The Future of DTP, Illustration and general GFX/Audio using RISC OS desktop computers (Is 64-bit!)
Stuart Swales (8827) 1357 posts |
Might help to add the memory clobber to the example in the ‘Determining Registers’ section for the unwary! |
Simon Willcocks (1499) 519 posts |
I’ll get back to it when I work out where my task switches are buggy. Probably. |
David Pilling (8394) 96 posts |
For completeness and not because it helps at all, this appeared yesterday on YouTube: #RISC_OS on #64bit #RISCV ? Is that even possible? Raspberry Pi Zero alternative! Wi-Fi Sheep Tech Channel |
Sveinung Wittington Tengelsen (9758) 237 posts |
Hm. Does https://github.com/Simon-Willcocks/RISC-OS-Kernel-in-C mean that this thread (which Mr. Willcocks has contributed to) is part of the solution? Call me slow, but since the last contribution was half a year ago, may this project has continued quietly under the auspices of the creators of this forum? Inquiring minds would very much like to know. Could it be? Is it? |
Charles Ferguson (8243) 427 posts |
I’m not going to pitch in on the many other things that have been said on this thread because, quite honestly, some of this stuff is quite batty – the idea that either RISC OS is going to be world beating, or that the current maintainers of the OS have sufficient vision to take it anywhere is clearly insane. However, on this I can at least offer a comment and opinion. That repository has been around a little while. The author’s questions on basic RISC OS things haven’t filled me with huge confidence, but then they’re actually doing something towards the future, which I can applaud and wish them all the best. Certainly it’s one of the more effective developments in the RISC OS world in recent times. On the other hand, my own version of RISC OS – RISC OS Pyromaniac – runs on macOS, Windows and Linux, both locally and in the cloud, works on 64bit systems. The whole point of RISC OS Pyromaniac is to make it possible to test, debug, experiment and try things out that would be tedious and difficult with regular RISC OS. Experiments with handling the underlying system to AArch64 and x64 have been interesting. As suggested by the commenters on this thread, the number of people actually willing to try things out and do development work on RISC OS is so low that expecting anything special from the community or from ROOL is unreasonable and unfounded in reality. |
Simon Willcocks (1499) 519 posts |
Sveinung, the last commits were actually last month, in the Multithreading branch, and it’s a one-person part time hobby project, probably not a serious contender. Charles, the questions about RISC OS are because I’ve never needed to do any low level coding on it and I’m trying to get my head around things. At the moment, the problems are due to SWIs insisting on enabling interrupts while in SVC mode which doesn’t play well with my threading model. At least qemu and physical hardware are behaving the same way, which is nice, but crashing, which is not. (I got fonts working, by the way. I was misled by the documentation for the fast horizontal line draw routine referring only to the screen, but it’s used on sprites generally.) |
David Thomas (43) 72 posts |
Two Risc PCs running side by side: 64-bit RISC OS. |
Clive Semmens (2335) 3276 posts |
Here you go: http://clive.semmens.org.uk/Photos/Assortment/RiscPCs.html !! |
Sveinung Wittington Tengelsen (9758) 237 posts |
Mr. Willcocks, thought I smelled a cigar. Maybe with a little help from your friends? Only it should probably be on a slightly organized and goal oriented footing to become ArmV9-compatible. Great job so far from what I can tell without having used it. Saving the GUI/WIMP paradigm is paramount, what’s beneath will necessarily be “alien” to original RISC OS because of the radical ArmV9 architecture changes. So, maybe cigar. :) |