Extending the Pi partition size?
Pages: 1 2
Jon Abbott (1421) 2601 posts |
Is it possible to extend the Pi RISC OS partition size from 2GB? Now that I’m testing CD based games on the Pi, I need that wasted xGB that’s on all my SD cards for CD images. If extending isn’t possible, can an additional partition be created? If neither of these are possible, how does one format a blank SD and add in the DOS partition for the bootstrap? |
||||||||||||
Steve Fryatt (216) 2046 posts |
SystemDisc is usually the way to do this. You can do it manually, but it’s a lot less straightforward. |
||||||||||||
Rick Murray (539) 13406 posts |
You can? I did a bit of investigation into the disc formats, both FileCore and FAT… …and in the end just went and purchased SystemDisc which made the whole process a doddle. |
||||||||||||
Jeffrey Lee (213) 6046 posts |
No. I think Ben Avison was looking into it at one point, but the structure of the FileCore format makes incredibly difficult.
Not one that would be recognised by RISC OS, no.
SystemDisc is the “official” way of doing it. I think you can set up dual-partition cards manually, but if you want the Loader DOSDisc hack, then SystemDisc is the only tool capable of setting that up properly. |
||||||||||||
Steve Fryatt (216) 2046 posts |
You can do it manually, but it’s a lot less straightforward. “…but it’s a lot less straightforward.” :-) |
||||||||||||
Clive Semmens (2335) 3130 posts |
My fingers are too fat to handle the individual bits, sadly. |
||||||||||||
Rick Murray (539) 13406 posts |
The thing that stopped me is that this wasn’t some module where a bit wrong would risk stiffing the machine until the reset button gets a prod… I felt that if I did not completely understand both filesystems right down to bit level, then attempting to build up a disc would be foolhardy. I could study the coffee and learn, certainly, but that takes a lot of time and effort and potential experimentation. Far quicker and simpler just to buy a tool already designed for exactly this task. |
||||||||||||
Rick Murray (539) 13406 posts |
We have all heard of Mondegreens, probably experienced a few ourselves. What we ought to have in this modern age is a website that lists the weird things “people write” thanks to their technology choosing a word that wasn’t quite the word they wanted. FWIW, I could study the code and learn… though to be honest I like the coffee interpretation better. ;-) |
||||||||||||
Steve Pampling (1551) 7932 posts |
And there was me thinking you were contemplating a Java implementation :) |
||||||||||||
Jon Abbott (1421) 2601 posts |
I’ve creating a FAT32 partition with the spare space and used Fat32FS to mount it via fat32fs:mount -fp3 :16 – free and painless. |
||||||||||||
Colin Ferris (399) 1748 posts |
Could something like a FAT32 partition on a RiscPc RO3.7/RO5.xx be created to get Long file names? Is it possible to boot from a DOS formatted HardDisk connected to a RiscPC? |
||||||||||||
Jon Abbott (1421) 2601 posts |
My softload boot sequence uses two drives, a legacy boot drive which softloads the OS I’m testing and also loads a version of FileCore, FileSwitch and ADFS that support long filenames for older OS’s. The second drive contains the !Boot etc and is a new format drive. End result being that RO3.7 thru RO6/5.x all support long filenames. |
||||||||||||
Colin Ferris (399) 1748 posts |
What a mini boot – loading the basic filecore etc from the floppy? |
||||||||||||
Jon Abbott (1421) 2601 posts |
Booting from floppy might work, if a little slow as it has to RMReInit the filesystem several times to load all three Modules. Everything you need is in the ZIP attached to the linked post, just put the !Boot on a floppy and update as per the README. If you want to softload an OS, you’d need to modify it to perform the softload after its loaded the updated filesystem Modules. |
||||||||||||
Colin Ferris (399) 1748 posts |
I StubG the ‘Softload’ prog (does anyone want to try it) – doesn’t need the ‘CLib’ module – now using with VRPC-DL. My RPC is RO4.02 – but other old machines are RO3.71. |
||||||||||||
Steve Pampling (1551) 7932 posts |
I don’t suppose there is any chance people will drop StubsG in the obscure bin that it really ought to have been in from day 1 of the “yah, boo, sucks”1 era. 1 Every time I think of the ‘risc os wars’ it’s what comes to mind. |
||||||||||||
Rick Murray (539) 13406 posts |
