azonenberg_work has quit [Ping timeout: 252 seconds]
unixb0y has quit [Ping timeout: 244 seconds]
<TD-Linux>
basically custom map drawers?
unixb0y has joined ##openfpga
azonenberg_work has joined ##openfpga
GenTooMan has quit [Quit: Leaving]
<azonenberg_work>
welp, back online
<azonenberg_work>
cell tower must have crashed and rebooted or something?
<azonenberg_work>
me and $wife both lost service at the same time for a couple of hours in a location that normally has 4+ bars of LTE
<mithro>
Anyone know how to tell nextpnr where to find ice40 db files?
<mithro>
Seems to be ICEBOX_ROOT....
<awygle>
ugh can't believe I forgot to check the alignment holes in the stencil. on the plus side I'm never doing this without an assembly drawing again.
Bike has quit [Quit: Lost terminal]
azonenberg_work has quit [Ping timeout: 246 seconds]
rohitksingh_work has joined ##openfpga
rofl_ has joined ##openfpga
jcarpenter2 has quit [Ping timeout: 252 seconds]
rofl_ is now known as jcarpenter2
<awygle>
whitequark: built a board except for the iron-y bits, turns out i don't actually have my iron. so i'll go get it tomorrow and finish up.
<whitequark>
awygle: ooh you can test it already!
<awygle>
whitequark: i could! eeeexcept there are solder bridges on the cypress :p
<whitequark>
lol of couorse
<awygle>
selfishly it's probably for the best as i can now go to bed and be well rested at work in the morning lol
<whitequark>
hah
<awygle>
so you know how we forgot to put the alignment holes in the actual stencil for rev A? i forgot to add them for rev B too.
<rqou>
lol you even warned me about this exact problem :P
s_frit has quit [Remote host closed the connection]
s_frit has joined ##openfpga
rohitksingh has joined ##openfpga
genii has joined ##openfpga
emeb has joined ##openfpga
wpwrak has quit [Read error: Connection reset by peer]
wpwrak has joined ##openfpga
rohitksingh has quit [Quit: Leaving.]
rohitksingh has joined ##openfpga
m_t has quit [Remote host closed the connection]
Miyu has joined ##openfpga
GuzTech has quit [Quit: Leaving]
<shapr>
GOOD MORNING!
<sorear>
Morning!
rohitksingh has quit [Quit: Leaving.]
<awygle>
Morning!
<awygle>
Happy Monday
rohitksingh has joined ##openfpga
<awygle>
isn't there something weird about the ice 40 reset? Like it's actually a set or something?
<awygle>
Oh nope found it, there's no initial value after global chip reset everything comes up 0.
<sorear>
right, synthesis works around that by inverting D and Q
<TD-Linux>
are there any downsides to that? if not, seems like a useful saving of bits
<awygle>
i'm not sure it's always free? although you'd think it could be
<sorear>
if you need inverted and non inverted from the same LUT, it’s not
<awygle>
ah, true
azonenberg_work has quit [Ping timeout: 252 seconds]
rohitksingh has quit [Quit: Leaving.]
azonenberg_work has joined ##openfpga
azonenberg_work has quit [Ping timeout: 252 seconds]
azonenberg_work has joined ##openfpga
rohitksingh has joined ##openfpga
digshadow has joined ##openfpga
rohitksingh has quit [Quit: Leaving.]
rohitksingh has joined ##openfpga
digshadow has quit [Ping timeout: 250 seconds]
m4ssi has quit [Remote host closed the connection]
rohitksingh has quit [Quit: Leaving.]
mumptai has joined ##openfpga
digshadow has joined ##openfpga
<shapr>
AT one point I heard that the polarfire FPGAs were unlikely to be RE'd because the bitstream is encrypted, and that could lead to DMCA problems
<shapr>
is that correct?
<shapr>
azonenberg_work: I think I heard that from you? polarfire reverse engineer unlikely because of encrypted bitstream?
<azonenberg_work>
i havent touched polarfire whatsoever
<shapr>
ok thanks, any idea who I could ask?
mwk has joined ##openfpga
<rqou>
i think awygle was concerned about DMCA
<rqou>
i personally don't think it's that high a risk but whatever
<awygle>
lol I mean yes, if you break encryption it triggers dmca
<awygle>
I didn't claim PF had mandatory encryption tho
<shapr>
hm, now I want to know
<rqou>
ianal but i believe you don't trigger anti-circumvention if you break the encryption on your own bitstreams
<shapr>
and won't have time to investigate until after ICFP
<azonenberg_work>
rqou: i think that is correct, because you own the copyright you can do whatever you want - it's not circumvention at that point
<rqou>
although a tool implementing "decrypt with default keys and convert to verilog" might be illegal because it allows you to circumvent the protection on somebody else's bitstream
<azonenberg_work>
correct
<azonenberg_work>
That would be
<shapr>
soo, I wouldn't get into legal trouble if I tried to reverse engineer a polarfire FPGA?
