OS2 World Community Forum

OS/2, eCS & ArcaOS - Technical => Internet => Topic started by: Pete on July 19, 2025, 02:03:18 pm

Title: Mypal68: A possible alternative browser?
Post by: Pete on July 19, 2025, 02:03:18 pm
Hi All

While browsing osnews.com I saw this:-


Mypal68: Firefox 68, maintained for Windows XP
Thom Holwerda 2025-07-17 Mozilla, Gecko 9 Comments

Do you have a Windows XP retro virtual machine or, god forbid, run Windows XP on your primary machine? You’re going to need a sort-of up-to-date browser, and it turns out Mypal68 offers just that. Terrible name aside, it’s Firefox 68 ported to and maintained to run on Windows XP SP3; SP2 and lower are not supported, but some people do seem to have some success getting it to run on those.

There are issues, of course: there’s a 1.5GB memory limit, and the browser will crash when it reaches that limit, and 64bit builds simpy don’t work at all, so there’s only a 32bit build. Version 74.1.0 was released a few days ago, but that version number doesn’t actually mean the browser is now based on Firefox 74; they had to change the reported version number for extension compatibility.



Could "a sort-of up-to-date browser" currently "maintained for WindowsXP" be of interest to us OS/2 users?

Would it be easier to port than Dooble + qt5/6?

If any of our "Mozilla experts" fancy taking a look: https://codeberg.org/Theodor2/Mypal68


Regards

Pete
Title: Re: Mypal68: A possible alternative browser?
Post by: Roderick Klein on July 19, 2025, 03:54:10 pm
Hi All

While browsing osnews.com I saw this:-


Mypal68: Firefox 68, maintained for Windows XP
Thom Holwerda 2025-07-17 Mozilla, Gecko 9 Comments

Do you have a Windows XP retro virtual machine or, god forbid, run Windows XP on your primary machine? You’re going to need a sort-of up-to-date browser, and it turns out Mypal68 offers just that. Terrible name aside, it’s Firefox 68 ported to and maintained to run on Windows XP SP3; SP2 and lower are not supported, but some people do seem to have some success getting it to run on those.

There are issues, of course: there’s a 1.5GB memory limit, and the browser will crash when it reaches that limit, and 64bit builds simpy don’t work at all, so there’s only a 32bit build. Version 74.1.0 was released a few days ago, but that version number doesn’t actually mean the browser is now based on Firefox 74; they had to change the reported version number for extension compatibility.



Could "a sort-of up-to-date browser" currently "maintained for WindowsXP" be of interest to us OS/2 users?

Would it be easier to port than Dooble + qt5/6?

If any of our "Mozilla experts" fancy taking a look: https://codeberg.org/Theodor2/Mypal68


Regards

Pete

A extremely big chance that this game over. This Firefox 68 version most likely contains RUST code. The question is how far it will help as Firefox 68 was end of life
August 25, 2020.  Before OS/2 VOICE started on the Dooble browser BWW and I looked at Firefox 52 ot 54. That would take at least 6 months of full time work to get from 45.9 to 52.
My initial thought with the very limited resources the OS/2 community is that it will most likely not help to port a 5 year old browser to the platform.

Dave what are your thoughts ?

Roderick
Title: Re: Mypal68: A possible alternative browser?
Post by: Dave Yeo on July 19, 2025, 05:48:30 pm
52ESR was pretty well the last Mozilla that built without Rust so anything newer is out of reach. Going the 52ESR route, there is PaleMoon which was forked off of 52ESR and in theory buildable on OS/2 but it would be a lot of work and while a lot of  modern stuff has been back ported to it and it will handle a lot of sites, it is still limited.
Title: Re: Mypal68: A possible alternative browser?
Post by: mauro on July 26, 2025, 08:58:26 am
I use latest release of Mypal68 since its first version has been released, on a XP SP3 vm.
I'm very satisfact with it; stable, responsive, well working on any webpage I need.
Guess by now that if it ever worked on Warp 4.52 / ArcaOS as it does on XP SP3, seems to me a good progress* under the item "browser".
 if

* Dooble QT5 still engages me with instability, random traps and hidden characters in the webpage
Title: Re: Mypal68: A possible alternative browser?
Post by: Dave Yeo on July 26, 2025, 07:31:53 pm
It's an interesting project. Looking a bit closer, it claims to be based off of PaleMoon but as far as I know, PaleMoon was forked from 52ESR and promised no Rust. I haven't kept close track to see if the no Rust is still there.
Mypal68 has a source.stamp file that points to the last checkin for ESR68, so yes based on ESR68 with I guess a lot of code from PaleMoon's Goanna web engine. Goanna is a fork of Gecko. Yet https://www.mypal-browser.org/ (https://www.mypal-browser.org/) says based on Quantum. Confusing, especially as it says it can use the old XUL based add-ons as well as PaleMoon add-ons.
I couldn't quickly find any build instructions but it seem to use some version of the MS compiler and the GitHub page says 1.2% off the code is Rust.
Out of the various Gecko forks, PaleMoon is the only one that might compile with our tool chain. There's also lots of sharing of code so they all should have similar capabilities. I mostly use SeaMonkey, both on OS/2 and Linux. The Linux version also displays most pages but was forked from 60ESR and contains Rust.
The problem is that we need a knowledgeable developer to spend a lot of time porting even PaleMoon. We have the advantage of all the OS/2 code is in history and our forks but it is a big code base. Also most of these projects have gone 64bit only, including Qt6. Not a show stopper but upstream developers  stop worrying so much about memory.
Title: Re: Mypal68: A possible alternative browser?
Post by: mauro on August 28, 2025, 09:09:18 am
here is another 32 bit browser recently appeared:
 "Supermium is a Chromium-based web browser designed to support older Windows versions while offering enhanced security, performance, and customization options"
Now, said that my knowledge in porting software is near to zero and this browser fits to Windows only, it raise to me two basic user questions:

