09:16:30 <Sompi> Is it true that Seamonkey is abandoning 32-bit x86 target?
09:17:56 <Sompi> That means that I also have to stop using it
09:20:11 <frg_Away> yes. ou can roll your own build. Support will not be taken out
09:22:27 <Sompi> So it still builds normally?
09:23:17 <frg_Away> yes
09:23:45 <Sompi> Are there any building instructions online? One friend also needs a non-SSE2 version
09:24:41 <frg_Away> but at least I won't test it any longer. If something fails to build I wont try to fix it but have no problem integrating patches.
09:25:28 <frg_Away> IanN provided linux buiild instructions on the web page. Non sse2 will not be possible I suspect.
09:25:45 <Sompi> Why wouldn't it be?
09:25:58 <Sompi> Because of Rust?
09:27:17 <frg_Away> probably built build config changes needed.
09:52:57 <Sompi> Linux's I/O has definitely became a lot slower than before
09:53:27 <Sompi> I'm installing updates to a virtual machne that runs Devuan Linux and it's been doing that for 20 minutes
09:53:33 <Sompi> And there's not even that many packages to update
09:58:29 <Sompi> stuff like this used to work very fast on a Pentium II, now it is very slow on an i7-based Xeon
10:09:11 <frg_Away> Make sure VT-d is on ifor the vm.
10:11:11 <Sompi> it's I/O that is slow
10:15:22 <frg_Away> VT-d is for io No vt-d = slow io in a vm.
10:16:12 <frg_Away> https://www.tenforums.com/virtualization/112764-what-intel-vt-d-bios-do-i-have-enable-well.html
10:18:57 <Sompi> the I/O is also slow on native hardware nowadays
10:18:59 <Sompi> a lot slower than it was
10:28:23 <Sompi> the mach command does not find python
10:39:23 <Sompi> How do I solve this? Every time when something requires python it means problems
10:40:32 <frg_Away> You need pytin 3.7+. 3.12 is probably broken. Downloa the firefox esr115 source and do a mach bootstrap. Then you should probably be set.
10:40:47 <frg_Away> ^python
10:41:24 <frg_Away> 64 bit host 8GB minimum
10:41:28 <Sompi> What Linux distribution should I use?
10:41:36 <Sompi> How does compiling and linking need that much?
10:41:57 <frg_Away> rust and libxul.
10:42:12 <frg_Away> central now needs 32GB thanks to rust.
10:42:47 <Sompi> doesn't that ruin the whole idea of "open source" if compiling the program requires 10 times more memory than using it
10:42:57 <Sompi> no, 100 times more
10:43:54 <frg_Away> can'T help it. need to follow mozilla alt least part way. No devs no independent real development.
10:44:33 <Sompi> I don't have that much memory
10:46:26 <Sompi> how large are the modules that are written in rust?
13:29:36 <Sompi> How am I even supposed to make any changes to the program if compiling it requires 32 gigabytes of memory and I only have 22?
13:29:56 <Sompi> Shit like that ruins the whole purpose of free software
13:31:00 <frg_Away> 2.53 only needs 8. cewntral is comm-central and SeaMonkey is broken there anyway. Just compiles.
13:31:01 <Sompi> SeaMonkey should ditch all rust modules and use only C and C++ code that is written in such way that it is easily portable to different 32-, 64- and maybe even 16-bit target platforms
13:31:05 <nsITobin> these are questions I asked in 2015 regarding chromium
13:31:19 <nsITobin> Sompi: not possible with the current codebase
13:31:22 <frg_Away> 16 and 32 bit are dead
13:31:46 <nsITobin> 32bit has been less relevant outside specialist usage and old hardware for a decade
13:31:50 <Sompi> 16-bit is inevitably going to be the future when China attacks Taiwan
13:32:04 <nsITobin> but seamonkey isn't 16bit
13:32:39 <tomman> ha, just updated to 32GB here
13:32:43 <tomman> costed ~$96
13:32:50 <Sompi> After all it's mostly just about the default int size, and avoiding pointer offsets larger than 65536
13:33:02 <nsITobin> but no i am upset about this too because it means I can't do mozilla in a VM
13:33:05 <tomman> Sompi: I've been using 64-bit OS since 2006
13:33:09 <tomman> it's a biiiiiit late to go back
13:33:18 <tomman> ...wow, 19 years!
