Hi Kim,
So, going back to what has been discussed in several threads and as well part what can we do next when knocking on IBM's door?
- Well, do we have any valid business cases that we can point to that actually would benefit from that IBM opens up the code?
The insurance industry (IIRC, State Farm used to be entirely OS/2)
The banking/ATM industry (with a solid business, support and distribution model in place, I am sure there would be a growing number of banks who wanted to give up their current headaches in support and maintenance). I'd focus on ATMs first... then move up to client and server machines. ATMs are easy... a few new drivers for the ATM components (if/where needed) such as the cash dispenser, deposit unit... and a good flash video handler - or better yet, a good shockwave/flash engine.
The medical industry for various specialized computer and recordkeeping systems (from experience, they dont care - or even want to know - what their systems are running, as long as it works, and works, and works - and it would also alleviate numerous problems they are having with "stolen" customer records on compromised machines),
The Point of Sale marketplace (there are apparently still a bunch of solution providers who are using OS/2 on these systems - including for dual display cash register setups hooked to dual bar code scanners, serving ads... Giant (MD, DC, VA) was running theirs on Warp 4 as of 2005 (and may still be) - and didnt install it until the early 2000's) - the only "big" thing it would need is updated touchscreen drivers - and the company that makes those screens (there's pretty much only 1 big one here - namely Elo) are willing to provide what's needed to make new drivers (which pretty much amounts to creating a USB version - and maybe different handling to allow RMB actions - but that (RMB) isnt needed for POS based stuff anyway).
- Are there specific patents that we can point to that involves OS2 that we would like to get our hands on?
The WPS. Patent wise, I dont know what else. Technology wise, the APIs (or documentation to make clean room replacements for them) for the WPS and other subsystems.
- IBM claimed that there are legal matters with open up the code; do we know what components that they might try to hide behind that we can exclude from the scope?
Believe it or not, I used to have the list of many of them. They came with some developer kit or compiler or with IBM's Directory and Security Server for OS/2 beta (I think that was the name of it). I'll check my VAC++ docs when I get the chance, but...
- Much of the graphics engine (MicroGrafix I believe - and others)
- Much of the MM codecs (numerous contributors)
- LanMan (MS - obsoleted anyway)
- Probably parts of PMShell (MS and others - though there is the PPC implementation)
- Some of the (non-NetBIOS) networking components
- Parts of HPFS (I
think)
- There are things online that suggest that parts of the kernel as well (MS and others) - but then again, there is the PPC and other kernels
- Legacy code in WPS (MS) and possibly parts of the API
- Of course the DOS and WinOS2 stuff
- Various core drivers (much of which have been replaced though - and the rest could probably be replaced or licensed)
The list of companies that provided portions of OS/2 was very long. The amount of work they did may have been minimal - or not. I dont know. I do know that MS was just one of many companies that wrote code for OS/2 - and in less sections than the remaining companies.
The problem I see is, in many cases, IBM may outright own the code - while in others, they may be licensing the code, and there is probably no way to determine which is the case without their assistance.
All in all; can we create a list of specific components and patents that we really would need to make our life easier and as well on the same time - make it hard for IBM to say no and hide behind legal matters and patents.
API documentation may be the way... but would require a lot of work. Some of the info could be gleaned from some of their developer kits and compilers. The bugs and errata that numerous programs have come to count on could not be.
I'd ask for the PPC kernel and abstraction layers, OS/2 REXX code, PPC WPS and PM. At the very least, from the PPC kernel/abstraction layers, a replacement should be able to be written for Intel architecture - assuming that the PPC kernel was finished. If so, then it would just be a matter of adding the new API calls added to those subsystems since the PPC days. It would be a start - if they have and can find the code.
Robert[/list]