> Personal Edition $129
> ArcaOS may not install or run on all systems, and Arca Noae makes no claim that it does. If you are unsure as to the suitability of your hardware for ArcaOS, please review this wiki page or contact us. *There is no trial or demonstration version available.*
Wow, that's some "You must really, really want to run OS/2" level marketing right there
There are other industries beyond finance, but that was a large one.
Personally, I’m happy for it existing. I hope that at some point open source efforts succeed or that Arca can secure access to the source at some point in the future.
I have been an OS/2 user back in the days, the system was great at that time, much better than the Windows counterpart, with a great UI and file system (which for example did not need defrag). It was however affected by one major issue: at the UI level it had a single event queue (http://www.os2ezine.com/v1n5/queue.html) and if some application would hang while processing its events, it would freeze the UI for good.
Also, how does that work given their $129 price tag? I have to separately buy a license to OS/2 and then their software modifies the distributed binaries?
No - when you buy a copy of ArcaOS you get an OS/2 license. Arca Noae have an agreement with IBM to resell OS/2 licenses, they just don't have access to the source code.
This is very easy to believe. In Summer of '95, I interned at a tester for a group in IBM Austin that was developing a product called OS/2 Lan Server Enterprise. This was a version of Lan Server where all the underpinnings were to be replaced with the DCE product they were developing concurrently. (So we'd write scripts to do things like create a user with an LS command and verify that the user was created with a DCE command.)
One of the other concurrent goals at the time was to remove Microsoft code from OS/2. Developers on the team were classified into those who had seen Microsoft code and those who had not. Even seeing the MS code was immediate disqualification from writing anything new, and this was actively enforced.
> I have been an OS/2 user back in the days, the system was great at that time,
I'd always wanted to be excited about OS/2, and the IBM job was about as close as I got. My day to day PC at the job was an ancient (1988-era) PS/2 Model 80, and it ran even the current versions of OS/2 surprisingly well.
Aside from IBM's odd marketing ("Get Warped" ?!?!?), Microsoft did a better job meeting customers where they were... both Windows 3 and Windows 95 are great examples of this. OS/2 was technically better, but never quite aligned with the market.
(The first release of OS/2 cost about triple what DOS did, had significantly higher system requirements, and needed specific software to enable the new functionality... which did not initially include a GUI. It was a very difficult sell. It might have gone differently if Microsoft had been able to convince IBM to require a 386... but OS/2 was part of the PS/2 strategy, and only the highest end of the four PS/2's had a 386.)
IBM has always been super cautious about intellectual property. I am an opensource committer at Eclipse Foundation (which was created by IBM) and there is a strict IP review process. A copy/paste from stackoverflow is not allowed.
> I'd always wanted to be excited about OS/2, and the IBM job was about as close as I got. My day to day PC at the job was an ancient (1988-era) PS/2 Model 80, and it ran even the current versions of OS/2 surprisingly well.
I started using OS/2 Warp on a 486 66Mhz with 4GM RAM, it was running pretty smooth for me.
IBM originally decided to call OS/2 3.0 “Warp” as in “warp speed”, and market the OS with a Star Trek like theme, but forgot to run this by Paramount before announcing the name. Paramount wouldn’t give them permission and the name was already public so IBM had to settle with using “warp” as in “bent or twisted” rather than “extremely fast” like I’m sure they would have preferred.
https://web.archive.org/web/20171029031903/https://www.insea...
My first job at Novell in 1988 was testing their OS/2 driver. In order to get OS/2 up and running, they had to buy a 4MB memory expansion card for my 386 box. IIRC it cost about $2000 at the time.
Whilst there are Linux workarounds for most Windows situations, Wine, alternative software, LibreOffice etc., often these can be messy or inconvenient to implement (in my case they often occur when longtime Windows users ask me to upgrade their Windows machines and they don't want Win 10 or 11 installed on them.
Here, the switch to Linux is often a step too far (and often they've legacy Windows apps that won't run under Wine—for instance, certain Windows CAD programs that have no Linux version (and their dongles only complicate matters)).
This brings me to alternatives: Acra Noae/OS/2 is too old and unsuitable, this only leaves ReactOS, which is a project going nowhere—its development speed makes a snail look like a hare (why there's so little interest in this project is odd given that the payoff would be great if it were successful).