13:33:35 <nsITobin> and can barely do it on bare metal where the build host's lifetime is limited and I will never install windows 11 on here
13:33:37 <tomman> that Core 2 Duo box is old enough to vote, drink alcohol, and maybe even use assault weapons
13:34:27 <Sompi> I use 32-bit operating systems on computers that have less than 5 gigabytes of RAM because all binaries are smaller on 32-bit
13:35:05 <Sompi> even if the computer supports 64-bit
13:35:10 <tomman> we now have a even bigger problem: compilers
13:35:35 <tomman> one of the reasons Debian is dropping 32-bit releases (but not binaries!) is that GCC doesn't even know what's  a Pentium anymore
13:35:49 <Sompi> How would it not know that?
13:35:52 <tomman> ask for -march=i586, get... conditional moves (a Pentium Pro feature)
13:35:57 <Sompi> wtf
13:35:58 <tomman> and Rust compiler is even worse
13:36:01 <tomman> yes, that's what I said
13:36:16 <nsITobin> I see these as non-issues
13:36:27 <nsITobin> because it is NOT too late to fork off all this shit
13:36:30 <Sompi> I thought that gcc is consistent about target platforms
13:36:33 <nsITobin> tho it is damned close to the wire
13:36:42 <tomman> ask 32-bit code, get... "something, yes, it will run on a Nehalem or Excavator so who cares"
13:36:44 <nsITobin> gcc hasn't been consistant since version 5
13:36:49 <tomman> Sompi: not anmore
13:36:51 <tomman> --not anymore
13:36:53 <tomman> for a long time
13:37:19 <tomman> that bite me recently when trying to crosscompile a Super Mario 64 executable to run on a ol' Socket 7 box for lulz
13:37:37 <tomman> if you ask for i686/Pentium III, you're kinda safe... for now
13:37:50 <tomman> but that's GCC, as I've said, Rust assumes "32 bit? It must have SSE2, amirite?"
13:38:07 <Sompi> Then what's the point of having the i586 target there if it outputs instructions that only exists in i686...
13:38:18 <nsITobin> historical scripts
13:38:21 <Sompi> tomman: Yes and the rust maintainers have no intention of fixing it
13:38:24 <nsITobin> and configure systems
13:38:42 <Sompi> nsITobin: But if i586 and i686 still aren't the same thing
13:38:43 <nsITobin> i mean somehow x86-64 is now versioned
13:38:47 <nsITobin> when did this happen
13:38:48 <tomman> I can understand the rationale: few of those systems are surviving, and developers have long ago moved over
13:38:50 <nsITobin> who decided it
13:38:54 <tomman> nsITobin: Intel
13:38:55 <nsITobin> and why was I not consulted
13:38:56 <tomman> blame Intel
13:39:01 <Sompi> i586 uses only some Pentium Pro instructions, but i686 uses them all?
13:39:05 <Sompi> That makes no sense
13:39:07 <tomman> now we have LINUX DISTROS dropping original Nehalems!
13:39:12 <nsITobin> how does INTEL decide AMD64?
13:39:14 <tomman> Sompi: i686 _is_ Pentium Pro
13:39:16 <nsITobin> explain
13:39:45 <Sompi> tomman: I know
13:39:46 <tomman> i586 uses no i686 instructions - Pentium MMX is basically i586 with MMX, while Pentium II is i686 with MMX
13:40:01 <tomman> but these days: "i586? Must be a Core 2 Duo, right?"
13:40:07 <Sompi> So how i586 has some Pentium Pro instructions too and how is that not considered a bug?
13:40:28 <nsITobin> because no one should be coding for that
13:40:33 <tomman> nobody has bothered reporting it, and filing a bug report at GCC (and all the involved protocol) is... not for the faint of heart
13:40:34 <nsITobin> coding for that IS a bug
13:40:45 <Sompi> But the current situation really is crazy. I have a relatively powerful Xeon computer and every time when I update software I fear that the new version does not support my CPU anymore
13:40:58 <Sompi> nsITobin: So I'm mostly only coding bugs?
13:41:09 <nsITobin> Sompi: to some.. not me
13:41:26 <Sompi> What about clang then?
