digshadow1 has quit [Quit: Leaving.]
digshadow has joined ##openfpga
digshadow1 has joined ##openfpga
digshadow has quit [Ping timeout: 268 seconds]
digshadow1 has quit [Ping timeout: 255 seconds]
digshadow has joined ##openfpga
zino_ has joined ##openfpga
scrts has quit [Ping timeout: 260 seconds]
scrts has joined ##openfpga
cr1901_modern has quit [Read error: Connection reset by peer]
promach has joined ##openfpga
<cyrozap-ZNC> nats`: What exactly are you looking for? If you're looking for something that says one way or the other "X HDL produces netlists that use less area than Y HDL", I don't think you'll find one, especially since that would be more dependent on the synthesis tool than the language itself (I think).
cyrozap-ZNC has quit [Quit: Client quit]
cyrozap has joined ##openfpga
digshadow has quit [Ping timeout: 260 seconds]
digshadow has joined ##openfpga
digshadow1 has joined ##openfpga
digshadow has quit [Read error: Connection reset by peer]
digshadow1 has quit [Ping timeout: 260 seconds]
digshadow has joined ##openfpga
<davidc___> I'm sure its availability 2032, at $20k/pc
<balrog> about the time to buy REcon tickets
<cyrozap> pointfree: It's just an "RF-processing subsystem" added to a Zynq UltraScale+, not a PSoC competitor (unless it's < $20).
scrts has quit [Ping timeout: 240 seconds]
scrts has joined ##openfpga
<cyrozap> How is that the false path?
<cyrozap> clang-analyzer is trying to tell me that the function returns from here by taking the false path, which doesn't make any sense. If it takes the false path, it will return much later in this function. If it takes the true path, it'll return false at this point.
digshadow has quit [Quit: Leaving.]
<cyrozap> I inverted the conditional and it actually shows itself returning later in the function after taking the false path, so I have no idea what's going on there.
digshadow has joined ##openfpga
Hootch has joined ##openfpga
<cyrozap> Oh, wait, I may have misinterpreted the message...
<cyrozap> Oh, yes, I was, and now I see where the bug is.
scrts has quit [Ping timeout: 268 seconds]
scrts has joined ##openfpga
GenTooMan has quit [Ping timeout: 260 seconds]
DocScrutinizer05 has quit [Disconnected by services]
DocScrutinizer05 has joined ##openfpga
GenTooMan has joined ##openfpga
GenTooMan has quit [Ping timeout: 260 seconds]
_whitelogger has joined ##openfpga
GenTooMan has joined ##openfpga
amclain has quit [Quit: Leaving]
pie_ has quit [Ping timeout: 268 seconds]
Lord_Nightmare has quit [Quit: ZNC - http://znc.in]
<rqou> offtopic: i wonder if the three burned-out light bulbs in my apartment might be related to our "unusual usage patterns" :P
Lord_Nightmare has joined ##openfpga
m_w has joined ##openfpga
cr1901_modern has joined ##openfpga
m_w has quit [Quit: leaving]
<rqou> woot, after tremendous yak shaving:
<rqou> $ dig +short AAAA osx-build-machine.dyn.rqou.com
<rqou> 2601:644:203:1472:5054:ff:fe90:955b
DocScrutinizer05 has quit [Read error: Connection reset by peer]
DocScrutinizer05 has joined ##openfpga
kuldeep has quit [Remote host closed the connection]
massi has joined ##openfpga
kuldeep has joined ##openfpga
flaviusb has quit [Quit: Leaving.]
flaviusb has joined ##openfpga
<rqou> offtopic: finally deciding to fix my backups after them being f***ed for a while and it seems crashplan has gotten a _lot_ faster
<nats`> cyrozap it's what I'm searching and it's exactly about the quality of synthetizer
<nats`> something puzzle me
<nats`> VHDL is more strict so it should be easier to optimise but in the mean time I think most of the market is verilog oriented so more effort are made on this one
<lain> rqou: based on prior crashplan experience, my guess is they're just discarding the data. that's much faster and cheaper than actually storing it, which they weren't doing before anyway :U
<rqou> so which backup vendor do you recommend?
<lain> rqou: unfortunately I don't have a good recommendation right now. I used to recommend JungleDisk (backed by Amazon S3, with local encryption), but they're changing everything around and it's fucked now too
<rqou> whellp
<rqou> at least this is the second layer backup
<rqou> the crappy raid-1 is the first layer backup
<lain> for unencrypted, I backup to amazon cloud drive with acd_cli
<rqou> (except the part where i haven't been consistently backing up to the nas :P )
<rqou> i kinda want encrypted backups and i have a lot of data
<lain> you can apparently use duplicity with acd_cli backend, with encryption
<lain> it's on my list to evaluate
<rqou> not for a good reason but rather because my data is a huge mess
<lain> haha
<rqou> apparently 3mil files / 1.5 TB
<rqou> contains fun gems like disk images, or 2tb sparse files with 0 bytes allocated :P
<rqou> which i created for lulz and apparently trolled myself
eduardo__ has quit [Ping timeout: 240 seconds]
<lain> but do you have an infinite zip file!
