| |
Count Zero
Registered: Jan 2003
Posts: 1922 |
Release id #212202 : 21 Second Backup V4.1
<Post edited by Count Zero on 23/12-2021 23:53>
Wont move all the crappy comment - just this and force further talk to here:
Quote:
I call BS on the screen loading time!
But thank you for confirming the SCREENSHOT I posted was REAL! aha!
So please let the real one stay
But maybe it just FEELS fast.
Anyhow, here's the ACTUAL load sequence in REAL time... with the REAL screens!
https://zerofusion.com/21_second_backup_cable/21_second_backup_..
It's the FASTEST copier ever made, full stop. NO other copier can copy a full disk in under 21 seconds.
Dual drive is even faster (no disk swaps)
NOTHING can touch this, not by a mile!
Loading speed TO the program != speed the program achieves on copying. ANYONE here just expects a friggin 20 blocks tool load from a D64.
Why do you mix these things so hard around?
And and -- screenshots - READ on the forums and LEARN about reasons - dont fucking just argue around (with me or others). |
|
... 10 posts hidden. Click here to view all posts....
|
| |
chatGPZ
Registered: Dec 2001
Posts: 11352 |
Quote:paperback writer works just fine in 2.4 SPS
The images you linked are for C128 and use double sided g64. 2.4 can not even attach them. Are you sure you know what you are talking about?
Quote:as Vice Team removed the extra G64 handling sometime after v3.2
You obviously don't want to help, but keep repeating your nonsense instead. Please go back to where you came from, you are wasting my time.
Quote:This should be handled in the Vice Support Forums!
Is that the same place where your buddies reported those non existing bug reports? Perhaps as project lead, i should know about those forums too? *shrug*
Quote:It doesn't work in newer Vice versions due to a 1541 VIA timer issue that crept in at some point.
Sure it's a VIA issue? And more importantly: any idea how to test for it?
Quote:PBW doesn't work in newer versions due to, ironically, better emulation in Vice 1541.
That phenomena is something ppl like mr.minihacker really need to understand, indeed.
Quote:In newer versions changing the density setting seems to adjust how bits clock into the shift register and they come in wrong if it doesn't match what the track is defined as, like the real thing does.
I'd have to check the code again to make sure - but i dont think anyone touched that code since the SPS patches. Actually a lot of the "working" g64 support relies on the bits being "valid" regardless of the speedzone, changing this would break so much more :)
So this boils down to: a test program is needed, as so often :) I wouldnt even be surprised if the problem is something completely different, not even drive related. Wouldnt be the first time that happens. |
| |
chatGPZ
Registered: Dec 2001
Posts: 11352 |
OK, did a quick code check... As expected a) any speedzone info from the g64 is ignored (see https://sourceforge.net/p/vice-emu/code/HEAD/tree/trunk/vice/sr.. ). Other than that, the speedzone set in the drives VIA is only used to compute the size of a bitcell (search for "speed_zone" in https://sourceforge.net/p/vice-emu/code/HEAD/tree/trunk/vice/sr.. ) - So very unlikely this is what changed that broke PBW :) Perhaps try setting both "wobble" settings to 0 (and if that does the trick, a bug report wouldnt hurt =P)
As for 21sec backup, in the initial loader, the c64 code doesnt even receive garbage - it gets all $FF. That smells like an easy fix... if we had this code standalone in a test program :) A bug report wouldnt hurt regarding this one either (on the bug tracker please, not this mystical place that apparently some people use) |
| |
LordCrass
Registered: Apr 2023
Posts: 8 |
I could've sworn I'd seen data read incorrectly when the density bits in $1c00 were wrong for what the track was set to, but I just tried it in Vice 3.7 and it doesn't seem to matter what those bits are, GCR still comes in correctly. huh.
As for the VIA timers, no I'm not sure that's what it is, as I just tried Vice 3.7 with Galaxian which has a routine that uses these, and it still works. The issue in 21sec is a decrypt loop that uses the timers and some unintended opcodes and at the end the timer values are correct in Vice 3.1, but wrong in 3.7, result is garbage decrypt and crash.
If I get time, I'll pull the routine out and make a test from it and submit a bug report. |
| |
chatGPZ
Registered: Dec 2001
Posts: 11352 |
Quote:If I get time, I'll pull the routine out and make a test from it and submit a bug report.
That would be really nice! There are really too many possible places where this could go wrong, debugging it without a minimal testcase is a bit too annoying :) |
| |
chatGPZ
Registered: Dec 2001
Posts: 11352 |
Just for the records: what broke 21 sec backup has in fact _nothing_ to do with G64 support, or even drive emulation - i added the dummy accesses to the "not cycle exact" 6502 cpu core a while ago, and that unfortunately resulted in some dummy accesses being in the wrong order. Which apparently confuses this decryption code (perhaps other things). The commit is r37549, some time after 3.4 was released. Tricky to fix unfortunately.
Having a test program - or knowing which opcode exactly is the problem - would be still nice, of course :)
https://sourceforge.net/p/vice-emu/bugs/1859/ |
| |
master_hacker
Registered: Feb 2009
Posts: 9 |
Groepaz, it'd be nice if it could be fixed up enough so that the Paperback Writer, Paperback Filer, and Paperback Planner would also work.
There are quite a number of people who would very much like to see this.
I know this all works in Vice 2.4 SPS, 3.0, 3.1, and 3.2
but recent version capability would be nice.
Much appreciated anything that can be done for Paperback titles, and many thanks in advance! |
| |
chatGPZ
Registered: Dec 2001
Posts: 11352 |
Step 1 to make it happen: make a bug report with testcase
PS: r43647 should fix the 21 sec backup |
| |
LordCrass
Registered: Apr 2023
Posts: 8 |
Indeed, it does. |
| Previous - 1 | 2 - Next |
