rem

So for example if i want to use OPL3 card i would do this:

SET BLASTER=A220 I5 D1 H5 P330 T6 (OPL3)
rem SET BLASTER=A220 I3 D1 H5 P330 T1 (OPL2)

And than save & run?

Yep - 'rem', it stands for 'remark', as in, 'what follows is just a comment, don't run it'

You could of course add those seperate 'SET BLASTER' lines into individual .bat files and run whichever one you want before a specific application. So run OPL2.bat or OPL3.bat, then the game/software you want to run.

If you're talking about DOS 6 or higher you don't need separate BAT files - you can use the menu commands to choose which SET lines you want to execute. So which DOS version?

Hmm, didn't think of that, i thought autoexec.bat was a master and only it could activate what i need.

Thanks Megatron, and thanks for explaining the command!

Oh, and, one last question, if i make 2 bat files, autoexec.bat should be empty?

[Edit]

@Norton

I use windows 98SE command mode. That is restart in MS DOS.

Norton Commander wrote:

If you're talking about DOS 6 or higher you don't need separate BAT files - you can use the menu commands to choose which SET lines you want to execute. So which DOS version?

Stojke wants to change on-the-fly, ie before running a particular programme. Choosing which one to use at boot-up via the config.sys/autoexec.bat menu is confusing the issue - all that needs to happen in this use-case is to change the environment variable based on which app/game/util Stojke wants to run next.

megatron-uk wrote:

Yep - 'rem', it stands for 'remark', as in, 'what follows is just a comment, don't run it'

Here's an interesting, useless bit of trivia:

Instead of using the REM statement, prefix your comments with two colons.

:: Like This

Using REM, the command interpreter reads the whole line before reading the next line. Using the double-colon method, the command interpreter hits the second colon, determines that this is an invalid label, and immediately jumps to the next line.

Probably 0.000001% speed difference on a system with more than a dozen MHz of horsepower, but could theoretically speed things up a bit on an old 8088 or something, if you have long blocks of comments, and you're running off a floppy disk.

<-- nerd.

Nerd!!!
😉

But didn't know that back when it was useful to me...

Hmm, seems that it doesn't matter, with or without the commands, tracker program still instructs both sound cards...

Assuming entering the SET environment for a card at the command line just before loading the tracker program doesn't help, :

The tracker program might be accessing the port 388 directly instead of looking for the SET parameters. If that's the case, both OPLs on the cards will be called. Unless, a program accesses the 388 port via the I/O port, for example, the common 220 or 240 I/O ports, which you can specify in programs that have proper setup or install program, it goes back to square one.

I once selected the Adlib sounds in Colonel's Bequest (Sierra Game) when I had 2 sound cards installed - a CT1350B and a CT2760 (SB1.5 & AWE32), and both FM chips started making music.

If the program does not have a utlity or setup to select the proper I/O address of a sound card, :

If it's a PnP card, the PnP utility might help - I'm using the Creative Labs' PnP utility to "shut off" the FM synthesis in my AWE64 Gold card, to route the FM requests to my SB16 CT1740's true Yamaha OPL3 chip.

To "shut off" the said chip, you have to select another "Configuration number" which does NOT have the FM address selection. When this is done, and when the card is initialized with the CTCM program, the PnP card will be given only the parameters of the selected configuration number. Purportedly, this disables the FM synth on the PnP card. You can double check the output when this is done.

Of course, the above is easier to follow if using a PnP card with the Creative's PnP utilities in DOS. In Windows, the same configuration tinkering will help.

[quote="SquallStrife"]Here's an interesting, useless bit of trivia:

Instead of using the REM statement, prefix your comments with two colons.

:: Like This

Using REM, the command interpreter reads the whole line before reading the next line. Using the double-colon method, the command interpreter hits the second colon, determines that this is an invalid label, and immediately jumps to the next line.[/quote]

No, that's not true.
When DOS sees a colon in the second position, it assumes that the first character is a drive letter. A colon is not a valid drive letter, so the rest of the line, which is treated as filename and parameters, will fail to execute.
However, for DOS prior to v7, it still attempts to *execute* that line (result: "Invalid drive specification" and "Bad command or filename"). Later versions of DOS know that it can't run, so they skip the line silently.

edit: fix my spelling mistakes so that it makes sense. :-)

peterferrie wrote:

No, that's not true.

Sure it is.

peterferrie wrote:

When DOS sees a colon in the second position, it assumes that the first character is a drive letter.

Why would it consider the second colon, when it knows the first one denotes a label?

