The last nail in OS/2's coffin

[Martin Iturbide]

Hello Dave, I get your points about some modern issues with Linux, but my point is that open source is better than close source.
For me open source is the good things about Linux, not the linux kernel or architecture. I think it is clear that OS/2 will not turn into Linux by turning it open source. Open source and Linux are different things, I guess that is clear.

Consider my phone and the majority in the world. Sure I can download the source code for the Linux kernel in my phone, what can I do with it? It's not like I have root and while my particular phone has instructions for rooting, unlike many, the kernel is still locked down by being signed and interacts with binary blobs.

This is a different subject from open source. Yes and No, there are communities that provides ROMs for different phones since they have the time, source code (even with binary blogs) and skill to recompile the Android image. But here the issue is that the monokernel merge together with binary blobs drivers, even with that, the kernel source code is still available for the community.

This took me to a different topic, why there is/was Android fragmentation and you can not easily update the Android OS (being open source) in phones.
1) Manufacturer don't care about software, just selling new phones, they don't offer straight forward software update path.
2) Monokernel, having all drivers compiled within the kernel gives you an issue if you only want to update the kernel and let the drivers alone.
3) Processors in phones and tablets are not as standardized as Intel and AMD chipsets.  This may be why it was easier to update Windows or OS/2 on your PC by going to the store and buying a CDROM box back in the 90's, than updating today your Android's phone.

Also consider IBM owned RedHat, you have to sign a contract to get access to the source, including the kernel which is likely patched.

This is an issue now IBM's Redhat is becoming an bad open source citizen, finding legal tricks to do not share the source code as a the GNU GPL demands.

Since the Open Source won world wide and almost now every system has an open source base or components, the trick companies are doing is putting open source solution on the cloud, extend those applications, and since they are offered as a service they try to use that trick to don't share the source code back.

What should be the solution? Moving to a license like "GNU Affero General Public License" ? or going back to the 90's close source model ?

Even with the dirty tricks that companies are trying with Open Source today, I still prefer it over close source that will became future abandoware. Today ArcaOS uses a good quantity of open source software that had provides a lot of help. We have to endorse to have more open source replacement in OS/2, that does not means getting ArcaOS for free. ArcaOS already have the crown of the jewel drivers (ACPI, UEFI, etc), so there is not risk for them.  Software development in OS/2 (the little that we have) should go from a license selling to a subscription support selling strategy.

Sorry for my passion on this subject.

Regards

[Sigurd Fastenrath]

OS/2 will survive for some years to come in the industrial business.

Private Customers will "die out" due to

- their age
- the uselessness of the System to private Users
- Memory Problem, no WLAN, no Localization,
- ArcaNoae showing no interest in private users

I would guess the number of German Users decreased during the last 3 Years, as there has allways been this tiring shifts of WLAN and German Local Version, to less of 10% of what have been before.

Yes, there are some people around using it for daily business but 99,9999999% of the Users do not even know what ArcaOS is.

It is impossible to persuade someone to have even a closer look at the system, because so many things are missing.

OS/2 had its time, that ended way back in 2005. It will never get opensourced nor will there be more private users, unless a miracle happens and ArcaNoae may solve the above problems, and add Bluetooth support and and and... But as David A. told me, there is no way to solve the memory problem.

It is much easier to use a Linux for free and install and use OS/2 in a Virtualization. There it can get it's tailored hardware.

Nor will ArcaOS ever be some kind of "Retro Plattform", as there are other, much more comfortable, customizable and free Solutions around like DosBOX, ScummVM etc.

It may sound to negative for some others, but that is my point of view, I just describe the facts without emotions, it is the way it is.

Personally I will try to have more success with the Lenovo Z13 Gen 2, just for fun, just to answer my eternal question I do have with OS/2 on modern hardware: will OS/2 run on it? And even if I get the Sound, with help from Paul and others, to work, the device will still be a "Technical Demo", nothing you can reliable work with.

What I would really like to see, was an OS/2 like Interface on Top of a Linux Kernel and System, building a clone of the WPS on Linux. In my opinion the only way to keep the OS/2 feeling alive in the future. But this is unlikely to happen as well.

All the best!

[Martin Iturbide]

Hi Sigurd

What I would really like to see, was an OS/2 like Interface on Top of a Linux Kernel and System, building a clone of the WPS on Linux. In my opinion the only way to keep the OS/2 feeling alive in the future. But this is unlikely to happen as well.

