OS2 World.Com Forum

Subject  :  Call to Action: Open Source OS/2 Abandonware
Author  :  Martin
Date  :  17 Sep, 2006 on 16:05
Please post your comments on the Call to Action: Open Source OS/2 Abandonware Article.

Subject  :  Re:Call to Action: Open Source OS/2 Abandonware
Author  :  BigWarpGuy
Date  :  17 Sep, 2006 on 22:27
If alot of people call for the open sourcing of abandonware, the auther (if they respond) will either open source it or realize there is a market for it and continue to develop it. It seems like a win-win situation.

---
BigWarpGuy
- - - - - - - - - - - - - - - - - -
OS/2-eCS.org
Director of Communications
- - - - - - - - - - - - - - - - -
supporting the past OS/2 user and the future eCS user
http://www.os2ecs.org

Subject  :  Re:Call to Action: Open Source OS/2 Abandonware
Author  :  obiwan
Date  :  18 Sep, 2006 on 03:39
Frequently, when a developer or company releases the source for a piece of commercial software, they want to find a worthy party to be responsible for it. If you look at OpenOffice.org, Mozilla, and other such projects, you see that they were not just dumped in an archive someplace, but they set up a Foundation with certain goals, to assure the quality and robust continued development of the project.

The abandonment of OS/2 software in many cases was because of lack of interest. A prime example is Post Road Mailer. When the developer decided to abandon even Postroad's Java successor, J-Street Mailer, he offered to release the source to both. A maintainer came forward for J-Street, which is now known as Polarbar Mailer, but no-one answered his call for a maintainer of Postroad. Since that time, many people have sought its source code from the developer but he can no longer be reached. Being utterly ignored when he initially offered the source likely was insulting, considering the countless hours and care he had put into developing it. I personally am a big fan of Postroad Mailer and am sorry I (we/all of us) missed that window of opportunity, as it was a mature application with much potential for further development.

Some companies went out of business and disappeared. I forget, what was the name of that word processor everyone loved? I regretted not buying it when I had the chance. If we had the source to it, we could today be using it with the latest Word files, without the bloat (sorry!) of OO.org or Lotus.

Then there are those companies that ceased development on OS/2 but developed later versions for Windows and other platforms. SlickEdit and BlueCAD are examples that come to mind. They would not release their source, because some of it would be in common with the later versions and, if ported, could compromise their sales.

These are the larger obstacles to open-sourcing OS/2 abandonware. I am skeptical of the idea that the developers do not know how to publish their source. Perhaps the action needed is to set up a Foundation, or to support and enhance existing organizations, to negotiate with former OS/2 developers, addressing their concerns within a professional structure (instead of being queried by yet another OS/2 user about that old project again).

This is what happened with PM Mail. In this case, the source and rights were actually bought by OS/2 Voice in order to develop and sell it. Not exactly open-sourced, but that fits well with the fact that PM Mail is still developed and sold for Windows.

For other abandonware, where the developer has disappeared from the known OS/2 universe, it could take quite some detective work to find the custodian of the source. That might benefit from collective action as well.


Subject  :  Re:Call to Action: Open Source OS/2 Abandonware
Author  :  Martin
Date  :  18 Sep, 2006 on 19:42
I complety understand that is desirable for developers (and to the OS/2-eCS community too) to have a responsable party for their software if it turns open source, but I think first we need to know which software can be open sourced.

I'm sure that many OS/2 ex-developers don't even know that there is a OS/2-eCS community that can be interested on their source code. This is a general call to let know OS/2 ex-developers that his source code can be usefull and that there is a community interested.

If there is a developer that want to open source his/her software only if there is a responsable party supporting it, it will be good if they can contact me to search for some interest on the community.

But I think we must find first the original developers that want to make their OS/2 software open source and them focus on try to find a developers party to continue it.

For the moment I want that the OS/2 abandoware gets an alternative to continue living. Open Source it will be the first step. It will take more time to find a developer party to continue it, but withtout the first step it will get harder to find this party.


Subject  :  Re:Call to Action: Open Source OS/2 Abandonware
Author  :  obiwan
Date  :  19 Sep, 2006 on 02:09
Maybe the trick is to set up something similar to the Bounties project. We need a list of abandonware projects that people are interested in, and some kind of petition, optionally with donations attached, to indicate how much interest there is. From there the developers can be found, and perhaps some private negotiations opened. This way, if they really do not know there is an interested community of users, they can find out just how interested people are.

What do you think?


Subject  :  Re:Call to Action: Open Source OS/2 Abandonware
Author  :  Fahrvenugen
Date  :  19 Sep, 2006 on 18:25

obiwan (18 Sep, 2006 03:39):
...

The abandonment of OS/2 software in many cases was because of lack of interest. A prime example is Post Road Mailer. When the developer decided to abandon even Postroad's Java successor, J-Street Mailer, he offered to release the source to both. A maintainer came forward for J-Street, which is now known as Polarbar Mailer, but no-one answered his call for a maintainer of Postroad. Since that time, many people have sought its source code from the developer but he can no longer be reached. Being utterly ignored when he initially offered the source likely was insulting, considering the countless hours and care he had put into developing it. I personally am a big fan of Postroad Mailer and am sorry I (we/all of us) missed that window of opportunity, as it was a mature application with much potential for further development.