<azonenberg_work>
Using it on your own bitstream would likely fall under the bona-fide security research exemption
<shapr>
or is that a different pile of worms?
<azonenberg_work>
publishing anything you learned from doing that would be dicey
<shapr>
ah
<shapr>
even source code to make it work?
<azonenberg_work>
I also dont know anything about how polarfire works
<azonenberg_work>
with xilinx parts the bitstream is cleartext by default and has optional encryption
<azonenberg_work>
it's possible, but i dont know for sure, that polarfire is always encrypted
<azonenberg_work>
and they either use a default key or a per-design key
<rqou>
whitequark why are you reversing shitty gba pirate carts? aren't psram-based "flash"carts a dime a dozen?
<shapr>
docs say polarfire has support for flash based bitstream storage, with encryption key in the non-volatile ram, so upgrades can be crypto signed
<shapr>
hmm
<shapr>
sounds like it's per-design, but I'm still not sure I want to spend $2k on this board
<rqou>
xilinx also "supports" that but doesn't do it by default
<shapr>
Is there a todo/priority list for which FPGA to RE next?
<shapr>
hmm, clearly I need to learn more
<rqou>
everybody has their own personal one apparently
<shapr>
well, I can understand that
<rqou>
mine is public but a lot of people's aren't: finish max v, max10, cyc10lp, cyc10gx
<rqou>
cyc10lp and max10 may or may not trade places
[X-Scale] has joined ##openfpga
X-Scale has quit [Ping timeout: 252 seconds]
[X-Scale] is now known as X-Scale
<azonenberg_work>
shapr: my #1 target would be anything with SERDES on it
<azonenberg_work>
Both xilinx 7 series and lattice ecp5 have serdes on them, and have been partially reverse engineered
<azonenberg_work>
but all of the work to date has been on the general logic fabric
<azonenberg_work>
nobody's touched the serdes ip on either of them
<azonenberg_work>
the lattice transceivers are likely to be less complicated and thus easier to RE
<shapr>
so, sounds like ecp5 is low hanging fruit
<shapr>
every month or so I'm tempted to go dig into the wiki and clean up the status docs, but then I go back to learning verilog
* shapr
shrugs
<awygle>
yeah help out with ecp5 would be my recommendation for new contributions
<awygle>
I haven't seen daveshah in a while but I'm sure you could figure out what needs doing pretty quickly
<daveshah>
Is there anything in particular that you want to work on?
<daveshah>
Adding BRAM support to Yosys and nextpnr would be amazing
<daveshah>
Or you could fuzz the gigabit SERDES, or the IOLOGIC pin gearboxes
<sorear>
polarfire docs said it was always encrypted when I looked at them earlier
<azonenberg_work>
shapr, awygle: yeah ecp5 is a top priority for me because right now i am stuck in xilinx-land
<azonenberg_work>
i cant use any of the foss-friendly fpgas because they're too tiny for all of my work
<azonenberg_work>
ecp5 would be useable for some of my smaller projects
<azonenberg_work>
And the xilinx effort is a lot further from being useable than the ecp5 project
<awygle>
Oh hey daveshah lol
<zkms>
yeah i'd appreciate ecp5 a *lot* too
<sorear>
If I acquired a versa board and a diamond install in the next month or two could I realistically help?
<qu1j0t3>
what _is_ the friendliest open one? ice40?
<qu1j0t3>
best supported*
<qu1j0t3>
i suppose i could check topic...
<qu1j0t3>
greenpak4? LOL. i bought GP4 because of azonenberg_work 's amazing blog post. but
<qu1j0t3>
in terms of ordinary fpga
<sorear>
yes ice40
<qu1j0t3>
sorear: thx
<sorear>
support for that is near-completely, and it’s small but a “normal fpga”
<azonenberg_work>
greenpak is the best supported because they documented everything
<azonenberg_work>
no reversing, just toolchain dev
<azonenberg_work>
it was slow progress because i did it solo, but it was straightforward
azonenberg_work has quit [Quit: Leaving.]
azonenberg_work has joined ##openfpga
pie_ has joined ##openfpga
<qu1j0t3>
ooh, cypress psoc5lp, i got one because it's interesting on paper
<qu1j0t3>
the future looks bright!