Just a similar GUI or OS/2 applications running over Linux?

Ryan C. Gordon tried that in some way that, but his goal was to run OS/2 software/games over Linux (interpreted like Wine or Odin), but he found out that cloning all the API was too much for him.
- https://github.com/icculus/2ine
Read "You have no idea how much effort went into getting this stupid white square on the screen."
- https://www.patreon.com/posts/project-2ine-16513790
He made a OS/2 application of white screen to run on Linux.

What it would be interesting it to don't got for all the API in one shot. Maybe starting only with the base CPI and see how can the rest of API (PM, SOM, WPS) can think they are still running under an OS/2 kernel. Of course, all OS/2 driver will be render useless in this kind of project.

But I'm a "hybrid kernel man" (That's part of the new masculinity  ), that why I dream of running the OS/2 Personality over Zircon.

Regards

[Lars]

What OS/2 applications do exist where you would not find an equivalent under Linux (or Windows)?
And more importantly: what developer would bother with OS/2 development tools?
I can tell you from my own experience that this becomes an increasing problem. One version of Watcom being incompatible with the next and unfortunately picking the "official" version does not help because it is the one that is buggy.

[Martin Iturbide]

Hi.

One version of Watcom being incompatible with the next and unfortunately picking the "official" version does not help because it is the one that is buggy.

I split the "OpenWatcom Discussion" to other thread.

Regards

[Mentore]

What OS/2 applications do exist where you would not find an equivalent under Linux (or Windows)?
And more importantly: what developer would bother with OS/2 development tools?

I got the feeling this is the main question and the main topic keeping the OS/2 architecture alive in the community (apart from the big commercial customers, which have economic interests in keeping their OS/2 installed base up-to-date).
It's the user experience.
I for myself have used OSes starting from the ZX Spectrum, saw and played with the first Mac and Amiga 1000, I still have three Spectrum(s) and a Sinclair QL, a Commodore 64, a 128 and an Amiga, I use Windows on a daily basis starting from Win 3.1, used Linux and still use Linux on the PC placed under the family TV, installed many unix flavors and window managers from fwm to KDE, GNOME, Ubuntu-based, Free BSD, Slackware, OpenSuSE, I even tried Sun Solaris 10 (still have three install DVDs for it).

I always find myself more comfortable with OS/2 GUI and overall with its system, even though I often get annoyed by its shortcomings.
Just an example: our CMD shell is lacking many features, like file and directory completion, a good command history, a complete 32 bit environment (it's still 16 bit IIRC), but nonetheless I feel pretty much at home into it.

So yes, I know I'll somehow keep at least an OS/2 machine at home, virtualized or on a bare metal machine. This, and the big investment I made in all these years, are my reasons.

I can tell you from my own experience that this becomes an increasing problem. One version of Watcom being incompatible with the next and unfortunately picking the "official" version does not help because it is the one that is buggy.

And this is a typical user community problem. We are shrinking, there's no doubt about it... Sadly I don't have real answers here.

Mentore

[andreas]

Sometimes i feel like I am the only one who uses OS/2 really "as a user". I'm not a programmer, just using it for my work daily. And I'm very happy with it!
Many people complain and i understand why. But i don't need my sytem for any internet-/game/-programming stuff. I use it mainly for writing, calculating, reading and organizing. also for handling my music files. For this OS/2 is perfect. Using it since vers. 2.11.
I wouldn't want any other OS to work with.
Am I really the only one?

Hope, OS/2 will not die!
Thanks to everyone who keeps it alive! Hope you guys don't get extinct...

[Per E. Johannessen]

Am I really the only one?

No, you're not.
I'm using it for both personal and business, and if we had a working browser and odbc (dBase) driver it would be nearly perfect.

[Dave Yeo]

Same here, though once again the browser is becoming a problem. While Linux is nice in that the hardware just works, it works well enough under OS/2 and Linux just feels klunky, and of course Windows just gets in the way thinking it knows best.

[Pete]

Hi All

I have to join in to say that I have been looking for a replacement OS since IBM announced OS/2's demise in 2000(?) and have not found anything that I am as happy to use.

The main problem I am currently bumping into is the same as everyone else: We do not have a capable browser. Seamonkey (and, I guess, Firefox) no longer work very well - if at all - on a lot of websites; Doodle (qt5) is simply too flaky to use due to crashes and system hangs.