I also am a big fan of Post Road Mailer, and I know that I've emailed Dan Porter at Innoval asking about the source. I'd love to get this, even if just to fix a few of the bugs and add some new features (such as the ability to use SMTP servers which require authentication, maybe add IMAP, etc). Unfortunately I have never received a response.



Some companies went out of business and disappeared. I forget, what was the name of that word processor everyone loved? I regretted not buying it when I had the chance. If we had the source to it, we could today be using it with the latest Word files, without the bloat (sorry!) of OO.org or Lotus.

You're thinking of Describe. It was a great word processor.


Subject  :  Re:Call to Action: Open Source OS/2 Abandonware
Author  :  obiwan
Date  :  20 Sep, 2006 on 06:13

Fahrvenugen (19 Sep, 2006 18:25):
I also am a big fan of Post Road Mailer, and I know that I've emailed Dan Porter at Innoval asking about the source. I'd love to get this, even if just to fix a few of the bugs and add some new features (such as the ability to use SMTP servers which require authentication, maybe add IMAP, etc). Unfortunately I have never received a response.

That's exactly what I'd like to do, and I bet we're not alone.



You're thinking of Describe. It was a great word processor.

That's right, DeScribe. Thanks. The sources for programs like DeScribe have to be kept by someone somewhere, don't they?


Subject  :  Re:Call to Action: Open Source OS/2 Abandonware
Author  :  Martin
Date  :  27 Sep, 2006 on 00:05
Well, I can not find some other post on the forums when several people put some abandonware that will like to see it open source.

So, it is a good idea, lets post the OS/2 abandonware that we will like to see open source.


Subject  :  Re:Call to Action: Open Source OS/2 Abandonware
Author  :  Fahrvenugen
Date  :  29 Sep, 2006 on 19:01
Okay, here's my list of stuff that I'd love to see open sourced. Some of these I doubt will ever be open sourced, but hey, you never know.

1. Post Road Mailer
2. Innoval's Hacksaw
3. Faxworks (PMFax)
4. BackAgain/2000 (or another similar professional backup product)
5. Ray Gwinn's SIO 2000 driver (I still use that driver)
6. Ceres Sound Studio (or instead of this, I'd love to see a port of the open sourced sound application Audacity... it is far past the capabilities of SoundStudio)
7. The CW-MMClasses (although I understand and respect Chris's reasons for not open sourcing them)

I'm sure there are others, but that's just what comes to mind right now.


Subject  :  Re:Call to Action: Open Source OS/2 Abandonware
Author  :  obiwan
Date  :  29 Sep, 2006 on 21:00
I came up with 8 I feel strongly enough about to think of:

1. Post Road Mailer
2. IBM Workframe, comes with VisualAge C++ 3
3. Blackbook, a freeware address book (hobbes)
4. Any and all EWS, but particularly I'd like to tinker with Alpha (an editor)
5. DeScribe word processor
6. Netscape 2.02 (to help fix elusive OS/2 performance problems in Mozilla ports)
7. Object Rexx for OS/2
8. Object Desktop


Subject  :  Re:Call to Action: Open Source OS/2 Abandonware
Author  :  El Vato
Date  :  04 Oct, 2006 on 06:45

Martin (17 Sep, 2006 16:07):
Please post your comments on the Call to Action: Open Source OS/2 Abandonware Article.

The most important application to open source is OS/2 itself. Once that is done, the applications will follow naturally.

All the major players in the OS field are moving to 64-bit and OS/2 should follow suit if it is to remain a viable alternative. Unless Serenity REALLY develops and comes up with a 64-bit kernel for OS/2 --PRONTO, an alternative open source replacement kernel should be the focus of the efforts to revitalize the OS/2.

It is no surprise that the applications that you mention continue to perform well under eCS --it is still in its essence old OS/2. What has merely been accomplished by eCS is a refinement to the IBM coarse and difficult installation procedure and a screen-scraping a la WinXX mode --nothing essentially revolutionary despite the marketing hype at Serenity.

In this article, Eric Raymond provides insight about what an OS (in his case Linux) is required to have during this transition from 32-bit to 64-bit hardware to challenge the MS monopoly: <http://www.eweek.com/article2/0,1895,2009201,00.asp?kc=EWLINEMNL082906EOAD >

Hence, unless OS/2 is to be regarded as "viable" as an Commodore and/or Amiga in this fast moving technological area, Serenity has to do some REAL development and come up with a 64-bit kernel. And if Serenity can not, a serious effort to fund an open souce kernel replacement (to begin with) should be the aim and not the peripheral applications. I, for one, possesing many of those applications continue to run them even on OS/2 Warp 3 with no need of eCS.


Subject  :  OSFree or FreeOS?
Author  :  BigWarpGuy
Date  :  04 Oct, 2006 on 15:47
Alternative to open sourceing OS/2 is the OSFree project? They need people to help with it.
http://www.osfree.org ?

Perhas restarting the FreeOS project? I think it has been dropped?abandoned? The site is gone and the Yahoo group for it is too(?).

---
BigWarpGuy
- - - - - - - - - - - - - - - - - -
OS/2-eCS.org
Director of Communications
- - - - - - - - - - - - - - - - -
supporting the past OS/2 user and the future eCS user
http://www.os2ecs.org


Subject  :  Re:Call to Action: Open Source OS/2 Abandonware
Author  :  lpino
Date  :  04 Oct, 2006 on 16:52
Sascha is trying to solve problems booting the L4. Anyone willing to take on the challenge?, he will help people getting the sources and make compile them.