<rqou> probably
<rqou> 42.zip?
<lain> haha
<rqou> i'll look
<whitequark> um
<rqou> oh right, that's the other one
<rqou> not 42.zip
<lain> well
<lain> there's a slightly better 42.zip implementation, more recent
<lain> ah yes, zipquine, that's the one
<rqou> it's for feeding to norton, right? :P
<lain> it's for emailing to people you hate
<lain> :3
<rqou> unless you're @taviso and "people you hate" is symantec :P
<rqou> i think "PoC to symantec security team exploits symantec's email scanner" deserves an award
<rqou> great, crashplan is now choking on a 30gb tarball
<rqou> lain: this is what i meant by "my file organization is totally f*cked"
<rqou> this 30gb tarball is itself a backup of another system somewhere
<rqou> except i didn't label what it was of
<lain> haha
<lain> yeah I have an "old backups" folder that is full of fuck
<lain> rqou: so thanks for reminding me about this, I'm adding the duplicity port to my build box so I can test this crap out
<rqou> you do photography; you should have an even worse file situation than me
<rqou> i've _only_ found about four random DCIM folders so far :P
<lain> I use lightroom to organize my photos
<lain> keeps everything sorted nicely :P
<rqou> i was going to do that, but that involves another round of yak shaving
<lain> just have to press import :D
<rqou> i need to set up the "how to actually access files on the NAS" part :P
<lain> apparently I have 1.5 TB of photos
<rqou> it used to use samba, which had a race condition with network bringup and didn't work half the time
<lain> this is what happens when you're a megapixel whore
<rqou> also samba is slow
<lain> uhhhhhhh
<lain> so configuring the duplicity port
<lain> oh, nvm, was going to ask a stupid question
<rqou> hmm i should test samba again now that the nas<->desktop link is 10gbps
<rqou> iirc samba was pretty terrible at permissions, but NFS was even worse
<rqou> certain people swear by 9p.2000L as a network filesystem, but i'm a bit wary because it's pretty uncommon
<lain> never heard of it
<rqou> huh, apparently people other than me also thought that "nfs server in kernel" was a little bit dumb: https://github.com/nfs-ganesha/nfs-ganesha
<rqou> 9p.2000L is a hacked version of the plan 9 network file system
<rqou> L for Linux
<rqou> because it's been realigned with how the linux vfs layer works
<rqou> certain users have made a 9p transport over virtio and over rdma
<rqou> supposedly 9p is a lot simpler and easier to understand than nfs
<whitequark> ugh nfs
<rqou> ime nfs barely works
<lain> I just use cifs, it works. only really notice slowness when I have a dir with thousands of files, which is a problem anyway so meh
<lain> I wouldn't run cifs over the internet, but it's fine for lan
<rqou> i've tried running cifs over the internet
<rqou> it doesn't work super great
<lain> yeah it's a bad plan
<rqou> i did cifs over ssh though
<lain> I've had to cifs over vpn before to grab files in an emergency and yeahhhh it's slow
<rqou> er, maybe it was cifs over openvpn
<lain> hm
<rqou> don't be enterprisey and run cifs over plain tcp over wan :P
<rqou> especially with highly enterprisey passwords :P :P
<jn__> like <company name><year>?
<rqou> you joke, but i've actually seen that
<lain> man what website did I have to deal with recently that was just impossible to create an account with...
<rqou> G+? :P
<lain> oh, it was the indiana medicaid site thing
<rqou> oh god, a us state gov website
<lain> the password entry field simply shows this: https://i.imgur.com/xAYdRME.png
<lain> but
<lain> it literally took reading the js to figure out what it wanted
<lain> see, each of those is required
<lain> but "special characters" is their own magical set of special ch ars
<lain> not ALL special chars that, say, lastpass will use when you enable special chars
<lain> and no, there is no text on the page explaining any of this other than what you see in that image
<lain> nothing in any FAQ, nothing anywhere. only the source knows.
<rqou> i wonder if they have problems with passwords like "'; drop table users; --" :P
<lain> "You cannot use a password you used in the last 24 times"
<lain> fucking lol
Bike has quit [Quit: leaving]
<jn__> i.e. they have your last 24 pw
<lain> yeah
<rqou> they might just have a hash of them
<rqou> one can hope :P
<jn__> (or hashes thereof)
<lain> oh holy shit
<lain> there IS text explaining it on the change password page (but not on the initial sign up page)
<lain> BUT ITS COMMENTED OUT IN THE HTML
<rqou> unlike a Certain Major Brokerage Firm that only accepts 8-char case-insensitive passwords
<jn__> koalaty
<rqou> but this Certain Major Brokerage Firm supported 2FA!
