TSS/8, OS/8, BASIC8 Discord group?

[Rick Adams]

After doing quite a bit with TSS/8, I've begun playing around with OS/8, and find the BASIC8 programs I developed previously with TSS/8 have some problems with this newer OS.

Is there a nice, friendly Discord for this subject where I might find some answers?

For one thing, I find that RANDOMIZE does not seem to actually randomize: RND(0) always returns the same sequence of random numbers.

[Vincent Slyngstad]

I can see RANDOMIZE not working correctly if there's no clock or similar variable with which to seed the generator.

I think you might see more action over at

https://forum.vcfed.org/index.php?forums/dec/

though I can't say if anyone will be able to answer off the top of their heads.

Vince

[Steve Tockey]

I've seen it elsewhere, but also in one or more versions of PDP-8 BASIC that the RANDOMIZE function executes as a NOP (no operation). The instruction is only there to be recognized and thus provide compatibility (i.e., so you can run the same source code) but it doesn't actually do anything. I did look into the OS/8 Language Reference Manual and it claims to actually do randomization.

-- steve

[Vincent Slyngstad]

OK, I took a few minutes to look at the manual, and then at the implementation.

The manual hints that the "seed" planted by RANDOMIZE is in fact, the count of loops during TTY I/O waiting.

So, if you just RANDOMIZE and ask for a random number, you'll always get the same sequence.  However, if you interact with the terminal in some way, particularly for input, RANDOMIZE will set up a new sequence.  A look at the implementation in BRTS.PA bears this out.

Hope that helps!

Vince

[Rick Adams]

I’ve tried RANDOMIZE, INPUT A$ and then generating RND(0), and I’ve tried RND(-1) and RND(N) where N is any number of positive integers, and it’s the same sequence of random numbers each time.

It’s curious that the earlier and more primitive BASIC8 with TSS/8 gets this right with no problem. It also has an ON GOTO and OS/8 doesn’t, even though OS/8’s BASIC obviously has a boatload of more advanced features that TSS/8 lacks.

It’d be nice to be able to access OS/8 via telnet like you can with TSS/8, but placing a ttix command in the boot script at various places doesn’t work.

Trying to port a random maze BASIC program from TSS/8 to OS/8 is giving me a FM ERROR… memory full? TSS/8 can do a 8x8 maze; even a 5x5 maze gives me this error with OS/8, which is puzzling. Is there a list of error codes for which I can look up FM ERROR?

[Vincent Slyngstad]

Sorry.

TTY I/O waiting affects RANDOMIZE, which then optionally affects RND.  So, you need to do the I/O before the RANDOMIZE or it has no effect.

In general you need some kind of sort-of-random counter thing for RANDOMIZE to work.  On TSS, there is a clock, so that's easy. On OS/8, they used the spin-waits for TTY input and output, since there is no clock.

Vince

[Rick Adams]

Okay, I got that to work, thank you. Of course, what if I don’t NEED to input anything? ^_^;

Still trying to figure out my FM ERROR.

[Vincent Slyngstad]

Over here: http://www.bitsavers.org/pdf/dec/pdp8/os8/OS8_Handbook_Apr1974.pdf

Chapter 6 is all about OS/8 BASIC.  On page 6-117 it mentions that "FM" is "Attempt to fix negative number", meaning that an attempt was made to convert a negative floating point value into an index or subscript.

Hope that helps!

[timr...@gmail.com]

Where did you find this BASIC random maze program?  I remember having one in BASIC, but probably later than my TSS/8 days, but I don't remember for sure.  It could decent sized mazes too.  Much bigger than 8x8.  If we can see the code we might see what the issue is.

[Vincent Slyngstad]

Is it this one?

https://svn.so-much-stuff.com/svn/trunk/pdp8/src/decus/basic8-55/amazin.ba

https://svn.so-much-stuff.com/svn/trunk/pdp8/src/decus/basic8-55/101_BASIC_Computer_Games_Mar75.pdf (page 16)

https://svn.so-much-stuff.com/svn/trunk/pdp8/src/decus/basic8-55

https://so-much-stuff.com/pdp8/software/decus.php#decus-basic8-1 (scroll down to basic8-55)

[Rick Adams]

No, it's one I wrote myself.  Please see:

https://github.com/yggdrasilradio/b8pp/blob/master/maze.txt

https://github.com/yggdrasilradio/b8pp/blob/master/maze.bas

Here's what I have, altered to be more compatible with OS/8 BASIC:

--
You received this message because you are subscribed to a topic in the Google Groups "PiDP-8" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/pidp-8/D-HPJ1Uk3kA/unsubscribe.
To unsubscribe from this group and all its topics, send an email to pidp-8+un...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/pidp-8/93e35c82-89ce-468b-8540-e5a83085e763n%40googlegroups.com.

[Rick Adams]

Nevermind, I seem to have fixed it!

On 3/12/2024 7:44 PM, Vincent Slyngstad wrote:

--

[Rick Adams]

It’s a maze program I wrote myself. I managed to figure out what I was doing wrong, now it works fine, and I’m able to generate pretty much as large a maze as I want. Thanks for everyone’s help!

Still not sure what FM ERROR was, it was complaining about a RETURN statement of all things, but my error was unwisely reusing a variable name.

[Vincent Slyngstad]

You are welcome.  I got your program working here, too.  I had to go get the PiDP version of OS/8 BASIC -- the archived V3D systems seem to share a mangled version.  The copy I got of your program from github also needed to be updated. At this point I'm reading it in from PTR:.

Not sure about that FM error either.

At some point in the future I want to put together a working RK05 image which can run all the DECUS games.