It's time to remind people that support for OS/2 is three months away from being over, forever. That means they are dropping the other shoe.

In my opinion if we don't organize and make a sound plan to take control of the future of OS/2, the platform will be irrelevant even for us.

All the "dead" OSs have some form of open source of their OSs. Haiku (BeOS), FreeDOS (DOS), AROS (Amiga), REACTOS(Windows). Why can't we.....

Are there more BeOS users than OS/2?

Think about it...

Leonardo Pino


Subject  :  Re:Call to Action: Open Source OS/2 Abandonware
Author  :  Fahrvenugen
Date  :  04 Oct, 2006 on 21:24
While I'd love to see OS/2 open sourced, it is pretty clear to me that IBM isn't going to do this. The bottom line is there isn't enough of a business reason for IBM to go through the process to open source OS/2.

What that leaves us with is either trying to develop an open source version (clone) of OS/2, or (in the interim) seeing if it is possible to replace specific components of OS/2 with open source alternatives (some have suggested seeing if the OS/2 kernel can be replaced with a modified version of the Linux kernel, although I have no clue how realistic this idea is), or to just continue using it in maintenance mode (new device drivers can still be developed, new applications can be written, it even could be possible to develop new API's) , but with the core of the OS being static.


Subject  :  Re:Call to Action: Open Source OS/2 Abandonware
Author  :  BigWarpGuy
Date  :  04 Oct, 2006 on 22:35
Would another option be supporting the OS/2 support in ReactOS?
http://www.reactos.org

---
BigWarpGuy
- - - - - - - - - - - - - - - - - -
OS/2-eCS.org
Director of Communications
- - - - - - - - - - - - - - - - -
supporting the past OS/2 user and the future eCS user
http://www.os2ecs.org

Subject  :  Re:Call to Action: Open Source OS/2 Abandonware
Author  :  obiwan
Date  :  04 Oct, 2006 on 22:51

El Vato (04 Oct, 2006 06:45):

The most important application to open source is OS/2 itself. Once that is done, the applications will follow naturally.


Please explain how that is a natural consequence. It is not self-evident.


All the major players in the OS field are moving to 64-bit and OS/2 should follow suit if it is to remain a viable alternative.

Please explain how a 32-bit kernel that performs well on 64-bit hardware is not viable. It is a technical question, and the fact that other OS's have a 64-bit kernel is not a technical answer.


Unless Serenity REALLY develops and comes up with a 64-bit kernel for OS/2 --PRONTO, an alternative open source replacement kernel should be the focus of the efforts to revitalize the OS/2.

While this seems a good idea, it is not clear why it is so imperative. What would it really achieve? What are the costs and risks, and are they outweighed by the benefit? Is there nothing else about eCS that is more worthy of focus?


It is no surprise that the applications that you mention continue to perform well under eCS --it is still in its essence old OS/2. What has merely been accomplished by eCS is a refinement to the IBM coarse and difficult installation procedure and a screen-scraping a la WinXX mode --nothing essentially revolutionary despite the marketing hype at Serenity.

You consider it an improvement to break compatibility with older programs? No improvements to eComStation are "real" unless only the latest software can perform well? Bootable JFS, ACPI support, current accelerated video drivers, and native OpenOffice.org are nothing essentially revolutionary?

If you wouldn't mind, please clue me in on how to do with Windows the customization and beautification of the screen that I can so easily do with eComStation.


In this article, Eric Raymond provides insight about what an OS (in his case Linux) is required to have during this transition from 32-bit to 64-bit hardware to challenge the MS monopoly: <http://www.eweek.com/article2/0,1895,2009201,00.asp>

This article is actually by Steven J. Vaughan-Nichols, and he quotes Eric Raymond. I don't really see any reasoning given there for the 64-bit kernel either. In fact it says that the Linux kernel has supported 64 bits for years, and it has not been the deciding factor over its share of desktops, which is what Raymond is concerned about. The only reason given for the urgency of 64-bit computing is that Eric Raymond says so, and he's very smart. If that is the reason for doing things then we all should have dropped OS/2 for Linux a decade ago.


Hence, unless OS/2 is to be regarded as "viable" as an Commodore and/or Amiga in this fast moving technological area, Serenity has to do some REAL development and come up with a 64-bit kernel. And if Serenity can not, a serious effort to fund an open souce kernel replacement (to begin with) should be the aim and not the peripheral applications. I, for one, possesing many of those applications continue to run them even on OS/2 Warp 3 with no need of eCS.

IMO that's one of the beautiful things about OS/2, and now eComStation, that you don't have to upgrade, but all your programs still work the same if you do.

I am not disagreeing with you necessarily, but I can't agree with you either, because you have given no substantive reason why. An open-source 64-bit kernel sounds exciting, but I don't have any reason to expect it to perform faster or more reliably than the very mature, stable, and fast kernel I have now. If anything, unless it used the same IBM codebase, I wonder if such a radical change would destroy the whole stability of the platform.

In contrast, I know exactly what I would do with the source to these applications we've listed, to improve my own productivity. That really is what it is about, isn't it?