- Chromium is open source - multi platform browser ?
- does multiplatform browser 32 bit means theoretically portable in OS2/ArcaOS ?

even if I realize that both "yes" answers do not mean automatically that it can be simply ported to OS2/ArcaOS, I'd like to learn -if possible-  comments from you.

https://www.supermium.org/ (https://www.supermium.org/)

their contact email has a funny address : emmawatsofficial54@gmail.com

and the street address 1600 Grand Avenue, London, NW1 6XE, UK  brings several companies link if you research it by Google.
Most of them shares same email address : prorevitaa@gmail.co  , but some others also have emmawatsofficial54@gmail.com.

who are them ?

https://github.com/win32ss/supermium

provided by

https://win32subsystem.live/

please donate to

https://www.patreon.com/win32

......
Title: Re: Mypal68: A possible alternative browser?
Post by: Martin Iturbide on August 28, 2025, 07:56:18 pm
Hello mauro

Thanks for sharing it, I didn't knew about this project.

Quote
- Chromium is open source - multi platform browser ?
As far as I know is open source.

Quote
- does multiplatform browser 32 bit means theoretically portable in OS2/ArcaOS ?
Yes and No. But I prefer someone with more experience to reply this.

For what I see this project support 32bits and 64 bits Windows XP, Vista, and 7. I think we should check the source code of it, the libraries that it uses and see how portable it is to OS/2.

The source code seems to be here: https://github.com/win32ss/supermium

Maybe, as a first step, we can get the Windows 32bits version and see if it runs with Odin.

Regards

Title: Re: Mypal68: A possible alternative browser?
Post by: Dave Yeo on August 29, 2025, 12:04:42 am
Some browser basics first. Most browsers have a backend and a frontend. Mozilla for example uses Gecko and the SpiderMonkey JS engine, with Gecko doing the all the rendering stuff. We use Gecko 45 in our Mozilla apps. The front end with Mozilla is Firefox, also SeaMonkey and Thunderbird, all using the same rendering engine and JavaScript engine. Web pages look basically the same whether viewed in FF, SM or even TB. SM and TB also use MailNews, another backend.
So the frontend is basically the UI. To be portable, the frontend uses OS specific widgets to draw the graphical stuff. Mozilla has widgets for each OS it supports to look native. Mozilla's license is MPL, similar to GPL or LGPL. The backend is usually more portable, though now a days uses some advanced OS stuff.
Chromium is BSD licensed with the source available. Chrome adds a bunch of closed source Google stuff. Chrome/Chromium is the frontend, the backend is Blink and the V8 JavaScript engine. There are a lot of browsers that use Blink and are usually considered Chrome based. This includes our Dooble port.
The problem with Chrome is the widget set. In open source there are 2 main widget sets, GTK, currently usually GTK3 and Qt, mostly Qt6 today.. Chrome and most of the other browsers based on the same backend uses GTK3, which we do not have a port of and porting it would take a lot of time and energy by a knowledgeable developer or team. So most of these Chrome based browsers are not going to work.
The Qt people have ported the backend and some of the Widgets to Qt, latest to Qt6, which also has the SimpleBrowser as a proof of concept. We do have the Qt widget set thanks to years of work by Bitwise, with current work on Qt6 being done by Paul. Wasn't that many changes between Qt5 and Qt6.
So basically most Chrome based browsers including Supermium would need GTK ported to OS/2, a huge job and is not going to happen without a lottery win or such.
We can use the same backend with Qt based browsers, we're doing it with Dooble, which sadly seems the only Qt based browser being worked on. And it only has one developer behind it. The frontend works great on OS/2 but we still need more work on the widgets, drag'n'drop for example needs to be implemented.
Bitwise, and now Paul have also done work on porting the backend but we've run into limits of our OS, including being 32 bit. Until someone very knowledgeable, like dmik of Bitwise can figure out how to proceed, we're stuck.
Another advantage these Win32 browsers have is they can be built on a 64 bit system. We don't have that option, which is one of the big problems, lack of address space for building.
Title: Re: Mypal68: A possible alternative browser?
Post by: Roderick Klein on August 29, 2025, 12:27:16 am
here is another 32 bit browser recently appeared:
 "Supermium is a Chromium-based web browser designed to support older Windows versions while offering enhanced security, performance, and customization options"
Now, said that my knowledge in porting software is near to zero and this browser fits to Windows only, it raise to me two basic user questions:

