Thinking of moving to PowerBASIC For Windows?

[Michael Mattias]

> You have to receive the product by snail mail to qualify.

And?

See if you can get that same deal on Microsoft Office.

And it's not like this 'need' came up just this past Thursday and the finished application has to ship this week.

And as long as I am posting.....

I thought of an alternate solution: engage a consultant to advise and/or design the software and recommend a program development product. This is why consultants exist.

[Jim Padgett]

I don't think it is a question of if the compiler will be around in a
coupld of years, but if the current OS will support DOS in a couple
of years.

------------------
Warped by the rain, Driven by the snow...

padgettjatcomcastdotnet

[Chris Holbrook]

Originally posted by Michael Mattias:

I believe the PB/DOS compiler is "inexpensive", not "cheap."

This language which divides us. To me, "cheap" means inexpensive, with no pejorative overtone. But then, I have a great imagination.



------------------

[Michael Mattias]

>The compiler is really cheap

I believe the PB/DOS compiler is "inexpensive", not "cheap."


MCM

[Bob Zale]

You have quite an imagination, Chris. Best of luck finding what you want.

Bob Zale
PowerBASIC Inc.

------------------

[Chris Holbrook]

The compiler is really cheap, if I thought that the it would still be around in a couple of years I would buy it just to evaluate, but Steve Rossell's reply avoids making any tangible committment to it. Maybe no other vendor offers a comparable supported product, I should really check this first!


------------------

[Steve Rossell]

Originally posted by Chris Holbrook:
What exactly is PowerBASIC's committment to the DOS product? Will it still be supported in two year's time?

PowerBASIC 3.5 For DOS is a supported product and will remain so for the foreseeable future.

Sincerely,
Steve Rossell
PowerBASIC Staff

[Ian Cairns]

I believe you should check first. I thought that electronic downloads of PowerBasic software were not refundable. You have to receive the product by snail mail to qualify.

regards, Ian.

------------------
IRC

[Michael Mattias]

>Not having a PBDOS licence, I can't even look at the PBDOS online manual

AFAIK, the 30-day money-back no-questions-asked policy applies to ALL PowerBASIC compiler products.

Surely you can decide in 30 days if you can accomplish your task using the PB/DOS compiler.

You can confirm current sales policies by writing mailto:[email protected][email protected]</A>

[Chris Holbrook]

For Steve Rossell

Originally posted by Steve Rossell:
Thinking of moving to PowerBASIC For Windows?

Sincerely,
Steve Rossell
PowerBASIC Staff

No, I wasn't - I already use the PB Windows and CC compilers, and I hope to use the DOS one too, for an application which I am developing - messages above refer. The prototype (non-PB, but developed using an unsupported compiler) works.

What exactly is PowerBASIC's committment to the DOS product? Will it still be supported in two year's time?




------------------

[Chris Holbrook]

My earnest thanks.


------------------

[Michael Mattias]

"NO" is the answer to the innterrogative form of your statement, " Actually I was thinking about moving to PowerBASIC for DOS, if it can do multithreading."

As far as accomplishing something, well, all things are possible given imagination, skill and earnestness.

[Chris Holbrook]

So it's "no" then?



------------------

[Michael Mattias]

Well, OK, here's the answer to the original question:

There are no intrinsic verbs, functions or compiler options offerend in the PowerBASIC/DOS compiler designed to support multi-tasking, task switching, asycnronous execution or real or simulated multi-threading.

[Chris Holbrook]

Paul,
OK, we are talking about the same thing, but "threads" doesn't imply "windows".
What I want is to create a 16-bit DOS application which will run under an MSDOS-like OS and can also run in a Windows environment (as a console application in XP, etc etc). The application has to support task-switching to allow concurrent threads, which will provide a multi-user capability of sorts on the DOS machine. More than one of these "users" will require a physical console of some description for IO. I'm considering this an alternative to licencing an OS which allows concurrency.
I know that it can be done, my question is really to determine whether PBDOS (I don't have a licence for this product) has any features which will help. Built-in assembler may be all that I need for the task switching.

Not having a PBDOS licence, I can't even look at the PBDOS online manual, so I'm prototyping with another 16-bit compiler (shame!).



------------------

[Paul Dixon]

Chris,
I mean that you are thinking in terms of how you might do this job on Windows. DOS isn't Windows but may also be able to do the job, just not the same way.


<<Presumably to start a thread you need to .. >>

Since you'll not be running in Windows you can choose how you want to do this. You're no longer restricted to the way Windows would have you do it so you don't need to presume anything about the way it works. YOU decide how you want it to work and write the program accordingly.

I don't know what these events are so it's difficult to be more specific.

If they are external hardware events like key presses, serial port activity, changes on external I/O ports, then DOS can handle these events using interrupts far better than Windows could ever handle them using threads.

If your asynchronous events are people accessing remote websites which you are monitoring or arrival of e-mails or activity on a USB device then it's more complex for DOS because the functionality to do these tasks is built into Windows and isn't built in to DOS.

Paul.

------------------

[Barry Erick]

If you simply want it to "look like DOS", use PBCC. That is the console
window version. And the console window is what looks like a DOS
box. But if you want it to run on a DOS machine, then you are trying something very hard to do in
DOS. For communications in the background, there is event driven statements.. ON COMM and others, although
in DOS, I hated those and simply polled, quite often, the buffers
to see if anything came in, and then go to the routine. That looked like multasking, but wasn't.

------------------

[Chris Holbrook]

Originally posted by Paul Dixon:
Now you're in trouble. If you're going to do the job in DOS...

Paul, what on earth do you mean?



------------------

[Chris Holbrook]

Originally posted by Michael Mattias:
>MS-DOS does not support asynchronous anything.

er, it's the events that are asynchronous.

------------------

[Paul Dixon]

Chris,
<<What sort of product life do you envisage for PowerBASIC for DOS?>>

I bought my PB/DOS 10 years ago and it has shown no signs of wearing out. I expect it to last at least that long again so the life of the product should not concern you.


<<Probably six or fewer concurrent threads. Presumably to start a thread you need to grab memory for a stack segment and copy the registers..>>

Now you're in trouble. If you're going to do the job in DOS then you have to throw off the shackles of Windows and stop using Windows language or you'll be forced into thinking Windows while programming DOS which is not the way to go.
Forget "threads" and think of what your program needs to do.

Mike,
<<You can't even load a 16-bit application under 64-bit versions of Windows>>

That doesn't mean DOS is dead. It means a future Windows will drop support for it. DOS and Windows are 2 different things, they can live independently of each other. DOS will continue to live on in millions of computers for many years. It has already outlived 16-bit Windows. It's still going now 32-bit Windows is in decline and it'll likely outlive 64-bit Windows.


<<TSRs don't count.>>

If they do the job then they count. Whether they match your expectations of a Thread as you might be used to under Windows is a different matter.
There's nothing to stop a DOS program running multiple, independent pieces of code and switching between them far more efficiently than any Windows system ever has. A little ASM and an understanding of interrupts comes in useful here.

Paul.

------------------