Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!usc!wuarchive!psuvax1!rutgers!cbmvax!jesup From: jesup@cbmvax.commodore.com (Randell Jesup) Newsgroups: comp.sys.amiga.programmer Subject: Re: Mike Farren Tutorial. Message-ID: <20191@cbmvax.commodore.com> Date: 29 Mar 91 00:25:41 GMT References: <1991Mar24.204206.11145@starnet.uucp> <20115@cbmvax.commodore.com> <1991Mar27.012717.11541@starnet.uucp> Reply-To: jesup@cbmvax.commodore.com (Randell Jesup) Distribution: comp Organization: Commodore, West Chester, PA Lines: 116 In article <1991Mar27.012717.11541@starnet.uucp> sschaem@starnet.uucp (Stephan Schaem) writes: >>I published in an AmigaMail article and devcon article (low-level drive >>access). The basic spec is there are (with MFM) 6250 nominal bytes of >data >>per track. Combine that with the amiga nominal timing (2.3% fast), >disk >>speed 5%), and amiga timing slop (5%), and you can store 5999 to >>6812 bytes per track, depending. So if you remove ALL sector headers, >>etc, you can get 959840 bytes on a disk safely >Sory I must have missed something. 6812*2*80=1089920.... >OK. 300RMP, 4 bytes every 63.5 microsecond THATS IT! You dont >need to do fancy calculation! No, that's 5999*2*80. You CANNOT assume you can store more than 5999 bytes per track when you write. 6812 is the maximum you can ever write (so you need a slop (gap) at the beginning of a write of 814 bytes minimum). See the article I published in AmigaMail on low-level drive access, with all the specs (and drives DO vary wildly). Other important numbers are 3ms step time, 15ms settle time (on top of the 3ms), 1200 microseconds post-write delay before deselecting or stepping or changing heads; 1000 microseconds side select delay before reading or writing. Motor on delay of 500 milliseconds OR until DSKRDY indicates the drive is up to speed. It's slightly different when mass-producing disks: so long as you don't create disks that exceed paula's sync rate, since you control the drive in the duplicator you can make sure you can write the number of bytes you want to (by running the drive slow or the duplicator fast). That number should not exceed 6812 (from above). >I CANT say anything publicly, but we dont do 'illegal' tricks... >We test on various drive (from 8 diferent compagnies) 4 diferent amigas >(NTSC,PAL,68000 to 68030, A500,A1000,A2000 (A2630), A3000). Ever check one where the owner had attached a genlock? And in any case, 4 amigas or 8 drives is a small sample. Some drives that have the most problems were only used for a short while, or a long time ago. >We do ALWAY checsums and everything we read... >Right now the ALL INDUSTRIE should supose you have 300RPM (speed >ajusted) >drive.So you designed for 300 rmp drives... >I dont have acceptable margin, but the tolerances are 'wide'. 300rpm, but we (and the manufacturers only guarantee) +- 1.5% (297.5 to 304.5 rpm). It can vary due to drive age, media type (some are rougher), temperature, etc, input voltage changes, etc. >If you want to talk about the subject, E-MAIL only, please... I think others should know these things (disk timing issues). >>some barely handle our spec. Some can read past track 79, some can't. > >Sory to say that but you should have a standart drive!Now the industrie >produce ONLY drive with 'internal' motor speed control and ajustment! >So you have 300 RPM drive. >I cant speak for them! but if our software dont load, you have a VERY >bad >drive! We try with LARGE speed variation and it alway pickup the data >corectly. >No, people to custom track sizing! messing up with MFM coding!!! >Using table coding. Note my signature: I work at Commodore, and I'm the person responsible for the trackdisk device in 2.0 (which is rather more efficient than in 1.3 - most of the extra speed is that FS buffers no longer have to be in chip ram). We have a set of drive specs (as mentioned above). Drives may vary within those specs, and they do. >>Then they should do it. > >Well PIRACY is the key... From my part its a maybe! if more than 2 >people in >the ALL AMERICA want HD support, I might considered it... Why do you say 2? Go to comp.sys.amiga.games and search the archives (if you archive everything there - most places don't, we do). You'll find hundreds of people saying they want HD support. You're over-exaggerating your position. BTW, people are less likely to flame you if you avoid making inflammatory statements (statements that are likely and intended to annoy or anger people). It's like yelling at someone (face to face) that they're entirely wrong and don't have any idea what they're talking about. Rational, reasoned discussion (and disagreement with reasons, not absolute statements) works far better. >>things (I've gotten personal letters (snail-mail, no less) from people >like >>this. The company wouldn't even pay for RKMs, let alone joining the >>developer program.) > >Are we talking about quality amiga product? Well, when the amiga will >sell more then 20,000 copies world wide for the best game of the moment >they might be some consideration!? This person had done several "hits", and gets a salary instead of royalties. >Also! stunning has it can be! th WORST EVER amiga developers go to >DEVCON! perosnally receive the 2.0 doc from CBM president of the >concerned >contrie! etc etc... And only do PC port over the amiga that dont use >anything amiga at ALL! and trash the OS and rewrite BAD disk function >that cause return in LARGE number!.... That's the sort of obviously over-stated inflammatory comment I was describing above. I'll resist toasting you to a crisp (this time). Next time I may not be so nice. -- Randell Jesup, Keeper of AmigaDos, Commodore Engineering. {uunet|rutgers}!cbmvax!jesup, jesup@cbmvax.commodore.com BIX: rjesup Thus spake the Master Ninjei: "To program a million-line operating system is easy, to change a man's temperament is more difficult." (From "The Zen of Programming") ;-)