13:41:32 <tomman> Outside of retrocomputing, there is really no interest these days
13:41:38 <tomman> Sompi: Clang is on the same boat, I'm afraid
13:41:54 <nsITobin> I mean I am on version 4 of my self.. and it still isn't what the world wants.. and at this point.. the world doesn't KNOW what it wants that's why shit is collapsing
13:42:00 <tomman> but then LLVM came muuuuch after Coppermine
13:43:47 <nsITobin> Sompi: besides what good is a modern compiler that won't even run on the hardware that if the worst happened tech wise we'd be forced to
13:43:52 <Sompi> In the past you could always use the "unofficial" bug reporting protocol for most free software: Just join their IRC channel and complain about the bug, the developers were usually operators in those channels
13:43:57 <nsITobin> we'd NEED period compilers or a new one for that hardware
13:44:07 <nsITobin> then you just have to bootstrap.. once back
13:44:32 <Sompi> Nowadays you usually need at least Chromium or Firefox and a Microsoft Github account
13:45:03 <nsITobin> yep
13:45:06 <nsITobin> guess what I don't have
13:45:08 <nsITobin> anymore
13:45:11 <tomman> Sompi: or a Hackernews account and a Xitter account and some loud social media following
13:45:14 <nsITobin> a github account i can use
13:45:38 <nsITobin> i HAVE another github account but if i do anything not read-only it will be banned cause it is obviously me
13:45:51 <tomman> Even if the project doesn't use Github, social media is basically a must for many projects these days
13:46:04 <nsITobin> listen to ourselves tho
13:46:11 <tomman> hell, I know projects (mainly in the gaming realm) whose bugtrackers live in... goddamned Discord
13:46:19 <nsITobin> we ARE saying you need to be approved by microsoft to be in most open source development
13:46:24 <nsITobin> tell me how FUCKED that is
13:46:39 <tomman> worse, they host code on Github (or Gitlab), but don't use the bugtracker on those at all
13:46:42 <Sompi> tomman: rust also uses discord. not the game, I mean the programming language
13:46:54 <tomman> Jesus Fried Christus
13:47:07 <tomman> yeah, that plague has been spreading across vast surfaces on FOSSland
13:47:28 <Sompi> rust is not free software
13:47:41 <nsITobin> rust is compiler and ecosystem as a service
13:47:47 <UsL> codeberg is almost the only viable git platform now days.
13:48:19 <nsITobin> i am on gitlab because codeburg is too small and special intrests would get me.. on gitlab i just have to worry about the automated system
13:48:26 <tomman> Anything centralized by definition is vulnerable to enshittification and similar woes
13:48:41 <tomman> but sadly selfhosting anything these days is expensive, difficult, and generally considered "a threat"
13:48:46 <nsITobin> and even then gitlab isn't NEARLY as socially connected as github
13:49:04 <nsITobin> we all kinda live in our own little worlds and actually have to make a choice to go elsewhere
13:49:12 <nsITobin> on github you can link around everywhere
13:49:15 <UsL> yes, we desperately need a anti-enshitification movement
13:49:21 <tomman> That too, these days the newest crop of devs are very... "social" folks, AKA they need their likes and reposts
13:49:30 <nsITobin> OKAY
13:49:33 <nsITobin> guys
13:49:42 <Sompi> the program is not free if its developers intentionally make it dependent on closed source software and centralized services and do it so that the user cannot work around it
13:49:44 <nsITobin> we all have known each other for years.. are we socially normal?
13:49:47 <nsITobin> any of us?
13:49:54 <nsITobin> i mean i am an extreme case
13:49:59 <nsITobin> but is anyone plain normal?
13:50:07 <UsL> nothing is normal today  anyway : )
13:51:36 <UsL> but, yeah,I guess it's NOT normal to  have a heap of social media trackers in your phone and life today
13:51:37 <nsITobin> maybe IanN .. he might be normal or just learned long ago to be less abby-normal
13:51:50 <UsL> in that case I am very not normal
13:52:47 <UsL> uh, I  meant not normal to not have. I should stop double negations
13:52:54 <tomman> what's "normal"?
13:53:04 <tomman> I'm a human, but that's it
13:53:22 <UsL> to be on every social platform ever. If not you're either a bot or a cave man : )
13:53:53 <UsL> Probably not a bot btw, since they actually are on every social platform..
13:54:19 <Sompi> things that were considered antisocial behavior are now considered social
13:54:33 <tomman> And viceversa
13:54:39 <UsL> indeed. We have come full circle
13:55:04 <nsITobin> i dunno why *I* needed to go on that trip but i guess I learned ... something?
13:55:30 <Sompi> nowadays it's normal to refuse socializing with people who don't use your favourite centralized closed source messaging spyware app (that also does not work if the other person does not have exactly the same type of phone that you have)
13:55:45 <nsITobin> guess get it over with last decade so i can deal with this now this decade I GUESS
13:56:25 <nsITobin> I'm feeling very orsen wells atm
13:56:33 <nsITobin> being always past that
13:58:04 <nsITobin> We know a little open source project where developers work...
13:58:06 <nsITobin> :P
14:16:34 <tomman> hmmm, Greasemonkey is borked here
14:16:45 <frg_Away> you can bitch complain fork or trying to build a time machine but it won't do any good and it is just a waste of time.