- Chromium is open source - multi platform browser ?
- does multiplatform browser 32 bit means theoretically portable in OS2/ArcaOS ?

even if I realize that both "yes" answers do not mean automatically that it can be simply ported to OS2/ArcaOS, I'd like to learn -if possible-  comments from you.

https://www.supermium.org/ (https://www.supermium.org/)

their contact email has a funny address : emmawatsofficial54@gmail.com

and the street address 1600 Grand Avenue, London, NW1 6XE, UK  brings several companies link if you research it by Google.
Most of them shares same email address : prorevitaa@gmail.co  , but some others also have emmawatsofficial54@gmail.com.

who are them ?

https://github.com/win32ss/supermium

provided by

https://win32subsystem.live/

please donate to

https://www.patreon.com/win32

......

Its a fair question and it comes up more often. But the first issue with such a port is that its missing OS/2 specific code. I do not know how much work it would be to apply the patches from a Firefox browser to get to compile or write code. (A none Firefox or QT based browser port to OS/2 is a none starter in my mind).

We have something like WINE for OS/2 called ODIN, a system that can run Windows sources on OS/2. But yet again this was last updated oooh 20 years ago. Using Wine is most likely a none starter and debugging the code will be pretty time consuming in my opinion. So we are back to porting the code as the option above.

But even when you get the platform specific code added to the current code base of any browser then the debugging starts.  A compiling browser is not a warranty for a stable browser.
Now I only skimmed the pages of this browser but what do I see:
https://github.com/win32ss/supermium
There is a V8 directory. V8 is a javascript engine also used by the current Dooble browser.
That is one of the things Dmitry is going to work on again to fix (from BWW).

So its possible if you port this browser we bump into the same V8 issue.

Also thinking more about this I think this talk about a 32 bit browser is partly, well not very helpfull. You can maybe trim the memory usage but the other issue is that webpages have become WAAAAY more complicated in terms of code your PC and how much CPU is needed to display it on your system.  I am not saying a 32 bit OS is not useable with an older CPU. But a 32 bit browser still needs memory to store the webpages.

My overall assessment is that with the limited human resources we have in the OS/2 community its a waste of time to look at other browsers to port. We would simply be spinning our wheels.
I am gladd to be proven wrong.  Any port would take 1 person full time and take about 1 year at least to get something workable (beta stage). And it might even be much longer!

Roderick Klein
President OS/2 VOICE


Title: Re: Mypal68: A possible alternative browser?
Post by: Dan Eicher on August 30, 2025, 02:09:42 pm
Thank you Dave for the summation.

You really helped improve my understanding of browser tech, and the challenges for older platforms.

Also, I loved:
>without a lottery win or such.
Title: Re: Mypal68: A possible alternative browser?
Post by: Dave Yeo on August 30, 2025, 06:30:50 pm
Other challenges.
Browsers have gone multi-process where each tab or at least each domain runs in its own process. They communicate with local sockets and our local sockets, meant for things like syslog, isn't up to it. If you run Dooble, you might notice it is in single process mode, which is only there for debugging purposes and is unsupported. If you put Dooble into multi-process mode, it becomes unstable. Multi-process also uses more memory, not too bad as most of the code is shared in DLL's.
Compiling. Firefox moved to Rust, memory intensive and a really crappy language for compatibility with itself. For example, something compiles with version 1.0.16 won't compile with 1.0.17. numbers might be off but minor version changes break stuff. Probably unusable on a 32 bit platform.
Modern C++ also can be pretty memory intensive, I've seen 2GB of memory used to compile one file. Makes it hard to run multiple jobs without running out of memory. Memory also gets fragmented where sometimes only a reboot will fix the fragmentation.
Linking the large DLL's, same problem with memory. Often need the whole address range (VIRTUALADDRESSLIMIT=3072) to link.
Actually it is pretty amazing how well OS/2 does in the memory department considering it was optimized to run on 4MB (really needed 8MB to use the WPS) and was developed in a time when 16MB of ram was a lot.
Title: Re: Mypal68: A possible alternative browser?
Post by: Mathias on January 15, 2026, 04:36:01 pm
Hmm... just a question out of curiosity.

> Often need the whole address range (VIRTUALADDRESSLIMIT=3072) to link.

Here's a naive question from a non-developer: Would cross compiling be a solution?

I have seen that you can use dist-cc on 64bit Linux to build ARM software on an x86_64 compiler.
On Gentoo Linux I have used dist-cc on a 64bit machine to build 32bit software on an x86_64 compiler for that other 32bit machine.
I have seen that you can build Windows NT executables with your Watcom OS/2 software. Would that also work the other way around? Compile a 32bit OS/2 application on a 64bit Windows 11 GCC *for* OS/2?

Let's say you have a 64bit machine with a huuuge ammount of memory. In my naive thought, wouldn't that mean, you could utilise as much memory as 64bit adressing would allow, and cross-compile a 32bit application, that (once compiled) is working within its 32bit memory range? - And if so.. then may the target not be a Windows executable, but OS/2?

