Get modem status?

[Clay Clear]

How would I get the modem status (Line Status Register /
Modem Status Register values) for external serial port modems?
In an attempt to find the answer to my previously posted topic,
I was perviewing the COMMUNIT.BAS file that ships with PB/DOS 3.5.
I tried the DsrStatus function and it ALWAYS returned the same value,
regardless of whether the applicable modem was turned on or off.
I assume that this is because the modem is an external one, so
the modem status register is in the modem, not in the system memory?
If that's the case, how would I go about retrieving its data?

Thanks for any replies.

P.S. If a workable solution is uncovered, I will write my OWN
version of DsrStatus (and Carrier), because PB included
a copyright notice in the BAS file. 'Sides, I prefer to use
my own rolled routines, and I have already found the applicable one
under INT 14 in Ralf's Interrupt Lists.

ADDED: I just tried the COMMUNIT.BAS Carrier function in a test
program, and it DID correctly return the status of the connection.
So, obviously, the Modem Status Register can be culled even for
external serial port modems. So, why does the DsrStatus function
ALWAYS return the same value? It does so with PB's version, and with
my own. My version uses the INT 14 call to do it.


------------------
Clay C. Clear
mailto:[email protected][email protected]</A>

[This message has been edited by Clay Clear (edited October 07, 2002).]

[Tom Hanlin]

The status registers are a function of the serial port, not the modem.
It doesn't matter whether it's an internal modem or an external modem.

I don't remember what the DSR state means, but it may not be
supported here. Not all serial lines are needed or used by all
serial devices.

------------------
Tom Hanlin
PowerBASIC Staff

[Clay Clear]

Tom,

All I know is, the API GetCommStatus (or whatever it's called)
works correctly in my weather bulletin retrieval program. In
other words, if one of my modems is turned off, it bumps down to
the next avaliable serial port. If that modem is not available,
then it errors out. The 32-bit program checks the DSR status
to determine if the modem is turned on. First, though, it checks for
erros in opening the commport, which will be generated if the
port is already in use by another program. I can't seem to get either
of these functionalities in a DOS program. <sigh>

Oh, well, the workaround, 'though quite "inelegant", is obvious.
Simply have my DOS program create a plain text file that simply
lists the comm ports to be queried, then call a 32-bit program that
both checks the availability of the modem, and the DSR status.
Then it will replace the lines in the TXT file with their respective
stati. In fact, I can disregard having the DOS program initially create
the file - it'll simply call the 32-bit program, which will write out
the stati of all the commports currently on the system, and THEN the DOS
program can read in the file.

DSR stands for "data set ready". In other words, is the modem
available and ready for use? ("data set" refers to the device
connected to the serial port, NOT the computer ). I *think*,
anyway. It's been so long since I've read up on modem jargon,
I can't be certain.

Thanks for your reply.


------------------
Clay C. Clear
mailto:[email protected][email protected]</A>

[Lance Edmonds]

To test DSR you need to test the 6th bit in the Modem Status Register... precisely what is done in the "DsrStatus" function in COMMUNIT.BAS.

However, if you are running under NT/2000/XP, then you're probably getting virtualized port values only, since those OS's virtualize all hardware access to Ring-3 (application-level) programs.

What level of success do you have on the same hardware but when running in plain DOS mode?

------------------
Lance
PowerBASIC Support
mailto:[email protected][email protected]</A>

[Mike Luther]

Lance,

I too spotted this a long time ago... mentioned it then and got no
reponse at all when I did.

P.S. If a workable solution is uncovered, I will write my OWN
version of DsrStatus (and Carrier), because PB included
a copyright notice in the BAS file.

Technically, I think he's absolutely right as to the fact that even
though the stuff is provided, you actually can't use anything!

Is there a place anywhere in the documentation, or on the Web site,
where a disclaimer is provided that the routines may be used by
anyone who is a valid or licensed user of the PowerBASIC products?

Thanks ..

------------------
Mike Luther
[email protected]

[Tom Hanlin]

I'm not sure what confusion there may be here. If you've paid for
PB/DOS, of course you can use the samples that come with it, just
like you can use the PB/DOS compiler, which is also copyrighted.
They'd be of no value at all if you couldn't use them.

Considering that you can download the samples right here on the
PB web site, I doubt redistribution terms are onerous. I can look
into it, if you'd like.

------------------
Tom Hanlin
PowerBASIC Staff

[Mike Luther]

Tom ..

I'm sure that what PowerBASIC intends with respect to re-distribution
rights for the source code samples are obviously not what's shown.
I think all you folks need to do is to post a blanket policy release
as is really intended on the web site and on the distribution disk(s)
that covers it.

Clay is only concerned on getting it absolutely right and PowerBASIC
is far too good an outfit to fuss with such a notice glitch-a-poo.
If legal goes along with the issue, it's trivial to cure what he saw...
If they don't ... then what he saw really is of concern, one would think.




------------------
Mike Luther
[email protected]

[Tom Hanlin]

The copyright just means that PowerBASIC has the right to set the
distribution terms, Mike, no more or less. I don't see where this
is cause for concern or what you mean by "not what's shown".

