<nekobot>
• OscarL (107b3976): git: update to version 2.50.1.…
janking has joined #haiku
jibsaramnim has quit [Read error: Connection reset by peer]
jibsaramnim has joined #haiku
janking has quit []
OscarL has joined #haiku
chilledfrogs has quit [Quit: connection reset by purr]
<OscarL>
Begasus[m]: in case you read the logs while I'm asleep... left some notes about the consistency report here: https://bpa.st/BVKA (short version, rebuilding openexr-2.4.1 with newer cmake will clear at least 8 packages in one go from the report)
<OscarL>
added some notes on what I think needs to be done to clear each type of issue on the reports (left only the directly affected packages, removed the ones that should clear up automatically once those are fixed).
<OscarL>
did fixed a couple myself (not many, but, hey... worst is nothing :-D). Also, merged git after testing a build on 32 bits.
<OscarL>
Now... off to sleep. See ya!
OscarL has quit [Quit: zzzZZZzzz]
janking has joined #haiku
<janking>
:)
chilledfrogs has joined #haiku
mmu_man has quit [Ping timeout: 480 seconds]
janking has quit [Quit: Vision[]: i've been blurred!]
janking has joined #haiku
janking has quit [Quit: Vision[]: i've been blurred!]
janking has joined #haiku
janking has quit [Quit: Vision[]: i've been blurred!]
AlaskanEmily has quit [Remote host closed the connection]
janking has joined #haiku
janking has quit []
janking has joined #haiku
janking has quit [Quit: Vision[]: i've been blurred!]
<Anarchos>
Begasus[m] now i am trying to find in the official doc of my intel_extreme graphics chip how to set brightness
<Anarchos>
it seems i have an old card from 2011 and the driver mixes the register for backlight control
<Begasus[m]>
nice Anarchos we all could benifit :)
<Begasus[m]>
now if someone would do some work on Radeon drivers :)
<Anarchos>
Begasus[m] i helped nephele for the radeon_hd brighness
janking has quit [Ping timeout: 480 seconds]
<Begasus[m]>
at least someone (and a few others) that are helped out :)
janking has joined #haiku
<Anarchos>
Begasus[m] time to go back to work
Anarchos has quit [Quit: Vision[]: i've been blurred!]
mmu_man has joined #haiku
nephele has joined #haiku
MisthaLu has quit [Quit: Leaving]
xet7 has quit [Remote host closed the connection]
<nephele>
Habbie: do you know of any good guides for self signed certs for mail, my head is spinning with all these guides and commands (and people offering "cool" unrelated solutions... xD)
xet7 has joined #haiku
nephele has quit [Quit: Vision[]: i've been blurred!]
<nekobot>
• Begasus (10c090b7): fastfetch, bump to 2.49.0 (#12714)
<Begasus[m]>
with dbus fixed for that ... :)
<OscarL>
Begasus[m]: silly /me went: "how hard can it be to patch openexr3.2 to make it work with gcc2" (for the exrtranslator). wasn't expecting a 17 MB+ tarball, only to find too late it requires gcc2 libdeflate :-D
<OscarL>
Maybe it makes more sense to just drop the gcc2 version of the exrtranslator. I'm not sure the use case of "needs to use exr formated images in a gcc2 Haiku application" is worth of having to keep that old openexr version "alive".
<Begasus[m]>
nah!
<Begasus[m]>
that one is still required for 32bit translator :)
<OscarL>
(the x86 one still would work for "modern" apps)
<Begasus[m]>
yeah, have been checking that a lot on the changes to move the translator out of the Haiku tree
<OscarL>
how many users reallistically can't live without a gcc2 exr translator? not that hard to use the x86 translator to turn that into some other format, if they *really* need to work with gcc2 apps.
<Begasus[m]>
so unless the gcc2 version isn't ready for the newer EXR translator we should keep the old version
<Begasus[m]>
maybe create a ticket for that at the bugtracker?
<Begasus[m]>
I'm no Haiku dev :)
<OscarL>
I understand the technicailities, but don't agree with it being worth it. Yes, will probably open a ticket about it to see what other's opinions are in the matter.
<Begasus[m]>
+1
<OscarL>
welp... exrtranslator ain't a Haiku issue anymore! :-)
<OscarL>
(wonder who we can blame for that? :-P)
<Begasus[m]>
well the translator moved to HaikuArchives
<OscarL>
(j/k, was a good move)
<Begasus[m]>
at least something I had a small part in for Haiku :P
<OscarL>
the ticket might be more general, as in... can we try to consolidate openexr versions? (we had 5 untill just a few minutes ago).
