Author Topic: About eComStation Pricing  (Read 48105 times)

eirik

  • Guest
Re: About eComStation Pricing
« Reply #15 on: October 13, 2014, 10:00:55 am »
Hi,

one way to increase the number of persons who wants to try eCS is to have a 30 day free period where people can install and test.  Many software companies use this approach successfully to avoid loosing potential users.  More eCS users is key for the future development of eCS.

Otherwise, I agree with much of what Fr4nk writes on prioritizing work. The key issue here is drivers, where net access (particularly WiFi) is where eCS struggles the most.  Qt apps are nice, but without proper drivers, people will not even get to that stage.

In economics where I work we have a saying "Poor money drives good money away".  Attempts to sell apps that are available for free is pretty close to this economics saying.  Point well taken Fr4nk.

Eirik

Martin Iturbide

  • OS2World NewsMaster
  • Global Moderator
  • Hero Member
  • *****
  • Posts: 4754
  • Karma: +41/-1
  • Your Friend Wil Declares...
    • View Profile
    • Martin's Personal Blog
Re: About eComStation Pricing
« Reply #16 on: October 13, 2014, 05:21:54 pm »
Quote
[...] at least share some ideas of what can we do about the future of the platform.

- set priorities and sort important things
- make it running perfectly on a reference hardware (support a small set of hardware but then perfectly)
- make killer apps (niche apps that don't exist on other platforms)
- get back to stability instead of desktop enhancers
- don't port 1000 (qt) apps and don't test but port 2 apps and make it big
- @developers: make as much as possible software free (Open Source or even freeware)
- @developers.ru don't try to sell mini apps which are better and free on all other platforms for 20 years.
- avoid things like "you can only use this software when you're a member of my club"


Thanks for sharing your ideas, those are very good.

In my wild dreams I will like to suggest a development strategy based in 2 pillars.

Pillar 1: Short Term Strategy:
- Drivers  - (Have a good reference hardware for eCS/OS2 to run)
- Stability
- Porting apps
- Find Killer Apps

Pillar 2: Long Term Strategy:
- Clone replacements for OS/2 close source components like Presentation Manager, SOM, WPS, Loader, etc.
- Fix the things that are not broken.... yes, fix the things that works.... Start experimenting replacing things. Ex: REXX vs. Regina REXX,  try to replace OS/2 close source DLLs produced but OSFree. This is very polemic since there is resistance to do modify the legacy, but it has to be done. The first tries will not be something stable, but if it is open it will allow to be improved over time.
- Some focus on replacing every single IBM close source EXE. Caring about little "quickwins" like "hey!!.. today I replaced bldlevel.exe with an open source one".

But the two pillars needs to be under an open source basis. It is not worthy to support "future abandonware". It is good that developers asks for money to do things, but it is not good to produce close work that can not allow derivative works. The idea is that things produced on the "Pillar 1: Short Term Strategy:" can also be reused for the second one.

About fixing what is not broken: I remember that some guy from an evil corporation told me "Good code produces a user community, bad code produces a developer community". I think that our main issue is that OS/2 is good code, it still runs after 12 years of not touching the PM, SOM, WPS, Kernel, Loader code.  That made us stop worrying about those components because it is good enough for our needs, but when some of it got broken or misbehave the real problems starts. The developers needs to do tricks to keep applications running, without any possibilities of fixing the source of the problem. That is why we need to move slowly caring about cloning the things that are working.

But there is also the part that can do the "Non-Developers".
- Share content/articles under a license that allows derivative works (OS2World Wiki, EDM/2 with Creative Commons)
- Help document the software and share the documentation with the public.
- Upload source code on the Github for other developers to find out what we have to create derivative works.
- Post pictures/screenshots/experiences on social networks and on the community.
- Keep collaborating on forums.
- Help posting news and team up with developers to help them manage announcements and requirements of the community.

The thing is not to leave everything for developers, the trick is for everyone to collaborate for our common benefit.
And any efforts that we made should be focused that can be added or included on something bigger, that's why I think open source/Creative commons are important.

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

Fr4nk

  • Guest
Re: About eComStation Pricing
« Reply #17 on: October 14, 2014, 08:18:46 am »
Well good ideas but if you make big plans (like rewrite the whole os) you need to consider the resources! Who should do all that work? It is  Here are only a handful of developers left while on other platforms like Linux are 100.000 of developers and some big companies behind.

