Hello all, after a nice week in Italy with my parents and sister and then two weeks partying in the Spanish Lloret de Mar I am finally back behind a screen. That should explain my quietness the last few weeks.
Therefore I could also not send an insulting message to Ian about my list of instruction timings. First of all, the list that appeared on Fred (many years ago!) was simply a basic indication that you couldn't use the official timings. I then went along and tested all the instruction speeds as compared to the speed of a NOP (with the screen on!) I then converted the relative speeds which I got from the tests to "t-state values". For this I am sorry since it is the improper use of the word t-state. If you check out the speed of a LD A,7 compared to a NOP then you should see that this instruction takes 9.3/4=2.3 times more time. Later on I also wrote in a letter to Fred (which was included) that my values were not totally correct and that I would send a new list as soon as I got it finished. I never finished this list. What the list comprised of was basically the time (in t-states!) taken by an instruction during screen and border time. I then saw that tendency for 4 cycle instructions existed and that some other funny things were going on that I aborted. When writing the burstplayer (as part of the SAM mod player v2.02 and the soon to be released WAV player) I discovered that I couldn't exchange little routines which theoretically had the same t-state length and chucked in the timing towel and aligned everything with palette colour changes - which was a hell of a hassle (5 different devices with 4 different sound routines). I wish I could get this damn Pentium to sample the sam's output at a massive frequency so that I could get timings aligned that way - but alas, no pc programming for me. BTW I have go a Z80 cross assembler for dos which is public domain - it also does 8086 and something else. Haven't used it yet since Comet is still handling nicely but the documentation makes it seem very good, it even includes auto-optimization of instructions! I must say that your (Ian's) explaination of t-states looked very impressive and thorough. I must say I still prefer the try and cry method. Still on the insult line, I don't see why I should get a slam for "recently" mentioning the sigma/entropy logo thing - since the last I talked of it was with Andrew at the Gloucester show (some 4 months ago!). I can't help it that when I see a graphic of a twisted sigma that I thought it looked like a twisted E. What's the big deal with it anyway? I couldn't care less if it was the Entropy E itself (as a manner of speaking). Looking forward to seeing the demo though. Also looking forward to the Chaos 3, David. On the subject of demo's anyone incorporated a MOD tune in their demo yet? Right, end of subject. Next, the accelerator sounds nifty (hmmm, 20 times the speed - that should allow 16 channel mods playing at 44 khz in background mode) We would need move internal memory to store the better mods though. What would the costs of the accelerator be though? Next question, how's the hard-drive (or dos) coming along? BTW Simon, could you mail me a teledisk copy of the comet-ascii thing - it might work nicely together with this very rough disassembler I've got which disassembles to a text file (open-type via streams aaaaaaggggh!). Ummm, that sort of rounds things up for me, time to get to sleep after having sat in the bus for a bloody 28 hours (5 hours due to the bus breaking down!). CUL8R Solar Flare of Entropy (aka [EMAIL PROTECTED])