I mean if compiling and linking needs 64 bit adressing, since stuff grew so huge over the years.. okay.. be it. But running the application later .... does that imply to have a 64bit environment also? Because if not, we could have all that big stuff on 32bit also. - And if that could be compiled then for OS/2 also on that 64bit machine... then this whole memory thing (while compiling and linking) would be solved.

Mathias
Title: Re: Mypal68: A possible alternative browser?
Post by: Dave Yeo on January 15, 2026, 07:38:58 pm
Yes, cross-compiling is possible.
Currently OpenWatcom runs on Linux and Windows and can build OS/2 programs fine. The official OW is still 32 bit, but running on a 64 bit platform gives a full 4GB address range rather then the (if lucky) 3.5 GB or so on real hardware.
There is also Jiri's 64bit fork of OpenWatcom which should also work. I'm not sure if he has ported the latest OS/2 OW fixes to his fork so possibly it might not work quite as well as the official OW but will give endless memory for building.
The problem with GCC is we use a fork of GCC so that would have to be ported and we also use various OS/2 binaries when building, which would also need to be ported. And the whole build environment needs to exist on the 64 bit platform. Both Windows and Linux should be possible with Linux perhaps the best choice.
So in theory with some work, GCC and the EMX tools can be rebuilt on Linux etc with some work. These would still be 32 bit but have access to the full 4GB's of address range.
Porting to 64bit could also be done but would likely turn out harder, perhaps much harder as all the basic types double in size.
Power is getting shut off so I'll continue
Edit: Hydro is finished doing whatever they did at the pole and I've lost my train of thought.
Basically, an OS/2 developer environment needs to be ported, things arranged and such. Possible that perhaps OS/2 in a virtualbox or such and using Samba could work with the host system.
Big job that is possible if a knowledgeable developer spent a bunch of time.
Title: Re: Mypal68: A possible alternative browser?
Post by: Mathias on January 16, 2026, 09:38:11 am
Sounds worth the effort! : ) - Cool stuff. Thank you for your time! - Great explaination!
Title: Re: Mypal68: A possible alternative browser?
Post by: Jochen Schäfer on January 16, 2026, 10:07:31 am
A cross compile environemnt would really be neat, especially because more and more software move to build environment like meson, which also have to be ported to OS/2 otherwise.
Title: Re: Mypal68: A possible alternative browser?
Post by: TeLLie on January 16, 2026, 11:54:03 am
Hi

Ko ported meson..

Kind regards Tellie
Title: Re: Mypal68: A possible alternative browser?
Post by: Jochen Schäfer on January 16, 2026, 12:12:46 pm
Thanks for the info, but it's just one example. And, don't get me started on IDEs ;-)
Title: Re: Mypal68: A possible alternative browser?
Post by: Andi B. on January 16, 2026, 03:59:13 pm
Thanks for the info, but it's just one example. And, don't get me started on IDEs ;-)
Here maybe Dmitriys approach (SSH file sharing to OS/2 box if I understand it correct) may help. Or running OS/2 virtually in f.i. Linux. But until now I hate doing such. Of course you don't need a browser for OS/2 anymore when running within VBOX....
Title: Re: Mypal68: A possible alternative browser?
Post by: Dave Yeo on January 16, 2026, 04:56:34 pm
For compiling large projects like Qt, having a stand alone box would help as you can have the minimal other stuff running to save memory. Likewise with a VBox instance, though I find that with VBox, multiple CPU's doesn't work too well. Running multiple jobs when building can really speed things up if you don't run out of memory. Some of this C++ code uses a lot of memory. I've seen compiling one file take a couple of GB's of memory and even hit my swap file with 3.4 GB's of visible memory when running multiple jobs.
Cross compiling from Linux would be best and in theory quite possible. Linux is way faster at compiling projects then OS/2. OTOH, OS/2 is quite a bit faster then Windows compiling this stuff where there is lots of forking of processes.
Title: Re: Mypal68: A possible alternative browser?
Post by: CalYY on February 19, 2026, 08:44:11 am
Alternative to MyPal:
http://rtfreesoft.blogspot.com/

I use IceApe from there on XP. Less memory used than MyPal
Title: Re: Mypal68: A possible alternative browser?
Post by: Martin Iturbide on February 19, 2026, 04:29:30 pm
Hello

It is the first time I hear about IceApe. It is mentioned that IceApe is based in SeaMonkey.

Win32 - Version 1.8 : https://github.com/roytam1/iceape-uxp/tree/winbuild

I see references to Iceape 2.7  - https://user-agents.net/browsers/iceape/versions/2-7
But I can not found the source code.

I would need something with more knowledge on this area to know if IceApe is compilable under/for ArcaOS.