So we need to stick to important things and don't start too many construction zones especially where things are working. Focus on things we need today and for next 3 years. I understand your point but it is 20 years too late. You say 'long term' but computers are a moving target. E.g. look at ACPI, it tooks years and it is even not ready and will never be ready as along as hardware changes.

Btw if you want rewrite os2 it would be even easier to write PM/WPS for a *x system and put it on top, like apple did with their system.

Martin Iturbide

  • OS2World NewsMaster
  • Global Moderator
  • Hero Member
  • *****
  • Posts: 4754
  • Karma: +41/-1
  • Your Friend Wil Declares...
    • View Profile
    • Martin's Personal Blog
Re: About eComStation Pricing
« Reply #18 on: October 14, 2014, 01:53:36 pm »
Hi Fr4nk.

For the moment it is only an idea that needs support and a lot of work. But the thing is that making OS/2 clone is not just having the resources and clone everything, it is a process that will take time. I want to set a goal for the future, even if it sounds to crazy right now, otherwise our future is to be shut down when the last of corporate user migrates to other platform. So, if you want to dream, you need to dream big, if at the end get only PM cloned, that will be success too.

If I would have the resources I will only focus on cloning PM, SOM and WPS running on eCS first. PM looks as the priority today (under the 2nd pillar)... or course that there are more short term priorities like drivers, but I think that companies and the community needs to care about the long term. 

Plus, the idea is to grow, we will not be a 100K dev community in one day, but we need to show progress (and good attitude) for other developers to join.
Martin Iturbide
OS2World NewsMaster
... just share the dream.

Sergey Posokhov

  • Full Member
  • ***
  • Posts: 169
  • Karma: +8/-6
    • View Profile
    • OS/2 API Research
Re: About eComStation Pricing
« Reply #19 on: October 14, 2014, 04:11:52 pm »
I'm not sure that IBM would allow someone else to make a clone...

Martin Iturbide

  • OS2World NewsMaster
  • Global Moderator
  • Hero Member
  • *****
  • Posts: 4754
  • Karma: +41/-1
  • Your Friend Wil Declares...
    • View Profile
    • Martin's Personal Blog
Re: About eComStation Pricing
« Reply #20 on: October 14, 2014, 07:17:56 pm »
Hi Sergey.

I don't think IBM care, and making a clone is not illegal. Please remember that Linux started as a Mimix clone, which is a Unix clone.  Linux lift up as a product by itself, but being a clone was on his origins.  And we have ReactOS which is a Windows 32bits clone, and I haven't hear that Microsoft is suing them (because they don't see $$). 

IBM will not move a finger, not even for suing if it not a good business. For example a ZOS clone will freak out IBM because it is still a good source of money for them (like happened before - They sued Platform Solutions Inc. (2006) and on 2008 they bought it. ....and the problem was that IBM didn't gave permission to run ZOS on their hardware, so it is not the same as cloning) Thinking about IBM should not be an stopper for cloning PM/SOM/WPS.

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

Andi B.

  • Hero Member
  • *****
  • Posts: 821
  • Karma: +11/-2
    • View Profile
Re: About eComStation Pricing
« Reply #21 on: October 15, 2014, 10:47:45 am »
Compaq cloned the IBM PC Bios (in contrary to other PC makers which more or less illegally copied it) and IBM could do nothing against it.

eirik

  • Guest
Re: About eComStation Pricing
« Reply #22 on: October 15, 2014, 09:28:16 pm »
There is an old saying "if it ain't broke, don't fix it".  Fr4nk got it spot on - the scarce developer resources we have should be applied where demand is most pressing.  Based on past posts and comments, that appears to be drivers (wired, wireless, screens), porting key software (like continue developement of Open Office, ThunderBird and Firerox for eCS) and mulitmedia (in particular video which is becoming more common on the net).

Comments on my "priority list" based on my recollection only -- anyone with a better memory than me?  Or maybe we should redo a quick web-survey to see what preferences people have?

Eirik

Doug Bissett

  • Hero Member
  • *****
  • Posts: 1593
  • Karma: +4/-2
    • View Profile
Re: About eComStation Pricing
« Reply #23 on: October 15, 2014, 11:34:30 pm »
I agree that the scarce developer resources need to be used in a better way. Despite what some seem to think, RPM/YUM is wasting a tremendous amount of developer time, for very little gain (in spite of what they say). Recently, it seems that RPM/YUM is contributing to DLL H_LL, not fixing it, as promised.