<azonenberg_work>
qu1j0t3: yeah cyrozap and pointfree have been working on those
<azonenberg_work>
havent heard much in a while
<daveshah>
Haven't had much time to work on it atm, and the time I've had has mostly been other stuff (CSI-2)
<daveshah>
Will be back working on it after ORConf, which I'm preparing for now
<daveshah>
I believe that will be recorded, and I'll discuss the ecp5 stuff and next steps more there
<awygle>
woo orconf. too bad i can't make it
<awygle>
sounds like a good time
<daveshah>
I'm really looking forward too it
<daveshah>
My first time visiting Poland too
<shapr>
is there a suggested/recommended ecp5 dev board?
<awygle>
i have a Versa
<awygle>
that would be my recommendation personally
<awygle>
as it also comes with a diamond license
<rqou>
does diamond use flexlm? :P
<awygle>
yes
<rqou>
lolcrypto or ECC?
<rqou>
i have an lmcrypt for the old ise key
<rqou>
idk if vivado uses the same one or not
<awygle>
no idea
<rqou>
paste one of the SIGN= entries?
<awygle>
is there a way to get icetime to print out its timing estimate in a machine consumable way?
<awygle>
er i guess i can dump json, but i just want "283880000" on stdout
<rqou>
meh, underspecified json is the IPC of the future! :P
<awygle>
as opposed to underspecified byte streams :p
<azonenberg_work>
almost as useless and a lot bulkier :p
<rqou>
underspecified xml with xxe and a billion laughs :P
<rqou>
and/or generated with printf and parsed with a regex :P
<rqou>
azonenberg_work's life right here
<azonenberg_work>
rqou: i wish the vulns i was finding on this gig today were that sophisticated :p
<rqou>
wait what
<rqou>
that bad?
<rqou>
how about "type AAAAA....AAA and now PC is 0x41414140"? :P
<jn__>
rqou: have you seen the iLO4 auth bypass bug a little white ago?
<rqou>
lol yeah
<jn__>
s/white/while/
<awygle>
maybe i just don't think like a business person. before i hired somebody like azonenberg_work or his employers i would definitely have fixed the stupid easy vulns.
<rqou>
but that requires an understanding of what a "stupid easy vuln" is :P
<pie_>
there's the anecdote about the businessman asking the artist why he should pay a lot for the thing it took him 10 minutes to draw, "but it took me 10 years to learn how to draw in 10 minutes" or something :P
<rqou>
also i learned at $WORK that apparently a lot of companies that hire people like azonenberg_work for Compliance(TM)
<rqou>
not for actually not having vulns
<pie_>
that....makes sense
<rqou>
Compliance(TM) has a direct and obvious impact on the bottom line; having vulns has (historically) had a smaller and less noticeable impact
<qu1j0t3>
daveshah: oh, neat. if there ends up being a video link, pls drop it here?
<daveshah>
sure, will do
<azonenberg_work>
rqou: i found that actual bug (textbook buffer overflow, in a shell argument parser no less) in a piece of scada gear a while back
<rqou>
wtf
<pie_>
azonenberg_work, nahhh i bet thats how they loaded plugins
<rqou>
azonenberg_work: in your expert opinion, how plausible are the claims of the form "if you nmap this network segment, there will be loss of life"?
<azonenberg_work>
rqou: i once took down a client's entire corporate network running a sql injection scan against one server
<azonenberg_work>
so, very plausible
<rqou>
wtf
* pie_
turns off his eyes
<jn__>
medical IT is scary
<pie_>
see no evil read no evil
<azonenberg_work>
rqou: i had tested the entire network in mostly-manual mode
<qu1j0t3>
jn__: tru dat
<azonenberg_work>
found one server with, as far as i could tell
<azonenberg_work>
sql injections in every form field on every page
<azonenberg_work>
No filtering whatsoever
<azonenberg_work>
I exfiltrated the entire db, cracked the mysql root password
<qu1j0t3>
yep, i've seen that in medical records handling code.
<qu1j0t3>
that was PHP and open to the public.
* qu1j0t3
has seen some shit
<azonenberg_work>
anyway, then i asked the client for permission to run a sqlmap scan to test the rest of the form fields just to make sure i could get a complete list of injectable fields
<azonenberg_work>
He says sure
<azonenberg_work>
an hour later as the scan is finishing they go "Whatever you're doing, STOP IT RIGHT NOW"
<pie_>
you couldnt just exfiltrate the source code? :P
<azonenberg_work>
Turns out, they had a single security appliance that was a combination firewall, vpn server, border router, and IDS/IPS
<qu1j0t3>
LOL azonenberg_work
<azonenberg_work>
The IDS sucked so badly that i was able to do all this hacking without tripping it once
<azonenberg_work>
but it had signatures for a few of sqlmap's queries
<azonenberg_work>
The IDS also had a memory leak...
