cc65/AtaSimpleIO cursor?

[TGB1718]

20 hours ago, TGB1718 said:

If you use CIO to clear the screen you will probably end up with 2 cursors

I think that's the reason, the original cursor and the one you put in place

I tried messing with ROWCRS and COLCRS, but for some reason it didn't want to play games.

 

Edited June 10, 2023 by TGB1718

[Harry Potter]

Try setting ROWCRS  and COLCRS before clearing the screen.

[sanny]

On 6/10/2023 at 10:41 AM, Irgendwer said:

@sanny?

clrscr() is part of the conio library which is optimized for speed.

[Harry Potter]

I can write a version for AtaSimpleIO optimized for size. 

[TGB1718]

9 hours ago, Harry Potter said:

I can write a version for AtaSimpleIO optimized for size. 

I know you have an obsession for writing smaller code, but unless your writing a program

that's not going to fit in the RAM available (personally never reached that point, (well maybe once))

I think most people want best performance over size.

 

How about re-writing for MAX performance

 

[Wrathchild]

10 minutes ago, TGB1718 said:

How about re-writing for MAX performance

Isn't that typically app specific. Simple example, if cart based the you could utilise massive unrolled loops wheras might avoid that on cassette.

 

Personally I think HP is flogging a dead horse, do we have examples of others using these libraries yet?

Edited June 13, 2023 by Wrathchild

[TGB1718]

1 hour ago, Wrathchild said:

Isn't that typically app specific.

I think you missed my "smiley"

 

1 hour ago, Wrathchild said:

do we have examples of others using these libraries yet?

As many have said before totally lacking documentation etc.

[Wrathchild]

13 minutes ago, TGB1718 said:

I think you missed my "smiley"

let alone a question mark 😆

[Harry Potter]

Well...I recently fixed the documentation problems with my text adventure codes, and I believe the documentation and header files are sufficient for AtaDiak65, AtaSimpleIO and MemXAtari.  Wrathchild: you are kind of right: I am flogging a dead horse and am optimizing too early.  I believe I have a reason to optimize some of my software, though: that it is designed for other users and that, in the end, I want people to be able to write more efficient code using my libraries.  

[danwinslow]

Nobody is using your stuff. That is your actual problem. You are providing something that no-one needs. All of your utilities help with things that aren't actually problems.  Everything you handle is either already done in numerous ways, or trivial. In particular, no-one needs "cubbyhole" optimization software. AdvSkel is actually not a bad idea, but its executed and documented so badly that no one wants to use it.

 

Stop obsessing about sizes, and take the time to build AdvSkel out into a real toolkit for people writing text adventures. Document it well. Then, even in this late day of 8bit, somebody somewhere might use it.

[Wrathchild]

Exactly, you can write using Inform or even ScottFree type adventures and get them running on A8 and anything else with an interpreter with little to no coding skill

[TGB1718]

Just for fun, I tried using some of the libraries of Mr Potter, Hmmmm....

took the libraries atasimplieio.lib and simpleio.lib from the downloaded zip file

AtaDisk65-002.zip and had a conflict when I compiled a test program, it would appear

that printc.o is included in both libraries ???

Also even with a simple test program,

open a text file that does exist

read 40 characters into a buffer

I used printf to try print the input string as I couldn't use the atasimpleio library due to the conflict.

then close the file

really simple test

 

nothing was printed to screen and on every occasion

Altirra eventually crashed

 

Bit of a mess really

😬

D:\ATR_NEW\cc65\lib>ar65 t atasimpleio.lib 
getkey.o
getline.o
goline.o
home.o
printc.o <<<<<<<<<<<<<<<<<<<<<<<<< here
printi.o
prints.o
printscr.o
revers2.o
tab.o
wrtitle.o
printrs.o
printrscr.o

 

D:\ATR_NEW\cc65\lib>ar65 t ataricio.lib 
accessciov.o
loadimg.o
openfile.o
closefile.o
readfile.o
writefile.o
loadimg2.o
crt0.o
printc.o  <<<<<<<<<<<<<<<<<<<<<<<<<<<<<here too
saveimg2.o

Edited June 13, 2023 by TGB1718

[Harry Potter]

There should be a file in the distribution called "printc.s."  Include it in the command line for the Atari-port batch files' cl65 command line before the libraries and come back to me with the results.  I was having a bug with it, also, and had to supply it directly.  The bug was that the start-up code wasn't writing properly to the module's code to define the address of the OS service for AtaSimpleIO's printc() function.  I'm sorry. 

[_The Doctor__]

Sounds like one should try their stuff out on a clean machine to see if their stuff works and then document how to actually use. This is of course then to be tested on real hardware before distribution.

rinse repeat, no new tale to tell...