I'll be glad to see if I can get the license text posted, though.
If you're in a hurry, try running the PB/DOS setup-- I imagine it's
part of the licensing terms you agree to on installation.

------------------
Tom Hanlin
PowerBASIC Staff

[Mike Luther]

Thanks Tom ..

The issue is simple. The code in question I think, is:

'====================================================================
'
' Serial and Parallel Port Routines for PowerBASIC
' Copyright (c) 1995-97 by PowerBASIC, Inc. All Rights Reserved.
'
'====================================================================
'====================================================================

The statement "All Rights Reserved" means EXACTLY what it says it does
Tom. It means that the complete code in the file to which that
header is displayed may not be used by anyone for anything without
the expressed permission of PowerBASIC, as far as this non-lawyer
knows.

The key to the issue is what constitutes that expression of right
to use, where it is displayed, and, for the user of the product,
a reasonable declaration of where that modified right exists.

PowerBASIC, as far as I know, does not intend for that total reservation
of rights to be at all true at all in this case. It *IS* provided for
use by a licensee of the product to use the example provided. Heck, it
wouldn't be provided unless that were the case!! But the words do mean
exactly what they say, to the extent that a modification of that text
cannot be found in clear proximity of that restriction, to offer relief
from the restriction for any user.

No, it's not at all what is wanted. But it stands as written unless
it is clearly set forth otherwise. You know, a comma here and the
word "except" and so on ... grin.

Again, I'm no lawyer. Just an observer.

To do what is really wanted, I suspect either of two things must be
present in the medium of distribution or someplace in clear public
notice. You have to either deliberately modify the EXAMPLE code, file
by file, to somehow state that the code is proper for inclusion in source
for a licensed user of PowerBASIC code, or ... it's likely good enough
for the crew to add a common file in the READ.ME of the source directory,
or on the WEB site that provides that relief. That common partial release
statement would likely be something to the effect that, "The right to use
of all source code examples provided with this product is expressly
granted for any licensed user of the PowerBASIC compiler toolset to which
these source files are attached, as may be more expressly set forth in
the general terms of license for this product or elsewhere."

Your legal crew can help you draw it.

I hit this exact same thing with what Clay noted many months ago in
relation to some of these exact same routines and was forced to do
exactly the same thing; re-write the whole lot of them just to be
very clear the 'samples' were not included in my work. I asked at
the time for explicit relief from that restriction to use the samples.
No answer was ever given.

So, I had to do exactly what Clay noted; re-write every single one of
the examples, variable names, the whole works, just to cover the
bases. That's obvioulsy NOT what was ever intended by your very fine and
thoughtful outfit ... But the statement, unabated, is clear.

Please .. don't shoot the mule. I'm just trying to help.

------------------
Mike Luther
[email protected]

[Tom Hanlin]

I'm afraid you have some confusion about the nature of copyright
law. There's a good casual discussion of the subject in one of the
files in the Download section-- might be worth looking over:

http://www.powerbasic.com/files/pub/docs/shrware.zip

As I said, though, I'll see if I can get the distribution info online.

------------------
Tom Hanlin
PowerBASIC Staff

[Clay Clear]

Lance,

Sorry for taking so darned LONG to follow up. It's just that
I've been up to my armpits in programming projects of all kinds.

I followed your idea of trying the same test DOS EXE both under
Windows and then under pure DOS.

The INT 14H routine I had written did not work correctly in
either case. So, instead of debugging it, because of time
contraints, I simply completely rewrote it using PB's COMMUNIT.BAS
DsrStatus's FUNCTION. Once I did that that, it worked correctly.
Under Windows, it ALWAYS returned the same result for the serial
ports, regardless of whether the modems were turned on or off.
It always returned TRUE for viable serial ports, and always returned
zero for nonexistent serial ports. Under pure DOS, the same EXE
correctly returned zero for both nonexistent ports AND for existing
ports where their modems were turned off. For exisitng ports where
the modems were turned on, it returned TRUE.

So, I think this proves that your theory is correct, about the
incorrect repsonses when run under Windows being caused by the
fact that Windows virtualizes the ports. This should also
explain why the API GetCommModemStatus DOES work correctly
from within a 32-bit program, because, since Windows basically, "owns" the
ports, it should "know" what their stati are.

Thanks for your insight, Lance. Once again, I apologize for my tardy
response. I posted this followup so other Forums members can
see what might be causes to their problems of a similar nature.

Note that my version of the CarrierFUNCTION also failed to work.
When I rewrote it, this time using PB's COMMUNIT.BAS's version, it
worked correctly. Also note that it works correctly also under
Windows, unlike the DsrStatus function. I do not know why
my versions did not work. As I said, I did not take the time
to debug them.

Tom,

Thank you very much for your clarification of the copyright
status of the COMMUNIT.BAS' routines. I am quite happy
that you commented on them, as I now know that I may use
them in my production code, which I wish to do, since my own
versions did not work correctly, whereas PB's did.


ADDED: The test DOS EXE was run on my Win98SE box.

------------------
Clay C. Clear

[This message has been edited by Clay Clear (edited October 17, 2002).]