I agree that driver development is at, or very near, the top of the list. The heavy lifting for wired network adapter drivers has been done (as promised), but NOBODY has picked it up from there. Paul Smedley did do some work on that, and managed to make updates for the Realtek, and Intel, devices, but nothing else has been added to the list. I took a look at doing some of that, but I cannot get the current drivers to build. Something is missing, and I think it is something that I cannot get (sorry, I don't remember the details). It also seems that Wireless drivers are under construction, but delayed because of other things that are happening. The longer it takes, more people will be forced to abandon the platform, and they are not likely to return.

I also have a Lenovo ThinkPad L530, where UNIAUD will make noises, but nothing like they should be. I am not too concerned about that, because I didn't get a laptop to play music (they aren't that good at it, even in windows).

OpenOffice seems to be under some sort of development. Firefox seems to be progressing, but far to slowly (partly because there are not enough developers working on it). Thunderbird, and Seamonkey, should follow Firefox, since they are from the same code base.

I am not too sure what is meant by "Multimedia". PM123 seems to work well for audio, and VLC seems to work with video (I haven't tried streaming video).

Surveys have been done too often, and always seem to say much the same thing. Basically what has already been mentioned here.

The cost of eCS has been mentioned, and is probably the main reason why we see few new users. I doubt if anything can be done about that, since it would cost IBM a penny or two to get the lawyers to change the rules, and Mensys needs to cover their costs. IMO Mensys has done a pretty good job of managing all of that, and getting updates done, to keep OS/2 actually working. Now Arca Noae seems to be taking on part of the support (although the specifics have not, yet, been made clear). As usual, there are not enough resources (including people, and money), to do what needs to be done.

Martin Iturbide

  • OS2World NewsMaster
  • Global Moderator
  • Hero Member
  • *****
  • Posts: 4754
  • Karma: +41/-1
  • Your Friend Wil Declares...
    • View Profile
    • Martin's Personal Blog
Re: About eComStation Pricing
« Reply #24 on: October 16, 2014, 12:01:48 am »
Hi

The old saying "if it ain't broke, don't fix it" is good for some things, but I think we need to bend that rule for some things.

We have a lot of "Short Term necessities", drivers, apps, etc. But they are only short term, after four years we are going to have that same necessities again, because hardware will change and applications will be updated. While we cover the "Short Term necessities" more users are going to leave the platform because some do not see a future for it. This happens because the copyright limitations we have with IBM close source binaries and that there is no way to improve some components of OS/2 (because only IBM has the source code of it). For example Presentation Manager, GPI, REXX, SOM, WPS are abandonware right now, nobody is coding them, nobody can fix them, only do workarounds.

If we want to really look in the future we also need to start worrying about a "Long Term Strategy". I personally believe that a long term strategy is to try to clone the things that are working but that are abandonware like PM, WPS, SOM, etc, etc. Making them open source will open the doors for other people to join freely to the platform and collaborate.

It means to start worrying about the things are working but are close source, and how they will be replaced in the future. It does not mean "let's stop everything and just exclusively worry about this" (this is not binary). The idea is to start having conscious that we need to clone PM, WPS, SOM, etc... before they break complete without a way to fix them, or before we loose all of the people of the community and Fr4nk turn off the light :)

I don't want it to sound to easy, because it is not, but our "short term necessities"  does not allows us to see a future for the platform.  There was some spark to try clone more things with Voyager and OSFree on the past, but now I can not find it anymore. I blame myself in part for supporting them more in those days.... they saw a future a possible for the platform at that time while I didn't.

I just hope that this spark can be ignited on the companies that are still making some money with OS/2 and on some developers. It is important to also start thinking on those components, even if those are still working today.

Regards
« Last Edit: October 16, 2014, 01:22:36 am by Martin Iturbide »
Martin Iturbide
OS2World NewsMaster
... just share the dream.

Dave Yeo

  • Hero Member
  • *****
  • Posts: 4808
  • Karma: +100/-1
    • View Profile
Re: About eComStation Pricing
« Reply #25 on: October 16, 2014, 01:31:36 am »
Compaq cloned the IBM PC Bios (in contrary to other PC makers which more or less illegally copied it) and IBM could do nothing against it.

That was in the era of no software patents so all they had to do was prove they reversed engineered it instead of copying it which just meant documenting the process very well.