Regards
Title: Re: Mypal68: A possible alternative browser?
Post by: Dave Yeo on February 19, 2026, 04:41:22 pm
It should compile on OS/2, it uses UXP, the Gecko fork that PaleMoon uses. PaleMoon itself should compile on OS/2 but it would take a lot of porting. We do have the widgets etc from Gecko (Firefox) but it would be quite a bit of work from someone knowledgeable. And that's the problem, an interested programmer with a lot of time.
Title: Re: Mypal68: A possible alternative browser?
Post by: CalYY on February 25, 2026, 02:37:20 am
Iceape= Seamonkey, Icedove= Email, Iceweasel= browser. Originally rebranded Mozilla by Debian Linux

 Source code:
https://mirror.cedia.org.ec/hyperbola/sources/other/

Windows version has updated security and javascript.  Should be able to compile for OS/2
Title: Re: Mypal68: A possible alternative browser?
Post by: mauro on February 25, 2026, 07:42:13 pm
32 bit?
Title: Re: Mypal68: A possible alternative browser?
Post by: Dave Yeo on February 26, 2026, 12:34:05 am
It should compile to 32 bit, above the poster says he is using it on XP, which is generally 32 bit.
Problem is that it would need a lot of porting. As I mentioned, we have some of the needed stuff such as the widgets but there are classes that need porting and who knows what else. I don't have the skills to compile iceweasel, if someone else did, I could probably compile iceape and icedove. Just like with our Mozilla where once Firefox was done, I could do SM and TB, though they still have a few bugs such as oauth2 being broke.
Title: Re: Mypal68: A possible alternative browser?
Post by: David Kiley on March 13, 2026, 07:15:10 am
It seems to me from what you guys have all said (that are more knowledgeable than me) that the 4GB/32Bit limit is pretty much a ticking time bomb.

So I'm kind of thinking now it's a waste of resources to try to port any browser at this point since it will be doomed to fail eventually anyways. (And I don't think ArcaOS being made 64bit is an option? Could be wrong)

So i've come over to the side that maybe another solution like winflector or some other form of RDS solution might really be the only future - since that breaks away from the 32bit limitations by thinking out of the box.

Like, for example, maybe providing funding to winflector for them to make a native client app so people can get the full functionality of winflector (if they could be convinced there is enough interests from the os/2 community). That would be a lot more productive since it would actually solve the problem of breaking the 64bit barrier through virtualization.

One question I was wondering about though: So if you are using a solution like winflector, which is letting you virtualize a 64 bit browser while you are still on a 32bit/4gb machine - Does the fact that your machine is 32 bit and limited in ram effect the performance of the virtualized session?
I haven't had a chance to fully test it myself yet.
Title: Re: Mypal68: A possible alternative browser?
Post by: Andi B. on March 13, 2026, 09:00:43 am
Why do you think about winflector? Isn't this another remote access software where the real program runs on another hardware on another OS and only the screen output is mirrored to your machine?

You can do such thing easily with our RDP port. Setup a Linux machine, install xrdp/xvnc and connect to it from your ArcaOS desktop. Either freerdp (much preferred) or vncclient. Works quite good. Probably will work the same with a remote windoze machine. But no need bothering with that crap.
Title: Re: Mypal68: A possible alternative browser?
Post by: David Kiley on March 13, 2026, 09:21:01 am
Why do you think about winflector? Isn't this another remote access software where the real program runs on another hardware on another OS and only the screen output is mirrored to your machine?

You can do such thing easily with our RDP port. Setup a Linux machine, install xrdp/xvnc and connect to it from your ArcaOS desktop. Either freerdp (much preferred) or vncclient. Works quite good. Probably will work the same with a remote windoze machine. But no need bothering with that crap.

Not sure why you consider it crap honestly. It might not be your preference but it's still a good product.
It has several advantages. Instead of having to login to the whole server as you do with RDP - you can login to just a single app automatically in it's own application window, so it feels like a real application.
It's also much faster and optimized for speed then a standard RDP console.
And if you offered it as a hosted service to others it has better security controls.
But you seem to feel strongly about it :).
Title: Re: Mypal68: A possible alternative browser?
Post by: Andi B. on March 13, 2026, 10:23:59 am
Why do you think about winflector? Isn't this another remote access software where the real program runs on another hardware on another OS and only the screen output is mirrored to your machine?

You can do such thing easily with our RDP port. Setup a Linux machine, install xrdp/xvnc and connect to it from your ArcaOS desktop. Either freerdp (much preferred) or vncclient. Works quite good. Probably will work the same with a remote windoze machine. But no need bothering with that crap.

Not sure why you consider it crap honestly. It might not be your preference but it's still a good product.
It has several advantages. Instead of having to login to the whole server as you do with RDP - you can login to just a single app automatically in it's own application window, so it feels like a real application.
It's also much faster and optimized for speed then a standard RDP console.
And if you offered it as a hosted service to others it has better security controls.
But you seem to feel strongly about it :).

Maybe I wasn't clear enough, mind English is not my mother tongue, but with crap above I meant windoze. Can't comment on winflector. But the mentioned OS I know long enough and good enough and know how it evolves. So I've a lot of arguments why I know it's crap. At least this days and for serious computer users. But I've already wasted too much time in my life with this OS.
Title: Re: Mypal68: A possible alternative browser?
Post by: David Kiley on March 13, 2026, 10:39:26 am
Maybe I wasn't clear enough, mind English is not my mother tongue, but with crap above I meant windoze. Can't comment on winflector. But the mentioned OS I know long enough and good enough and know how it evolves. So I've a lot of arguments why I know it's crap. At least this days and for serious computer users. But I've already wasted too much time in my life with this OS.

