Author Topic: Adding an extra thing to the ArcaOS strategy ?  (Read 1931 times)

Martin Iturbide

  • OS2World NewsMaster
  • Global Moderator
  • Hero Member
  • *****
  • Posts: 5659
  • Karma: +50/-1
  • Your Friend Wil Declares...
    • Martin's Personal Blog
Adding an extra thing to the ArcaOS strategy ?
« on: February 09, 2026, 04:30:47 pm »
Hello

Today I feel like brainstorming, so I'm sorry, and please don't get mad if the idea is not good.

I want to talk to Lewis about adding an additional thing to the ArcaOS strategy. The idea is to start replacing (when possible) IBM OS/2 close source components (binaries, .exe) with open-source alternatives as part of their strategy.

The idea is to embrace, when possible, every little replacement for OS/2’s EXE, DLLs, and even MSG files, etc. in favor of open-source alternatives that does the same, and possible the user will not even notice the difference.  This will be following the idea of Lynn Maxson (ex SCOUG OS/2 member) trying to get each component replaced little by little instead of big changes in architecture.

Yes, Arca Noae are already doing it, they replaced view.exe with newview for exampre, but it is not part of their strategy. The idea will be that if someone comes with a replacement of “bldlevel.exe”, Arca Noae can took the time to evaluate the technical and legal aspects, and if it is approved to be included in ArcaOS.

"Don't fix what it is not broken". The advice is good, but we are still having the dependency on close source that can not be improved. They idea will be to have fully tested replacements so the user doesn’t notice the change, or even get new features.

I think that as a community we can provide suggestions of what to replace, testing of the software and even creating documentation. While some other with higher technical level can even create the little replacements.
 
Please notice that the idea is to add this extra thing to the Arca Noa strategy, I don’t mean it is the only thing to do, it is just to add something else to what they are already doing. 

What do you think? Should we try it?

Regards

Martin Iturbide
OS2World NewsMaster
... just share the dream.

David Kiley

  • Full Member
  • ***
  • Posts: 159
  • Karma: +3/-0
Re: Adding an extra thing to the ArcaOS strategy ?
« Reply #1 on: March 15, 2026, 11:50:43 am »
ArcaOS could embrace the apple strategy when they created OS/X of using an open source base system (in their case Darwin) with thier own commercial propriety system on top of it (keeping the OS/2 UI for example like the workplace shell/WPS). And, they had created the "classic" mode to allow old os 9 apps to still work.
Would require them embracing a totally new direction though.

Sounds kind of complicated to just replace single parts with open source technology - so at some point it makes more sense to fundamentally change the rules of the game like Apple did I think.
I mean, I know it takes less resources to just keep going with what they have.
But at the same time, at some point it makes more sense to jump onto a whole new ship if the old barge is starting to rust and getting holes and taking on water, as much as I love the 90s OS/2 :).

If ArcaOS embraced a base system like Darwin or some of unix, programming a browser would be simple.
I know a lot of users woud prefer ReactOS or something Dos based since it's closer to the Os/2 Core. Yet ReacOS has struggled themselves for a decade to get beyond beta and don't have a functional browser themselves.
So it might be time to embrace reality and create something new based on what is, if ArcaOS has the guts to do so.

I mean, do I need yet another version of os/2 warp that works on a 2026 motherboard? Not at all.
Do I need a functional browser? Yes.
They are already spending a ton of resources but just in the wrong direction.

I think it's time to relegate the old IBM license to a virtualized session inside a totally new OS, while keeping WPS on top of it, which is where the user has always spent 99% of their time anyways.
My 2 cents.
« Last Edit: March 16, 2026, 01:18:45 am by David Kiley »

JTA

  • Full Member
  • ***
  • Posts: 106
  • Karma: +2/-0
Re: Adding an extra thing to the ArcaOS strategy ?
« Reply #2 on: March 17, 2026, 07:42:32 pm »
David Kiley's comments:  "I mean, do I need yet another version of os/2 warp that works on a 2026 motherboard? Not at all. Do I need a functional browser? Yes. They are already spending a ton of resources but just in the wrong direction. I think it's time to relegate the old IBM license to a virtualized session inside a totally new OS, while keeping WPS on top of it, which is where the user has always spent 99% of their time anyways."