Subject  :  Re:Call to Action: Open Source OS/2 Abandonware
Author  :  El Vato
Date  :  05 Oct, 2006 on 06:42

El Vato (04 Oct, 2006 06:45):

The most important application to open source is OS/2 itself. Once that is done, the applications will follow naturally.



obiwan (04 Oct, 2006 22:51):Please explain how that is a natural consequence. It is not self-evident.

Context: potential developers are distributed across the globe and not everyone of them can afford even the cost of an eCS license (about the same cost for a WinXPP license but with NO modern applications). Keeping the OS/2 beast (source code) in the dark, chained, and only accesible by the few elite "priests", effectively keeps a cross-fertilization from happening inside the bowels of the OS/2.

For potential open source developers who might want to suggest extensions and/or modifications to the OS/2 kernel and/or API to accomodate compatibility with current open source projects/applications, it is not simply possible.


All the major players in the OS field are moving to 64-bit and OS/2 should follow suit if it is to remain a viable alternative.


obiwan (04 Oct, 2006 22:51):Please explain how a 32-bit kernel that performs well on 64-bit hardware is not viable. It is a technical question, and the fact that other OS's have a 64-bit kernel is not a technical answer.

Perspective...Do OS/2ers want a niche/hobby OS or do they want an workstation/enterprise class OS to compete with the big boys?


Unless Serenity REALLY develops and comes up with a 64-bit kernel for OS/2 --PRONTO, an alternative open source replacement kernel should be the focus of the efforts to revitalize the OS/2.


obiwan (04 Oct, 2006 22:51):While this seems a good idea, it is not clear why it is so imperative. What would it really achieve? What are the costs and risks, and are they outweighed by the benefit? Is there nothing else about eCS that is more worthy of focus?

Perspective, perspective...the window of transition is upon us (if we agree with Mr. Raymond). When a customer like a Small and Medium Business (SMBs) is evaluating an entry level OS workstation/client for its newly acquired 64-bit AMD or Intel based hardware, OS/2 should be among its potential choices. A 64-bit workstation/enterprise class where applications like the Apache Geronimo application server and the Derby small-footprint DB2-like database, at the very minimum, should be able to execute and SCALE well --can eCS do that???


It is no surprise that the applications that you mention continue to perform well under eCS --it is still in its essence old OS/2. What has merely been accomplished by eCS is a refinement to the IBM coarse and difficult installation procedure and a screen-scraping a la WinXX mode --nothing essentially revolutionary despite the marketing hype at Serenity.


obiwan (04 Oct, 2006 22:51):You consider it an improvement to break compatibility with older programs? No improvements to eComStation are "real" unless only the latest software can perform well? Bootable JFS, ACPI support, current accelerated video drivers, and native OpenOffice.org are nothing essentially revolutionary?

Perspective, perspective, perspective...do not superimpose vos perspective upon mine, my dear fan of Chewbacca.

In going from a bootable FAT formatted partition to an HPFS one, only one file is necessary to make the newly formatted hard disk/partition bootable: OS2BOOT. Accordingly, a hack to that same file is likely only necessary to make an JFS formatted disk/partition bootable. That, I hardly call revolutionary, as any OS/2 hacker with access to the source code of the file system might have done looong ago. Additionally, that booting method may not be exactly too stable for the general OS/2 environment --otherwise IBM might have done it in a whiff.

There are issues that SNAP team has not resolved due to lack of access/modification to the OS/2 video subsystem.

OpenOffice runs excellent under Linux right now and users do not have to pay $250.00. Are you sure the OpenOffice port is not a Win32 hack like the Innotek Java port and the VirtualPC/2 that run so unstable under the current OS/2 environment?


obiwan (04 Oct, 2006 22:51):If you wouldn't mind, please clue me in on how to do with Windows the customization and beautification of the screen that I can so easily do with eComStation.

This is very OS dependent and your question has nothing to do with the METHODOLOGY to "upgrade" an OS a la WinXX. Again, you are reducing my statements to your notions --preconceived or not.


In this article, Eric Raymond provides insight about what an OS (in his case Linux) is required to have during this transition from 32-bit to 64-bit hardware to challenge the MS monopoly: <http://www.eweek.com/article2/0,1895,2009201,00.asp>


obiwan (04 Oct, 2006 22:51):This article is actually by Steven J. Vaughan-Nichols, and he quotes Eric Raymond. I don't really see any reasoning given there for the 64-bit kernel either. In fact it says that the Linux kernel has supported 64 bits for years, and it has not been the deciding factor over its share of desktops, which is what Raymond is concerned about. The only reason given for the urgency of 64-bit computing is that Eric Raymond says so, and he's very smart. If that is the reason for doing things then we all should have dropped OS/2 for Linux a decade ago.

Perspective, perspective, perspective, perspective...server vs desktop/workstation --that is the focus of Mr. Raymond. Based on the previous pattern of transition from 16 to 32-bit architecture, this 32 to 64-bit phase in the Intel and or AMD customer/SMB and or enterprise client/workstations represents another chance to fight for turf on the desktop/workstation area. Mr. Raymond knows that the server class OS on 64-bit is old news for Linux.


Hence, unless OS/2 is to be regarded as "viable" as an Commodore and/or Amiga in this fast moving technological area, Serenity has to do some REAL development and come up with a 64-bit kernel. And if Serenity can not, a serious effort to fund an open souce kernel replacement (to begin with) should be the aim and not the peripheral applications. I, for one, possesing many of those applications continue to run them even on OS/2 Warp 3 with no need of eCS.


obiwan (04 Oct, 2006 22:51):IMO that's one of the beautiful things about OS/2, and now eComStation, that you don't have to upgrade, but all your programs still work the same if you do.

