Here you can find side projects of the staff and great projects which we think should get extra support. (Note that featured products are not products of MaxCoderz)
kalan_vod wrote:Well I just meant that could you fill the screen. Like not shrinking it 50% but shrinking it 75% or so?
That's what I tried. The problem is that you lose pixels in doing so; if you look at the SCHIP "Blinky" (Pacman) you'll see the dots and walls are one pixel thick and they end up being erased.
Well I thought it would look ok if that had happened, but you say you tried it and it looks crappy so I don't mind . It is still a great program !
benryves wrote:The best looking method was to scale the entire thing down to 50%, but add an offset of either one x or one y (so four different frames) each loop, creating a flickery greyscale view. For some games this looked OK, for others (such as Ant) it made the graphics a horrible mess.
In the end, I removed that code completely and just use the pan-and-scan method.
The former method sounds interesting. Perhaps you could make that an option in the settings?
Ben, do you still have no Internet on your own PC? I think that could explain how you can be so productive I shouldn't be wasting time online as much as I do, and actually code stuff
Timendus wrote:The former method sounds interesting. Perhaps you could make that an option in the settings?
I might, but the extra "video modes" added quite a lot of code and were very ugly.
Ben, do you still have no Internet on your own PC?
Nope, no internet access. This is why I only learned about I being modified by the FX55/FX65 instructions today, as the docs I downloaded on Friday evening didn't mention it.
The reason I hadn't heard about I being incremented after copying the registers to memory was because... it's wrong.
I added scaling the LCD: here's a "before and after" shot
I also reorganised the "ROMs" folder of the zip so it should be easier to find the right files.
I haven't done anything with the sprite routines. I'll probably add more conventional wrapping, and possibly an option to change the interrupt speed, then call it a day (the interrupt is used on the timers - notice how Joust runs at roughly the same speed, whatever options you set. That's because it sensibly uses the CHIP-8 timer).
sound awesome, I am installing a few files on my calc now.
Only problem tho you might want to fix for people who can only use an USB cable: in TI connect only the emulator, IBM and UFO can be sent, all other calcs are flagged as incompatible type. Works fine with tio graph link tho
xlibman wrote:sound awesome, I am installing a few files on my calc now.
Only problem tho you might want to fix for people who can only use an USB cable: in TI connect only the emulator, IBM and UFO can be sent, all other calcs are flagged as incompatible type. Works fine with tio graph link tho
Argh, TI-Connect. It's because the files have lowercase characters in the name.
Come to think of it, if you send the 15puzzle file, you won't be able to delete it off your calculator without a RAM clear, as the OS won't display it (urk).
kalan_vod wrote:Very nice scaling! Btw it looks like a litte gs in there?
When a black and a white pixel pair are compressed to a single pixel, it is drawn every other frame. Hence the grey, which is very noticable in SCHIP Blinky with the one-pixel wide dots/walls.
I rewrote the sprite routines to speed them up and allow for wrapping around the screen (title screen of Field, paddle in Brix).
There's a dedicated program to convert CHIP-8 programs included, and I've reconverted all the programs to work with TI-Connect. I also removed X-Mirror (calls 1802 functions) and boot (requires loading at $000 rather than $200). You can also set the speed of the interrupt timer between 0Hz and 255Hz (defaults to the normal 60Hz).
tr1p1ea wrote:The ant game looks really good ... and changing the int timer to 0Hz aye? .
0Hz is the slowest (in fact, the timer is stopped). The screenshots were taken at 60Hz, with a highish frame skip. I culled every other frame, then extended the delay, but it's still fractionally faster than the calculator... depending on which browser/GIF viewer you use