<pie_>
uh lol
<azonenberg_work>
boom, entire network down until soembody walked into the DC and unplugged it
<azonenberg_work>
then plugged it back in
<azonenberg_work>
that was a finding in the report :p
<awygle>
LOL
<awygle>
that's... wow.
<rqou>
have you ever experienced a client saying "ok, thanks for the report. we have checked the Compliance(TM) checkbox. we accept the risks of all findings"?
<azonenberg_work>
awygle: tl;dr you could dos their network with a few thousand http requests over a fairly long time
<azonenberg_work>
didnt even have to be sql injectable
<azonenberg_work>
it just had to let you send http to it
<azonenberg_work>
and it had to look like sql to th eids
<awygle>
that's incredible
<awygle>
i ran into some code written by a contractor a while back which just didn't contain "free" at all
<azonenberg_work>
rqou: worse, we've had clients that actively fight us getting things done
<awygle>
and which malloc'd 4k on every like, received packet
<azonenberg_work>
tl;dr user of equipment requests a pentest
<azonenberg_work>
they ask OEM, who they have a support contract with, to provide some docs etc for us
<azonenberg_work>
OEM stonewalls
<awygle>
and the solution was "if memory usage gets above X, reboot"
<azonenberg_work>
awygle: o_O
<rqou>
awygle: works fine if it's a missile :P
<jn__>
awygle: was that also in a "security" appliance?
<awygle>
rqou: yeah basically lol. iirc they ended up fixing the problem (just use the stack instead of malloc-ing) but keeping the high water mark test. which is fair i guess
<awygle>
jn__: no thank goodness
<rqou>
i assume you've all seen the "the missile will explode before too much memory leaks" solution? :P
<awygle>
yes lol
<azonenberg_work>
rqou: they do that with thermal management too
<azonenberg_work>
Tiny heatsink, FPGA/CPU will be max of X degrees ambient at launch
<azonenberg_work>
max range of missile is Y
<azonenberg_work>
heatsink will reduce rate of temp rise to Z degC/sec
<awygle>
batteries too, obviously
<rqou>
also o-rings? :P
<rqou>
too soon still?
<azonenberg_work>
as long as X + Y*Z < Tjmax you're good
<awygle>
and turbopumps, sometimes
<awygle>
there are rocket engines whose turbopumps are constantly failing, just slow enough to complete the mission
<awygle>
i really need to find this paper i read a while back where they fired linksys routers at each other on rocket sleds
<awygle>
it sounds ridiculous enough that i'm starting to worry i made it up
<rqou>
there was a hilarious discussion on birbsite about "if you receive a piece of UXO that has GPL code, can you demand the source code from the aggressor country?" :P
<qu1j0t3>
you dreamed it awygle
<qu1j0t3>
rqou: this doesn't even sound far-fetched :<
<rqou>
so, ##openicbm when? :P
<rqou>
we can both into space _and_ into global thermonuclear war :P :P
<awygle>
how about a nice game of chess?
<gruetzkopf>
ask joshua about that
<qu1j0t3>
that should be the lichess.org slogan
<rqou>
awygle: e4 :P
<awygle>
rqou: g3 :p
* awygle
is very bad at chess
<rqou>
yeah me too
<awygle>
i guess it's g5
<awygle>
oh well
<azonenberg_work>
rqou: you can demand it
<azonenberg_work>
But its unlikely you will realistically be able to compel them successfully to provide it
<awygle>
"national security"
<awygle>
i wish people giving lfsr polynomials would also give it in hex form. i always screw up the conversion.
<pie_>
i have not <rqou> i assume you've all seen the "the missile will explode before too much memory leaks" solution? :P
<qu1j0t3>
pie_: doesn't take much imagination
<qu1j0t3>
pie_: pretty sure it was the original model for PHP memory management
<qu1j0t3>
pie_: "The request will end before we run out, don't worry!"
<qu1j0t3>
pie_: *never reclaim, leak everything* in effect
<pie_>
haha
* qu1j0t3
aside: I'm not convinced PHP doesn't still use this model ;-)
<awygle>
lol i coredumped iverilog
<awygle>
ten points to me
<pie_>
awygle, your point counter has overflowed
<awygle>
segmentation fault core dumped
<jn__>
hmm, how do i write a disclaimer that says that some RE'd documentation is not written by the company that made the RE'd hardware but by independent people?
<qu1j0t3>
phone a lawyer? :)
<jn__>
qu1j0t3: that's the correct solution, i know, but... i don't even know any lawyers
<prpplague>
jn__: as qu1j0t3 said, get a lawyer, don't be stupid and try it yourself
mwk has quit [Ping timeout: 272 seconds]
<pie_>
whats the worst that could happen
<qu1j0t3>
jn__: what country?