<rqou> by appending TOTP keys to the end of the password
<lain> "You must change at least 4 characters in your password from your last password" hashes don't exiiiiist
<rqou> WTF
<rqou> even PHP idiots can do better than that
<lain> the funny thing is EVERY TIME YOU LOG IN to this site, it gives you a big scary warning where you must agree that you're entering a secure website and for your safety it's super duper uber secure and you must be secure when using it
<lain> whatever that means
<jn__> they don't check against the last 24 passwords on the client side, right?
<lain> jn__: they don't appear to
<rqou> lain: do they block ssllabs like Certain Other Shitty-Security Websites? :P
<lain> didn't mcdonalds recently get caught storing the user's password (encrypted) in a cookie for some dumbass reason?
<lain> rqou: good question
<lain> let's see... it's a comodo OV cert
<lain> sha256, 2048 bits
<rqou> "super duper uber secure" and OV...
<lain> "Each time you login to your account to access case information or report changes, you will be required to accept the Confidentiality Agreement below.
<rqou> must be the budget cuts :P
<lain> By clicking on the ‘I Accept’ button, you are accepting the confidentiality, acceptable use and other privacy policies as mandated by the State of Indiana.
<lain> Also, note that it is your responsibility for printing and keeping copies of this sensitive information. You should never provide your password to a state worker or anyone else. Click the ‘I Do Not Accept‘ button to end this session and log out.
<lain> Click here to review or print the User Acceptance Agreement you accepted when creating your account. That agreement governs your use of this site.
<lain> "
<lain> I wonder what happens if I click I Do Not Accept
<lain> I'm not going to try it though because it'll probably permaban me until I discuss my problem with a human, and I have enough trouble with their system as it is :P
pie_ has joined ##openfpga
<rqou> hmm, i wonder how the government can break out of the cycle of "gov websites suck" --> "_I_ wouldn't work as a programmer for the government" --> "gov websites continue to suck"
<lain> good question
<lain> healthcare.gov is a disaster as usual
scrts has quit [Ping timeout: 240 seconds]
<lain> I'm currently in some kind of health insurance limbo, have to remember to call them today between 0800 and 1600
<lain> well, they got an A-
<lain> that's pretty astonishing for a .gov :P
<rqou> wow damn
<rqou> that's actually higher than some banks :P
<lain> yeah
<lain> I love my bank
<lain> I have no idea if their site is secure, and I don't care
<whitequark> rqou: lol banks
<lain> it can't do anything
<rqou> or did PCI-DSS check-the-box security kick some banks in the butt a bit?
<whitequark> gov and bank websites are like race to the bottom
<lain> I can view my balance, literally anything else requires an in-person visit and forms filled out by hand
<lain> which is /fantastic/ as far as I'm concerned, not even being sarcastic here
<lain> because so what if someone hacks my account
<whitequark> lain: wow, that's dysfunctional
<whitequark> I mean it's great if you never travel or move I suppose
<lain> haha
<whitequark> personally I don't even use 2FA, availability is more important to me
<rqou> i'm not sure my bank even does 2fa
<lain> I do have a PIN setup so I can do transfers via phone
<whitequark> they can't bottom out my credit cuz I dont have any and I have cash on hand almost always
<lain> yeah I'm mostly cash, I do have a credit card now but mainly because they offer better fraud protection than a debit card, and this one gives me 5% cash back on amazon which saves me a TON of money :P
<lain> had my debit card stolen at a fast food place once
<lain> krebs is always writing about new fast food and hotel and gas station breaches
<whitequark> I had my debit card stolen at a fast food place once
<whitequark> as in literally stolen
<lain> I think it's irresponsible to use anything but cash in modern day US, except online where cash isn't acceptable (then make sure your card has a good fraud policy)
<lain> whitequark: LOL
<whitequark> never had any trouble otherwise
<lain> I miss paypal's debit generator thing
<lain> where you could generate a new CC number for each transaction
<lain> and set a time and/or transaction limit on it
<rqou> my father's bank offered that
<lain> god that was great
<lain> BoA also has (had?) it
<rqou> maybe they still do
<rqou> hrm really?
<whitequark> I don't do credit because if you forget to pay like 1 RUB then you may never get out of the country before the bank lets you
<whitequark> well, RU credit
<jn__> cash is great, i love cash
<lain> whitequark: how did someone steal your card
<whitequark> lain: they stole my wallet
<whitequark> that had card in it
<lain> bummer
<rqou> lain: another krebs article that apparently has something to do with vidya: https://krebsonsecurity.com/2017/02/who-ran-leakedsource-com/
<whitequark> oh that was long ago
<lain> rqou: yeah that was a good one
<rqou> apparently vidya is srs business
<whitequark> SC in HK unfortunately doesnt have debit
<whitequark> so I'll have to do credit now...