Thanks to all who have kept OS/2 alive to date.


Regards

Pete

[Dave Yeo]

The trick with Dooble is to mark as many DLLs that it uses to load high as possible. Also the profile can get corrupt after crashes and hangs so needs to be wiped out. Good trick is to save a copy of the profile after some personalizing, before a crash and use that to replace it if it gets corrupt.
Helps a lot to have the blocklists etc installed as well, a lot of problems come with ads and the JavaScript they use.

[ivan]

Hi All,

I have a way round the browser problem.  I use an ITX board with Linux and a USB/HDMI KVM switch, most of which was salvaged from the local dump (I did buy the KVM new).

When our Firefox won't display a site I just switch over to the Linux unit and a few seconds later the site is up (always assuming NoScript allows it).

[Edmund Wong]

Hi All

I have to join in to say that I have been looking for a replacement OS since IBM announced OS/2's demise in 2000(?) and have not found anything that I am as happy to use.

The main problem I am currently bumping into is the same as everyone else: We do not have a capable browser. Seamonkey (and, I guess, Firefox) no longer work very well - if at all - on a lot of websites; Doodle (qt5) is simply too flaky to use due to crashes and system hangs.

Thanks to all who have kept OS/2 alive to date.


Regards

Pete

It's not just the OS/2 version of SeaMonkey is having issues with websites.  The windows version (2.53.18.x) is becoming a continual chore as websites no longer work properly.     

I don't know much about all this hardware talk(I'm only a user); but I'm guessing it'd take a huge dev group to 'modernize' OS/2.  Having a 64bit OS/2 system would be nice; but since from what I'm seeing in this thread,  the number of people that can actually do that is slowly decreasing by the year.  Am I right?   Is it possible to even modernize the OS/2 code or does it require a rewrite? 

Ed

[Martin Iturbide]

Hello Ed

Having a 64bit OS/2 system would be nice; but since from what I'm seeing in this thread,  the number of people that can actually do that is slowly decreasing by the year.  Am I right?   Is it possible to even modernize the OS/2 code or does it require a rewrite? 

There had been some discussion about the 64bit kernel for OS/2. - "OS/2 - ArcaOS 64Bits Kernel Discussion"

I'm not an expert but for what we have discuss I can say:
- We don't have the required labor resources to create a 64bit kernel at this moment.
- If we create an OS/2 64bit kernel, it has to be compatible with 32bits, otherwise all the OS/2 software won't work.
- There is no source code available (legally) of the OS/2 kernel, which make it more labor intensive to implement.
- In the case we get an 64bit kernel, we need to have the development tools to start creating 64bit OS/2 applications.
- If we remember the old days, the market had a difficult change to 64bits. Intel Itanium failed, Windows versions were incompatible (Windows XP 64bits), they have to fall back, wait more years to have a easy going 32bits to 64bits application transition.

So, yes it is kind of hard right now. Buy maybe if AI evolves for software development, some development can be automatized requiring less labor.

What is OS/2? the kernel? the applications? the WPS?. My idea is that maybe OS/2 is the desktop experience and the applications we use. That why my approach is more to have an already available 64bit kernel and create the OS/2 interpretation over it.

Regards

[Eric Erickson]

The other thing that threatens OS/2 is the aging of its users. I predict that will be the actual cause of death.

I'm scared to interpret this like my age group (40's) is going to be last one using OS/2   
Let's try to get people on his 30's to the community !!!   

Regards

This is a very similar problem that many of my clients experience, albeit on a different platform (IBM z/OS Mainframe). There is still a very large demand for mainframe developers (hence why one of my clients dragged me out of retirement). The problem is that how do you introduce and get the current crop of developers in their 20's and 30's excited about the mainframe. We are exploring the use of tools like ZOWE Explorer run inside Visual Studio to give them an interface that is more familiar to your average developer than the old 3270 screen and ISPF.  The other big problem is that pool of IBM z/OS Assembler programmers is getting smaller every year, but there are literally millions of lines of code still out there running in production environments. As one of the fish in that diminishing pool, I get about half-dozen inquires a week asking if I'm interested in doing some assembler programming work.

Not sure what the solution is for OS/2. But I will also say that I have not put a lot of thought into what the future might be. Professionally I have not done any OS/2 work in about 15 years, but personally I still run a few copies in VMs and 1 physical machine, which will bring me to the new thread I'm about to start.