14:16:53 <tomman> trying to add a userscript to bypass Imgur brokennes, and it fails to paste scripts from clipboard
14:17:00 <tomman> instead I get two JS errors on the error console
14:17:11 <tomman> Fecha: 10/5/25 10:15:15 a. m. VET
14:17:12 <tomman> Error: TypeError: window.opener.messageManager is undefined
14:17:14 <tomman> Archivo fuente: chrome://greasemonkey/content/newscript.js
14:17:15 <tomman> Línea: 32
14:17:18 <tomman> Fecha: 10/5/25 10:15:16 a. m. VET
14:17:19 <tomman> Error: TypeError: aSource is null
14:17:21 <tomman> Archivo fuente: chrome://greasemonkey-modules/content/extractMeta.js
14:17:22 <tomman> Línea: 8
14:17:36 <nsITobin> frg_Away: but can I do all that and deal with today's issues.. and all i can say is I can try
14:18:05 <tomman> the first one comes when hitting "New Script"
14:18:25 <tomman> the second one when pressing the "Use Script from Clipboard" button
14:18:29 <frg_Away> I still use an old Pentium for an eprom burner but 64 bit is so much better. memory allocation in 16 bit was a nightmare and 32 bit is too limited for anyting big today.
14:19:04 <tomman> butbutbut segments are PHUN!
14:19:52 <tomman> wonder what the hell is window.opener.messageManager
14:20:11 <frg_Away> Even keeping SeaMonky alive is a struggle and forking anything like a compiler or os needs dedicated people who are not there. You end up with a half finished one man project. Really no offence but that is the reality.
14:20:13 <nsITobin> message manager for some window opener async jazz?
14:22:00 <Sompi> integers that don't fit in 16 bits are very rarely needed
14:22:14 <nsITobin> a part of me still feels that isolation of people with opinions has been systatic in some areas .. and primary distraction has been the trivialities we are all on about as if one can't have more than one system or thing.. as they lay the monoculture I my self fell into the trap of the inverse monoculture instead of an anti-monoculture.. i am sure i am not the only one who has made this mistake
14:23:06 <tomman> I can't even find documentation about messageManager on MDN
14:23:15 <nsITobin> tomman: use udn
14:23:24 <nsITobin> realityripple's semi-decent clone
14:23:32 <nsITobin> linked on xulorg and cci
14:24:09 <nsITobin> mostt firefox internal docs were moved off to readthedocs site for firefox source or just vanished
14:24:35 <tomman> oh, so it's not messageManager that is dead, it's "window.opener" that somehow is null
14:24:49 <tomman> and of course trying to access a property of a null... is bound to do Icky Things
14:24:59 <tomman> then why the hell is "window.opener" null?
14:25:23 <Sompi> people with IQ less than 115 tend to see everything in a dualistic way
14:26:10 <Sompi> or to be more specific, people with an IQ less than one standard deviation over the median
14:27:33 <Sompi> I haven't yet figured out how to fix it. They always see only two options and cannot comprehend that there might actually exist more alternatives than those two
14:27:45 <Sompi> No amount of reasoning is enough to fix it
14:28:28 <tomman> Hmmm, the only clue I find regarding window.opener being null is a bunch of stuff that only seems to apply to websites, not XUL content
14:30:53 <tomman> OK, saving the script to a "blah.user.js" and opening that from the browser works
14:31:28 <tomman> still... why the hell window.opener is null on this dialog? And what's causing that paste from clipboard to eventually get null text too?
14:37:11 <tomman> well, imgur is trying its hardest to become unusable for me
14:37:22 <tomman> none of the userscripts I've tried manage to redirect me to the actual images
14:37:46 <tomman> and now their HTML junk wants JS dynamic imports, so all I get is a black screen (that still loads spy scripts from some .ai domain, cute)
14:38:02 <tomman> the userscripts I've tried so far get caught into endless redirect loops
14:40:04 <tomman> any attempt to redirect to the actual image URL just get redirected to the HTML site
14:40:47 <tomman> apparently they now enforce referer checks
14:41:59 <frg_Away> use ibb.co
14:42:53 <tomman> problem is... dealing with 3rd-party links
14:43:15 <tomman> Thankfully I have a personal web host I pay for (in local currency, even!), but people still use Imgur, like developers are addicted to Github
14:49:14 <nsITobin> i don't have an image uploader on linux else i'd go back to using my personal subdomain.. tho I suppose I could get a drop and paste-drop script
14:49:30 <nsITobin> then it would be like imgbb which I been using
15:56:47 <tomman> I'm a dinosaur. I just use FTP :P
15:57:09 <tomman> apparently there is even FTP for cellphones nowadays!
15:57:12 <tomman> tech is wonderful
16:29:07 <nsITobin> tomman: you saying you can do things on the telephone you could only do on com-PUters?
21:29:50 <nsITobin> Sompi: btw i wasn't disagreeing with you earlier if you thought so