We all want a "functional browser", and most want it within OS/2 (ArcaOS). But, no need to beat up on AN, because it's their ball (their OS), and they've chosen not to do apps. This probably makes sense, if you assume their primary focus is all the business users of OS/2 licenses.

So, the problem is ours (the community of retail OS/2 or ArcaOS users) ... we just need a solution, and it isn't to wait on developers to develop the "os/2 browser", or move "os/2 to x64", as discussed in too many threads. That will either never happen, or will take many more years/decades ...

IMHO, for "right now", the quickest solution here is to AToF your new hardware (which does put OS/2 or ArcaOS into a vm) ... AToF can happen now, if you choose to do it. It doesn't require AN to do anything for us, so, we should stop beating up on them (for things they've already stated "they aren't doing").

AToF does everything you describe in your last sentence of the commented text above. You'll have a modern browser on your PM desktop. And, any other x64 and/or modern app (mostly written for windows) on your PM desktop.

(detailed threads of AToF in the virtualization sub-forum ... and yes, I'm working on a simplified version of it. I might've presented it at Warpstock 2025, but they still don't have a good "remote attendee" solution, among other problems)

Mathias

  • Full Member
  • ***
  • Posts: 203
  • Karma: +3/-0
  • using ArcaOS
    • IRC
Re: Adding an extra thing to the ArcaOS strategy ?
« Reply #3 on: March 18, 2026, 08:42:41 pm »
Dear Martin,

I think .. in theory yes, it is an interesting idea to replace as many components as possible with opensource alternatives, BUT.. besides all the porting work, this will kill the OS one day.

Reason why? - Imagine you have replaced a good chunk of tools and maybe even main components.
Next day the open source community decides ProjectX (or some dependencies) move on to Rust or some other compiler, that we are not able to compile for OS/2 with.
Now you're operating system is busted, if you can't move back to the old closed source components. (maybe due to dependencies)

Of course you'd say the components are old anyway; Why would it hurt being unable to update the newly replaced elements also.
Well.. opensource software usually comes with loads of dependencies. In the end it's very well possible to end up at a point where you can't update anything anymore since everything depends on anything else.

These dependencies are fine as long as you can satisfy them. But which each dependency failing the risk raises that the whole thing is breaking apart.

So.. I'd strongely suggest to remain in the own environment, and NOT to depend on opensource too much.
It is a nice idea in a world, where the are no 32bit barriers and where you could move along with all the requirements the opensource platform has and will have, and in a world where there are enough developers for the platform. Could work out then : )


--> On the other hand, I do understand the need to chance "something" in order to become more modern. And somewhere we need to start.
OpenSource has lots of developers and OS/2 only has a few. - So shouldn't hurt to move to OpenSource more. Correct. But the world spins faster than we could replace components. (Currently we realise environment is aging by the day, and we need to get creative to overcome the next obstacle and again and again..)
Somewhere you'd always hit the 32bit barrier, and with even more linux distributions moving to 64bit only, dropping 32bit completely, air is getting thinner for 32bit ports. (Yes up2date 32bit software won't be around forever..)

Since OS/2 is closed source, logic dictates, to cut the ropes at a certain moment in time. The future might be an OS/2-look-alike operating system that 100% is opensource, and therefore is compliant with all the changes opensource brings in the future.
An operating system that is 64bit, that is at the pulse of the current time and updated to current software versions available in the opensource world.
An operating system that looks and feels like OS/2 (if I could wish for Warp 3 style), but actually is native linux.
For me, there would NOT even need to be compatibility with native OS/2 applications. Who needs that may buy ArcaOS please.

The main focus of that opensource-compliant OS/2-look-alike would just to have the look and feel of OS/2 that we all love so much. Same dialogs, same shortcuts, same window decorations, etc. All the settings need to be at equal places.
Since it is current opensource, you could have an up2date dosbox, qemu, vbox, etc to run your retro stuff.