I am not disagreeing with you necessarily, but I can't agree with you either, because you have given no substantive reason why. An open-source 64-bit kernel sounds exciting, but I don't have any reason to expect it to perform faster or more reliably than the very mature, stable, and fast kernel I have now. If anything, unless it used the same IBM codebase, I wonder if such a radical change would destroy the whole stability of the platform.

In contrast, I know exactly what I would do with the source to these applications we've listed, to improve my own productivity. That really is what it is about, isn't it?


Again, perspective. If OS/2 is to be relegated to a hobby like exercise for which you are willing to pay at least $250.00 is fine since all you will be running is oldie applications and thus I have not provided any substantive backing. I will be happy not spending $250.00 to run those same applicatioins if I remain with my OS/2 Warp 3-4.

On the other hand, some of us want to run/recommend to business the same applications that other OSes run and --why not?-- compete with WinXX for vendor applications. I would like to run/recommend WebSphere (Java based) workstation/enterprise class applications like Rational's Studio or at least its equivalent Eclipse development environment.

It was with great difficulty --trial and error-- that I was successful in having IBM CloudScape/Apache Derby, a Java based DB2-like personal developer's database, to execute under OS/2 with Golden Code Development (GCD) port of Java. And I see no reason not to unleash the OS/2 beast so that others may contribute source modification and/or applications to compete with WinXX than that of an equal footing with 64-bit OSes..

Closing one's eyes complacently and repeating, "we are an elite pool gene --we do not need cross-fertilization from other than Serenity/IBM" when suddenly the major playeres move to 64-bit and BOOM! we are gone..

OS/2 was designed as an integrating platform but in the current context it appears as if it were ok to be another DOS.


Subject  :  Re:Call to Action: Open Source OS/2 Abandonware
Author  :  obiwan
Date  :  05 Oct, 2006 on 07:52
Good response, El Vato. Just a few comments/questions.


El Vato (05 Oct, 2006 06:42):
For potential open source developers who might want to suggest extensions and/or modifications to the OS/2 kernel and/or API to accomodate compatibility with current open source projects/applications, it is not simply possible.

I think everyone agrees it either needs to be open-sourced, upgraded by IBM, or replaced, eventually. The question is how urgent is it? The API can be extended without changing the kernel.

It still isn't clear how releasing the source to it would affect whether application developers release their source.


The window of transition is upon us (if we agree with Mr. Raymond). When a customer like a Small and Medium Business (SMBs) is evaluating an entry level OS workstation/client for its newly acquired 64-bit AMD or Intel based hardware, OS/2 should be among its potential choices. A 64-bit workstation/enterprise class where applications like the Apache Geronimo application server and the Derby small-footprint DB2-like database, at the very minimum, should be able to execute and SCALE well --can eCS do that???

Maybe not, but is that really dependent on replacing the kernel?


In going from a bootable FAT formatted partition to an HPFS one, only one file is necessary to make the newly formatted hard disk/partition bootable: OS2BOOT. Accordingly, a hack to that same file is likely only necessary to make an JFS formatted disk/partition bootable. That, I hardly call revolutionary, as any OS/2 hacker with access to the source code of the file system might have done looong ago.

The source for JFS has been available for years now, and any OS/2 hacker didn't do it, Serenity did.


Your question has nothing to do with the METHODOLOGY to "upgrade" an OS a la WinXX.

Sorry, I misunderstood you to mean the screenscraping was to make OS/2 more like WinXX. You actually meant it was an upgrade of the same sort you might find in WinXX.


I will be happy not spending $250.00 to run those same applicatioins if I remain with my OS/2 Warp 3-4.

I can't install Warp 3 on even semi-modern hardware for $250.

The rest of your points are well taken. It reflects a problem as old as OS/2 itself.

Perhaps that is yet another reason to open-source these abandonware applications, so that they can ultimately be ported to other platforms.


Subject  :  Re:Call to Action: Open Source OS/2 Abandonware
Author  :  Blonde Guy
Date  :  05 Oct, 2006 on 23:32
OK, so open-sourcing OS/2 is stupid. Don't waste your time.

1. The code is really, really old. Much of the OS/2 kernel (over half of the lines of x86 assembler code) is devoted to processing the config.sys file. Quite a bit more is devoted to handling PS/2 hardware (and I don't mean Play Station). Lots more of the essential code of OS/2, for example the named-pipe code, is 16-bit.

2. The code is totally compromised. There are something like 300 vendors, most of which are no longer functioning corporations. It would cost millions in legal fees.

3. The function of the OS/2 kernel is pretty well documented. The point of the open-source OS/2 kernel clones are that it isn't that hard to duplicate the function of the closed-source parts of OS/2.

4. I like the Serenity/Mensys/Netlabs approach of replacing one piece at a time until all of the IBM parts are gone. Want a 64-bit kernel? There's one under construction. If it works, and it's correctly licensed, Serenity/Mensys will be including it in eCS.

---
Expert Consulting for OS/2 and eComStation


Subject  :  Re:Call to Action: Open Source OS/2 Abandonware
Author  :  obiwan
Date  :  06 Oct, 2006 on 01:23

Blonde Guy (05 Oct, 2006 23:32):
OK, so open-sourcing OS/2 is stupid. Don't waste your time.