<jn__>
.de
<jn__>
pie_: some variant of getting sued, i guess
<prpplague>
jn__: you can be personally sue
<prpplague>
sued
<qu1j0t3>
hm. I only know an IP lawyer in .ca, but if it's important, you shouldn't have much trouble tracking one down
mumptai has quit [Read error: Connection reset by peer]
<pie_>
for giving bad documentation of for copyright enfringement?
<pie_>
*or for
mumptai has joined ##openfpga
<jn__>
pie_: "clean" documentation without obviously copyrighted content
<pie_>
@ personally sued
<jn__>
<rant> this stuff needs standard text, similar to open source licenses </rant>
<pie_>
jn__, ask CCC?
<jn__>
pie_: i don't think that'll work
<pie_>
¯\_(ツ)_/¯
mwk has joined ##openfpga
<jn__>
:|
<pie_>
jn__, what are you trying to protect against
<jn__>
1. people thinking i leaked $VENDOR-internal documentation
<jn__>
2. legal action from $VENDOR (which i don't know crap about)
<rqou>
there's no way to really defend against 2
<pie_>
1 sounds relatively easy 2 is lawyer category i guess
<rqou>
they can always try to sue you into bankruptcy even if they know they will lose
<pie_>
well maybe you can just upload the stuff super anonymously? is that reasonable
azonenberg_work has quit [Quit: Leaving.]
<rqou>
you can hide if you feel that is important but it doesn't actually absolve you of anythhing
<prpplague>
pie_: no such thing
<awygle>
they are both lawyer category
<awygle>
although be aware a lawyer may try to tell you "just don't do the thing"
<awygle>
so you really need to know what your risk tolerance is
mumptai has quit [Quit: Verlassend]
<rqou>
you may or may not be able to hide behind corporate limited liability, but *) that doesn't always work *) that doesn't actually ensure that your thing can _stay_ released
<prpplague>
jn__: i have no clue what you are RE, but i can tell you three of the big corps i have worked for in the past have dedicate groups inside their company specifically for tracking and defending their IP from people who RE
azonenberg_work has joined ##openfpga
<rqou>
the "safest" thing i know of so far for a basic hobbyist is to stay inside the EU :P
<pie_>
prpplague, huh.
<awygle>
the conventional wisdom seems to be to do what clifford did, and reverse engineer the stuff, release the documentation, and then have somebody else totally unrelated to you do the software based on the documentation
* jn__
nods
<rqou>
all while being in the EU :P
<jn__>
the EU part is easy for me :)
<rqou>
yeah, despite the EU's current fuckery with copyright, the "reverse engineering for compatibility" thing seems to have much stronger protections
<gruetzkopf>
and reimplementations of certain system service to get rid of certain restrictions
<jn__>
you don't need an nvidiablob on linux
<gruetzkopf>
(this stackup should be portable to linux though)
<rqou>
wow the scene's been pretty busy
<gruetzkopf>
armada and fincs did most of the gl stuff from what i've read
<gruetzkopf>
and we now have nouveau people hanging out in efnet/#switchdev :D
<rqou>
what's the workflow for running linux?
<gruetzkopf>
enter rcm, upload uboot, boots
<rqou>
oh wow
<rqou>
nice
<rqou>
i should actually unpack the switch systems i bought with my earnings from XBT going up right before the crash :P
<gruetzkopf>
enter rcm: either pull emmc, short pin10 of right joycon port to gnd (ANDROID_HOME), hold VOLUP, press power, or intentionally corrupt /dev/mmcblk0boot0 and /dev/mmblk0boot1
<rqou>
does dual-booting work?
<rqou>
can we run horizon under kvm? :P :P
<gruetzkopf>
if you're not doing the last method then yeah, simply don't short it
<gruetzkopf>
i did the last method, there's this handy android app which performs the required usb stuff and then uploads a payload
<shapr>
anyone coming to strangeloop?
<gruetzkopf>
i use hekate for that payload, which should also be able to load a linux kernel and provides a boot-time selector and backup/restore for emmc
<gruetzkopf>
the ususal setup after that replaces the fairly useless 'album' applet with the homebrew loader, (applets get less memory, get killed on home-button-press, but also don't get recorded to the play log and then reported to nintendo)
<rqou>
wtf
<rqou>
so this is "inject h&s" redux
<gruetzkopf>
not really :D
<rqou>
why can't people patch out the recording of logs instead?
<gruetzkopf>
because it's not fully understood yet
<gruetzkopf>
the crash-report process is replaced by something that dumps them in text-form to sd
<gruetzkopf>
but there's other stuff
<gruetzkopf>
this replacing happens using a filesystem-proxy at runtime
<gruetzkopf>
they turned the sigchecking up to 11 this time
<gruetzkopf>
doesn't help them if that has already been replaced at the time the binary is accessed..