Everything would work out-of-the-box since basically it IS linux and no longer OS/2. (but feels like OS/2)

For the OS/2 community, this argument is invalid though, since the OS/2 folks want their native OS/2.... which we can only have either one or the other.
Mixing up the two worlds is only possible to a specific extent. At some point in the future you just need to cut the ropes and recreate the whole thing. Just you won't get the under-the-hood-functionality.. lilke natively run OS/2 stuff.. or Windows 3.x or DOS, since we do lack the code.
But you would get an up2date operating system, that feels like OS/2 and which you could ACTUALLY use on a daily basis for EVERYTHING that comes to your mind. Like in the old days..

For ME, OS/2 and ArcaOS is NOT for daily use. I see it as an enthusiast operating system. Retro is the top thing at the moment, and it is wonderful to NATIVELY run ArcaOS (OS/2) on a current machine you can buy in a store. ArcaNoae does an excellent job with that, and I cannot thank them enough for their time and effort!

What do I do with OS/2 and ArcaOS?
- Mainly retro stuff like running old games
- trying out old development environment
- trying out CoBOL
- feeling young again when I run (and change) my own QuickBasic code from back in the day
- doing disk forensics
- text work like translations
- listening to CDs (yes compact discs) like I did in the 90ies with the native OS/2 CD player (yes feeling younger right at the spot. :D)
- trying to keep an ssh client working in order to connect to  linux machines
- connecting to Windows VMs and physical machines on the network via RDP

So.. as you can see. no browsing needed in my case. Also I do not need OS/2 to run my current favourite games (just the old games from the 90ies need to run fine! (I need DTA and joystick/pad translators please :D)
For all the above stuff I will always keep a native ArcaOS (current machine) and OS/2 Warp 3 (486 DX2-66) running.

Wouldn't hurt though to have an up2date 64bit OS/2-look-and-feel-like linux distribution which's window manager comes as close to the WPS as possible (including all the well known settings folders and so on).

Mathias
« Last Edit: March 18, 2026, 08:49:17 pm by Mathias »

David Kiley

  • Full Member
  • ***
  • Posts: 159
  • Karma: +3/-0
Re: Adding an extra thing to the ArcaOS strategy ?
« Reply #4 on: March 20, 2026, 06:38:53 am »
We all want a "functional browser", and most want it within OS/2 (ArcaOS). But, no need to beat up on AN, because it's their ball (their OS), and they've chosen not to do apps. This probably makes sense, if you assume their primary focus is all the business users of OS/2 licenses.

I'm not trying to beat up AN in anyway. They have done a great job. And yes it's up to them what they want to do. It's just a question (for me) if it's worth paying to upgrade for what it's currently giving me, direction wise.

Martin Iturbide

  • OS2World NewsMaster
  • Global Moderator
  • Hero Member
  • *****
  • Posts: 5659
  • Karma: +50/-1
  • Your Friend Wil Declares...
    • Martin's Personal Blog
Re: Adding an extra thing to the ArcaOS strategy ?
« Reply #5 on: March 21, 2026, 08:05:08 pm »
Hello

First of all, thanks for the feedback, and sorry I don't reply that fast, since I had been busy with the server DDoS attacks and it slows me down on forum replies (like I guess is happening to all of us interacting with OS2World)

My idea is for "replacing (when possible) IBM OS/2 close source components (binaries, .exe) with open-source alternatives" is to be added to the strategy when possible and not becoming their main strategy.

I want to put a little example.
- Let's replace bldlevel.exe with the OSFree bldlevel.
-- https://github.com/osfree-project/osfree/tree/master/OS2/CMD/bldlevel
- It uses the The 3-Clause BSD License

ArcaOS could embrace the apple strategy when they created OS/X of using an open source base system (in their case Darwin) with thier own commercial propriety system on top of it (keeping the OS/2 UI for example like the workplace shell/WPS). And, they had created the "classic" mode to allow old os 9 apps to still work. Would require them embracing a totally new direction though.
Even that I may like that strategy, It is basically grabbing a new operating system (maybe Linux) and create a VM that runs OS/2 on it, and move forward to migrate to Linux. I personally don't like that.

Or creating a 64bit based OS/2 OS from scratch, but we don't have the resources to do that.