Doug Bissett

  • Hero Member
  • *****
  • Posts: 1593
  • Karma: +4/-2
    • View Profile
Re: About eComStation Pricing
« Reply #26 on: October 16, 2014, 08:01:25 am »
Quote
If we want to really look in the future we also need to start worrying about a "Long Term Strategy". I personally believe that a long term strategy is to try to clone the things that are working but that are abandonware like PM, WPS, SOM, etc, etc. Making them open source will open the doors for other people to join freely to the platform and collaborate.

The main problem with that, is that every resource is working flat out, just to try to stay where we are now. There are no people, and no money, to do any more. Taking resources out of the current mix, to even think about most of what you propose will end OS/2 sooner, rather than later. If you take even one knowledgeable person off of driver development, to put them on the problem of trying to define the WPS (for example). we will lose even more ground in driver development. Without drivers, OS/2 and eCS, cannot use new hardware. Without hardware to run on, the user is forced to run it in a virtual machine. When you do that, you might as well pick a host that will already run what you need, and use OS/2 only for those things that the host system can't do. After a while, maintaining a virtual machine becomes a chore that is no longer necessary.

The bottom line is that the "long term strategy" is to make OS/2 (whatever it may be called), last as long as possible. There  is no hope that it can last beyond being useful, other than as a footnote in history. DOS has already passed that stage. Some people still use it, but not enough to bother counting them. DOS has been cloned, and expanded, but it still isn't used much, and DOS is a LOT easier to clone than OS/2. The amazing thing is that OS/2 is still viable, even with no serious development for too many years. Those who choose to work on it are doing wonderful work, but they cannot perform miracles, and they cannot be working on two projects at the same time.

The other side of the equation is that hardware, as we (and OS/2) know it, is going to go away sometime in the next few years. Your average cell phone is now capable of doing almost everything that a desktop computer can do (my new cell phone has a quad core processor). Why would anybody buy a desktop computer? That, in turn, will drive up the cost of desktop computing to the point where there isn't going to be enough market for anybody to bother building desktop machines. Servers that are useable as a desktop machines are well on their way out of existence, being replaced by the "cloud", which requires large amounts of storage and processor power (far more than any desktop user could ever dream of using). That too will drive up the cost, because there will be no need to have enough machines to make it economical to build them. Of course, OS/2 will live on in old hardware, as long as it will still run. Will it be useful? Probably not any more than DOS is today. There will also be an accelerating exodus, of developers and users, as the old hardware, and software, become less, and less, useful.

The light may remain on, for many years yet, but to accomplish a clone of OS/2, would require many hundreds of millions of dollars, and many hundreds of developers. We have little hope of getting enough people interested, even if somebody donated the money. I suspect that those who would be willing to tackle the software development, would come from the Linux world, and they already think they know more about software than IBM did (and that may be true, today), so they would be reluctant to try to reproduce what IBM  produced, all those many years ago. At best, you would end up with a kludge that might look like OS/2, but it would probably not have the solid base that we know, and love. Even if they do reproduce OS/2, it is not suitable to run on a "wristwatch computer" (which I expect would be common by the time the task was accomplished). Even Linux, and windows, are not suitable for technology like that. Something new will be developed (perhaps that should be the target of your quest), making all of the current software, and hardware, obsolete. All we can hope for, is that whoever does it will plan very carefully, and make it much better than what we have today. So far, I don't see the light at the end of the tunnel. Even if I did, it would be unclear whether the light is an approaching train, or clear blue sky.

Dave Yeo

  • Hero Member
  • *****
  • Posts: 4808
  • Karma: +100/-1
    • View Profile
Re: About eComStation Pricing
« Reply #27 on: October 16, 2014, 08:27:44 am »
To add to what Doug said, the desktop itself is moving away from being OS/2 compatible. Hard drives have grown to big for the old way of partitioning, which has led to new ways to boot an OS. In theory we can adapt with new drivers but it would be a big job by itself, especially since I believe the kernel itself contains some of the functionality that allows booting.
Even worse for us is the move to 64bit computing. Needed programs such as the browser are growing and eventually won't fit into a 32 bit address space, we're already close to the limit. The office suite is likely going in the same direction. Even things like video are starting to move that way. I loosely follow X264 development and they don't have much interest in supporting 32bit, especially as the newer processors have extensions that work much better on 64bit CPUs,  use the various SSE implementations for math instead of the primitive math co-processor is one example. Another is AVX which our kernel will never support as bigger registers need to be saved when doing a context switch, and once again the 64bit version is much more superiour to the 32 bit implementation. And 64 bit CPUs have totally dropped support for 16 bit code when in 64 bit mode so not only the operating system would need rewriting but most every device driver.
It has been a good run and there are a few years left but not many for our 1990 technology.

