question about role of modules in riscos

TVTropes said:

Possibly the best-known book by Neal Stephenson, Snow Crash is the tale of a sword-slinging hacker who teams up with an extreme skateboarder in a Post Cyber Punkdisincorporated USA to fight “Snow Crash” – a computer virus for the brain. Oh, and there’s a badass biker with glass knives and a nuclear bomb strapped to his motorbike, too.

And somebody delivers pizza.

Ho hum.

Experience? Paprika, Masamune Shirow, Hackers (film), and anything ever written by William Gibson. ;-)

aaaaa

aaaaa

In any case, Jon makes massive contributions to RISC OS – I’d be surprised if he has a lot of time for anything more!

I wouldn’t go that far, my contribution to RISCOS itself has been minimal to date. Besides the usual bug reporting and suggestions around improving/fixing certain areas I’d say I’ve steered clear of actual RISCOS contribution.

I was really talking about your contribution to RISC OS in the broader sense – not necessarily to the code at ROOL. You’ve certainly uncovered a few bugs, missing features, etc., and helped test the new GraphicsV API. And the JASPP project is hopefully going to bear fruit for the OS as you’ve mentioned recently on the Hypervisor thread. Anyway, just wanted to say that your efforts are much appreciated. Very sorry to hear that you’re unwell. I really hope you get some good news about that soon.

aaaaa

At first the dude deliver pizza for the mafia,

Okay, the dude, not the dudette. Still, I appear to have got it reasonably correct for having never heard of the book. It looks like it would make a pretty good movie¹, even if it would seem a bit “been here, done that”.

¹ So long as they don’t mess it up like they did with Johnny Mnemonic.

aaaaa

aaaaa

It’s not infinite and you’re not going to achieve much on an hospital bed =)

That’s not the impression I got from some nurse romance books I flicked through in a doctor’s waiting room (which must surely have been some sort of meta joke?).

I’m saggitarius and rooster with astrology, i don’t have a specially quiet temperament and have hard time to stay inactive,

I’m Saggitarius and… something. I don’t remember. I had a chart once that said I was a lively and outgoing person.

Ha ha.

Anybody that knows me will know I’m the sort of introvert that borders on being just plain antisocial. It also amuses me when people describe me as “shy”. Not quite. Shy people want to be a part of things but don’t quite know how. Strongly introvert people do not want to be a part of things and are generally happier if they are just left the hell alone (see what I mean? ;-) ).

So, yeah, don’t put too much faith in people that make prognostications based upon what a bunch of stars are doing. Somebody much more famous than me said There is no fate but what we make for ourselves.

aaaaa

aaaaa

http://www.4corn.co.uk/articles/acornc5/
Specifically the C manual

Or is that not old enough?

Not old enough I’m afraid, need to back to 87/88 rev. 1 of C for APCS-A / M

I’m going to need a copy of the very first C manual, as APCS-A / M were deprecated when RISCOS 2 came out.

The Google-fu is not strong in this one…

Certainly isn’t. Everything I’ve managed to piece together from the web, debugging CLib itself and watching what few game I have that use it are doing, I’ve documented here

That covers the register differences in APCS, the CLib initiation SWI, and I’ve managed to figure the two chunk sizes. I’ve yet to confirm the functions match later CLib’s and have no information around any functional differences in the CLib functions themselves.

If people could keep an eye out for an original CLib rev.1 manual, I’d really appreciate a copy so I can document it properly.

Not old enough I’m afraid, need to back to 87/88 rev. 1 of C for APCS-A / M

Hmmm, I have found references to the old original Norcroft compiler in the old Beeb related forums. Really speaking it’s the rollover from Beeb to Arc era you’re looking at isn’t it?

You could ask Codemist – being academics they probably keep all manner of things in dusty cupboards. :)

Back on the original subject of the thread This manages to reproduce information from the ROL branch with the nice 26 bit module header that won’t work on RO5.x

Understandable after a fashion since the modules link here is a dead end.
Even the PDF’s in DDE24 document the 26 bit version header.

Would a slightly updated version of the text from the DDE chapter 14 be in order here, or has someone done a newer version already that hasn’t made it to the web documentation?

Understandable after a fashion since the modules link here is a dead end.
Even the PDF’s in DDE24 document the 26 bit version header.

When I get a chance I’ll add the documentation to the Wiki, I’m surprised it’s not there already though. I guess not many people are writing Modules these days.

“Not old enough I’m afraid, need to back to 87/88 rev. 1 of C for APCS-A / M”

There is a small amount about aps-a/m on page 475/476 of the acorn c3 (1989) manual here.

http://www.4corn.co.uk/articles/acornc3/

I guess not many people are writing Modules these days.

