[Featured][Dev] PindurTI (Best Thread 2005)

Jim e · Post by **Jim e** » Wed 23 Aug, 2006 9:07 am

oh i remember some thing. Doesn't pti.conf have to be in the working directoy?

CoBB · Post by **CoBB** » Wed 23 Aug, 2006 9:16 am

Exactly. In other words, it needs to be where PTI creates it.

Post by **tr1p1ea** » Sun 27 Aug, 2006 6:04 pm

I dont know if this has been reported but:

Code: Select all

  ld b,(ix + 0)

Shows up in the debugger as:

Code: Select all

  ld b,(hl)
  nop

It is misrepresented no matter what the ix offset, treated as 'ld b,(hl) \ opcode corrosponding to ix offset'

Doing:

Code: Select all

  ld a,(ix + 0)

Works fine though.

Spencer · Post by **Spencer** » Sun 27 Aug, 2006 7:54 pm

Resizing the "Interrupts" pane in the debug window, then resizing the entire debug window consistantly causes a crash.

CoBB · Post by **CoBB** » Sun 27 Aug, 2006 7:55 pm

Thank you, tr1p1ea, I fixed it now. The same error would also occur with c, d and e too, but not h, l or a. Note the changelog for another little development: setting debugger slots (note that breakpoints are still universal [!], but I'll change that soon).

I'll look into your bug, Spencer.

Spencer · Post by **Spencer** » Sun 27 Aug, 2006 8:02 pm

One point worth mentioning is that I just copied the new exe over the old one. I didn't clear save states or config files.

CoBB · Post by **CoBB** » Sun 27 Aug, 2006 8:17 pm

You don't need to. When there's a change in any of the file formats, I always make it clear.

Post by **tr1p1ea** » Mon 28 Aug, 2006 5:05 am

I did a bit of fiddling and discovered that the debugger resizing bug occurs when you try to make the debugger windows *width* smaller, without making it larger first. If you resize it so the width is larger, then make it smaller, it appears to work ok.

It also doesnt matter if you fiddle with the other info frames (interrupt, registers etc), the crash will occur if you open the debugger and try to reduce its width.

Adjusting the height appears to work fine.

CoBB · Post by **CoBB** » Mon 28 Aug, 2006 7:28 am

I fixed it now. The height was just as bogus. You could simply crash the debugger by dragging the right or bottom border of any component at the edge of the window.

Post by **qarnos** » Thu 31 Aug, 2006 11:18 am

This isn't so much a bug as it is a request.

Whenever I alt+tab to PindurTI, I inadvertantly enable the screen capture. It's a bit of a pain when you don't notice it for the next hour or so and end up with a rather large GIF file!

If you could find the time to correct this somehow, it would be much appreciated!

CoBB · Post by **CoBB** » Thu 31 Aug, 2006 12:40 pm

To or from? I can only start or finish a screenshot when leaving PTI this way.

Saibot84 · Post by **Saibot84** » Thu 31 Aug, 2006 6:47 pm

bug report: setting a breakpoint and loading a different ROM maintains the same breakpoints, but if these breakpoints now fall "inside" of an instruction, there's no way to clear the breakpoint w/o closing PTI

CoBB · Post by **CoBB** » Thu 31 Aug, 2006 6:51 pm

Thanks! The breakpoint handler is just an ad hoc hack that needs to be cleaned up anyway...

Saibot84 · Post by **Saibot84** » Thu 31 Aug, 2006 8:57 pm

in regards to the stack viewer, I was thinking that PTI could (upon first writing to the screen or whatever buffer you use) assume that the stack is full of 2-byte values... however, I thought maybe you could make some way for the user to specify (manually) that certain values are of a different length. I'm not sure how or why, but I was imagining a programmer doing ld hl,$FFEE \ ld a,$DD \ ld bc,$CCBB push hl \ push af \ inc sp \ push bc and it would be cool if PTI could not only show that as

$FFEE
$DDCC
$BB00

but also, you could right-click on the CC and select an option that'll make the stack viewer show this instead:

$FFEE
$DD
$CCBB

i don't know... maybe it's just a stupid idea

Post by **qarnos** » Fri 01 Sep, 2006 6:01 am

CoBB wrote:To or from? I can only start or finish a screenshot when leaving PTI this way.

Oops - my fault. You're right - it's only when tabbing away from PTI. Perhaps you just need to check if ALT is being pressed when TAB is hit. That should take care of it.