<whitequark> probably can set the limit on it to zero or something
<rqou> my experience dealing with HK banks was that they're even more annoying than US banks
<lain> my paypal debit is ridiculous
<lain> the per-transaction limit is $10k
<lain> ten thousand fucking dollars
<rqou> ostensibly more secure, not exactly sure if it is
<rqou> but quite a bit more difficult to do stuff remotely
<lain> I'm perfectly happy that my business debit (through my bank) has a limit of like $400 per day or per transaction whichever comes first
<rqou> maybe this is only a problem because i'm in HK rarely
<whitequark> rqou: well HSBC has flat out refused to open an account for me for months
<lain> I can call them to temporarily raise it
<whitequark> because theyre scumbags
<whitequark> SC has been... acceptable so far
<rqou> SC is ?
<whitequark> standard chartered
<whitequark> they opened an account in like a hour
<rqou> ah
<rqou> my HK account is with citibank, and it also involved a lot of stalling
<lain> right, I should shwr and head into the datacenter. bbiab
<rqou> this was before FATCA too
<lain> hopefully can get mike to install the dish today :D :D :D
<whitequark> HSBC (and Hang Seng, they seem to have exact same req's) asked for a lot of bullshit paperwork
<whitequark> some of which I didn't just not have
<whitequark> it simply has never existed and could not possibly exist in this world
<whitequark> SC just wanted my ID and employment contract
<whitequark> and visa
<rqou> my parents purposely set up my HK account to break the "bullshit paperwork" infinite loop in case I ever want a different (better) HK account
<lain> my bank does have some issues though, like one time I was waiting in line in the lobby and someone was resetting the pin on their debit card, and the teller read the new pin and their entire debit card number aloud to the entire room to confirm
<rqou> because apparently paperwork infinite loops are relatively well-known in HK
<rqou> :P
<whitequark> lol
<lain> but I've worked with banks before doing server and teller machine upgrades, and I know the tellers just have zero security training/knowledge
<lain> they mostly exist to be a friendly interface to what should be automated
<lain> for people who are allergic to computers
<rqou> the only time i interact with tellers is to get quarters :P
<lain> :P
<whitequark> rqou: why do ATMs in HK say "advice"
<whitequark> instead of "receipt" or w/e
<rqou> idk, i've never used an ATM in HK :P
<whitequark> DO YOU WANT ADVICE? YES/NO
<lain> >yes
<lain> >get a job
<lain> thanks atm
<whitequark> fukken yes but that's not what you mean do you
<lain> lol
<whitequark> also this SC ATM is bizarre
<rqou> machines in the EU getting confused with US chip-and-signature cards is worse
<whitequark> you insert your card into it and it starts to print something
<rqou> or rephrased "the US is gratuitously different from the rest of the world again" :P
<whitequark> except nothing comes out
<whitequark> I suspect it is actually printing a syslog or something
<whitequark> on paper
<lain> well
<lain> rip the front off and find out!
<lain> and/or attempt backscatter x-ray on it
<lain> just roll up with a backscatter machine in the back of a van
<whitequark> oh I could just wait until it gets refilled
<lain> nobody will notice
<whitequark> walk up to it and take a photo or two
<whitequark> but im too lazy
<lain> I mean
<lain> that's the boring way
<lain> but sure
<rqou> hmm i saw an atm being refilled in tuen mun once and they actually had an armed guard standing around it
<rqou> i don't know if they would appreciate photos being taken
scrts has joined ##openfpga
<lain> don't taze me bro
<rqou> no, with an actual gun
<whitequark> yes
<whitequark> I've seen that
<whitequark> well, what's the worst thing that would happen? this is not US, they ain't gonna shoot me
<rqou> lool
<rqou> get the triads to disappear you? :P
<whitequark> pffff
<rqou> yeah, the triads have more profitable things to be doing :P :P
<whitequark> exactly
scrts has quit [Ping timeout: 260 seconds]
scrts has joined ##openfpga
talsit has left ##openfpga [##openfpga]
pie_ has quit [Ping timeout: 240 seconds]
kuldeep has quit [Ping timeout: 260 seconds]
rqou_ has joined ##openfpga
pie_ has joined ##openfpga
rqou has quit [Quit: ZNC 1.7.x-git-709-1bb0199 - http://znc.in]
rqou_ is now known as rqou
<lain> hm
<lain> this is interesting
<lain> using PSL properties (or similar) to automatically instantiate assertion logic in an fpga design
<rqou> offtopic: anyone (lain?) want to buy this vanity domain? http://i.imgur.com/SD1sTUL.png
<lain> basically, instead of or in addition to simulation, and instead of instantiating a logic analyzer in the design, automatically turn your design property assertions into logic in the design itself
<lain> rqou: nop
<rqou> yeah, kinda expensive to buy just to put a photo of a 555 timer on it :P
<lain> rqou: btw which registrar are you moving to?