Uh-hu… Right. http://gerph.org/riscos/ramble/clibrary.html Okay, I get it. It is reprehensible that a softload CLib does not refuse to die, happily allowing another softload CLib to replace it (and thus causing a core meltdown for boring technical reasons that roughly amount to doing a skydive and suddenly finding your parachute has been replaced by a piano). But… StubsG is only a tiny bit larger than Stubs. So what problem is it actually trying to solve? Reading through the documentation for StubsG from ROLtd’s site, it looks as if it is a “generic” munge to get a C program working regardless of what incarnation of CLib was installed. But – wait. WHAT? Two observations. The painfully obvious one is that newer versions of CLib will provide newer functionality and maybe also bug fixes. To encourage the idea of not updating because it might be difficult… that’s surely the height of stupid. And, of course, I have been using RISC OS since RISC OS 2. Back then, almost everything came with a better CLib than the one in ROM. Ditto RISC OS 3.10. It isn’t as if I’ve never updated CLib in my life. Actually, I think I have something like 6-8 versions of it on my old A5000’s harddisc. When a new version turned up in something (often on a cover disc), I would rename the current with a version number suffix, copy out the new one as CLib, then reboot to get it going. This was just something I did from time to time. I’m not such a big snowflake I said “can’t, won’t”. And, of course, even with StubsG one can still bring down the machine by softloading over a softloaded CLib, so I wonder what the point was? |
||||||||||||
Steve Pampling (1551) 7932 posts |
RO wars remember? Feel free to find a nice solid, but wipe clean, wall to bang your head on. |
||||||||||||
Rick Murray (539) 13406 posts |
Actually, I kind of missed most of that. I dropped out of the scene early in 2002, and didn’t come back until… what was it… 2009 or somesuch. Ah, I can look at my blog. Duh. August 2009. So I missed the fun stuff. Got it second and third hand, and even from that I often thought “what the f….udge caramel sundae”.
No, I reserve that bloodied dented wall for the fact that even now, over a decade later, and the war utterly decided, this crap is still a really big deal in some people’s lives. I kind of want to say “’nam is over” but the exact level of scathing might be missed, given it is referencing a left-pondian thing. But hey, I’m now enlightened as to what StubsG actually is. I wish I stayed in bed. Time to go make (yet) another dent in the wall… right next to the gaping hole marked “hit here for Great Repeal Bill”. |
||||||||||||
Steffen Huber (91) 1945 posts |
The point of StubsG was to leave the decision of using which version of SharedCLib with the user. There was broken software that was broken by the Clib update, integrating the new CLib into !Boot was not really that simple (after all, many users fiddled with !Boot, mostly those who didn’t really understand it), and there was that Select version that refused to get its CLib replaced. For a software developer like me, StubsG was a great thing. I didn’t need to distribute the updated SharedCLib to users (would have ruined the single floppy idea, and the licence was unclear for people who didn’t own Castle C/C++), I didn’t need to put complicated instructions into the manual on how to update the SharedCLib, and the software ran just fine on anything from RISC OS 3.10 up to RISC OS 5.xx out of the box. |
||||||||||||
Rick Murray (539) 13406 posts |
PS – when I got my Internet in 2009, it was one megabit down and quarter up. This was bumped up to 2 megabit (more or less) with quarter megabit down, which was about the limit of what’s physically possible with somewhere between 4.5km and 5km of line, some overground, some under, and a lot of extra patches and joins in recent years (most of them last year – fallen tree and backhoe’d road). Now?
I’m currently pulling 3.2 megabit. It varies between 3.7 and 3.2 (so far) depending on the weather. It’s chucking it and blowing a gale, so hardly optimal. |
||||||||||||
Patrick M (2888) 115 posts |
Is there a webpage (or other thread on this forum or something) about this war, the details, why it happened, etc? I vaguely know about there being two versions of RISC OS (5 vs 6) but I don’t know much else. |
||||||||||||
Rick Murray (539) 13406 posts |
SIX is really 4.something, but SIX is better than 5, right? Yeah. About that. It’s probably best not to ask. Go look at the ROLtd’s history on each version (of RISC OS), and notice the obvious absence, like five is a cursed number (cough, actually it’s four, cough). |
||||||||||||
Rick Murray (539) 13406 posts |
https://www.theregister.co.uk/2004/06/17/riscos_dispute/ (it made it into The Register!) http://www.drobe.co.uk/article.php?id=2373 Etc. |
||||||||||||
Chris Mahoney (1684) 2100 posts |
This thread may amuse/shock/fascinate (there are some photos of the whole setup on page 5). |
Pages: 1 2