Oh god

Hi TobinWithTooManyNames

good morning

nsITobin greetings

frg_Away: i threw a bunch of links related ssuao from UXP into the log.. as for a UI that shouldn't be too difficult with XUL technology especially seeing as I can reference one of i think moonbat's extensions.. it's a bit crappily designed but I hear it works well luckily I know a few things about UX design .. especially SeaMonkey's

well scratch that extension.. it's GPL for some reason

no matter pref manipulation isn't hard..

nsITobin swamped as usual. All on the backburner unless I get a patch which applies. There is also Bug 1862395 which i would handle together with it.

that would need a check in useragentoverrides to not do anything if that is set

oh

wait wtf

well you know you have the firefox version for release 56 set to 60 right?

i say.. subst in a value set in confvars

get rid of this hacky math

nsITobin yeah thought about setting a constant.

or that

frg

uint32_t firefoxVersion = appVersion.ToInteger(&rv);

appVersion not platformVersion

maybe

2.whatever -3 ?

= -1

thunderbird matches firefox versions so it would work as expected

if it mattered to thunderbird

i am almost sure that is the cause of the math screwup

appInfo->GetVersion

is that the app or mozilla version tho

checking

has to be appVersion

not platformVersion

yep

appInfo->GetVersion gets version which is app version so ye seamonkey's version is being used to calculate a firefox esr version and that is why it is broken

yeahh has been broken for ages

I dunno why this isn... yes i do

i know exactly why this isn't using the xpcom interface

and calling xreappdata directly

or whatever

of course even if fixed it will operate on a value of 60 not 56 and even then the esr version is so old hardly anthing is fixed so... i return to recommend trashing it and just having a constant

hardly anything would be fixed by it*

it sure is one of those roadblocks to having not-thunderfox

But i defer to an old Mythbuster's saying.. when in doubt? C4.

frg_Away: so which do you want "fixed" or fixed

fixed version as in constant

one being as intended and the other being as actually needed

ye

Its not really supported and snakeoil too. I would rather rip all this out but this will start a ton of discussions. You will be tracked anyway with or without resist fingerprinting.

frg_Away: as long as you make sure to unapply the patches before switching using multiple queues on workbench is pretty smooth

i dunno if it will error out or qfinish it for you if you switch with applied patches

revisios are unknown. It will not end well. You then need to delete the status file and do a purge and clean up.

Happend to me a few times when I accidently overwrote the status file. Can be recovered but time consuming.

frg_Away: If it was Pale Moon, pitch forks and violant forum posts unless there was something Latest and Greatest to take over for it plus a long winded post explaining it no one will read.. Luckily this isn't Pale Moon

but it is classical mozilla so always a risk

and of course normally I'd be deployed to intemidate users into stopping complaining for no reasons creating a distraction larger than the issue its self.. thus resolving it..

i doubt that will work on mozillazine LOL

or for me

sooo

wrong vm

frg_Away: do you want just a simple hard coded constant or do you want it tied to something else? and what value to start?

What does mozilla do currently? I think they set ti to a constant or not?

I'll check

That is why I wnated it to fix with the ua overrides. VCurrently setting 91.0 but scattered over in the code in 2 places.

refactored

MOZILLA_UAVERSION

oh there it is

yeah

seems they are using the current version so would be 91.0 for now as in our ua.

'MOZILLA_UAVERSION': '"60.0"',

milestone.txt controls those defines

values*

we need it tied to general.useragent.compatMode.version

uarghhh...

and it was me :)

that should also be tied to the pref

lol

yeah or constant used in the file

was something quick and dirty.

yeah

some of this stuff I can just rainman it other stuff confuses the HELL out of me

in UXP we did a bunch of other crap to the UA like app can be build id instead of version and added a gecko compat slice cause the gre slice was renamed to goanna but you don't need all that crap I figure set a confvar value which sets the pref default value and then the pref prevails.. simple and clean

ugh frg_Away waiting for a build i didn;t actually start...

I am new to virtual desktops :P yeah I know right?

not the concept but using them effectively

nsITobin all depends on the host too. With vbox and 2 cpus I can actually do 2 builds at once becuase of 2 more or less independent numa nodes.

currently still slap dab in the middle of fedora 39 unfortently

its only a matter of me building an entire distro from source

and vmware

I have attempted multiple builds before but the bottleneck for me is disk io

not ram so much or cpuz

6 cores, 16gbs ram

cause I also run paint shop pro and some webdev off tthe SM one.. the clone for central is just that

config.status and a cursory objdir check is a great way to figure something out about the build system without actually waiting for it to build as expected

Google Photos don't work with Seamonkey anymore. It only asks to log in and doesn't let see the pictures

With Firefox it just shows the photos without asking to log in, even in privacy mode

So they are showing different content to different browsers

And it only seems to happen with certain photos, and some photos are displayed normally also on Seamonkey...

