-
nsITobin
tomman: is there ice cream yet?
-
Sompi
-
Sompi
Those "woke" people really seem to infiltrate everything, now even LibreWolf is in their hands
-
Sompi
Somehow they always manage to reach the leading positions
-
Sompi
And always they are so secretive
-
Sompi
does anyone know who thas "ohfp she/her" is?
-
Sompi
that*
-
tomman
uuuugh, the Dolphin emu folks switched to AVIF images for their blog entries. WHY.
-
nsITobin
so webp but with patents
-
nsITobin
WHY
-
nsITobin
oh it is av1 based k
-
nsITobin
still.. why
-
nsITobin
png is best
-
nsITobin
a one frame movie != a single image format
-
nsITobin
ok so what was the logic of backing out a rollup to get a patch in mc only to reland in comm..
-
nsITobin
and why are thunderbird people jumping the gun on m-c changes before those have happened
-
tomman
nsITobin: don't worry, I've spotted HEIF images being used in the wild too
-
tomman
because once again WHY
-
nsITobin
singularity of stupidity where the sheer mass of the stupidity will collapse and then explode with all our hopes and dreams.. or so goes the theory
-
nsITobin
explode showering us with our hopes and dreams*
-
nsITobin
granted its a stupid theory but there it is
-
nsITobin
it is*
-
nsITobin
works well for social climbers because once it starts coming apart at the seems it is someone elses problem
-
nsITobin
seams*
-
nsITobin
i can't type today.. perhaps an indication? lol
-
tomman
-
tomman
"CLOSED WONTFIX NOTABUG UPDATESAREGOOD"
-
tomman
"Yes - bincompat is one-way. Old programs can use new redists, but new programs can’t use old redists. This allows us to add functionality over time - features, fixes, and performance improvements."
-
tomman
the Linux madness, now on a Windows box close to you!
-
nsITobin
yeah well everything else is shifting and flipping.. and i suspect if several ones act quickly a linux distro could be created with stability and as I was saying the other day an actual idealized windows-level de
-
tomman
"you need the Visual C 2022 runtimes... no, not those, those other, no, not newer enough"
-
tomman
it's like places rejecting $100 bills prior to the colorized 2006 series
-
tomman
because "no, they're too old" despite still being legal tender
-
nsITobin
that note is legal tender for all debts public or private
-
nsITobin
if an older bill that is still legal tender will not be accepted then money has no value.. which it doesn't
-
nsITobin
anyway practically speaking windows has been a dead platform since windows 10 and its free upgrade
-
nsITobin
tomman: if microsoft hedges its bets on linux virtualization and a chromium browser engine.. what need is there of Windows except as a brand name..
-
nsITobin
that people have been told for 40 years they have to have
-
nsITobin
that concept of Windows in your mind as a sort of entity an established thing of defined attributes and characturistics and ways of doing things.. that's all just legacy and on the block
-
nsITobin
and it is indeed the Mozilla model. Micrsoft IS Mozilla at scale now.. It is why I wish Mozilla was only doing this to take out a much bigger threat.. but I know that almost certainly hasn't been even the tiniest fraction of it and it makes me sad.
-
tomman
so you mean MICROS~1 has become a overweight San Francisco startup? :D
-
nsITobin
yes
-
nsITobin
at scale
-
tomman
someone should remind Redmond that they're in Washington, not California :P
-
nsITobin
it was all over when Microsoft chose Wordpress instead of their own CMS.
-
nsITobin
That's how it starts
-
nsITobin
What did Mozilla first do.. they added google ipc
-
nsITobin
and the codebase has been a 15 year trainwreak ever sense.. it is a small wonder ANY of us can do ANYTHING with any of it
-
nsITobin
google ipc and webidl were the first steps in removing Mozilla tech from the playing field and it has worked brilliantly despite my best efforts.
