beebasm putbasic problem

[sydney]

Every time I try to use PUTBASIC, even with a very simple source and BASIC file, I get the following error:

Code: Select all

Error: test.bas: Unspecified file error.

My source file (test.asm) is:

Code: Select all

PUTBASIC "test.bas", "test"

My BASIC file (test.bas) is:

Code: Select all

10PRINT"SYD CANT CODE"
GOTO10

and my command line is:

Code: Select all

beebasm -i test.asm -do test.ssd

I'm running linux mint 18.1 and I'm pretty sure I'm running an up to date version of beebasm, it reports beebasm 1.08.
I can INCBIN and INCLUDE files with no problems. What am I doing wrong?

[ctr]

You can't omit the line numbers in your basic code. This is a nuisance.

[sydney]

So so simple!

I'm sure I'd read line numbers were optional. Thanks for the quick answer.

[ctr]

I'm sure I'd read line numbers were optional.

That rang a bell so I had another look. Github has a bunch of proposed updates including automatic line numbering. Stardot has a thread about it.

[tricky]

The one that catches me out, is no abbreviations.
P. bad
PRINT good

[Rich Talbot-Watkins]

jsbeeb's BASIC tokeniser actually uses the BASIC 4 ROM code to tokenise lines! As in: it creates a fake Beeb, puts the ASCII string into the right place in memory, sets up the necessary zero page, and then emulates the ROM call until it exits.

https://github.com/mattgodbolt/jsbeeb/b ... okenise.js

That kind of approach would of course work perfectly for BeebAsm, were it to have a full 6502 emulation engine inside (something I had considered at some point, in order to get decent subroutine timing information).

[SteveF]

Well it's not the swanky approach suggested by Rich, but here's a quick-ish hack to add support for BASIC abbreviations. I *think* this is equivalent to the proposed-updates branch in Tom's repo plus this change, but frankly I'm not entirely sure...

https://github.com/ZornsLemma/beebasm/t ... eviations2

Fair warning: I have only given this the most cursory testing. Let me know how you get on if you try it...

[simonm]

Im reminded that it would be great to get a BeebAsm repo going in the stardot team; it's such a useful tool and there's a number of useful patches around now.

[kieranhj]

Im reminded that it would be great to get a BeebAsm repo going in the stardot team; it's such a useful tool and there's a number of useful patches around now.

Indeed. And we never got to the bottom of the Teletextr bug when I was getting different behaviour at runtime whilst using a locally built binary from one of the branches.

[SteveF]

Im reminded that it would be great to get a BeebAsm repo going in the stardot team; it's such a useful tool and there's a number of useful patches around now.

Indeed. And we never got to the bottom of the Teletextr bug when I was getting different behaviour at runtime whilst using a locally built binary from one of the branches.

My best guess here would be that you were running into one of the bugs documented in the RetroSoftware thread here: http://www.retrosoftware.co.uk/forum/vi ... 57ff9c9a9a The very last post in that series mentions a bug I introduced while fixing an earlier bug which caused assembly to succeed but generate invalid output, rather than fail with a mysterious/incorrect error.

But it's just a guess, of course. As far as I know the latest proposed-updates version has no bugs in this area; if you do manage to reproduce this behaviour with the latest build let me know and I'll take a look...

Cheers.

Steve