<rqou>
huh
<rqou>
no equivalent to apple's "just patch libmis and everything is pwn"? :P
<shapr>
pie_: it's probably the same as on the one ring
* shapr
looks
<gruetzkopf>
you can also replace any arbitrary file in any game with this method
<shapr>
pie_: One ring to rule them all, one ring to find them, One ring to bring them all and in the darkness bind them.
<rqou>
hmm, highly invasive patches like this just make me nervous
<rqou>
e.g. the 3ds "replace half the sysmodules" patch is really scary
<gruetzkopf>
that's at-runtime patching
<gruetzkopf>
at-loadtime, sorry
<gruetzkopf>
reboot and you're back to 100% normal
<TD-Linux>
shapr, I also guessed the same from experience looking at japanese DOS error messages and guessing what english one they were shaped like
<rqou>
but i want a system to be haxed all the time while still being able to play games
<rqou>
while also not being detected of course :P
<shapr>
I could read and write Sindarin in high school, but that was thirty years ago
<shapr>
TD-Linux: heh, I've just branched out into more languages :-P
<rqou>
invasive patches just seem much more likely to trip a detection
<shapr>
TD-Linux: 𐑖𐑱 𐑺𐑰𐑖𐑩𐑯?
<shapr>
KirinDave got me interested in Shavian last week
<gruetzkopf>
which is why you blacklist the reporting hosts?
<rqou>
lol
<rqou>
can you do that?
<rqou>
and still be able to bank your pokemans? :P
<TD-Linux>
shapr, 何？
<shapr>
TD-Linux: don't know about that, sorry... In descending order of skill I can read American, Swedish, Dutch, German, Portuguese, Spanish, Croatian and probably some that don't come to mind
bcoppens has joined ##openfpga
<pie_>
shapr, ah heh
<shapr>
Used to read lojban and esperanto and a bit of klingon
mwk has quit [Ping timeout: 240 seconds]
<pie_>
lol
<shapr>
Elvish is hard to read at a distance, the letters are too similar; I found dwarvish easier/faster.
<TD-Linux>
oh man, lojban *and* esperanto. you can make extra informed arguments on wikipedia discussion apges
<shapr>
xu do tavla mi bau la lojban
<shapr>
there's even a quenya mode for lojban, I wrote that for awhile :-P
<awygle>
i gotta get back into japanese. i can still _say_ "i'm not very skilled in japanese but i'll try my best", but i can no longer write it or read it at speed.
<rqou>
weeb :P
Bike has quit [Ping timeout: 252 seconds]
<gruetzkopf>
my latin teacher in middle school once busted someone who had transliterated a proable solution to the test to elvish lettering
<shapr>
It's weird, when I was a kid tolkien nerds were the weebs, now you can get a university job studying tolkien
<prpplague>
hehe
<gruetzkopf>
pointing out the mistakes in that transliteration
<prpplague>
shapr: yes, and roll playing games like were completely nerdy... now they are called "Fantasy" baseball and football
<shapr>
I was the worst of the weebs, staying up till 4am on the internet in the 80s. I'd tell my classmates about it, and they'd say "What the heck is the internet? Why would you do this thing?"
<shapr>
prpplague: blood bowl :-P
<prpplague>
shapr: internet in the 80's ?
<shapr>
yeah, I got on the internet (and IRC) in 1988
<prpplague>
shapr: what part of the world?
<prpplague>
shapr: access was extremely limited during that time period
<shapr>
I agree, I was lucky enough to discover that UTK had a block of modems for student access
<shapr>
you could telnet anywhere from there
<prpplague>
shapr: ahh
<shapr>
I was not a UTK student, never have been
<prpplague>
oh, hehehe
mwk has joined ##openfpga
<shapr>
but before there were 'computer crimes' I spent lots of time exploring the network
<shapr>
prpplague: I got lucky, grandkid visiting his grandparents for the summer won a brand new maxed out Timex Sinclair, and left it at my house for a year when he went home.
<prpplague>
shapr: mmmmmm TS, one of my favorites
<shapr>
I was 11 in 1982, tomorrow I turn 47! whee!
<rqou>
wow you people are all oooooold :P
<qu1j0t3>
shapr: ZX-81?
<prpplague>
hehe
<qu1j0t3>
shapr: one of the first machines I used too
<qu1j0t3>
shapr: my uncle had one , later my dad bought one
<prpplague>
qu1j0t3: TS-1000 was the US version of the ZX
<qu1j0t3>
yeah i have no idea about the north american nomenclature
<qu1j0t3>
prpplague: was that the -80 or the -81 though?