<Begasus[m]>
haven't checked repology "yet" to see if there is a newer version :P
<OscarL>
oof.
<Begasus[m]>
heh
<Begasus[m]>
python, poppler, boost, ICU .... all with different versions, it's good to keep up on newer version, the ones in use should then graduatly be updated on recipes using them
<OscarL>
these are my notes while looking around openexr last night: https://bpa.st/LJIA
<Begasus[m]>
updating recipes using openexr24 can't always be easely be moved to openexr3
<OscarL>
don't mind multiple versions, so much as I dislike having depedencies on them grow too "wide".
<Begasus[m]>
there are already plenty patches around to prove that :)
<Begasus[m]>
on updates those "should" be checked for working with newer openexr, no new version has been added for a while, so safe to assume our current ones are stable
<OscarL>
re: 2.4 -> 3+.... good to know. thus why I put that with a question mark in the notes :-)
<Begasus[m]>
poppler releases 3 to 4 versions each year, not compatible with each other, so when I hit those I try to move them to our latest version in tree
<Begasus[m]>
OscarL: I wouldn't worry too much on those for now :) (should look at what the issue is with those on the consistency report)
<Begasus[m]>
fastfetch passed CI checks upstream :)
<OscarL>
rebuilding openexr24 against newer cmake should clear most(all?) the openexr related entries in the report.
<OscarL>
seems some recipes have: "cmd:cmake >= 3", and that causes them to appear in the report (as there's no such cmake 3.x .hpkg for install now)
<waddlesplash>
OscarL: does anything even use that v8 recipe? it might be able to be dropped
<waddlesplash>
nodejs compiles its own for example
<OscarL>
(at least that was what I understood last night)
<OscarL>
waddlesplash: I'm always happy to swing the axe! :-)
<OscarL>
will double check.
<Begasus[m]>
+1
nipos has left #haiku [Disconnected: Received SIGTERM]
<Begasus[m]>
OscarL: it's the same with doxygen, now there's only 1 version
<OscarL>
"libresprite" requires libv8.
nipos has joined #haiku
<Begasus[m]>
calligraplan hasn't seen any recent updates
<Begasus[m]>
ktorrent* can be updated on next gear
<OscarL>
ah.. that's "libre-sprite" and not a lib-resprite, heh-
<OscarL>
(provides cmd:libresprite)
<Begasus[m]>
I think that v8 recipes has been broken (build) for a long time
<Begasus[m]>
-s
<Begasus[m]>
k, patch is still good for cmake
<OscarL>
libresprite --> "animated sprite editor & pixel art editor"
<Begasus[m]>
maybe I should try to upstream some of that
<Begasus[m]>
rings a bell here OscarL although not for what :P
<Begasus[m]>
k, cmake build is running, biab
bbjimmy has quit [Quit: Vision[]: i've been blurred!]
bbjimmy has joined #haiku
<Begasus[m]>
re
<Begasus[m]>
OscarL: jmairboeck does a way better job on that openexr thingy :)
nipos has left #haiku [Error from remote client]
<Begasus[m]>
I had to use a secondary arch app to check that for the _x86 translator
<OscarL>
I still think it is a too niche of a use-case for us to keep that old lib around, but... won't push it too much if everyone disagrees with me.
tuaris has joined #haiku
<Begasus[m]>
requires "lib:libfltk_x86 >= 1.3.9" of package "lmms_x86-1.2.2-4" could not be resolved ... still an issue atm
<Begasus[m]>
ps OscarL , I won't fight on it too :)
<Begasus[m]>
maybe tag korli in the issue
<Begasus[m]>
he was pretty much involved and aiding on the move
<OscarL>
you want me to get yelled at, don't ya? :-P
<Begasus[m]>
lol
* OscarL
was trying to be sneaky about this!
<Begasus[m]>
korli doesn't yell, he's a good guy! :P
<Begasus[m]>
normaly I wouldn't fight at all, but in this case it touches some "core" element in Haiku, a Translator
<Begasus[m]>
personaly I don't use thumbnail view in Tracker, so I'm not hurt by it :)
<OscarL>
given that was dropped from Haiku already...
<Begasus[m]>
you get me drift :)
<OscarL>
any way... back to consistency report here :-)
<Begasus[m]>
+1
<Begasus[m]>
tackling opencascade here
<Begasus[m]>
lib:libtcl8.6 this could be tricky :/
<OscarL>
time in 32 bits is an easy fix (missing $secondaryArchSuffix in the defineDebugInfoPackage.