-
nsITobin
this is anti-widows development btw
-
nsITobin
this vs thread
-
nsITobin
it goes against EVERYTHING Microsoft had in decades past done regarding visual studio and the hordes of DEVELOPERS DEVELOPERS DEVELOPERS DEVELOPERS
-
tomman
hmmm, the benchmarks for the latest AMD GPUs are on
-
tomman
and... well, the answer is "don't buy yet"
-
nsITobin
AMD can't innovate they purged too many minds.. all they can do is crappily copy intel without the last decade's worth of research development and manufacturing
-
tomman
half-baked drivers, as usual. But then, if you buy ANY new piece of hardware close to launch date, you deserve what you get
-
nsITobin
all AMD products are suffering post AM5 and since they ran out of ATI technology to endlessly push
-
nsITobin
gfx suck too
-
tomman
wait 6 months at least, let drivers mature and prices normalize
-
tomman
eh, I'm happy with my lowly Zen 2 APU
-
nsITobin
that isn't am5
-
tomman
the era of ATi driver hell is LONG gone - now you just get half baked (but stable) drivers at launch date
-
tomman
yep, but still...
-
tomman
I don't regret having jumped ship
-
tomman
Intel? VINTAGE ONLY.
-
nsITobin
well i prefered it when drivers were model or at least series specific.. amd unified driver has made it so i can't run multiple amd cards outside its range.. because the devices are semi-identical according to driver
-
nsITobin
i guess comm people are done for now screwing with build configuration.. time to see what all there is..
-
nsITobin
frg_Away: Good morning.
-
nsITobin
So obviously with these new patches I have to fix one of the mass conversions because it altered confvars.. so I am going to have to exclude that bit during this conflict.. I hope i don't screw it up.
-
nsITobin
0:01.50 File "C:/seamonkey/src/comm/suite/moz.configure", line 91, in <module>
-
nsITobin
0:01.50 imply_option("MOZ_APP_VERSION", seamonkey_version.version)
-
nsITobin
0:01.50 File "C:\seamonkey\src\python\mozbuild\mozbuild\configure\__init__.py", line 1243, in imply_option_impl
-
nsITobin
0:01.50 raise ConfigureError(
-
nsITobin
0:01.50 mozbuild.configure.ConfigureError: Cannot infer what implies 'MOZ_APP_VERSION'. Please add a `reason` to the `imply_option` call.
-
nsITobin
first thing I think I will do is eliminate SEAMONKEY_VERSION as THUNDERBIRD_VERSION is long gone
-
nsITobin
... guess i should find the bug for that..
-
nsITobin
so i can reference it
-
nsITobin
that will better match current prevailing config
-
nsITobin
-
nsITobin
Let's all assume I have already ranted about the surrounding conditions contibuting to my task before me and let's do a thing!
-
frg_Away
nsITobin Hi. Verson was used for Lighting add-on. Now integrated.
-
nsITobin
nothing I haven't done oh.. about four times already :P
-
nsITobin
and i expect to do it once more on 253 one day
-
nsITobin
well now there are other issues
-
frg_Away
nsITobin not sure
bug 1507948 can be used. TB veesion is gone but ours is totally different and probably needs to be kept.
-
nsITobin
what needs to be kept?
-
nsITobin
mozilla just very recently began ripping autoconf out from under us.. confvars is basically done.. and the remaining rbanding vars left in I am sure is an error as well..
-
frg_Away
What is in needs to be kept and moved to python because SeaMonkey specific. Only MOZ_UPDATER and _PROFILE_MIGRATOR are probably elsewhere already.
-
nsITobin
sorry i was talking about central not 253
-
nsITobin
for 253 *_VERSION hasn't been requred for years
-
nsITobin
not since calendar left AMO
-
nsITobin
because why would you ever need multiple targetApplications for something that is built and bundled specifically for the app per build
-
nsITobin
just they didn't seem to bother until 65
-
nsITobin
it was only ever for install.rdf
-
nsITobin
and why was calendar de-extensionized? cause no more binary xpcom components for extensions.. and calendar uses em
-
frg_Away
nsITobin no was only done after 60. Binary components were already done for in 2.49 and compiled in when the option was set. xpi won't work if the underlying program was compiled without calender support.
-
frg_Away
-
frg_Away
this is gone in central.
-
nsITobin
SEAMONKEY_VERSION only exists in suite/ on central
-
nsITobin
but the build configuration is basically screwed now
-
nsITobin
and the underlying stuff has changed a bit as well
-
nsITobin
I am disorented again
-
nsITobin
within moz.configure
-
frg_Away
Thr installers need ot for the file at lest so moz_app:_Version and others need to be set.
-
nsITobin
would it be at all possible to get thunderbird contributors to stop updating SeaMonkey's build system?
