MaxCoderz

ticalc link is up

http://www.ticalc.org/archives/files/fi ... 42807.html

you can send feature requists to featprog[at]ticalc.org

triple post? i did only klick "submit" twice...

double post, maxcoderz is slow today...

sending %xxxxxxx0 to port $3 should do the trick

send %00011101 to re-enable every interrupt, see day 23 of asmin28days

more than a year after i started on this project, it is finally in releasable state. it isn't finished, (why? various reasons, most important being a lack of interest) but very playable, it is only missing 2 important features: you cant "die" yet. this means all the scissors are kind of useless ...

you missed a few....

i still need to fix the font, however

wow! verry nice

i already released a demo a few months ago, i think its still hosted on UTI

tell me any bugs you might find

King Harold wrote:Well be careful, you haven't compressed something until you can decompress it

bongo wrote:Actually I can't find the actual hexadecimal codes in this document. Where are they?

http://www.ticalc.org/archives/files/fi ... 26877.html

lz77, and then von-huafmann the lz77 decompresser already works, and von haufmann is pretty easy: shift everyting 3 bits to the right if the outcome is %111, push $30 on the "decompress stack" folowed by 8, and then 5 bits from the compressed data if the outcome is %110, get 4 more bits and look up ...

thread retrieval!

i finally tackled the compression:

bytes in grey are bytes i won by compressing the level, ratio is around 70%, average is a bit higher

next task: decompressing

well, vdu4 loads (Text.4.PutMapNoWriteBuffer) in (Text.PutMapNoWritebuffer+1), but vdu5 loads the adress of a routine?

sll is the same as scf\ rl

just 4 wasted clock cycles and a byte..