Thanks for chiming in Blonde Guy. Good stuff to know, and upbeat too.


Subject  :  Re:Call to Action: Open Source OS/2 Abandonware
Author  :  lpino
Date  :  06 Oct, 2006 on 02:02

Blonde Guy (05 Oct, 2006 23:32):
OK, so open-sourcing OS/2 is stupid. Don't waste your time.

1. The code is really, really old. Much of the OS/2 kernel (over half of the lines of x86 assembler code) is devoted to processing the config.sys file. Quite a bit more is devoted to handling PS/2 hardware (and I don't mean Play Station). Lots more of the essential code of OS/2, for example the named-pipe code, is 16-bit.

True, the goal shouldn't be to have the source of OS/2 Intel. That would be silly (stupid). None of the Open Source OSs have been ported using the original kernel. They all start from some other kernel.

2. The code is totally compromised. There are something like 300 vendors, most of which are no longer functioning corporations. It would cost millions in legal fees.

True, as I said before that's not the point. The point is to maintain and improve the OS/2 API, to extend it's life (64 bit extentions) and free ourselfs from IBM.

3. The function of the OS/2 kernel is pretty well documented. The point of the open-source OS/2 kernel clones are that it isn't that hard to duplicate the function of the closed-source parts of OS/2.

Very true.

4. I like the Serenity/Mensys/Netlabs approach of replacing one piece at a time until all of the IBM parts are gone. Want a 64-bit kernel? There's one under construction. If it works, and it's correctly licensed, Serenity/Mensys will be including it in eCS.


Had no idea there was a 64 bit kernel being developed with the intention of implementing the OS/2 API on top of it. Can you be more specific?.


Subject  :  Re:Call to Action: Open Source OS/2 Abandonware
Author  :  Blonde Guy
Date  :  06 Oct, 2006 on 04:21
OK, two places to look a possible future for OS/2 and eCS.

This is the direction some of the best OS/2 programmers want to take.

http://wiki.netlabs.org/index.php/Voyager_FAQ

This is the reconstruction of the OS/2 kernel, but using open source.

http://www.osfree.org/index.php

---
Expert Consulting for OS/2 and eComStation


Subject  :  Re:Call to Action: Open Source OS/2 Abandonware
Author  :  El Vato
Date  :  06 Oct, 2006 on 06:17

Blonde Guy (05 Oct, 2006 23:32):
OK, so open-sourcing OS/2 is stupid. Don't waste your time.

Perspective...when Netscape open sourced its jewel --the Mozilla code base-- back in 1998-99, there were many who thought it "stupid." Yet the code thrived by providing an alternative to the MS IE. And as you know, Mozilla is more than a browser; if pushed a little further it can even become an OS itself, as Netscape had envisioned.


Blonde Guy (05 Oct, 2006 23:32):1. The code is really, really old. Much of the OS/2 kernel (over half of the lines of x86 assembler code) is devoted to processing the config.sys file. Quite a bit more is devoted to handling PS/2 hardware (and I don't mean Play Station). Lots more of the essential code of OS/2, for example the named-pipe code, is 16-bit.

Do not underestimate the open source programmers. The Linux kernel is bigger and the programmmers will handle machine code as well a C modifications. I have met former IBMers who read assembly like reading a children's book.


Blonde Guy (05 Oct, 2006 23:32):2. The code is totally compromised. There are something like 300 vendors, most of which are no longer functioning corporations. It would cost millions in legal fees.

IBM has donated some 500 patents to open source < http://www.eweek.com/article2/0,1759,1750358,00.asp?kc=EWNKT0209KTX1K0100440 >; that easily is worth "millions." Hence, if the proper pitch were made to IBM about an angle that would bring the company revenue BUT in line with their current open source approach, "millions" would be petty cash for IBM.


Blonde Guy (05 Oct, 2006 23:32):3. The function of the OS/2 kernel is pretty well documented. The point of the open-source OS/2 kernel clones are that it isn't that hard to duplicate the function of the closed-source parts of OS/2.

It may be well documented and it may not be "hard" to duplicate; notwithstanding, judging from the sites that you referenced further in your post, it is not necesarily peachy to do it --nor the aim of what is to be achieved exactly specific-- that leaves the 64-bit kernel in Limbo.


Blonde Guy (05 Oct, 2006 23:32):4. I like the Serenity/Mensys/Netlabs approach of replacing one piece at a time until all of the IBM parts are gone. Want a 64-bit kernel? There's one under construction. If it works, and it's correctly licensed, Serenity/Mensys will be including it in eCS.

Where in Serenity's roadmap is 64-bit exactly? Is there a link where the company commits to its customers such feature in the futue? And if so, how far is the company from that goal? What proprietary OS/2 system components has the company replaced with open sourced equivalents by now? I am asking because not even the (hidden) OS2BOOT file --needed to modify whether a file system will be bootable or not-- has been made open or replaced.

What does correctly licensed mean? Judging from the links that you provided, there is nothing concretely stated and you are not providing anything better than the "stupid" but concrete petition by many of us to IBM: COME ON, BIG BLUE, OPEN SOURCE EVEN THE POWERPC VERSION OF OS/2.


