tjb@tadtec.uucp (Tim Bissell) (09/18/90)
I am not a great assembler user (yet -- but I hope to be when I get some time)
but I thought I would add these comments:
It would be nice if the Assembler could assemble the output of the
disassembler. This would involve a line format like this
<#optional_address>:<#optional_data> opcode operands <;optional comment>
e.g.
#01234:#4 sethex ; foo
The addresses and data could be ignored (is there any situation where
they would be useful?)
Obviously lines which could not be generated by the disassembler (e.g.
with labels defined on them) need not follow this format.
While I'm using up net bandwidth, I'll volunteer to modify sass 2.31
(or the latest version I can get) to conform to the assembler syntax
that is agreed upon. This is for people without access to (or disk
space to spare for) Perl.
And finally I would like to thank Andreas Gustafsson for a fine piece
of software in the chip-8 interpreter. I would love to see it posted
to the net - *very* few people in the UK have FTP access.
Would it not be a good idea to:
(1) augment chip-8 to better use the features of the HP48 (e.g. 131x64
dot screen etc.) we could call it chip-48 and retain chip-8 for
using the vast library of software lying around in the ROMs of video
games!
(2) standardise an "assembler" syntax for the language. I'm hacking
out a HP-48 hosted disassembler for chip-8 strings and will post
it when I have time to finish it (if no-one beats me to it)
Thanks once again to all you 48 hackers (and the HP crew) who have
shared your knowledge and software with us!
Tim
--
Tim Bissell Tadpole Technology | (tjb@tadtec.UUCP || ...!mcsun!ukc!tadtec!tjb)
Cambridge Science Park | "Is life worth living? DoD#174
Cambridge ENGLAND | That depends on the liver" CBR 600bson@rice-chex.ai.mit.edu (Jan Brittenson) (09/18/90)
In article <TJB.90Sep17230402@tadtec.uucp> tjb@tadtec.uucp (Tim Bissell) writes: >It would be nice if the Assembler could assemble the output of the >disassembler. This would involve a line format like this > ><#optional_address>:<#optional_data> opcode operands <;optional comment> > >e.g. > >#01234:#4 sethex ; foo > >The addresses and data could be ignored (is there any situation where >they would be useful?) If you've done enough work disassembling to have replaced all addresses with symbols, there is little need to keep the address and opcode info at all if you're going to modify the code and feed it back into an assembler. Also, note the following. 1. I bet *all* PD ML software (like CHIP-48) is going to be distributed with source code freely available. No need to disassemble stuff like that. 2. Since there is no realistic escape from the internal ROMs in the HP-48, understanding them is a key issue to ML programming. The better the understanding, the happier a sizable segment of HP's HP-48 customer base will be. 3. HP and third-party vendor software is a reasonable target for disassembly. It is only to understand the code, not to have people improve and redistribute it, since that would, well, lead straight to the slammer. It is so as to be able to interface your software with what exists, what you have (supposedly) paid for, and like to know how to take full advantage of. > Obviously lines which could not be generated by the disassembler (e.g. > with labels defined on them) need not follow this format. One of the biggest advantages of a multi-pass disassembler is the ability to generate symbols. > While I'm using up net bandwidth, I'll volunteer to modify sass 2.31 > or the latest version I can get) to conform to the assembler syntax > that is agreed upon. This is for people without access to (or disk > space to spare for) Perl. This sounds excellent to me. (Although I do believe you mean ASAP - but please adjust them both!) I think it's very important to adhere to a single syntax, and to do so at an early stage. I remember much to well what it was like with the HP-41... Especially when there exist such excellent means of communication as USENET and Internet. > Would it not be a good idea to: > augment chip-8 to better use the features of the HP48 (e.g. 131x64 > dot screen etc.) we could call it chip-48 and retain chip-8 for > using the vast library of software lying around in the ROMs of video > games! If anyone has old games lying around, how about posting some ROM dumps - surely the manufacturer must since long consider these games commercially obsolete?
alonzo@microsoft.UUCP (Alonzo GARIEPY) (09/19/90)
tjb@tadtec.uucp (Tim Bissell) writes: > It would be nice if the Assembler could assemble the output of the > disassembler. This would involve a line format like this > <#optional_address>:<#optional_data> opcode operands <;optional comment> > e.g. > #01234:#4 sethex ; foo I concur. This is the "standard" listing format.