* qu1j0t3
never touched an -80
<shapr>
rqou: bet I can do better cartwheels than you can :-P and better mountain unicycling
<prpplague>
qu1j0t3: TS-1000 was the zx-81
* qu1j0t3
nods
<shapr>
although I did lose my last cartwheel contest to a triathlon competitor...
<shapr>
rqou: I challenge you to a cartwheel contest next time I see you
Miyu has quit [Ping timeout: 245 seconds]
<rqou>
lol i can't even do a cartwheel
* rqou
is a lazy fatass :P
* shapr
wins
<TD-Linux>
I've been trying to discuss a custom video encoding ISA in japanese and it's... challenging
<TD-Linux>
rqou, it's okay I wasn't born in the 80s
<TD-Linux>
despite this z80 was my first assembly language because it will never die in the TI-83+
<prpplague>
qu1j0t3: i've been tinkering with a discrete ULA replacement for my TS-1000
<prpplague>
when i say tinkering i mean playing around for like 3 years, hehe
<prpplague>
sinclair's use of the ULA was a brilliant idea at the time
<rqou>
TD-Linux: hey, nspire is finally ARM (with a z80 emulator lol)
<rqou>
somehow TI refuses to acknowledge ixh/ixl/iyh/iyl opcodes :P
<TD-Linux>
the nspire was also horrible when I last tried it. terrible keyboard, terrible battery life, super fat
<TD-Linux>
really slow "mouse cursor"
<TD-Linux>
in some ways I wouldn't mind them making the ti-83+ forever. but at a $15 price point
<TD-Linux>
actually, the newest rev of the nspire looks a lot better. here was the original for reference
<rqou>
also, why would they lower the price point if they can keep milking the "SAT approved" marketing? :P
* awygle
's rant about the college board goes here
<rqou>
do you have one for pearson too?
<rqou>
actually i don't remember actually seeing your collegeboard rant, can you actually insert it? :P
<TD-Linux>
I'm from the midwest so we took the ACT instead of SAT
<rqou>
i mean, TI milks that too
<awygle>
i mean it boils down to "fuck those assholes and their ridiculous monopoly on a harmful system"
<rqou>
harmful system?
<awygle>
the SAT is like, obviously a net negative for the world
<awygle>
which exists only to enrich the college board
<awygle>
and make certain people feel superior
<rqou>
imho ap classes made the "harmful system" of the american k-12 educational system much more bearable
<rqou>
but yeah, the sat is kinda useless
<awygle>
AP classes are like, a hackish patch to a broken system
<awygle>
an improvement but way way downstream from the sources of the problems
mwk has quit [Remote host closed the connection]
<rqou>
sure, i can agree with that
* awygle
's californian is coming out, lots of extra "like"s in this one
<awygle>
AP tests should be free though, _obviously_
<awygle>
i also, personally, wish that it was clearer what the value proposition of AP tests was, because I took *counts* i think 11 AP tests and 6 IB tests? and got out of a single semester of math and a single semester of english.
<awygle>
but that doesn't really comment on the fundamental system, at least without getting into how college admissions work which is a whole other thing
<rqou>
hmm, i took a similar number (no IB) and it was useful for artificially inflating my academic standing
<rqou>
which made taming the TeleBEAR much easier
<awygle>
yes, that too, i did come in as a late sophmore
<rqou>
also for me there was the huge value of "now you aren't in the regular classes"
<awygle>
was that worth more than a thousand dollars? idk
<rqou>
i definitely felt it was
<rqou>
but maybe my situation is very different
<awygle>
not the "not in regular classes" thing, the "early registration" thing
<rqou>
despite not officially having tracks, my school basically self-segregated into two separate schools in the same campus
<rqou>
the "honors/ap people" and "regular people" basically never interacted
<awygle>
yeah we did that too, i just didn't have the problems with the "regular people" you seem to have had
<rqou>
well maybe the problem will gradually go away here in fremont as the indians and chinese take over :P
<rqou>
yes, that's actually how the demographics of the classes worked too
<rqou>
it's a not-insignificant part of the price premium of a house here too
<rqou>
the bay area is pretty cursed, isn't it?
mwk has joined ##openfpga
<rqou>
bay area protip: don't build your high school near the caltrain tracks
<rqou>
too soon?
<gruetzkopf>
my uni had (and will soon again have) this lecture hall called TEMP
<jn__>
and it has a different building called SemiTEMP
<gruetzkopf>
which is directly next to aachens freight yard
<rqou>
uh... that's not what the problem was in this case
<azonenberg_work>
rqou: somebody built a school literally next door to a gun club somewhere east of seattle
<azonenberg_work>
and now people are complaining about it
<rqou>
lol
<azonenberg_work>
(mind you the club had been there for decades before they broke ground on the school)
<rqou>
but is it a high-stress school full of asians? :P
<sorear>
something something grade separation
<azonenberg_work>
rqou: well as far as i know there have been no incidents
<azonenberg_work>
just people complaining about the noise
<azonenberg_work>
but still
<emeb>
shooting range distracted by the noise that the schoolkids make?