<rqou> name.com
<lain> lemme know how it is plz
<rqou> i'm a bit concerned because some of my domains are showing as just "ok" rather than "pendingTransfer"
<rqou> they all suck
<rqou> i mean, registrars are commodities
<lain> eventually, yeah
<lain> I've heard good things about gandi
<lain> or you could use that tucows service to become your own registrar on the cheap :P
<lain> a few friends did that
<lain> well, you become a domain provider or something
<lain> not an actual registrar
<rqou> a reseller?
<rqou> seems not worth the effort unless you hoard vanity domains even more than me
<rqou> :P
<lain> eh
<lain> there's like no real effort afaik
<lain> you can just sign up and sell yourself wholesale domains
<lain> but yeah
<lain> :D
<lain> I don't think there's much benefit either, so meh.
<rqou> wait, they also resell ssl certs?
<rqou> i wonder how many bugs you can find there? :P
<rqou> so skimming the symantec f*ckup thread on mozilla.dev.security.policy:
<rqou> >>So after reading this, the following auditors aren't trusted by Symantec
<rqou> >>anymore:
<rqou> >>- E&Y Korea
<rqou> >>- E&Y Brazil
<rqou> >>
<rqou> >>The following isn't trusted by Mozilla anymore:
<rqou> >>- E&Y Hong Kong
<rqou> >>
<rqou> >>This seems to be a worrying trend to me.
<rqou> >
<rqou> > It's OK, E&Y have offices in 150 countries, it'll take years to go through them all.
<lain> lool
<lain> also <3 greentext filter thing for weechat
<whitequark> lol opensrs
<whitequark> I might consider that
<whitequark> rqou: poke
<rqou> yesA?
<rqou> oh about your osx question that i missed?
<whitequark> yeah
<rqou> idk why, but occasionally passing "-v" in the boot args makes osx not throw the "prohibited sign"
<rqou> i think it forces caches to get rebuilt
<rqou> but in general i'm currently using qemu+ovmf+clover with all the acpi hacks turned off in clover
<whitequark> oh no that wasn't the thing I'm asking
<whitequark> I got it installed
<whitequark> or rather, I got the installer to finish and reboot
<whitequark> now whatever it installed doesn't boot
<rqou> how about using "-v"?
<whitequark> nothing helps
<whitequark> well to begin with there isn't a kernel on that partition
<rqou> hrm
<rqou> oh right, the way clover works is imho f*cked
<whitequark> if I extract the kernel from Essentials.pkg then it resets the system somewhere very early in boot
<rqou> so i used a real mac to install clover to a tiny (100mb or so) disk image
<rqou> for the installer i set that as the boot drive
<rqou> with the osx installer as a cd drive
<rqou> after it finishes installing i keep the clover 100mb thing as the boot drive
<whitequark> if I try to boot the prelinkedkernel I think the same happens?
<whitequark> mhm
<rqou> this boots into the real installation
<rqou> and install it on the real installation partition
<rqou> from there you can download the clover installer again
<rqou> and then you can remove the 100mb clover image
<rqou> so afaik how the boot process works now is
<rqou> qemu boots ovmf
<rqou> ovmf boots clover, which is a differently-hacked edk2
<rqou> (so this part is stupid)
<rqou> clover loads an hfs+ driver into efi
<rqou> it finds the osx bootloader boot.efi
_whitelogger has quit [Ping timeout: 240 seconds]
_whitelogger has joined ##openfpga
<whitequark> and the linux applesmc driver *doesn't* implement the osk-getting capability afaik
<whitequark> so they have disjoint sets of functionality
<rqou> the smc is supposed to return temperature and/or voltage stuff
<rqou> so clover has yet another mode where it can boot from seabios
<rqou> but i've never tested it
<rqou> it has different bugs
<rqou> clover is some weird duct-tape of edk2+refit
<rqou> and the code is a giant mess to the point where i can't tell what is what
<rqou> i am simultaneously amazed and disappointed
<rqou> for extra lulz, iirc the linux applesmc driver and osx use different mechanisms to discover if the smc exists
<rqou> afaik linux requires dmi to be set correctly
<rqou> also, different macs have _different_ SMCs
<rqou> the smc also gets firmware upgrades
<rqou> macs are weird
pie_ has quit [Ping timeout: 268 seconds]
<rqou> whitequark: oh i forgot
<rqou> you also need osxaptiofixdrv2
<rqou> don't ask me why
<rqou> as far as i understood what it was doing, this driver just allocates/frees/moves some efi memory arbitrarily
<rqou> this is even more extra bonus stupid because in qemu the memory map can be whatever you want
<rqou> but apparently whatever this thing is "fixing" in ami aptio seems to originate from edk2
<lain> phone almost had me worried
<lain> ssh client updated and started throwing warnings about a host key change on my irc jail
<lain> buuut it was just: ssh client app added support for ecdsa, and hadn't previously cached the ecdsa key, so it considered that a "change"
<qu1j0t3> x_x
<rqou> yeah i was about to guess that
<rqou> client learned how to do djb-crypto :P
<whitequark> oookay let's see
<rqou> lain: you're going to be doing uefi stuff at some point right? you should join hackintoshing :P
<lain> rqou: just talked to mike, he's gonna run the dish up the tower this week :D
<whitequark> "boot0ss: error" well that's encouraging
<rqou> wait is this with clover?