Subject  :  Re:Call to Action: Open Source OS/2 Abandonware
Author  :  flywheel flywheel@worldonline.dk
Date  :  06 Oct, 2006 on 10:36
There have been some discussing of a 64-bit kernel (Which AFAIR is the way to go), but AFAIR Voyager is going 32-bit and the eCS on-top-of-an-64bit-Opensource-OS discussions seems to have faded away.

Subject  :  Re:Call to Action: Open Source OS/2 Abandonware
Author  :  BigWarpGuy
Date  :  06 Oct, 2006 on 15:01
What might be helpful if Serenity Systems signs an agreement with IBM to view the source code of OS/2 with the promise of not releasing/open sourcing it. If possible, Serenity Systems could 'recruit' (unpaid) programmers to work on it and those programmers would sign an agreement not to release/open source the code they are using/seeing of the OS/2 kernel.

The above probably would only work in fiction. It would be geeky version of a covert operation; instead of 'cloak and dagger', it is more 'cloak and coder'.

---
BigWarpGuy
- - - - - - - - - - - - - - - - - -
OS/2-eCS.org
Director of Communications
- - - - - - - - - - - - - - - - -
supporting the past OS/2 user and the future eCS user
http://www.os2ecs.org


Subject  :  Re:Call to Action: Open Source OS/2 Abandonware
Author  :  Martin
Date  :  08 Oct, 2006 on 16:00

Blonde Guy (05 Oct, 2006 23:32):

4. I like the Serenity/Mensys/Netlabs approach of replacing one piece at a time until all of the IBM parts are gone. Want a 64-bit kernel? There's one under construction. If it works, and it's correctly licensed, Serenity/Mensys will be including it in eCS.


I also like the approch of replacing one piece at a time. And also like when Serenity puts as much of Open Source (OSI approved Licenses) software on eComStation.

There had been some discussion before on how it had to be open sourced, many said, from down to top, first the kernel, other said lets go from Top to Down, lets make open source replacement of the applications first.

I really don't care about it, as long as we can replace the closed components of OS/2 and eCS. Possible we can build a group that focuses on open source replacements for OS/2-eCS like the NetlabsWiki Idea. That can possible help to anybody to create an Open Source OS/2 (or partially open source) distribution and will help eComStation to keep including more open source replacements.


Subject  :  Re:Call to Action: Open Source OS/2 Abandonware
Author  :  obiwan
Date  :  08 Oct, 2006 on 20:42
I might add that we need not panic. We all have been using OS/2 long enough to know what it is, and what it isn't. We've had the same "close to death" prognosis for many many years. There has always been some component we seem to believe OS/2 will never have, and somehow it seems to come about anyway. Yes it takes hard work, and by now we should know that each of us has to put in some work to keep this going. But I think we can be positive and have fun with it. It exists because it works. It will continue to improve because people are working. Relax.

Subject  :  Re:Call to Action: Open Source OS/2 Abandonware
Author  :  El Vato
Date  :  09 Oct, 2006 on 03:28

Martin (08 Oct, 2006 16:00):
[quote]Blonde Guy (05 Oct, 2006 23:32):

4. I like the Serenity/Mensys/Netlabs approach of replacing one piece at a time until all of the IBM parts are gone. Want a 64-bit kernel? There's one under construction. If it works, and it's correctly licensed, Serenity/Mensys will be including it in eCS.



Martin (08 Oct, 2006 16:00):I also like the approch of replacing one piece at a time. And also like when Serenity puts as much of Open Source (OSI approved Licenses) software on eComStation.

For a business entity that depends on an open source community a give and take approach is a necessity --and not an option.

Noticing the success with which Red Hat's Fedora and Novell's OpenSuSE have in their business/development model, I see the need for a fundamental change in Serenity's business model (if the open source effort is to succeed) with Serenity remaining a catalyst for OS/2 developers.

I can be more explicit, but it goes without saying that if Red Hat and/or Novell develops a driver, they offer it back to the community that supports them --either company does not tell you: you have to buy a subscription if you want to have it.


Martin (08 Oct, 2006 16:00):There had been some discussion before on how it had to be open sourced, many said, from down to top, first the kernel, other said lets go from Top to Down, lets make open source replacement of the applications first.

The top-down approach will place unecessary restrictions on the specification/choice of the kernel/API for those who will take on the kernel task at a later date.

Hence, instead of tying one arm behind those who will take on the kernel aspect, why not give them some slack for creativity in optimizing the kernel for the 64-bit architecture? And to allow them to implement an API (subset or otherwise) that is most in line with proven and/or modern trends in OSes?

Modern application support should be the focus (if we are to remain OS competitive) instead of remaining whithin an inertial twilight zone by requiring backward compatibility with old applications.


Martin (08 Oct, 2006 16:00):I really don't care about it, as long as we can replace the closed components of OS/2 and eCS. Possible we can build a group that focuses on open source replacements for OS/2-eCS like the NetlabsWiki Idea. That can possible help to anybody to create an Open Source OS/2 (or partially open source) distribution and will help eComStation to keep including more open source replacements.

I, as yourself, like the wiki approach to concretely establish a plan to follow --instead of bringing out notions out of thin air to cover up for special interests. I have no problem with the GPL, notwithstanding. It simply forces companies to share with its contributors what themselves (community) have helped to create.

It is no wonder, though, that those open source communities thrive by developing applications for their OS: they know that the business entity that they support will respond in kind.