Oh OK I understand you now. I get your feeling but wouldn't it be worth it to use a fraction of Windows if it meant a functional browser?
Title: Re: Mypal68: A possible alternative browser?
Post by: Andi B. on March 13, 2026, 02:02:34 pm
Quote
but wouldn't it be worth it to use a fraction of Windows if it meant a functional browser?
No. There are better alternatives. Cheaper. More secure. And without reading and using (stealing) your data. And without giving permission to delete your data.
Title: Re: Mypal68: A possible alternative browser?
Post by: Dariusz Piatkowski on March 13, 2026, 03:18:36 pm
David,

...One question I was wondering about though: So if you are using a solution like winflector, which is letting you virtualize a 64 bit browser while you are still on a 32bit/4gb machine - Does the fact that your machine is 32 bit and limited in ram effect the performance of the virtualized session?...

I have been using the FreeRDP app in such a fashion for a long time now.

Yes, you need another box, be it Win or Linux, that's actually hosting the application itself, but beyond that (which honestly is a tiny price to extend OS/2 usage in real-world scenario quite significantly) this is a very realistic solution.

Sure, it's no native app, no doubt about it, and sometimes little quirks pop up here and there, especially brought on by the hosting OS updates/changes, but otherwise it's a pretty decent setup.

In my case I've been using ThinStuff (licensed) but the last time I checked they did not support Win10/11, not sure if that has changed or not though.

Winflector is in fact my next 'go-to' as I have a much beefier Win11 box on my LAN that can really handle multiple such RDP sessions with ease.
Title: Re: Mypal68: A possible alternative browser?
Post by: JTA on March 13, 2026, 03:34:22 pm
"the 4GB/32Bit limit is pretty much a ticking time bomb"

"Why do you think about winflector? Isn't this another remote access software where the real program runs on another hardware on another OS and only the screen output is mirrored to your machine?"

Both of these comments are currently "problems" with OS/2, but they are not showstoppers in any way. The whole point of "ArcaOS - Thrusters on Full" (AToF ... over in the virtualization subforum) is to solve these and other problems such that you can be very productive in OS/2, and *not experience* any of these problems. For example:

AToF runs on one machine (your desktop, your laptop), and utilizes all the reources of that machine (ssd's, 32 or 64gb ram, cpu's w/ multiple cores ... and makes all of these resources available to OS/2. It does this magic with a stripped down version of win10, such that the OS is a slave to OS/2's need ... win10 doesn't steal your data, it doesn't do anything other than provide resources to multiple OS/2 vm's on one box.

Multiple OS/2 vm's free us all up from nearly every single os/2 problem that this forum is full of:
- can't compile a massive app in one OS/2 vm? dedicate an OS/2 vm to compiling, or throw multiple vm's at it ... problem solved.
- can't access the occasional windows app? winflector, running on the win10 hostOS, is there to give any OS/2 vm a windowed win app, 64-bit, running on your os/2 desktop ... problem solved.
- can't access a piece of hardware, as there are no native OS/2 device drivers? win10 hostOS provides this, and makes everything available to the OS/2 vm's running ... all of them, all on one machine ... problem solved.

There's plenty more, just not enough room in a thread reply to show off all the features ... I demo'd this at Warpstock 2023, and wrote all of the AToF threads in the virtualization subforum, all with the goal of just using tools in the toolbox to solve all of OS/2's problems. Nothing magical, other than thinking outside of the box, and stringing tools together (win10 ameliorated, virtualbox, winflector, etc.) to ensure that OS/2's insurmountable limts aren't really a problem. These limits (4gb ram, 32-bit, no modern browser, never enough native device drivers, etc.) are likely to never go away, but it turns out that we don't need to worry about them anyway.

We can run 32-bit OS/2 forever, and get at any 64-bit feature we want ... just assemble those tools in the fashion that *you choose* ... if you choose not to, that's OK as well. It doesn't change the fact that OS/2 is not limited.

Just take one inexpensive box (my $200 ebay 2018 dell laptop, i7, 32gb ram, ssd), AToF it, and I can run multiple OS/2 vm's, on one physical machine. It isn't Windows, even though win10 is my hostOS (could also be linux, etc.) ... it's OS/2 (ArcaOS) all the way! I don't see the hostOS, I don't think about it at all, but Windows is slaved to OS/2 (the way it should be), and I get to keep using OS/2's very unique features ... forever!

Anyone else can, if they choose to ...

NOTE: we need to get ArcaOS and Winflector folks together, to produce an OS/2 native winflector client, which is a port of the linux 32-bit native client to OS/2. I've done much to request this, but it's been "crickets" from both parties. Everyone else should pile on as well, repeatedly, until these two sides finally get together to get it done.

Title: Re: Mypal68: A possible alternative browser?
Post by: David Kiley on March 13, 2026, 11:24:32 pm
Regarding:
"the 4GB/32Bit limit is pretty much a ticking time bomb"

I just meant coding a native web browser, even if achieved now, is going to run up against a memory limit eventually.
Whereas a virtual solution like winflector won't.
So I was saying I think the resources should go into another solution.