<Begasus[m]>
nice spot!
<Begasus[m]>
s/spot/catch :)
<Begasus[m]>
for now focussing on 64bit
<Begasus[m]>
Error: problem 1: package tk_devel-8.6.10-4 conflicts with xlibe_devel provided by xlibe_devel-0.3.3-1
<Begasus[m]>
Fetching package for devel:libx11 ...
<Begasus[m]>
bleh ..
<Begasus[m]>
have been fighting with this not that long ago, gave up then ...
<Begasus[m]>
ah, I guess it was this trying to enable TMM locally
Aedil has joined #haiku
<OscarL>
these gnu autotool checks are insane. takes 10x longer for the checks than to compile the damn /bin/time.
<Begasus[m]>
heh
<OscarL>
(which I bet doesn't use most of the checked stuff anyway)
<Begasus[m]>
I'll try a few days here with new cmake before pushing it
<Begasus[m]>
ow ps, that was TBB earlier, not TMM :)
cptaffe has quit [Ping timeout: 480 seconds]
xet7 has quit [Remote host closed the connection]
<OscarL>
mmm, something still off with time in 32 bits. how come "hp time_x86" gives me "Error: time_x86 not found in repository", but "hp time" works, when the time .recipe has "?x86_gcc2"
<OscarL>
wth?
<Begasus[m]>
always fun those sci-libs ..
xet7 has joined #haiku
<Begasus[m]>
hp time_x86 launches here
nipos has joined #haiku
<OscarL>
nevermind, I'm an idiot.
<Begasus[m]>
lol
<Begasus[m]>
not arguing on that ...
* Begasus[m]
ducks
<OscarL>
"rebooted into 32 bits..." turns out that I hit enter in the wrong BootMan menu entry, and just booted right into 64 bits again :-D
* OscarL
will change the background color in 32, to avoid this stupid mistake again.
<Begasus[m]>
rofl
<Begasus[m]>
OscarL: for those ffmpeg issues, I try to use ffmpegn${secondaryArchSuffix}_devel these days
<Begasus[m]>
saves the hastle on those libversions
<OscarL>
sounds good (even more so if there's a short comment about why using that). +1
<Begasus[m]>
although I'm going against my own advice to always use version strings where needed :D
<Begasus[m]>
but I guess for some we can make exeptions
<OscarL>
"practicality beats purity"
<Begasus[m]>
in this case it is :) (seeing those devel packages outside of ffmpeg6 only contain those) :P
<Habbie>
no 802.1x support in our wifi dialog it seems
<Begasus[m]>
fix it? :)
<Habbie>
sadly there's also open wifi here so i am in no rush ;)
<Begasus[m]>
:)
<Begasus[m]>
bah ... /sources/OCCT-7_8_1/src/StdPrs/StdPrs_BRepFont.cxx:460:30: error: invalid conversion from 'unsigned char*' to 'const char*' [-fpermissive]
<OscarL>
one would hope those would use shallow clones for the sub-modules, but "nah!, let's just waste resources just because" seems to be more common in many such cases :-(
<nekobot>
• OscarL (d56e712d): dbuspython: drop unneeded recipe.…
mmu_man has joined #haiku
<Begasus[m]>
afk for a bit, this digikam build will take some time
Anarchos has joined #haiku
<Anarchos>
are you there PulkoMandy ?
AD_Haiku_onPC_ has joined #haiku
AD_Haiku_onPC has quit [Read error: Connection reset by peer]
andreaa has joined #haiku
BrunoSpr has joined #haiku
<BrunoSpr>
hello all, is an update ok atm? Nightly 64bit
andreaa has left #haiku [#haiku]
<Anarchos>
let me reboot and i tell you
Anarchos has quit [Quit: Vision[]: i've been blurred!]
Anarchos has joined #haiku
<Anarchos>
BrunoSpr yes it seems :)
<BrunoSpr>
Anarchos, ah ok thank you
<Anarchos>
how can i compare two patchsets from a PR on gerrit ?
<BrunoSpr>
see you later
BrunoSpr has quit [Quit: Vision[]: Ich wurde eingeweicht!]