<lain> rqou: lol, maybe if I have a need to deploy osx software and don't want to invest in an actual mac
<rqou> you need to change some sooper sekrit options to make it install in efi mode
<rqou> otherwise it installs in bios mode and does different stuff
<whitequark> rqou: that's the MBR that came with clover
<whitequark> I think
<rqou> yeah so clover has an "mbr" version and a "uefi" version that are different
<whitequark> what
<rqou> on one of the pages of the installer (either right before or right after selecting the target disk) there is an options button in the bottom left
<rqou> and then there is an "install for uefi booting" option that then proceeds to completely change how the installer behaves
<whitequark> the install log says "Clover EFI installation"
<whitequark> wait
<whitequark> EFI not UEFI
<rqou> um idk
<rqou> did you expect this code to be sane? :P
<whitequark> trash fiiiiire
<rqou> i wonder how qemu officially tests this?
<rqou> especially -device isa-applesmc
<lain> test in prod #yolo
<rqou> even testing in prod
<rqou> the linux driver doesn't bind to the qemu device correctly
<rqou> osx is a giant pile of hacks that qemu doesn't natively boot
<rqou> so i don't know what is being used to test exactly?
<whitequark> ok hm
<whitequark> I don't think they test it
<rqou> afaik qemu's code quality could use improvement
<rqou> much better than hackintosh crap of course
<rqou> iirc landley had a great rant about how qemu's arm versatile emulation was very broken to the point it caused the linux kernel to also be broken
<rqou> because "versatile" essentially got co-opted to mean "qemu" because a) nobody has *real* arm versatile boards anymore and b) versatile has PCI
<whitequark> rqou: hm, clover doesn't work
<whitequark> I just get a black screen
<whitequark> how do I make it at least explain why it fails
<rqou> after clover's boot menu?
<whitequark> yes
<rqou> one of the options in the bottom row lets you pass "-v" to the osx kernel
<rqou> but various pieces of the stack that know about screen resolution need to agree or the screen will be garbled
<rqou> when i use it through virt-manager somehow qemu+virt-manager+clover pick a screen resolution
<rqou> but ovmf has a different idea?
<rqou> anyways, you can press f2 on the virtual tianocore logo to change its virtual screen resolution to match what clover thinks it is
<whitequark> nope, just black screen
<whitequark> after adding -v
<rqou> lain: how are you enjoying this discussion of this shitshow? :P
<rqou> er, try disabling all of the acpi hacks?
<rqou> basically in the same menu there is a whole crapton of "enable this hack"
<rqou> try turning all of them off
<whitequark> nope, black screen
<whitequark> immediately after trying to boot, anyhow
<rqou> hrm, idk
<rqou> is this still your nested qemu?
<whitequark> no, just a normal qemu
<whitequark> hang on
<rqou> did you include osxaptiofixdrv2.efi like i mentioned?
<whitequark> yes, I did
<whitequark> at leasst I checked it
<rqou> well, i'm out of ideas
<rqou> try an older/newer qemu :P :P
<rqou> my two qemus on debian sid and ubuntu xenial behave slightly differently
<rqou> hrm, you also need to tell the kvm host module to ignore unknown msrs if you're using kvm
<rqou> if you're not using kvm, all bets are off
<rqou> iirc landley encountered a f*ckton of fpu/sse bugs in qemu (non-kvm mode)
<whitequark> well the installer booted at least
<whitequark> without kvm
<whitequark> well
<whitequark> I found some random config.plist file for clover
<whitequark> now it "boots"
<whitequark> the graphics is garbled
<rqou> yeah, you need to f2 in ovmf and change the resolution
<rqou> and if it's running at lightning speed you need to recompile clover
<whitequark> christ
<whitequark> ok what was the darwin key
<whitequark> rqou: can you just give me your config.plist
<whitequark> rqou: also, your advice doesn't help
<whitequark> no matter what I change in ovmf, clover sets it to 1920x1080
<whitequark> and then darwin promptly barfs all over
<rqou> for me somehow if i change ovmf to match clover darwin no longer barfs
<rqou> i have no idea why
<rqou> or wait
pie_ has joined ##openfpga
<rqou> i'm using qxl
<whitequark> oh
<whitequark> I use stdvga
<rqou> maybe that's related?
