if this works we'll move to FACT, and then onto new projects after I've made a few FACT boards
<SolraBizna>
like the ARS!
<Bob_Dole>
maybe. ARS intentionally being super retro though would probably be easier. encapsulate some tinyfpga boards, plop on a big QFP ice40hx4k, get pdip SRAMs and some EPROMs..
<Bob_Dole>
intentionally seeming*
<Bob_Dole>
some greenpak breakout boards, too
<Bob_Dole>
and then we have the packages and available logic from the Era you've described with modern parts.
<Bob_Dole>
I mean, QFP has been getting used since the late 80's. I think even the 286 was available in qfp
rohitksingh has quit [Quit: Leaving.]
<SolraBizna>
I think all four of the big ICs in the ARS are DIP
<SolraBizna>
six if you count the RAMs
_whitelogger has joined ##openfpga
<Bob_Dole>
there's level shifters that are fairly fast. I think whitequark mentioned their glasgow project is using some, so looking at that board would be where to find them. And then the greenpaks are 5V tolerant
<Bob_Dole>
and open
<Bob_Dole>
greenpaks don't have a ton of logic on them or a ton of io
<SolraBizna>
that would probably fit the APU and the "hub" IC, but definitely not the PPU
<Bob_Dole>
Ididn't notice, can greenpaks do level-shifting themselves?
rohitksingh has joined ##openfpga
dingbat has quit [Quit: Connection closed for inactivity]
Bob_Dole has quit [Ping timeout: 260 seconds]
rohitksingh1 has joined ##openfpga
rohitksingh has quit [Ping timeout: 264 seconds]
lovepon has quit [Ping timeout: 260 seconds]
flaviusb has quit [Quit: Leaving.]
flaviusb has joined ##openfpga
rohitksingh1 has quit [Ping timeout: 244 seconds]
rohitksingh has joined ##openfpga
wpwrak has quit [Ping timeout: 252 seconds]
wpwrak has joined ##openfpga
s_frit has quit [Remote host closed the connection]
mumptai_ has quit [Remote host closed the connection]
dingbat has joined ##openfpga
<whitequark>
Bob_Dole: note that greenpaks only have OEs on half the pins
rohitksingh has quit [Quit: Leaving.]
<whitequark>
controllable via fabric
<whitequark>
I've actually considered greenpaks
<whitequark>
also they're kind of slow
<Bob_Dole>
"ARS" is supposed to be "the maximum thing achievable in the late 80s for game consoles"
<Bob_Dole>
so greenpaks being slow might be OK for some roles within the system
emeb has joined ##openfpga
drawkula has quit [Remote host closed the connection]
drawkula has joined ##openfpga
renze has joined ##openfpga
<sorear>
what does this have to do with the Argentine peso
<qu1j0t3>
very little i suspect
rohitksingh has joined ##openfpga
<Bob_Dole>
are there any ppc softcores out there? I've seen x86 (ao486) 68k (TG68), MIP, SPARC, and older ARM verions
<qu1j0t3>
don't see anything on opencores, but that would be a nice project.
* qu1j0t3
is also powerpc fan
<Bob_Dole>
I was thinking: why pair a x86 chip with desired project when there's x86 softcores that work? may need some patching up, but.. starting point. And an fpga would allow sticking some other arch on it for amiga, mac, or general acceleration purposes.
<whitequark>
SolraBizna: well i recently wrote an original 8086 plugin for binary ninja
<whitequark>
that includes semantic lifting
<Bob_Dole>
zet86 covers 16bit stuff
m4ssi has joined ##openfpga
<whitequark>
i don't remember why
<Bob_Dole>
and I think zet86 had patches within the past few months merged to github
<Bob_Dole>
and it also has been run on xilinx and altera.
<Bob_Dole>
and seems to be verilog I think
edmund__ has quit [Quit: Ex-Chat]
<Bob_Dole>
(I don't know enough about vhdl or verliog to look at a source file and say yep, that's verilog.. and remember having a hard time finding where it said what it was written in.)
<Bob_Dole>
SolraBizna, does that sound like a better plan to you though? not include x86 hard-core, but have an fpga with some dedicated RAM on it that can be reloaded without whole-system reboot for any ISA you can write a soft-core for?
<SolraBizna>
other than the general problems with heterogenous architectures, yes
<Bob_Dole>
need x86 software, just load ao486 and boom, accelerated x86 vm. need amiga software? boom fast amiga vm. need 68k mac? boom accelerated 68k mac vm
m4ssi has quit [Ping timeout: 264 seconds]
<sorear>
qemu on a smallish soc will be faster than a softcore for most of those combinations…
<Bob_Dole>
also whitequark, azonenberg_work is the ecp5 going to be better to try to get a high ipc with bigger slower clocked things or go full netburst high clock for future risc-v cores?
