In fact there are some companies still making money with DOS. I worked
for a company until mid 2016 building embedded 386EX systems, which they
still do up to now. They license Datalight ROMDOS as well as a BIOS from
another vendor, but for that I do not remember the name...
Once I was digging in documentation there out of curiosity, there were
also some prices listet in it. The BIOS SDK+Source took 12.500€
(something around 14k $) IIRC + license fee for every device.
Of course, we can discuss if an order of 14.000$ would make a big deal -
it is definitely not, if you have to run a company. But still there is
some money in this niche.
The company I worked for is building radio modems+control devices for
remote control of whatever you like, my job there was to put together
the parts for a device that controls elevators. I know at least 2
elevators running DOS in our place, if I go around and look out for
company-signs I will find more.
The product is being sold since early 2000s and since it is a industry
and not a consumer product it will be there for longer, butat least it
will be serviced (and therefore making some more money). I suspect the
hourly price for service of these devices to be at least 80$ without tax.
Some of these 386 boards support PCI and USB with some ALI chipset and
the drivers for this were written in that company. There are PCI-cards
for radio-communication, that they also wrote the drivers for. (or at
least, they bought drivers - which makes no difference - someone earned
money for DOS-drivers.)
(I still have some unfinished prototypes and gargabe from production
line, which is able to start the bootloader, but some parts are marked
with a red arrow on it - so something must be wrong with these.)
Nils
On 10/01/2018 01:00 AM, dmccunney wrote:
On Sat, Sep 29, 2018 at 7:51 PM Rugxulo <rugx...@gmail.com> wrote:
On Sat, Sep 29, 2018 at 8:17 AM dmccunney <dennis.mccun...@gmail.com>
wrote:
It's no loss to MS to make DOS 1.5 and 2.0 available under a
permissive license.
"No loss" might be inaccurate. While it may be trivial compared to
"newer technology", it's impossible to say that their (MSDN?) revenue
from such legacy software is so low as to be totally worthless.
Fine. So call it "no loss worth *caring* about". I don't know what
revenue from DOS related technologies MS may still generate, but the
amount will be so low it simply won't be *visible* on a balance sheet
or P&L statement. Far larger amounts will be attributable to
*rounding* errors.
Remember that they have hundreds of thousands of employees!
Yep. And I'd be quite startled if the continuing revenue from DOS
related stuff paid even *one* of those salaries. DOS is dead as a
commercial . It has been for decades. Keeping variants of DOS and DOS
apps running are usually hobby labors of love for those doing it. It's
why I shake my head when people talk about getting new drivers written
so DOS can support stuff created since after it was no longer
maintained and supported. The people who can *do* that tend to be high
level programmers who write code for a *living*, and expect to be
*paid* for what they do. They are extremely unlikely to do it for free
as a hobby, and who would *pay* them to do it for DOS? I don't know of
anyone.
DOS has been dead as a commercial product for a long time.
MS isn't the only vendor of a DOS-compatible OS. DR-DOS and ROM-DOS
are still sold online. (Do OS/2 variants also count? Maybe.)
Which OS/2 variants? The one I'm aware of is eComStation,
https://www.ecomstation.com/. The outfit that makes it got the rights
from IBM, and essentially services accounts that still have
substantial OS/2 deployments, and it's cheaper and easier to try to
continue to use OS/2 than migrate to a different architecture.
(Stardock, who does stuff like the Window Blinds and Object Desktop
enhancements for Windows, developed under OS/s, and tried to get the
rights from Microsoft but were unsuccessful. Not sure what they might
have done if they were able to get the rights, but support for 32 bit
apps would have been a major improvement for the OS. Not supporting 32
bit Windows apps effectively killed it.) (I was an OS/2 admin at one
point, running it on a machine that was a specialized telephony
server, communicating with a predictive dialer. ITtjust ran, and if
there was a problem - reboot and things worked. The company making the
dialer ported the controlling app to NT Server.) There are a lot of
aftermarket companies doing stuff like that. Corel WordPerfect is
essentially supporting the large number of companies that ran WP for
DOS back in the day, and moved to Windows versions. These days, Word
owns the word processing market, so *new* sales of Corel WP to folks
who weren't former WP users will be negligible. The same comment can
be made about the outfit that still sells and supports former Borland
IDEs and language products. There is revenue in supporting the
existing market, but that market is highly unlikely to grow.
Yes, DOS is unpopular nowadays, but it's still a well-known niche.