That's not only counter-intuitive, it's counter to the bare mechanics of how batch file execution happens.

peterferrie wrote:

A colon is not a valid drive letter, so the rest of the line, which is treated as filename and parameters, will fail to execute.
However, for DOS prior to v7, it still attempts to *execute* that line (result: "Invalid drive specification" and "Bad command or filename"). Later versions of DOS know that it can't run, so they skip the line silently.

It's understandable that REM would be slower when echo is on, but with echo off I'm skeptical that there is a performance benefit because the command processor has to locate end-of-line to find the beginning of the next line in either case. Also, I have to wonder if using :: for comment lines could slow down GOTO commands, because they will be compared, even if briefly, to the label being searched for.

ripsaw8080 wrote:

It's understandable that REM would be slower when echo is on, but with echo off I'm skeptical that there is a performance benefit because the command processor has to locate end-of-line to find the beginning of the next line in either case. Also, I have to wonder if using :: for comment lines could slow down GOTO commands, because they will be compared, even if briefly, to the label being searched for.

That first bit is entirely possible.

Although it shouldn't affect GOTO commands, because they are invalid labels, they should be rejected by the interpreter all together.

Edit: To elaborate, I don't think that the entire file is re-scanned every time GOTO is called. The way I believe it happens is that valid labels are stored in memory alongside a disk location, as kind of a rudimentary pointer. Then when GOTO is called, it check for a matching pointer, jumps to that disk location, and continues executing.

It would be interesting to watch the whole process in a proper debug session.

Well, labels can be before or after a GOTO that searches for them, and I doubt that an initial full pass is made through a batch file to pre-process labels into a table with file pointers... the DOS command interpreter isn't very sophisticated.

Appears you might be right!

Based on this, yeah, you probably would see a performance hit if you used lots of GOTOs and :: comments.

Interesting!

I dare say, though, the interpreter will move on from a :: comment faster than it would move on from a valid but incorrect label.

Also, I wonder if it goes top-down every time GOTO is used, or if it scans down then loops back to the top?

Now I'm curious.

Did I mention nerd? 😜

Edit: Just did a test, and yeah, it always goes top down.

OR... that could just mean that the interpreter ignores identically named labels after the first... Hmm...

Maybe something like this to test with:

1@echo off
2echo first line > outfile
3::a
4::bunch
5::of
6::comments
7::to
8::test
9::the
10::performance
11::of
12::goto
13:again
14echo another line >> outfile
15goto again

Stopwatch it for 10 seconds before pressing Ctrl-C and see how many lines it writes to the output, then try again with REMs. This gives the interpreter every chance to somehow memorize the position of the one valid label. I purposely omit a space after :: because it's possible that label comparison breaks at the first separator, so perhaps another test with an added space will be different.

Malik wrote:

Assuming entering the SET environment for a card at the command line just before loading the tracker program doesn't help, : […]
Show full quote

Assuming entering the SET environment for a card at the command line just before loading the tracker program doesn't help, :

The tracker program might be accessing the port 388 directly instead of looking for the SET parameters. If that's the case, both OPLs on the cards will be called. Unless, a program accesses the 388 port via the I/O port, for example, the common 220 or 240 I/O ports, which you can specify in programs that have proper setup or install program, it goes back to square one.

I once selected the Adlib sounds in Colonel's Bequest (Sierra Game) when I had 2 sound cards installed - a CT1350B and a CT2760 (SB1.5 & AWE32), and both FM chips started making music.

If the program does not have a utlity or setup to select the proper I/O address of a sound card, :

If it's a PnP card, the PnP utility might help - I'm using the Creative Labs' PnP utility to "shut off" the FM synthesis in my AWE64 Gold card, to route the FM requests to my SB16 CT1740's true Yamaha OPL3 chip.

To "shut off" the said chip, you have to select another "Configuration number" which does NOT have the FM address selection. When this is done, and when the card is initialized with the CTCM program, the PnP card will be given only the parameters of the selected configuration number. Purportedly, this disables the FM synth on the PnP card. You can double check the output when this is done.

Of course, the above is easier to follow if using a PnP card with the Creative's PnP utilities in DOS. In Windows, the same configuration tinkering will help.

i see, yes it searches for 388, but, before Adlibtracker 2 starts, with those 2 sound cards in the PC (CT1350 and CT2230) it can not find OPL3, so it asks me to manually tell it where OPL3 is. I do that by typing -> adtrack2 /a388

By doing that both chips make music. I will try the PnP tool.