-
nsITobin
on central
-
nsITobin
# version.txt content from the project directory if there is one, otherwise
-
nsITobin
# the firefox version.
-
nsITobin
it is no longer possible to alter the MOZ_APP_VERSION outside version.txt without a preexisting condition being set
-
nsITobin
it will automatically pick up version.txt and version display
-
nsITobin
it is no longer something the project.. does
-
nsITobin
or app
-
nsITobin
thunderbird uses firefox versioning
-
nsITobin
which is mozilla milstone versioning
-
nsITobin
okay.. here is my situation as I see it.. I have before me comm and mozilla central.. a recent botched thunderbird patch making changes the build system was never going to accept and obviously without giving it even a ./mach configure to test.. I either need direct instruction on what work to complete or just set loose and make it do SOMETHING so that everything isn't blind applied left untestable.
-
frg_Away
would just make it build using mozilla versioning and then we can see what goes later.
-
nsITobin
yeah well i have to undo what they did and work through EACH thing and handle it properly then I was gonna rewrite suite/moz.configure to be maintainable.. what they did was a blind port of everything they previously missed and they did it.. wrong.. so its worse now and all at once instead of working through it.
-
nsITobin
the nature of moz.configure is its all strangely async so consistant order of what will provoke an error first changes
-
nsITobin
so i have to move each item out of confvars one at a time and test via ./mach configure and checking values in config.status foreach
-
nsITobin
I just about had it the other day but I trashed it because I had no idea what they were doing if it would be a different patch or just the same reapplied after m-c was patched
-
nsITobin
i was up to about five bugs before their patch landed then was reverted and landed again
-
nsITobin
I am gonna have to make a decision.
-
Sompi
With lossless compression PNG almost always ends up in a smaller file size than AVIF. With lossy compression it seems to be about 50/50, sometimes AVIF is smaller and sometimes JPEG
-
Sompi
AVIF just means more compatibility problems and security vulnerabilities
-
nsITobin
i mean god forbid moz.configure was as nice as this
dpaste.org/msoNq
-
nsITobin
oh something to look forward to
-
nsITobin
-
nsITobin
config.status looks correct
-
IanN_Away
i'm getting "502 Bad Gateway" on bmo
-
nsITobin
confirmed
-
nsITobin
502
-
nsITobin
ffs
-
nsITobin
did bugzilla fall over?
-
nsITobin
my conspriacy senses are tingling.. I am fighting them tho.. too much to do
-
IanN_Away
it's back
-
nsITobin
dunno why I am so conserned .. never needed bugzilla before to figure anything out.. just additional metadata i guess
-
frg_Away
nsITobin happens pretty often lately. hg is affected also. comes with the cloud
-
nsITobin
well that's all the reason I need to shut down bugzilla and hg and just switch to github
-
nsITobin
:P
-
nsITobin
i mean it's soooo old and unreliable
-
nsITobin
if it can't be stable on a poorly managed overengineered distributed network.. what use is it?
-
nsITobin
but seriously.. i wonder why isn't azure supposed to be good?
-
Tekk
Because no 'cloud' provider can be.
-
Tekk
They're all total garbage. Seems inherent in the industry.
-
nsITobin
no shit.. i bet they are being targeted because of that USAID bs
-
nsITobin
there is a comma in there
-
nsITobin
no comma shit
-
nsITobin
allow me to rephrase: no, i bet they are being targeted because of that USAID funding bs
-
nsITobin
anyway, gonna build this see what happens
-
nsITobin
on top of the adclick contraversy
-
frg_Away
nsITobin I would leave the tin hat at the door stop. This happend for weeks now.
-
nsITobin
i'll set it over there on the soapbox for now :)
-
nsITobin
using rust 1.82 as it seems to be behaving memory wise
-
nsITobin
-
nsITobin
0:33.13 C:/seamonkey/src/toolkit/xre/nsAppRunner.cpp(4162,38): error: expected ')'
-
nsITobin
0:33.13 4162 | "Maybe try to reinstall " MOZ_APP_DISPLAYNAME "?\n",
-
nsITobin
0:33.13 | ^
-
nsITobin
i suspect when empty it is fine.. this is on the mozilla side
-
nsITobin
either that or MOZ_APP_DISPLAYNAME doesn't work in moz.configure yet.. has to be set by branding/configure.sh
-
nsITobin
switch it to that see if it helps