You have a talent for understatement.
There's also still a fair amount of commercial DOS software being
sold (not just games but apps, even if they haven't been updated in
years).
How much *money* is in the niche? "Even if they haven't been updated
in years" is a telling statement. The former development efforts are
sunk costs. It's fairly trivial to keep selling existing DOS products
that have already repaid the costs of developing them. But are any of
those outfits doing *new* development? Show me one...
It's easy to trivialize the decades of DOS legacy that survives. But
certainly just because some hipster/geek somewhere declared DOS
"dead" didn't immediately make all DOS software freeware and/or "open
source". (If some government somewhere did that, there would be
complaints. Granted, a lot of stuff is in legal limbo and unused for
no good reason, so maybe that should be freed, if literally no one
can use it otherwise, but you know that will never happen, sadly.)
I don't trivialize it. My only point is that there is no longer
*money* in it, and no one has reason to do *new* DOS product
development. It's all about the money.
(Speaking personally, I'd love to see *FreeDOS* re-licensed under
something other than the GPL.)
I don't honestly know if that's even legally possible now that Pat
has died. (Gotta love legalese, ugh. No, I'm not a lawyer.)
I don't believe it is possible.
I also don't think GPL hinders many potential contributors (versus,
what, BSD two-clause??). I'll admit that GPL can cause some practical
problems, in rare cases, but it also avoids or solves some other
practical problems (again, in some rare cases).
The viral nature of GPL makes it a problem in many cases. The nature
of the license is that if your code links against GPLed code, it too,
becomes GLPed. An awful lot of the code you might want to link against
GPL code will be off limits because the folks who wrote that code
won't accept it becoming GPLed. My irony meter pegs off scale when an
open source project can't use code from another open source project.
The nail in that coffin for me was when GPL3 was incompatible with
GPL2. I increasingly wish someone had torpedoed Stallman in his crib.
(I've met him. He does not live in the same world I do. He reminds me
of a monk in the middle ages, supporting himself with alms donated by
pious followers and totally focused on his idea of who God is and what
He wants. He has only tenuous contact with the world the rest of us
live in. The last time I met him, he was distributing a screed against
eBooks. Conversation with him revealed his real problem wasn't that
they were electronic - it was that he could not pay for one in cash,
but had to have a credit or debit card. I don't think he has either,
and wonder what he will do as society becomes increasingly cashless.
What happens to him when cash goes away? (And it might in his lifetime.)
FreeDOS seems to mostly focus on "four freedoms" (free/libre), aka
GPL or OSI. As long as we're as "free" as possible, I think we're
okay. It gives us the most advantages, and it helps the most people.
But I don't think splitting hairs on that end will (practically)
improve anything much, if at all.
Agreed on being as free as possible, and the question is how free
FreeDOS *can* be. The bigger question is "Why use FreeDOS at *all*?"
No amount of freedom will compensate for no plausible use case to make
the effort worth expending. See above about "hobbyist labor of' love."
There aren't a lot of DOS contributors anyways. Heck, most people act
like they can't even install a compiler or figure out a "simple"
makefile. I do honestly wonder where all the decades-worth of
DOS-savvy developers went. Certainly not everyone forgot literally
everything, but it's such a complex world, and people have other
priorities. It's just sad that so much working software was
abandoned, deprecated, thrown away, left to rot. I think most people
just don't care (but certainly many act like they can't figure out
anything). I mean, when even a noob like me can get more done than
them, you know something's wrong! ;-) People give up too easily,
dismiss failure as "normal", they have really low thresholds of
patience and testing.
The DOS savvy developers went where the *money* is. People who get
paid to write code tend to do that.
I think software is overengineered. Simplicity is a virtue (says Dr.
Wirth). I think it takes a lot of hard work and effort to simplify
things (without losing functionality). It takes a genius to be simple
and elegant. But most people don't have the time, patience, or energy
to do it. Or maybe I'm idealizing too much. It's a complicated world.
(And no, obviously I'm no genius.)
Much software is over engineered. A lot of the stuff like that I've
seen is a result of not properly understanding the problem domain and
knowing what your software needs to do.
(Sorry for the ramble.)
No apology required. _______ Dennis
https://plus.google.com/u/0/105128793974319004519
_______________________________________________ Freedos-user mailing
list Freedos-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/freedos-user
_______________________________________________
Freedos-user mailing list
Freedos-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/freedos-user