<whitequark>
never go full netburst
<Bob_Dole>
I've thought about SoCs, I was thinking about vortex86 or Geode, because of the frequent-enough need of x86.
m4ssi has joined ##openfpga
<Bob_Dole>
sorear ^, but then those are harder to get these days. an extra fpga would be easier to source (just ordering more of the main cpu's), no NDAs, and give some versatility.
<SolraBizna>
I'll just buy sorear's gigabox when it's ready, and use that to emulate every architecture at once
rohitksingh has quit [Quit: Leaving.]
lovepon has joined ##openfpga
rohitksingh has joined ##openfpga
m4ssi has quit [Ping timeout: 252 seconds]
m4ssi has joined ##openfpga
massi_ has joined ##openfpga
pie__ has quit [Ping timeout: 252 seconds]
m4ssi has quit [Ping timeout: 264 seconds]
<gruetzkopf>
i do have some obscure 8086 (well nec v20) and 80186 software i should take a close look at
<SolraBizna>
is there a way to put an FPGA with way too many connections into my schematic without it being super awkward?
<azonenberg_work>
SolraBizna: i normally make one sch symbol per io bank
<SolraBizna>
... now I wish I had asked that before I packed 115 pins into one symbol
<azonenberg_work>
lol
rohitksingh1 has joined ##openfpga
rohitksingh has quit [Ping timeout: 268 seconds]
rohitksingh1 has quit [Client Quit]
pie__ has joined ##openfpga
massi_ has quit [Remote host closed the connection]
<Bob_Dole>
when getting Voltage Regulators, what company do people prefer? (need 3.3v to 1.2 and 1.8v)
<Bob_Dole>
<SolraBizna> the ideal one would be one that can have a source of 5V or 3.3V, and an output of 3.3V, 1.8V, or 1.2V
<SolraBizna>
Mouser's smart search is the way I normally find ICs I need, but it's almost impossible to use when it comes to voltage regulators
Guest47146 is now known as gruet
gruet is now known as gruetze
<m_w>
if you need less juice a combination of the buck and ldo might be easier
<m_w>
less passives to source
<m_w>
those pesky inductors
<SolraBizna>
for the OhGodASolderTest, a self-contained buck converter is probably what I should use
<Bob_Dole>
I probably would have been looking for MAX brand stuff if other parts weren't suggested. >.>
<SolraBizna>
using an LDO for 3.3V→1.2V seems like a tremendous waste
pie_ has joined ##openfpga
mumptai has quit [Quit: Verlassend]
pie_ has quit [Ping timeout: 252 seconds]
pie_ has joined ##openfpga
pie__ has joined ##openfpga
pie_ has quit [Ping timeout: 240 seconds]
pie___ has joined ##openfpga
pie__ has quit [Ping timeout: 245 seconds]
s_frit has quit [Remote host closed the connection]
s_frit has joined ##openfpga
emeb has quit [Quit: Leaving.]
<SolraBizna>
I can never remember which of N-MOSFET and P-MOSFET "switches on" with positive gate voltage and which "switches off" with positive gate voltage
<SolraBizna>
same with NPN and PNP BJTs
<whitequark>
NTR BJTs
<whitequark>
wait wrong channel
<sorear>
BJTs are current-mode devices
<sorear>
but yeah
<sorear>
analog electronics is complicated and intimidating ;;
<SolraBizna>
the ARS design has a big blank space where I just say "And then YIQ conversion and NTSC modulation happens here"
<zkms>
what is ARS?
<SolraBizna>
an 8-bit game console
<zkms>
oic
<whitequark>
zkms: do you have a highlight on ntr
<zkms>
w-wha why would I have such a thing, i'm not interested in nuclear thermal rockets or anything of the sort
<Bob_Dole>
I am
<SolraBizna>
my daily headache interrupted my schematic making with nothing left to do but hook up the non-directly-driven status LEDs
<Bob_Dole>
nuclear thermal rockets are fascinating
<SolraBizna>
if I can just remember which kind of MOSFET to use on the "active-low" vs. "active-high" statuses, I can muddle through the rest
<sorear>
just try to avoid your YIQ module upconverting and transmitting data-dependent noise from the digital domain
* SolraBizna
burrows deeper into the all-DC, all-digital bunker
ZipCPU has joined ##openfpga
<qu1j0t3>
whitequark | wait wrong channel // if you dont mind me asking, what channel did you have in mind