<rqou> this code really is a trash fire
<whitequark> what the fuck
<whitequark> that resulted in an uh
<whitequark> a 2048x2048 display
<whitequark> with an apple that has rendered correctly
<whitequark> but not the progressbar
<whitequark> actually
<rqou> wtf
<whitequark> something fucky happend and now config.plist is just empty
<whitequark> fsck says it's fine
<whitequark> oh, hm
<whitequark> what the ever loving fuck
<lain> XD
<whitequark> /EFI/CLOVER/config.plist Contains a free cluster (43). Assuming EOF.
<whitequark> /EFI/CLOVER/config.plist File size is 3155 bytes, cluster chain length is 0 bytes.
<whitequark> does it remove its own config?!
<whitequark> because I bet that's what happens
<pie_> the fuck?
<whitequark> actually
<whitequark> that happens even before clover.
<whitequark> OVMF does that.
<whitequark> oh.
<whitequark> I see.
<whitequark> that was some sort of cache-related thing with qemu-nbd and linux fs layer
<whitequark> no idea what it is
<whitequark> unmounting a filesystem might not have committed everything to the nbd device?
<whitequark> rqou: oh my god
<whitequark> so I figured out how to boot it
<whitequark> you have to go into OVMF and set the resolution
<whitequark> but you have to do it on EVERY boot and you have to use the "Reset" option in OVMF
<whitequark> it does save its garbage in NVRAM *but* it doesn't affect the next cold boot
<whitequark> for some goddamn inane reason
promach has quit [Quit: Leaving]
Bike has joined ##openfpga
amclain has joined ##openfpga
mifune has joined ##openfpga
mifune has joined ##openfpga
massi has quit [Quit: Leaving]
azonenberg_work has joined ##openfpga
mifune has quit [Ping timeout: 240 seconds]
talsit has joined ##openfpga
Hootch has quit [Read error: Connection reset by peer]
Hootch has joined ##openfpga
<nats`> so it exists....
<nats`> there is ONE guy on this planet who is too dirty for breitbart !
GenTooMan has quit [Ping timeout: 260 seconds]
<azonenberg_work> nats`: lol
GenTooMan has joined ##openfpga
Hootch has quit [Quit: Leaving]
GenTooMan has quit [Ping timeout: 240 seconds]
pie_ has quit [Changing host]
pie_ has joined ##openfpga
GenTooMan has joined ##openfpga
<pie_> ugh, my surface 3's microusb port is loose and it looks impossible to disasseble without breaking the screen glass by acccident, fucking adhesives, apparently it can charge with something called "dockport", you guys know nay adapters by any chance? seems obscure as fck... https://www.reddit.com/r/Surface/comments/56ww5c/surface_3_non_pro_broken_micro_usb_port/
<pie_> *any
<pie_> m$ sells a dock for 200$ -.-
<whitequark> pie_: you can remove the glass first by heating it
<whitequark> there are specialized tools. or hot air guns work. or I think you could get one of those hand warmers.
<whitequark> (but never tried the latter)
<pie_> im not very well equipped
<nats`> it's risky
<pie_> and id need to solve this relatively soon. that too @ nats`
<pie_> hungary
<whitequark> which country are you in pie_
<nats`> did that on an iPod nano.... it died :D
<whitequark> do you not have guys in small shops doing all sorts of phone repairs?
<whitequark> I've seen them successfully do BGA reballing in a 2m^2 space...
<pie_> welll i guess i could give that a shot
<pie_> would mean no machine for a couple days probs :(
<whitequark> doubt it
<whitequark> that sounds like a one hour fix at most
<pie_> i guess
<whitequark> resoldering it is a piece of pie
<whitequark> microusb will be in stock
<whitequark> er, cake
<pie_> not that i do anything sensitive but not using FDE makes me uncomfortable
<pie_> lol
<pie_> good point
<pie_> i guess i should look into it tomorrow
<nats`> can I ask why you bought a surface ?
<whitequark> they needed a surface... to cut vegetables on
<nats`> :p
<whitequark> I did that with my Nexus 7 (2013)
<whitequark> thats the only thing it was uitable for
<whitequark> suitable
<whitequark> actually it excels at that
<pie_> nats`, i got it as a gift
<nats`> ok
<pie_> its not bad i just wish it had proper linux support
<pie_> :P
<nats`> I always prefere real desktop computer
<pie_> also microusb charger :I
<pie_> sure me to but i walk back and forth between the univ all day
<pie_> so its surprisingly convenient
<nats`> ah yes :)
<nats`> for that I have my x220 :D
<whitequark> >convenient
<whitequark> >x220
<pie_> i dont really have the money to spend on a good machine, though i probably should sometime...but anyway this works rather well, minus the windows reinstall that it needs because its buggy as FUCK
<nats`> in fact x220 is an old lenovo
<nats`> with a core i5
<nats`> it oftens cost about 100 euro on ebay :D
<nats`> but it's pretty powerfull machine
<pie_> huh.