<rqou>
loool
<sorear>
so in Boston people get hit by subway trains a few times a year, by commuter trains every other week, I suspect the difference here is grade separation
<rqou>
caltrain has a ton of level crossings
<rqou>
grade separation is expensive
<gruetzkopf>
it's the long-time target of all european railway infrastructure providers
<TD-Linux>
they are putting a ton of bridges in
<TD-Linux>
for caltrain too
<rqou>
i'm pretty sure in the US the target is "don't lose money" :P
<TD-Linux>
but it is expensive and slow
<sorear>
do they not use the same 10MM per fatality benchmark the FAA does
<gruetzkopf>
getting a new level crossing approved in germany is basically impossible
<TD-Linux>
the new trains they ordered are pretty nice (stadler kiss), rode them on s-bahn in berlin
<gruetzkopf>
KISS? only on those lines which mix up RE and S service
<gruetzkopf>
i also heard something about them wanting siemens desiro HC and then not being able to get federal funding
<gruetzkopf>
(which is why the local project (RRX) is getting them on time)
<TD-Linux>
gruetzkopf, yeah IIRC it was instead of some other out of service train
<gruetzkopf>
ah
<sorear>
Now I’d just like to not have level crossings between pedestrian paths and 40+ mph arterials
<rqou>
nobody wants to continue with some NIH indian gauge trains? :P
<TD-Linux>
the trains were ordered before federal funding was delayed iirc
<rqou>
oh wait caltrain isn't a strange gauge is it?
<gruetzkopf>
well, no KISS in my area, but FLIRT1 and FLIRT3 use basically the same tech
<sorear>
don’t think so
<sorear>
caltrain is the one that’s dragging their feet on electrification
<gruetzkopf>
FLIRT1 looks better, FLIRT3 fares better when colliding with standard buffer-and-chain wagons
<rqou>
well, i hope caltrain doesn't decide to go with some crappy ancient 1960s rectifiers :P
<rqou>
it's not like i'm hating on any particular piece of shit system :P :P
<sorear>
there are way too many rail standards
<TD-Linux>
caltrain is standard gauge
<rqou>
i'm talking about a particular indian gauge 1960s dc third rail system that always implies they needed to buy parts off of ebay to fix things
<TD-Linux>
rqou, it's going to be 25kv 60hz ac iirc
<gruetzkopf>
ah, so they're not bothering with new transformer designs
<TD-Linux>
hey man there's no reason they couldn't use modern rectifiers on BART
<rqou>
not certified? :P
<gruetzkopf>
25kV 50Hz is already a variant that exists on the stadler EMUs
<TD-Linux>
yeah it's the current preference for new high speed or heavy trains it seems
<gruetzkopf>
(of course germany is 15kV 50/3 Hz)
<TD-Linux>
most shinkansen lines are 25kV 60hz
<awygle>
isn't there a fault line in japan where 50 and 60 Hz meet?
<TD-Linux>
50/3 hz is nice for universal motors but not for inverter powered motors
<TD-Linux>
awygle, yes but I think the shinkansen lines are at 60hz all the way thorugh
<gruetzkopf>
yeah electrification here was mostly done before modern inverters were a thing
<gruetzkopf>
my favourite vehicle is the class 420 EMU, which sounds (and works) mostly like a vacuum cleaner slowly being turned up to max
<awygle>
insert 420 meme
<gruetzkopf>
it's not strictly a universal motor, rotor and stator are powered by different windings
<gruetzkopf>
(transformer windings)
<sorear>
25@60 is what Boston-New Haven has, yay
<gruetzkopf>
tell people to aquire a siemens class 406 "ICE3 multi-system" emu
<gruetzkopf>
those will go up to 330kph on that power supply
<rqou>
lol you're funny
<rqou>
US tracks aren't nearly good enough for trains to run that fast
<sorear>
if they ever electrify the rest of the suburban system here they’ll either use that or have exciting geometric problems on the Providence branch
<gruetzkopf>
okay, they'll run up to 200kph on terrible tracks
mwk has quit [Ping timeout: 245 seconds]
mwk has joined ##openfpga
<rqou>
124 miles per hour?
<rqou>
still dreaming :P
<rqou>
try like 80 mph / 125 kph lol
<gruetzkopf>
(and if you're somewhat careful, they'll also run on 600VDC (even though only really designed for 1500VDC, 3000VDC, 15000 / 50/3 and 25000 / 50