Martin Iturbide

  • OS2World NewsMaster
  • Global Moderator
  • Hero Member
  • *****
  • Posts: 4754
  • Karma: +41/-1
  • Your Friend Wil Declares...
    • View Profile
    • Martin's Personal Blog
Re: About eComStation Pricing
« Reply #28 on: October 16, 2014, 03:08:16 pm »
So, the strategy is "give me drivers now, because I know that I will drop the platform eventually" "We are doomed anyway" "There is not future".

For the moment I keep seeing projects like ReactOS and HaikuOS that are acomplishing more in open source cloning than us. Where do they get developers? Where do they get money? They have the same issues.. market drifting away from PCs, limited resources, no marketing.... but they keep going, why? It is just easy to put your arms down and do not think about this.

Again, this is not about taking away all the resoures to clone. This is about making every effort to "add on" to a common goal of being a free (as in freedom) platform.
Let's make it little steps, like making the little bldlevel.exe replacement open source and adding on.
Martin Iturbide
OS2World NewsMaster
... just share the dream.

Doug Bissett

  • Hero Member
  • *****
  • Posts: 1593
  • Karma: +4/-2
    • View Profile
Re: About eComStation Pricing
« Reply #29 on: October 16, 2014, 05:57:36 pm »
Quote
So, the strategy is "give me drivers now, because I know that I will drop the platform eventually" "We are doomed anyway" "There is not future".

You are catching on. Unless somebody drops about 500 million dollars into the pot, it isn't possible (be realistic). Even if that does happen, you are not going to find developers who even know what OS/2 is, never mind try to duplicate it. Then, there is the problem that OS/2, as it exists today, is not going to be useful on something like a cell phone (existing cell phones will be obsolete in 5 years). It would be easier, and more effective, to develop a whole new OS, aimed at hardware that will exist in another 5 years (it will take at least that long to create a prototype, and that will be a moving target).

Quote
I keep seeing projects like ReactOS and HaikuOS that are acomplishing more in open source cloning than us.

Are they? I don't see users rushing out to use them. The main "problem" with both of them, is that they are trying to be something else, that already exists. By the time they get there, the target has changed (we have that problem too).

Quote
This is about making every effort to "add on" to a common goal of being a free (as in freedom) platform.

Nice thoughts, but not necessarily a good way to go. Linux already has that locked up. Why would anybody want to start over with a 1990s OS, when they can jump to Linux. Sure there are a few things that OS/2 can do, that have not been implemented in Linux, but the developers are free to do what they want, and there are enough of them, with financial support, to get it done. Since nobody has bothered to make WPS for Linux, I have to assume that there is little demand for it. My guess would be that about 50% of what the WPS can do, is never used anyway. I know that I don't use much of what it can do. I doubt if the remaining community could actually define all of what the WPS can do.

Quote
Let's make it little steps, like making the little bldlevel.exe replacement open source and adding on.

That tool is useful when programmers include the information in their work (not all that difficult, as I understand it). In fact, the program could likely be implemented in REXX (which is automatically "open source"), but I fail to see the point in creating a new tool that does exactly what the old tool does. If it is to be done, one would need to clearly define the function, AND, whatever new things it should do. Then, you would need to convince programmers to use it (many don't even use what already exists). The EXCEPTQ tool seems to be catching on with programmers, but only because it benefits them. It does nothing for the average user.

Quote
I believe the kernel itself contains some of the functionality that allows booting.

I think that the QSINIT project may be part of the answer. It would take some hard work, but it may be possible to build a "front end" loader that will hide the modern ways of doing things, and implement a translation layer in the background for OS/2 to use (effectively, run a VM in QSINIT). QSINIT has already implemented a RAMDISK that uses memory above the 4 GB mark, that OS/2 can use. A more defined API could make that memory available to programs that are written to use it. It may even be possible to replace the current memory allocation API so that old programs will use it, without even knowing about it. Don't waste valuable developer time re-inventing the wheel. Add what we need now, and use the old code while it still works. Rewriting the old code, using the same restrictions that exist today, accomplishes very little anyway.