Time for a OS/2 box

[Leonardo]

After reading about the history and evolution of Amiga and now about Apple thinking on moving away from Intel, I came to the thought that maybe we should try to find a way to create a box specially design to run the latest OS/2 without having to worry about compatibility issues and it would focus efforts on one set of requirements (drivers and features like hardware 3D ... and some other dreams).

I know that some vendors offer boxes that are 100% tested to run OS/2, but this is more on the line of a BeBox or the Amiga example. Times are different now than back in the late 90's when Intel/Microsoft would crush any resistance...


What do you guys think?... what about PS/2 .... haha!

[Martin Iturbide]

Hi Leonardo.

I think it is not a bad idea. On these days  it seems very easy to create customized machines from china factories. If someone is interested on creating/selling this will have to think more about the strategy and what kind of box they want to offer. I personally don't see it as Bebox, but a more smaller scale (non big ATX case). But that is something that it will never get a consensus on the community   I notice that a people like to have ArcaOS on laptops. It will be interesting to know more about the personal preference of OS/2 users to know if they use more ArcaOS on Desktops or Laptops.

Regards

[Neil Waldhauer]

I think it's good not to overestimate the size of the market for these machines. Can you have a machine built in china if your total run is going to be 8 computers?

[Doug Bissett]

This topic has been discussed, many times, before. It is a good idea, until you start to analyze what needs to happen to actually do it (and, for now, we can ignore the laptop problem).

First, somebody has to put out some cash to buy a few different combinations of hardware (the case is the least of the problems) and determine what works best. Then, they need to determine if the devices are still available, substitute something else for those devices that are no longer available, test again (repeat as necessary), and then they need to put out enough to purchase parts for about 1000 machines (to get quantity discounts). Okay, after a month, or so, you get a truck load of parts, that need to be assembled, and tested. Then, you can offer them to prospective buyers, and each one will want different features. If you are lucky, you might sell 75% of them in the first year. By then, all of the parts will be obsolete, so there would be an expectation that the price will drop. You also need to determine if you will sell them with OS/2 (ArcaOS) pre-installed, or if you will just sell the hardware. Then comes the problem of whether you include an option for windows, or Linux. Do you pre-configure the HDD, or let a user do that? And so on. You would probably also run into problems trying to ship across international borders, and methods to get paid.

Laptops, of course, come as they are. you cannot just mix, and match, parts.

Keeping a database of current hardware (which is usually 6 months out of date, in a world where there is a 3 month turn around on availability) is probably the best that can be done. The advantage is, that by the time somebody says that a specific device actually works, it will likely be on the clearance shelves, at a discount (if you can find it), but the replacement part will still be cheaper. Then, there is the problem where a manufacturer changes the specs on a device, part way through a production run. The 1000 devices that you order may not work, while the one that you tested did work (this problem also exists while trying to maintain a database).

Good idea? You bet, Will it ever happen? Not very likely.

[Bogdan]

After reading about the history and evolution of Amiga and now about Apple thinking on moving away from Intel, I came to the thought that maybe we should try to find a way to create a box specially design to run the latest OS/2 without having to worry about compatibility issues and it would focus efforts on one set of requirements (drivers and features like hardware 3D ... and some other dreams).

This idea already failed with IBM's Power Series in 1996. In the end this machines were used as low-end AIX workstations.

I know that some vendors offer boxes that are 100% tested to run OS/2, but this is more on the line of a BeBox or the Amiga example. Times are different now than back in the late 90's when Intel/Microsoft would crush any resistance...

Unfortunately Intel and Microsoft have a bigger market share nowadays. No more Sun, Digital, SGI or HP.

[Eduardo Vila Echagüe]

I think an alternative could be that Arca Noae approaches some existing laptop manufacturer and work together the more needed drivers for one of their current models. Some way should be found for the coexistance of Windows and ArcaOS. Most OS/2 fans will buy that model. Perhaps some enterprises also. The total market I think will be a few thousands.

Eduardo Vila Echagüe
Santiago, Chile

[Bogdan]

I think an alternative could be that Arca Noae approaches some existing laptop manufacturer and work together the more needed drivers for one of their current models.

Good joke, there was no known hardware certification or vendor specific driver development in the last 11 years.

Some way should be found for the coexistance of Windows and ArcaOS.

Virtual PC is available for 14 years now and you can install OS/2 as a guest system. Is there a reason why ArcaOS will not install?

Most OS/2 fans will buy that model. Perhaps some enterprises also. The total market I think will be a few thousands.

By this calculation every fanboy has to buy 100 devices.

[Martin Iturbide]

wow.. I didn't expected this topic will be subject to such a fast reality check !!

The idea of having some brand/model to have somekind of "ArcaOS Certification" is also good. It is not like a company has to invest on an exclusive hardware for ArcaOS, it is more to test a machine where all the components works with ArcaOS and let the people know that XYZ model is running fine.

Bogdan,  saying that just because something haven't happened on the last 11 years it won't today is a good joke for me. Is this some example of an "Appeal to probability" fallacy ?

Regards

[Neil Waldhauer]

While I like the idea of specifying specific hardware that works, what would that entail?

1. USB 2.0 hardware: Not really available in current chipsets, so that's some custom work. Arca Noae may soon have a better driver.
2. Very old WiFi: Our latest support is via GenMac for 802.11b, and even that has some issues. That's almost 15 years old.
3. BIOS: We're going to need BIOS support, and I'm hoping that can be our own custom CSM
4. Disk support: Don't show us GPT or disks larger than 2 TB.

And the market is complex. End users might buy hardware like this, but how would they even know it exists. Likewise, for any industrial users, would they even know to look. Also, most Industrial users want support for Microchannel, ISA and other interfaces that are long gone. If you take industrial users out of the picture, I'll stick with my prediction of 8 sold units.

[Bogdan]

...
And the market is complex. End users might buy hardware like this, but how would they even know it exists. Likewise, for any industrial users, would they even know to look. Also, most Industrial users want support for Microchannel, ISA and other interfaces that are long gone. If you take industrial users out of the picture, I'll stick with my prediction of 8 sold units.

I would share this observations. The industrial field and military sector have enough suppliers for reliable and compatible hardware. For data center or networked end user applications it's easier to virtualize remaining environments.

Enthusiasts can download and run the OS/2 PCM Compatibility Test Program (still available from ftp://ftp.software.ibm.com/ps/products/os2/hwcert/). In fact the results will look disappointing to most of them.

Times are crazy nowadays - it's a bigger business to sell two used Microchannel-based systems from storage than 100 OS/2-compatible systems with more recent specs.