I suppose anyone trying in the BASIC assembler and following the documentation would fall at the hurdle of not including the header entries all the way through to 0×30 or possibly end up including 26 bit initialisation and finalisation code as per current documentation.

If they have the DDE and follow the examples in there they would be OK, but that presupposes the proto-programmer investing in the DDE just for a bit of hacking around.

I began this: https://www.riscosopen.org/wiki/documentation/show/File%20formats:%20Relocatable%20Module

Must get around to finishing it, but I wanted to take apart the OS to verify that what the documentation says is indeed what happens. Of specific note is the extended service call method.

TBH – I’m not sure a programmer “hacking around” with writing modules is a great idea. A certain poltergeist might tell me that is elitist, but given that modules run alongside the kernel with the same abilities, we can’t rely upon the OS to catch faults. The usual fault behaviour is a freeze. I don’t want to dissuade people from writing modules, but the RMA isn’t a place to go timidly and without a road map. And maybe guns too.

Must get around to finishing it, but I wanted to take apart the OS to verify that what the documentation says is indeed what happens.

I’m thinking the numbering being in decimal and no hex reference needs changing.
sort of like the table here

Referencing 26 modules as a legacy situation would be good and pointing out that 26/32 neutral is possible with proper coding. Simplest coding would be 32 bit only.
Headers are simple – always do the full set and indicate 32 bit or not.

TBH – I’m not sure a programmer “hacking around” with writing modules is a great idea. A certain poltergeist might tell me that is elitist, but given that modules run alongside the kernel with the same abilities, we can’t rely upon the OS to catch faults. The usual fault behaviour is a freeze. I don’t want to dissuade people from writing modules, but the RMA isn’t a place to go timidly and without a road map.

He’d be right, it’s elitist.
When you can make a copy of the install, including ROM image on a second card and clone across when you screw up, exactly why should a proto-programmer care whether they stiff the machine?
Ever thought how many times the names we recognise as key developers trashed their machines back then? and how long each recovery took?
We don’t have the latter problem.

I wonder how many times Jon has stiffed his test machine while working on ADFFS, or how colourful the language was…

Now when it comes to putting the “finished” item out there then it does need care, but initial work? Really?

The point isn’t that they might stiff the machine – God knows I have done it enough times for a variety of dumb reasons.
The point is that it isn’t a place that is good to go if you are not backed up with good documentation, indeed this is part of the problem – the best documentation is for a 26 bit system so will run into problems with modern RISC OS. Problems that if code is copied from current examples (except DDE), things may work or fail in unexpected ways which could make debugging difficult and it can be a real headache trying to work out why something has failed when everything grinds to a halt…

You don’t have to be a beginner to get things to fail in unexpected ways. The first thing I did on my pi was solder a reset button to it. The reset button on my iyonix is a small hole in the front of a shiny black box. The box is in the dark under my table. I had to put masking tape around the hole so I could see where to poke the matchstick I keep by my side.

Stiffing a machine is a fact of life when programming modules – how else would you know when they are working :-)

http://psychic.vibe.free.fr/pf/index.php/welcome/article/en/riscos%20modules%20and%20memory%20model_0/tech/osdev

http://psychic.vibe.free.fr/pf/index.php/welcome/article/en/riscos%20image%20filling%20modules_0/tech/osdev

It’s just a start it might contain errors and not finished =)

It’s just a start it might contain errors

You have references to 26 bit style module headers. The correct 32 bit header style is what you used in the filer module work and is below. Note the extra header entries, paticularly the entry for the flags word.

Header
+0 Offset to Start code
+4 Offset to Initialisation code
+8 Offset to Finalisation code
+12 Offset to Service Call handler
+16 Offset to Title string
+20 Offset to Help string
+24 Offset to Help and Command Keyword table
+28 Offset to SWI chunk base number
+32 Offset to SWI handler code
+36 Offset to SWI decoding table
+40 Offset to SWI decoding code
+44 Offset to Messages file name
+48 Offset to Module flags

Stiffing a machine is a fact of life when programming modules – how else would you know when they are working :-)

And we’re the ones who know (usually!) what we’re doing. My remark was not meant to be elitist, it was meant to be a suggestion that writing modules is not a thing to take lightly. Especially when there are documents that no longer (correctly) apply.

As for your Pi’s button…

Isn’t this normal? ;-)

The 32bit module header from ROOL’s PRM, File formats: Relocatable Module