Sompi: could be worse

Google Docs finally broke here, and worse, you get stuck at a reload prompt

Why are they serving different content to different browsers when the page itself clearly works on Seamonkey?

Sompi: check the actual image encoding

also play with ssuaos

Earlier I thought that the stylesheets of Google Photos are somehow broken on Seamonkey, but apparently it looks just the same on Firefox so it's supposed to look that way...

No contrast, no indication of what is a link and what isn't

How can I see the image format?

They are using JavaScript to show the image and the browser does not recognize it as a normal image element

Nevermind, I got the link to the actual image via inspector

It is just a normal JPEG image and Seamonkey can show it just fine

But if I go to the actual webpage with Seamonkey, it just redirects to the login prompt

And that only happens with some specific images, and with some other images Seamonkey users are allowed to view them normally

can you check the encoding on em

"Identifying as Firefox" seems to fix it...

nsITobin: What encoding?

on the image

maybe some are some new format

and others are webp and png

and jpg

does seamonkey have webp

it should?

The problematic image is this: photos.app.goo.gl/U6TDgEXAciwE1PqB8

It is just a normal JPEG image

If I try to load that page with SeaMonkey, it just redirects to a login page

Identifying as Firefox and then opening it in a private window works

"image/jpeg"

it does work in IBM Qauntumnewt aka Pale Moon

and works as-is on seamonkey on linux

Maybe it's then somehow tied to my identity and not exactly a browser thing

you have adverisize as firefox right?

in the preferences

I toggled it on and it started working in private mode

yeah google already discriminates against non-blink

firefox is a barely maintained exception

also works on wip

wasn't this checkboxes not too long ago?

I like it.

Sompi: identifing as JUST SeaMonkey it works on private and otherwise

on WIP

so ..

Why is modern web like this...

because we let it happen

I feel responsible

anyway

I don't use other Google services than Youtube

and I always write my own webpages so that they don't discriminate against any browser

historically just targeting basic mozilla compat was enough cause if it worked in Mozilla it tended to work everywhere save IE6 edge cases

logbot.thereisonlyxul.org/seamonkey/20240827#c5587 I wouldn't mind betting he really, really dislikes you too.

to be fair that was under the banner of the moonpeople so that MAY have had a thing cause I didn't become the asshole I am recovering from until well after he was a dick

oh he left

it was when Pale Moon changed its guid so very early on and he was like massively rude to me when I was reaching out to extension developers for future support he took all my posts and merged them into a single thread and then shoved it out of the way... after that he was hostile from then on

it's why I stopped posting on Mozillazine

I write my web pages mostly to the xhtml spec (only deviating when there are Good Reasons) and test in Gecko and Webkit. I stopped bothering about IE support years ago and have never cared about MS Edge.

If they support web standards, they should be fine.

jonadab: xhtml/1.0 transitional the best seems to quirk its way into cross-browser support nicely vs trying str8 html 5

HTML 5 doesn't add anything I care about.

i mean video and audio tags are cool but those have baggage as well

The one thing I want that XHTML doesn't have, is the ability to embed block-level elements in a paragraph. I work around that by using <div class="p"> as my standard "paragraph" markup. Which is annoying, but it's the best option I've found.

dunno why we needed several specialized divs and why it isn't xml-based

wouldn't a span do the job?

It's possible to style a span to be block-level, but why do that when div exists?

good question

why don't we have vbox and hbox in html

that would be useful

I can't fathom the logic behind why <p> is block-level but not allowed to contain block-level elements. What, the XHTML authors have never put a blockquote or bulleted list inside a paragraph, really? Dunno.

But that's the only thing XHTML lacks that I want.

My preferred method for audio and video files is <a href="/media/foo-vid.mpg">Get the foo video here</a>.

frg_Away: please confirm current latest patch in wip comm queue is Moonchild's 0.0.0.0 patch cause the normal stop is above the ipdl work

I understand why YouTube doesn't want to do it that way. They want to show advertisements to pay their bills. I get that. But I'm not them.

nsITobin it is. Put it into 2.53.19 too.

cool just wanted to make sure my script or system didn't screw up some how

I will move it around later. Need to align with 2.53 again.

no problem at all

this is how webpages should be: sininenankka.dy.fi/leetos

Similar examples: slackware.com openbsd.org

Sompi: is BinOC and co too fancy?

lol