The question I've never had answered successfully is why after all these years hasn't there been a successful clone of Windows. And by that I don't mean a clone of the grossly over-bloated 7GB kind currently availale from MS—the outrageous size of these behemoths is one of the reasons why users like me want a small streamlined alternative O/S.
ReactOS would fit the bill re small size and basic features but it just doesn't work and probably never will. Nevertheless sufficient of it exists to prove the concept is actually feasible—especially, if the principal aim was limited to just building a basic utilitarian Win API- compatible OS with a basic CUA-type GUI. For many of us, a minimalist GUI-based file loader/task switcher is all that we need (the O/S I envisage would be more like OS/2, as such, given its limited features, it'd hardly even qualify as a Windows clone).
It's perplexing that a viable alternative product hasn't emerged in recent decades and I'm not totally convinced that Microsoft's saber-rattling over copyright etc. is the full reason it hasn't.
MS hasn't made any serious moves to close down ReactOS, perhaps because there aren't any copyright breaches serious enough to prosecute, or it's worried about antitrust/monopoly action being launched against it—or that it doesn't need the PR problem of being accused again of being a bully, or that it thinks that ReactOS is too insignificant and too late to worry about—or it' a combination of these factors.
MS would likely have little difficulty in scaring off 'respectable' Western money for development but I doubt that the same pressure would be just as effective in Russia or China—remember the publicly ReactOS received when it was demonstrated to Putin several decades ago? It was quite a splash.
The question is why those countries didn't take that opportunity and run with it. The economic and strategic advantages seem obvious, so does the kudos that they would've gained. Clearly, I'm missing some important factor here.
Re OS/2, it would be interesting to know details of the agreement between IBM and Microsoft and what percentage of the code each contributed to the project. If IBM did most of the development then perhaps the MS code could have been stripped out but I'd suspect the reverse was the case.
I'm a lot more sympathetic to Microsoft than most. They used some highly anti-competitive tactics, but their products fit the market well, and the competition suffered from a combination of terrible luck and serious strategic miscues.
OS/2's a great example of this. If you bought it in 1987/8, it cost three or four times the price of DOS, had higher system requirements, and gave you no benefits at all without totally new software. Windows 3.0 did well because it was cheaper and worked with the hardware and software that you already had.
> It's perplexing that a viable alternative product hasn't emerged in recent decades and I'm not totally convinced that Microsoft's saber-rattling over copyright etc. is the full reason it hasn't.
It's an expensive effort, and the ability to fund an effort like that can be inversely proportional to the willingness to take on the risks you mention.
> Re OS/2, it would be interesting to know details of the agreement between IBM and Microsoft and what percentage of the code each contributed to the project. If IBM did most of the development then perhaps the MS code could have been stripped out but I'd suspect the reverse was the case.
There was enough Microsoft development that IBM was working actively to remove Microsoft code from OS/2 in 1995. The Lan Server Enterprise was a part of that, but there was also Workplace OS, which was a fully IBM effort to rebuild OS/2 into a portable OS on top of Mach. Workplace OS feels a lot in retrospect like a way for IBM to turn OS/2 into something more competitive with Windows NT.
I'd say you're right and replicating Windows now would be hellishly expensive but it wouldn't have been so some decades back. After all, ReactOS has essentially proved it's possible.
Despite the fact that it's not practicability usable it sort of works in limited circumstances (I periodically test it). Given that it's a smalltime backyard effort but can run MS Office on some hardware demonstrates to me that if serious effort were put into its development it would do most of my needs.
The other problem of there not being any other API-compatible OS in competition with Windows is that Microsoft has never fixed the many problematic issues with Windows—as it simply never had to. It's too complex to discuss them here but if MS had had fair competition from another OS then end users would have had a much better run.
https://archive.org/details/ecomstation2.1
Knock yourself out. :-)
Just want to look at it, not install it? Bootable demo CD:
https://www.ecomstation.com/demo.phtml
Mirror:
https://archive.org/details/ecs22demo_usa_beta5_202011
NOTE: it will not boot from USB. You need to burn this onto an optical disk, and boot from that disk in an optical drive, if you want to run it on hardware.
If you just want to run it in a VM, it'll be fine.
Who the heck buys software without a test drive?
The $129 are a lot cheaper than $250 in 1995-$, inflation-corrected probably nearly doubled since then. If someone needed it back at that time, today that's a real bargain.
https://www.nytimes.com/1995/03/09/business/company-news-vob...