...And the business entity will make its profit by architecting the most mature components/applications/etc. released (and tested by the community) into a coherent and funcional package that the company will support for a determined length of time (available to those who buy a subscription).


Subject  :  Re:Call to Action: Open Source OS/2 Abandonware
Author  :  El Vato
Date  :  09 Oct, 2006 on 07:30

obiwan (08 Oct, 2006 20:42):
I might add that we need not panic. We all have been using OS/2 long enough to know what it is, and what it isn't. We've had the same "close to death" prognosis for many many years. There has always been some component we seem to believe OS/2 will never have, and somehow it seems to come about anyway. Yes it takes hard work, and by now we should know that each of us has to put in some work to keep this going. But I think we can be positive and have fun with it. It exists because it works. It will continue to improve because people are working. Relax.

Since the creation of OS/2, the underlying architecture (fabric of the specific universe where OS/2 was designed to exist/compete) has remained a fairly constant variable. But that medium is about to experience a drastic change as we approach this transitional phase to 64-bit: the resources available to an OS will have multiplied (a parallel universe will be detected by other OSes).

If the OS/2 can not take advantage of the newly available resources, it will find itself in the ship(s) of DOS/Windows during the dawn of 32-bit computing some years back. Notwithstanding, the similararities end there when one realizes that OS/2 does not have the marketing Borgs to massively brainwash others into submission --nor the marketing power to impose conditions upon computer manufacturers.

And there the reason that I advocate the creation/adoption of 64-bit kernel as a priority --someone suggested the L4Ka::\Pistachio microkernel with a malleable API < http://l4ka.org/projects/pistachio/ > (I think)-- around which system developers will effectively focus their efforts; rather than have (higher level) application developers guessing what the kernel/API will be made available to their creations.

...and no, my dear Chewbacca fan, I can not relax...quite possibly you are experiencing the black hole effect: everything appears "normal" when you are caught by their powerful effect --what you have not realized is that the very fabric of the (OS) universe has even been warped


Subject  :  Re:Call to Action: Open Source OS/2 Abandonware
Author  :  Ben Dragon
Date  :  09 Oct, 2006 on 16:20
Since we're all delving into what if's and speculating on what would be good for an enchanced or supplanted OS/2 based system, then we might as well get right to it and stop playing patti-cake.

That is to say, why waste time on a 64-bit system? OS/2 was reknown for being double the bit rate of its nearest competitor. Why not jump to the forefront once more and shoot for a 128-bit OS?

By the time it gets done 128-bits will be the standard anways.

Let's plan to rob the limelight from whatever other inferior OS claims the spot? Jump the queque I say, and return to the glory days and make OS/2 shine once more!

One has to think ahead or we'll forever be suckin' on the hind tit.

* sits down and lights a cigarette *

Just my thoughts on the matter...

---


Subject  :  Re:Call to Action: Open Source OS/2 Abandonware
Author  :  obiwan
Date  :  09 Oct, 2006 on 23:14
El Vato.

If you want an open-source 64-bit kernel, GNU-licensed software, the RedHat/Novell business model, no compatibility with legacy OS/2 applications, disassociation fom IBM, and no upgrade fees, why not just use Linux?

Forgive me if I read you incorrectly. You seem to be dissatisfied with OS/2 as it stands in late 2006, and want its universe to transform immediately to this ideal.

There clearly is work to be done on OS/2, as we all have noted, and avenues have been referenced where we can contribute where we are most interested. Your passion for a 64-bit kernel is laudable, and one would hope it moves you to work on it.

However, it will not meet your deadline. Serenity will not fundamentally change its business model. No-one will instantaneously invent a new, free-of-charge, open-source, fully mature and stable, 64-bit OS/2, to dominate the desktop market in 2007.

So we all need to decide if that is acceptable, and worth investing in. One day, when it is possible, the L4 kernel could conceivably replace the old IBM kernel.

Edit: Removed wordiness and rudeness, while preserving the message. It's bothered me how I said this since I wrote it. My overexuberance was sparked by a desire to quell any panic about an end to OS/2. However I likely offended El Vato, and brought down the friendly atmosphere of the forums. My apologies to everyone who read it.


Subject  :  Re:Call to Action: Open Source OS/2 Abandonware
Author  :  El Vato
Date  :  10 Oct, 2006 on 06:44
My dear fan of Chewbacca,

You are "...read[ing me] .. incorrectly..."

We are not on the same page...

Man is in love with ephemeral creations,
What else is there to say???


Subject  :  Re:Call to Action: Open Source OS/2 Abandonware
Author  :  obiwan
Date  :  12 Oct, 2006 on 07:19
I think perhaps I understand you now, El Vato.

So Martin, did your article have anything to do with the release of the source of DrawIt/SketchIt, and (imminently) SIO?


Subject  :  Re:Call to Action: Open Source OS/2 Abandonware
Author  :  Martin
Date  :  13 Oct, 2006 on 01:31
Well, I think Ray Gwinn had plans to open source their development before I posting my article. (BTW - SIO is not open source yet, if anybody is willing to help Ray to remove some propietary dependancies on the code is welcome)

About Drawit, I contacted Jasper to know how the development was going and suggested to turn it open source. He replied me that he was willing to public the source, and later turn it GNU GPL in his website. But I'm not sure if he read my article.

I only can thanks them to open source their OS/2 software and give it away to the community.


Powered by UltraBoard 2000 <www.ub2k.com>