<OscarL>
would be nice if one could click on a path in Terminal and open a new tab with cwd set to said path. (similar to how the "cmd+click to open in Tracker" works)
<OscarL>
Anarchos: you can select the "diff" link in some comments, or choose specific versions of the patchset to compare (under the "FILES" header... notice the "Base -> Pachset" thingie)
<Anarchos>
ah yes i found, thanks
<OscarL>
** "view diff".
<OscarL>
no problem :-)
HaikuUser has joined #haiku
HaikuUser is now known as nukacal
<nukacal>
Just wondering, I have an Intel HD Graphics 530. It appears to be fully compatible with Haiku according to other user reports. Any hints why does not work for me? I gotta boot in safe boot
zardshard has left #haiku [Error from remote client]
<PulkoMandy>
Different wiring of the display? Or my machine the hdmi output doesn't work, but a displayport to hdmi adapter on the displayport output does
zardshard has joined #haiku
<PulkoMandy>
Our driver doesn't handle all situations yet
<Anarchos>
i have an ivy bridge intel_extreme graphics card. It boots fine, except the screen is black
<Anarchos>
i think it is due to the backlight_register being on the north display engine registers (Gen 7.0) contrary to what is done for haswell (gen 7.5)
<Anarchos>
PulkoMandy i saw you commit some things for the haswell card recently :)
<nukacal>
That's interesting. This PC does have display port output, gotta check for adapters since this monitor is HDMI only
<Anarchos>
haswell is northdisplay engine too. (gen 7.5) but sandybridge is south (gen 6.0)
<PulkoMandy>
Anarchos: Yes, registers often move around. Maybe we are missing some cases or we are setting them for the wrong generations of cards
<OscarL>
nukacal: I see device ID 191B in the driver, but wikipedia lists also device id 191B as "Intel HD 530", and I can't find 0x191B in the intel_extreme driver.cpp file.
<OscarL>
nukacal: running "listdev" on the command line should show the actual device id values, if curious. ("listdev | grep VGA" might be easier still)
<Anarchos>
I use « listdev |grep -n "Graphics" »
<nukacal>
OscarL: Let's try that out
<nukacal>
8: device 1912: HD Graphics 530
<OscarL>
(at least that 530 has separate IDs for mobile vs desktop, not like this HD 600 :-D)
<OscarL>
nukacal: welp, that saves some testing :-D
<nukacal>
hahah, it does not seem to be the variant B, 0x191B
<nukacal>
I used your command, Anarchos. Pretty useful, thanks!
xet7_ has joined #haiku
AlienSoldier has joined #haiku
xet7 has quit [Ping timeout: 480 seconds]
<OscarL>
ah! too bad that page has .pdf only back to Gen 4.0. (need to hunt down some Gen 3 docs in case I get bored and want to give a go at testing things on my GMA 3150)
<OscarL>
"This section describes Audio features." (from the Kaby Lake docs). Cool! Now I can spend a lot of time reading that and folling myself, pretending I might make a dent in adding "hdmi audio" to this thing! :-P
dalme has joined #haiku
<nukacal>
That would be powerful!
<OscarL>
too bad my most complex driver so far is for just reading temps/fan/voltages :-D
<Anarchos>
OscarL what is the model you look for ?
<OscarL>
Anarchos: I have a couple of Atom N450 and N455 based netbooks (both use the Intel GMA3150 iGPU). I barely use them these days (too slow for HaikuPorts work)... and the ones that still hav the 1024x600 screen work OK, with some "caveats": https://dev.haiku-os.org/ticket/19041
<OscarL>
so while having Gen 3 docs for them would be cool... doubt I'd really spend much time with them, other than "reading specs just because I'm bored", as I tend to do sometimes, LOL.
<Anarchos>
OscarL if you find me the doc i could maybe try
<OscarL>
alright! will keep that in mind!
xet7_ has quit []
<Anarchos>
OscarL for now i am building a driver for my card to trace the address of register used for backlight, to confirm it uses the wrong one
<OscarL>
wishing you the best of luck on that front, Anarchos. Working backlight control is just so darn useful.
* OscarL
looks with disdain a Lenovo C255 AIO that only allows to lower the brigness just a bit :-( (at least it has hardware buttons for it)
OrangeBomb has quit [Quit: Slacking off]
nukacal has quit [Remote host closed the connection]
HaikuUser has joined #haiku
<OscarL>
Mmm, so there's "PCH backlight control" *and* "CPU backlight control" (at least on Gen9.5). groovy.
HaikuUser has quit []
<Anarchos>
OscarL i have a N450 laptop whose i can control the brightness
<OscarL>
I can in mine too, via "hardware keys" that only work while on the bootman menu ("good enough" for my usage at least: setting it aaall the way down).