I've been in discussion with winflector about partnering with them to offer the hosted solution, and also talked to them about making a native client software for os/2, but it seems they are not interested:

"I'm afraid it's not so easy. To engage in software development for a niche environment, you need hardware, software, and application development tools. In addition, you need to learn a new, unfamiliar environment, and above all, you need to distract the team from the ongoing work of maintaining and developing the core versions of our software. It's probably safe to assume that a dedicated client for OS/2 won't be created, but you can try using the browser version, assuming there are modern versions of Opera, Chrome, or Firefox that work on OS/2."

And the last sentence is interesting.. as in the future it's possible the browser won't be modern enough for winflector to work.
Not sure about that though.
Title: Re: Mypal68: A possible alternative browser?
Post by: JTA on March 14, 2026, 01:06:42 am
That last bit from Winflector is interesting, because I offered to buy them an ArcaOS seat & such (afaik, the only $$$ part of a dev machine's software, as all else seems to be free or open source?).

I'm not sure, but I thought that an existing Winflector 32-bit linux client would mean a fairly easy port to OS/2.

So, the bigger problem is that they think OS/2 is too "niche". Unless AN gets involved, it may be hard to dissuade them from this view ... if there are no seat numbers, there's no way to judge interest.

How do we get AN, EComStation, and OS/2 Warp seat count? I can see potential seat numbers in the thread view counts on OS2World ... other sources?
Title: Re: Mypal68: A possible alternative browser?
Post by: David Kiley on March 14, 2026, 02:49:33 am
Yeah. I asked them if we achieved 200 virtual users and all 200 users were willing to pay something for the client app..and that was their response. That would have been over 6 thousands of dollars in "seats" plus the client license fees. That would be a pretty high bar and I would think enough financial return.So they are really not interested in developing a native app. in one sense I get it.. coding for os/2 isn't an easy venture. They don't even have a native app for os/x which has a lot more users

 The big question is if we can continue to make the browser compatible enough. Like if they come out with html6 will it be codable with what we have. I would think coding the base might be easier then the whole browser but I don't know.
Title: Re: Mypal68: A possible alternative browser?
Post by: Dave Yeo on March 14, 2026, 05:56:51 am

I'm not sure, but I thought that an existing Winflector 32-bit linux client would mean a fairly easy port to OS/2.

Not really, unless it uses Qt which is about the only widget set we share with Linux. The Windows 32 bit version would actually be easier as Windows and OS/2 are cousins. It would still be hard as hell, perhaps easier to go the Odin route. A small team might be able to do it in a year including testing etc to make it stable. Can't see it making business sense.

Quote
So, the bigger problem is that they think OS/2 is too "niche". Unless AN gets involved, it may be hard to dissuade them from this view ... if there are no seat numbers, there's no way to judge interest.

How do we get AN, EComStation, and OS/2 Warp seat count? I can see potential seat numbers in the thread view counts on OS2World ... other sources?

It's hard to say how many OS/2 users there still are, but I doubt there's many using OS/2, eCS and AOS full time. Even if we knew how many licenses Arca Noae has sold, it wouldn't mean much as a lot of them are likely tyre kickers. I've seen a few people from the past show up with an AOS license, play a bit and move on.
AN itself doesn't have much in the way of resources, otherwise they could have found a skilled developer to update our existing browser years ago, before we fell so far behind.
Title: Re: Mypal68: A possible alternative browser?
Post by: David Kiley on March 15, 2026, 03:45:45 am
they could have found a skilled developer to update our existing browser years ago, before we fell so far behind.

Did Ecomstation contribute to the browser? Or something changed that they didn't need to?
Title: Re: Mypal68: A possible alternative browser?
Post by: Dave Yeo on March 15, 2026, 08:58:53 am
they could have found a skilled developer to update our existing browser years ago, before we fell so far behind.

Did Ecomstation contribute to the browser? Or something changed that they didn't need to?

Yes, first they paid Rich to fix the printing. That was FF v4. Walter and I took it to v10, then Serenity paid Bitwise for 17 to 45, actually at the end the community was paying. I supported SM and TB, easy once Firefox built and ran as they all use the same engine. At that point I was left alone with a few resources from Arca Noae and I don't have the skills to do what was needed for 52 and it took awhile for me to even get it building to the point where we needed someone much more knowledgeable. I also ported it to the newer GCC and applied a lot of security fixes from 10four (Mac fork of 45).
In theory someone knowledgeable  could continue development to the latest PaleMoon with our source, though PaleMoon has really rearranged the source.
Meanwhile Bitwise thought that porting Chromium would be a better route using Qt. Turned out pretty hard, newer GCC was needed, lots of libc updates and got Qt5 to where it is now. Development kind of stopped due to issues updating Qt5 on our system and then war broke out and the developer left Russia and spent quite a bit of time stateless. Paul applied the Qt5 patches to Qt6 but wasn't skilled enough to really go on. One problem is that Qt6 doesn't officially support 32 bits, so no testing by the Qt folks and no worries about freeing memory at times.
At the beginning of the month, Dmitriy, Bitwise's lead developer, did post that he expects to get back to working on Qt6 in the next couple of months.
Meanwhile KOMH has ported our build system to 64bit Linux which should really help. Memory has been a problem compiling browsers since FF10 where I needed VAL=3072 and at the time with only 1.5 GB of actual ram, the swap file could overflow (2.1GB limit) and crash the system if I did anything else while building. Linking some of the Chromium DLL's is even harder due to the memory limits of our system.
It's actually amazing what we've accomplished in the way of compiling browsers. We were the last 32 bit OS that could build Firefox and considering that most of the design of OS/2 was done in the early 90's or earlier...