<pie_> funny thing is when i count them i have too many gadgets, all gifts, i just wish i had one good, durable laptop lol
<pie_> this guy was fine but idk wtf made the microusb bork
<pie_> ugh, but i hate this construction with adhesives bs
<pie_> though at least its not clips
<pie_> i cant decide which is worse
<nats`> 351986964985 <= check that on ebay
<nats`> adhesive for sure
<pie_> i always break clips
<nats`> when it's adhesive you easily break the screen :po
<pie_> i have one item listed "Lenovo Thinkpad X220 Convertible Laptop i7-2640M 2.80GHz WEBCAM 12.5" for 150$ (not counting shipping from the us to EU and taxes...so probably more like 200$
<nats`> look in UK :)
<nats`> there are a lot of x220
<nats`> one model is a tablet and the other one a laptop
<nats`> mine is a laptop
<pie_> oh
<nats`> in really good shape for 120 euro
<nats`> little low on ram
<nats`> anyway it's a machine I like :)
<pie_> ok. ill consider it i guess
<nats`> IIRC it's one of the most recent supporting libreboot but check to be sure
<pie_> libreboot is done by a caustic sjw though isnt it?
<nats`> yep
<nats`> this one
<nats`> I don't remember the name of the first one
<pie_> no, i dont care that much thanks :P
<nats`> which support bloated firmware
<pie_> id rather not
<pie_> wow, cant find a single minidisplayport charger adapter thing
<pie_> yea
<pie_> huh, amazon apparently has the official dock for the measly price of 35 + 35 shipping! woo! $39.99 + $33.34
<pie_> -.-
<nats`> I'm always suspicious about "compatible"
bibor has quit [Quit: WeeChat 1.0.1]
<pie_> its the official m$ one afaict
<pie_> "Microsoft Docking Station for Surface 3 (not compatible with Surface Pro 3) SC EN/XD/ES Hdwr (GJ3-00001)"
<pie_> i could buy a textbook for that price
<lain> lol
<balrog> pie_: heh, remember when GnuTLS split?
<pie_> no i dont >.>
<balrog> err
<balrog> apparently that led to the sed maintainer quitting
<pie_> said dock sells for 80 eur in europe :I so probably about the same or less after taxes
<pie_> ugh
<pie_> im not sure if i want to read that right now heh
<pie_> frustrated enough
<balrog> :/
<balrog> basically: RMS is often an asshole
<balrog> (film at 11)
<mtp> rms is suuuuuch an embarrasment
<pie_> so i dropped my charger down the stairs in the library
<pie_> seems to get hotter than it used to...
<pie_> hope its not shorted but it still seems to suply power and doesnt blow any fuses
<pie_> no obvious way to disassemble
<pie_> fuck me
<pie_> and now charging seems stuck at 6% for whatever reason :I
<pie_> though i just rememebred i have a tablet thats smaller but would work while this one gets repaired i guess
<pie_> mr tinyscreen
<pie_> f****ck plz charge
<pie_> if it dies i might make some progress with my books though heh
<cr1901_modern> pie_: The libreboot situation is because a FSF member was stalked by another member who didn't like her (b/c she's trans I believe). She reported it and got fired in retaliation.
<cr1901_modern> Unfortunately, this is actually fairly common in the US.
<nats`> uhhmmmm cr1901_modern...
<pie_> could be
<nats`> you provide a lot of proof like leah did
<nats`> I really think there is a lot of overreacting here
* nats` wonder if it's a good idea to use CH340
<whitequark> 'enough proof': just a bit more proof than I have right now
<cr1901_modern> nats`: There is no proof. So I have two options:
<cr1901_modern> 2. Examine my priors, and say to myself "huh, I've seen a lot of people who don't know each other say the same thing- they got fired as retaliation. Why could that be"?
<cr1901_modern> 1. I demand evidence that isn't going to show up b/c charitably that's private info I'm not privy to, or uncharitably such evidence doesn't exist.
<rah> just in case anyone's interested..
<nats`> cr1901_modern there are like 900000000 other reason to be fired
<nats`> it reminds me of some people reasoning like that about disabled people..... Yes disabled people can be dumbass too !
<nats`> anyway someone here used ch340
<nats`> I wonder how much trust I can have int hat chip
wpwrak has quit [Ping timeout: 240 seconds]
digshadow has quit [Ping timeout: 260 seconds]
digshadow has joined ##openfpga
bibor has joined ##openfpga
felix___ has joined ##openfpga
felix_ has quit [Read error: Connection reset by peer]
felix___ is now known as felix_