<OscarL>
I do have some old source code for a linux driver that allows to interact with those keys (are *not* exactly acpi events, IIRC).
<Anarchos>
where to put intel_extreme.accelerant in config/non-packaged ?
<OscarL>
Never got around to figure out if it could be ported/re-written for Haiku.
<Anarchos>
OscarL i can change brightness from Screen
<OscarL>
nice :-)
<Anarchos>
OscarL let me find you my exact model of graphics card
<OscarL>
mine should be 0xA011 according to Wikipedia.
<OscarL>
right: "device a011: Atom Processor D4xx/D5xx/N4xx/N5xx Integrated Graphics Controller"
<Anarchos>
OscarL i have a011 AND a012 (wtf !?)
<OscarL>
Anarchos: :-D yes, those machines report two VGA devices for some reason.
<Anarchos>
OscarL "listimage kernel" show me intel_extreme and vesa addons being loaded
<OscarL>
first is "Display controller (VGA compatible controller, VGA controller)", the other is just "Display controller"
<Anarchos>
OscarL same for me
<OscarL>
I guess different "wiring" in the motherboard (or different BIOS firmware) explains why you can use brightness control, while I can't.
<Anarchos>
i have only "mode 1024 765 32" in /config/settings/kernel/drivers/vesa
<OscarL>
I kinda assumed one device was for the internal display, and the other for the VGA output. not sure if that's actually the case.
<Anarchos>
OscarL i thought to that too
<Begasus[m]>
grabbing digikam-8.7.0-1-x86_64.hpkg and moving it to /Opslag/haikuports/packages/digikam-8.7.0-1-x86_64.hpkg
<Begasus[m]>
so far so good :)
<Anarchos>
OscarL i once encountered a weird update which forces me to erase old settings file (can't remember which app it was). Too bad pkgman update doesn't deal with old settings files !
<OscarL>
Anarchos: newer netbooks (the ones with a Celeron N2008 for example, only show one device, IIRC, same with the PowerVR based Atom N2600)
<Anarchos>
OscarL mine is a Fizzbook Zoostorm
<Anarchos>
OscarL that's why there is a "Fizzbook Zoostorm" key map in Keymap :)
<Anarchos>
"Fizzbook NL2" in Keymap
* Anarchos
did it.
<OscarL>
beats havnig a Windows one, I guess? :-D (this netbook has a "Huayra" key instead of a Win one... Huayra being the name of an Argentine Linux distro).
mmu_man has quit [Ping timeout: 480 seconds]
<Anarchos>
OscarL it was a windows one,, but i reset the disk to be haiku only !
<Anarchos>
OscarL that's the netbook i am porting the linux accelerometer driver on.
<OscarL>
nice, albeit Haiku make those N450 Atoms run too hot for me.
<OscarL>
(we miss the equivalent to Linux's "acpi_cpufreq" driver, so those poor CPUs can catch a break)
<Anarchos>
should i put my own intel_extreme.accelerant in "/boot/home/config/non-packaged/add-ons/accelerants" ? This folder doesn't exit,.
Anarchos has quit [Quit: Vision[]: i've been blurred!]
<OscarL>
that should work, I guess (assuming you're don't need a modified intel_extreme driver too)
<OscarL>
too late.
* OscarL
did tried some .accelerant mods trying to enable video overlay on the GMA3150... to no success.
jmairboeck has joined #haiku
mmu_man has joined #haiku
Anarchos has joined #haiku
Anarchos has quit []
Anarchos has joined #haiku
<Anarchos>
i am back
<scanty>
hooray!
<OscarL>
Begasus[m]: that `ceres_solver-2.0.1~git-1: requires "lib:libamd"` on 32 bits is odd, as ceres_solver is "!x86_gcc2", so no idea from where that comes from, heh.
<Anarchos>
OscarL git blame ?
<OscarL>
Anarchos: won't help, issue is not with those lines, but what HaikuPorter makes up from them.
dalme has quit []
<OscarL>
hey there, scanty, btw :-)
<OscarL>
we shouldn't be seeing "ceres_solver" at all there (at worst, it should be "ceres_solver_x86")
<Begasus[m]>
will have a look tomorrow OscarL , if you don't beat me to it :)
<Begasus[m]>
closing down here
<Begasus[m]>
cu peeps!
<OscarL>
see you!
jmairboeck has quit [Read error: Connection reset by peer]
<OscarL>
if the issue turns out to be because it has ARCHITECTURES="!x86_gcc2 all" instead of the more common "all !x86_gcc2"... I'll eat my hat.