I like a lot the idea of grabbing a different kernel (Zircon, ReactOS kernel or Linux Kernel) and create an interpreter for OS/2 over it, but I guess this is a different subject of what I try to discuss here.

I think it's time to relegate the old IBM license to a virtualized session inside a totally new OS, while keeping WPS on top of it, which is where the user has always spent 99% of their time anyways.
I will prefer to construct a "OS/2 like WINE" interpreter under a new kernel to keep running the OS/2 experience on real hardware, like it was discuss on this forum thread, but again it is not the subject I want to discuss here and we don't have the resource to move on that way.

...BUT.. besides all the porting work, this will kill the OS one day.

Reason why? - Imagine you have replaced a good chunk of tools and maybe even main components. Next day the open source community decides ProjectX (or some dependencies) move on to Rust or some other compiler, that we are not able to compile for OS/2 with. Now you're operating system is busted, if you can't move back to the old closed source components. (maybe due to dependencies)

Here I think we have different perception of open source. Open source does not mean using an specific dependencies libraries, some open source software evolved on that way, yes. But the idea is to ask Arca Noae and to the community to re-use open source and also to create little replacements for IBM OS/2 close source binaries, not necessary using a Linux project to get ported.

I think the warning you post is good. "Killing" the software by using a library that eventually evolves to Rust, I feel it is the better than using a close source software that can not evolve anymore (like all IBM OS/2 binaries). In the case of IBM's close source we don't have a way to continue it, in the case of open source software we can fork it. 

Regards
« Last Edit: March 22, 2026, 01:09:01 am by Martin Iturbide »
Martin Iturbide
OS2World NewsMaster
... just share the dream.

Dave Yeo

  • Hero Member
  • *****
  • Posts: 5735
  • Karma: +150/-1
Re: Adding an extra thing to the ArcaOS strategy ?
« Reply #6 on: March 22, 2026, 02:52:55 am »
Hi Martin, while your idea is a good one and I agree more open source components in our OS is a good goal, I don't think Arca Noae really has the resources to implement it. Binaries need to be checked for compatibility and bugs.
Something like bldlevel isn't critical and a buggy or incomplete version wouldn't be too bad, many other components are more critical. Plus there is likely a subset of users who expect the core system to be as it always has been.
One idea is to develop a parallel system. Perhaps under something like \OS5\ all these tools can be installed in bin, dll and such and users who are interested can install the system. With the right PATH, LIBPATH etc the open source binaries will have priority so you'd start with x:\OS5\bldlevel  with x:\OS5\ at the front of the PATH, or at least forward of x:\OS2. Same with x:\OS5\DLL, x:\OS5\help and such.
I picked OS5 to align with ArcaOS 5.x and avoid OS4 which is already taken.
Another addition could be FreeDOS or other open source DOS under x:\OS5\MDOS or such.
Once a few utilities and other binaries are working in the parrallel sysem it would be easier to suggest Arca Noae considering them.
As for worries about open source projects moving in weird directions such a using Rust, they can always be forked, so not a problem. Of course with our shortage of active developers...

David Kiley

  • Full Member
  • ***
  • Posts: 159
  • Karma: +3/-0
Re: Adding an extra thing to the ArcaOS strategy ?
« Reply #7 on: March 22, 2026, 07:59:36 am »
Just curious, does AN have any access to the os/2 closed source code that they use to produce their new versions of os/2? Or they just have to develop "around" code they don't have access to?

My programming experience is very "basic" as in Microsoft basic :).
But how would open source utilities or binaries help if the main closed source core isn't being touched either way?


Dave Yeo

  • Hero Member
  • *****
  • Posts: 5735
  • Karma: +150/-1
Re: Adding an extra thing to the ArcaOS strategy ?
« Reply #8 on: March 22, 2026, 05:43:59 pm »
No access to the closed code. They do have permission to do binary patches, which is how they've patched the kernel and such.
Note that the patched binaries have to be distributed with ArcaOS, not separately.
Personally, I wonder if IBM still knows where the latest source is. Rumor also says the PowerPC OS/2 source is lost, which is a shame as that would have been the best to open source, pure C code.