Title: Re: Mypal68: A possible alternative browser?
Post by: David Kiley on March 15, 2026, 11:40:36 am
Wow.. thanks for all of your (and others) work. It's fascinating to read everything that happened. And yes impressive what was achieved for so long on a 90s operating system. When I was actively using ecomstation as my primary desktop, the browsers sure kept up for a long time. I had no trouble with any sites back then.

I wonder how many sales Arcaos gets compared to the eCs days. I assume the numbers have gone down as enterprise companies might retire or upgrade their aging equipment eventually.
Title: Re: Mypal68: A possible alternative browser?
Post by: JTA on March 15, 2026, 07:46:02 pm
"I've been in discussion with winflector about partnering with them to offer the hosted solution, and also talked to them about making a native client software for os/2, but it seems they are not interested:
  "I'm afraid it's not so easy. To engage in software development for a niche environment, you need hardware, software, and application development tools. In addition, you need to learn a new, unfamiliar environment, and above all, you need to distract the team from the ongoing work of maintaining and developing the core versions of our software. It's probably safe to assume that a dedicated client for OS/2 won't be created, but you can try using the browser version, assuming there are modern versions of Opera, Chrome, or Firefox that work on OS/2." "

Winflector is a commercial company, with proprietary software/code. But, the above still seems like it's possible, it just needs:
1. a free ArcaOS license (possibly two ... I've offered to buy this for them)
2. an OS/2 (emx?) build VM ... we should be able to get this detailed and built for them
3. an OS/2 developer ... (without "distracting" their existing devs)

The last one is on us ... Winflector just need to bend a little. Sign up Paul S., or KOMH, or anyone else (with a non-disclosure and other legalese), and now they have a dev dedicated to the effort (of just porting the linux 32-bit to an OS/2 native). If it's not these two, and there could be plenty of reasons why not, then we need to find/develop others, starting with detailed "here's a build environment" (emx, gcc, whatever) plans.

All are big asks ... Winflector has to bend (open up that piece of software), existing devs have to participate (or we grow more of them), AN could help by revealing "seat count" data, if not restricted in some way (their business practices, IBM, etc.) ... we all have to help with $$$ (I have offered $$$, in licensing terms, and could offer other help).

A native Winflector OS/2 client is the last piece of the AToF puzzle for me ... the HTML5 client (Firefox v45) gets me 95% of the way, and everything works. But ... the OS/2 native client would be the finishing touch!



Title: Re: Mypal68: A possible alternative browser?
Post by: David Kiley on March 15, 2026, 11:01:11 pm
A native Winflector OS/2 client is the last piece of the AToF puzzle for me ... the HTML5 client (Firefox v45) gets me 95% of the way, and everything works. But ... the OS/2 native client would be the finishing touch!

Regarding the above. It would be nice if they were willing to bend but I think it's a dead end. We are just going to have to be happy with the html5 option. I've noticed from their forum that they are very firm when they don't want to do something.

Also consider that they don't even have a mac client..which has millions of users probably.

I ran a 9 line BBS (online service) in the 90s by an innovative company called PowerBBS. They had made the first graphical BBS that was like AOL and very cool but it also supported the dos like terminal services that 99% of people used at the time. Our big problem with them was that the support for "dos" bbs games didn't work very well. When I tried to talk to them about it they would just say "they are not going to reinvent the wheel" even though I was spending a ton of money with them and needed it to be competitive. Sometimes you just have to take a no as a no regardless if it shouldn't be a no.

Also, the goal I had presented them was at least 200 seats (which I calculated out of curiosity with my partner status) and that would have come to $8,398,00. So if that amount of money didn't motivate them then a free arcaos license means nothing to them. And I didn't even think the metric was actually achievable frankly. When I posted on this forum (the most rabid os/2 supporters) about selling one seat as a hosted solution I got crickets in response. I'm not upset by that.. you have to provide value regardless of if you think something is valuable. I sell on eBay and I could think my products are the greatest thing in the world but that doesn't mean anyone will buy it if they don't agree. It reminds me of an interview with Steve Jobs in which he said he had tons of products he loved but didn't motivate people. But when he saw the LaserWriter (mac lazer printer) he instantly knew millions of people would want it since it was one of the first that offered Adobe's PostScript fonts. He just knew everyone would love it.

The problem with os/2 is i'm not sure everyone really *loves* it.
A lot of os/2 clients are enterprise using legacy applications and they probably don't care about the internet. Only users like you and me care about that kind of thing. So even if ArcaOS has sold 20,000 seats it doesn't mean those users are motivated for a browser.