<scanty>
i just found an old usb thumb drive, and on it was a bootable r1beta2 install.
<scanty>
back from 2020
* OscarL
starts putting some salt on his hat... as changing "!x86_gcc2 all" for "all !x86_gcc2" seems to actually change things. sigh.
<OscarL>
I'm on 32 bits now, if ceres_solver has ARCHITECTURES="!x86_gcc2 all", then "hp ceres_solver" starts trying to build the recipe (it shouldn't).
<OscarL>
if I change that to ARCHITECTURES="all !x86_gcc2", then "hp ceres_solver" fails as intended, as the recipe is marked broken.
<OscarL>
we have 2001 .recipes with "all !x86_gcc2", vs only 4 with "!x86_gcc2 all"
Aedil has quit [Quit: leaving]
<OscarL>
scanty: nice to compare against nightlies :-)
dalme has joined #haiku
<OscarL>
just don't try to update it as-is (bootloader changes will render it unbootable)
<scanty>
well, i don't have the need for such a thing, so i'm going to find something else to do with 128GB
<OscarL>
scanty: nice size!, hopefully faster than the ultra cheap 32 GB ones I have here (sometimes go as low as < 10 KB/s, lol), So much for being "Kingston".
<scanty>
mine are sandisk
<scanty>
real sandisk
<OscarL>
Anarchos: i didn't. will take a look later this night when I have some more available data on the cellphone. thanks!
<Anarchos>
OscarL intel_extreme is really a rabbit hole
<Anarchos>
i can't make my mind how all this work
Anarchos has quit [Quit: time to sleep]
<OscarL>
there's a reason device driver developers doesn't grow on trees :-)
<OscarL>
s/doesn't/do not/
<jezek2>
though a lot of the complexity could be often improved at the HW level, there is no real reason to have extreme complexities, it's more like nobody cares and splitted teams or something
<PulkoMandy>
The hardware is complicted and does a lot of things
HaikuUser has joined #haiku
<jezek2>
yes, but sometimes quite simple changes could improve things a lot
<OscarL>
jezek2: what, you don't like reading a 3 bits data which value comes from 3 separate registers? (happened to me some days ago while trying to read a fan divisor, FFS)
HaikuUser has quit []
<jezek2>
once you get experience with building such systems (eg. using FPGA) you start to see it
<PulkoMandy>
jezek2: the intel engineers make a lot of changes to simplify things at each device generation. But then the driver has to supports the older generations of hardware as well
<OscarL>
I appreciate making use of every bit, but, man... that was a bit extreme :-D
<jezek2>
that making something much better doesn't cost you much resources, and can even improve the overall design for the HW side too
<PulkoMandy>
on an fpga? Yet. When making a new chip costs hundred of thousands of dollars to produce it? Not so flexible anymore
<PulkoMandy>
If your hardware experience is on FPGA, that is still very very close to software development
<jezek2>
PulkoMandy: no, only if you treat is as another "CPU" and do things wrongly
<PulkoMandy>
on an fpga, fixing your mistakes is cheap, you just update the bitstream
<jezek2>
when you design for FPGAs it's quite similar to ASICs, there are just different trade offs, eg. in FPGAs you have fancy memory blocks that are cheap, in ASICs they're costly
<jezek2>
it's not the point, the point was designing
<jezek2>
on ASICs you do respins too, and for both FPGAs/ASICs you spent most of the time in simulators
<PulkoMandy>
It is not at all like whut intel engineers are doing, iterating on a design that started 20 years ago witheour drivers supporting all generations of it
AlienSoldier has quit [Quit: Vision[]: i've been blurred!]
<jezek2>
there are two things, an architecture of single lineup (can be a single generation or multiple), I'm talking about the design of these, that it can be often improved a lot with minimal costs
<jezek2>
then you have the separate issue with multiple different generations
<jezek2>
that's unavoidable
<jezek2>
but still if you improve the former, the latter is also a lesser problem as the overall complexity is smaller
<jezek2>
and due to better design there are probably more similarities
<jezek2>
sometimes they do good designs, like the very forward thinking of Motorola CPUs which allowed them to use the same ISA for vastly different kinds of CPUs
nipos has left #haiku [Disconnected: Replaced by new connection]
nipos has joined #haiku
neoncortex has joined #haiku
Skipp_OSX has joined #haiku
<OscarL>
have been pulling my hair due to an odd permission error happening only on 32 bits. turns out I had a slightly out of date ncurses6 installed there, and that PulkoMandy had already fixed the cause of the problem.