Edited June 13, 2023 by _The Doctor__

[Wrathchild]

also really consider getting yourself on github and then issues can be tracked, people can view change history etc

Edited June 13, 2023 by Wrathchild

[TGB1718]

Did a bit of disassembling to try find out what's happening, have a look at this bit of code from the list file from cl65

in particular the JSR instruction highlighted, then look at the disassembler output at the same instruction

This was compiled with printc.s on the cl65 compile and the file in the project directory

 

000000r 1  A0 0C            ldy     #$0C
000002r 1  20 rr rr         jsr     subysp
000005r 1  A0 0B            ldy     #$0B
000007r 1  B9 rr rr     L0002:    lda     M0001,y
00000Ar 1  91 rr            sta     (sp),y
00000Cr 1  88               dey
00000Dr 1  10 F8            bpl     L0002
00000Fr 1               ;
00000Fr 1               ; printc(instr);
00000Fr 1               ;
00000Fr 1  A5 rr            lda     sp
000011r 1  A6 rr            ldx     sp+1
000013r 1  20 rr rr         jsr     _printc
000016r 1               ;
000016r 1               ; }
000016r 1               ;
000016r 1  A0 0C            ldy     #$0C
000018r 1  4C rr rr         jmp     addysp

 

206F   JMP  0000      4C 00 00 <<<<<<<<<<<<<<<<<<<<<<<
2072   LDY  #0C       A0 0C                    ^
2074   JSR  20D8      20 D8 20               ^
2077   LDY  #0B       A0 0B                    ^
2079   LDA  2108,Y    B9 08 21              ^
207C   STA  (82),Y    91 82                    ^
207E   DEY            88                           ^
207F   BPL  2079      10 F8                    ^
2081   LDA  82        A5 82                    ^
2083   LDX  83        A6 83                    ^
2085   JSR  206F      20 6F 20  DOES A JSR HERE i.e. to a JMP 0000
2088   LDY  #0C       A0 0C

[Harry Potter]

I just reinstalled and tried the latest published copy of AdvSkelVic65 and didn't see the bug.  The cartridge-based Atari version crashed after the intro screen, though, but the screen did display.

[Harry Potter]

Maybe you have the wrong setup in your emulator: AdvSkelVic65's Atari ports require an Atari 800 emulation.  BTW, the bug with the JSR $0000 instruction is fixed by a code snippet in the crt0.s file that writes the proper address to the instruction, as the JSR instruction calls the OS's print char function, and I need to write the proper address, and the address is different on different systems.

[TGB1718]

Ok, putting crt0.s fixed the problem, but again it's down to documentation, you need to describe the build process.

[Harry Potter]

That's better.    In the next version, I have to document it better.  I thought I documented it better in the last release.  Is there any specific issue you are having with the documentation?  Did you read the comments in the code?  Much of the build process is a fill-in-the-blank style.  For example, if you want to add the ability to read a certain item, you add the code to read the item to the vRead() function by adding a condition to the switch() there, adding code to display the necessary text and ending the condition with a return;.  I suggest putting the extra text in either WordsDB.c or WordsDB2.c and use printtok() to display the text, so that it can be stored in a system's extra RAM.  Does rthis help?

[_The Doctor__]

That's one of the problems!- 'in the next version I will document it better'.  If you say all of this stuff in the threads why not copy paste the things to the documentation now?

and for that matter, making people read endless threads to do the simple fixes blah blah shouldn't be the case... you should do the fixes immediately so people could build and use what you keep going on about. No one wants to have to put that much time into something that should be ready to use and ready to go.

Edited June 14, 2023 by _The Doctor__

[Harry Potter]

If there are any specific problems with the manual or the code comments, I'm all ears, as I'm currently updating AdvSkel65 now.  AdvSkelVic65 is having a problem with 32k Atari cartridge support, and, while trying to fix it, I broke its 32k C64 cartridge support.

[drac030]

On 6/12/2023 at 8:24 PM, sanny said:

clrscr() is part of the conio library which is optimized for speed.

clrscr() optimized for speed. Like to make it execute once per frame, 50 times per second. At the expense of compatibility, portability and everything else. Yes, that makes perfect sense. Especially in a C library.

Edited June 15, 2023 by drac030

[Harry Potter]

Good news: AdvSkelVic65 is currently working properly on all supported systems.    Now I need to work on the manual.

[sanny]

On 6/15/2023 at 5:36 PM, drac030 said:

clrscr() optimized for speed. Like to make it execute once per frame, 50 times per second. At the expense of compatibility, portability and everything else. Yes, that makes perfect sense. Especially in a C library.

Don't know what this has to do with C itself. It's there to speed up screen output. You don't have to use it.

https://cc65.github.io/doc/library.html#s5