Slow banks & other

Hi Yaroslav! YK> FA from the Czech Republic: Never (!) install YK> vector I to slow memory. YK> [skip] YK> Based on what FA wrote, you can do - YK> water: all odd-numbered banks are slow. YK> Is this so? - I don’t presume to judge, I’ll say YK> only: all programs on Scorpio, where YK> register I = #40-#7f (bank 5) pretty honey- YK> lyat. this is not true. Or rather, not always. On almost all our machines (with a common memory field) banks are equal in speed (Scorp, Kai, ATM, Profi, Leningrad, Pentagon, Composite). The vector can be placed anywhere. branded cars: SPECTRUM 48 #4000...#7FFF - brake then two groups: SPECTRUM 128, +2 - first +2a, +3 - second I don’t remember which one has what, but one group is slowing down on banks 1,3,5,7, and the other on banks 4,5,6,7. YK> Frame scrolling ... YK> 18.. well, 19 lines.. 1 bank down, 1 up. what are you talking about? what banks? YK> 20 lines won't work - music... it's necessary If we're just scrolling, we'll be able to do 24. YK> stack: POP HL: LD (xx),HL: LD SP,xx Not the fastest. YK> About frame scrolling, where is LDI, and even more so - YK> more LD (IX+n),R - no need to say. that's it for sure :) YK> Driver 3900 baud. YK> Instruction needed: where, what, where .. YK> 1. Functions performed YK> 2. I/O addresses of the driver itself YK> 3. Maximum memory YK> 4. The source is highly desirable ZX Format #6 OR #7. There's an MAC article about the driver. I need the source code for the XTR driver - I can give it to you. I'll probably post it in the FAQ... YK> 8 images for one arrow is not bold? YK> For her, two will be enough. YK> Along the edges; left and right sprite and mask YK> must have at least 7 free bits. YK> You can move the sprite directly in the buffer when- YK> yes, he moved further than the edge, just ne- YK> place the first image there and add - YK> those or subtract 1 to the sprite output address. Terrible brake. Nightmarish. The mouse should take 2-3 thousand clock cycles. this one will take five. YK> p.S. Why do people need knowledge? -To speak- YK>rit: We won't need them. YK> And neither they nor their knowledge will be useful YK> children or their grandchildren. too deep... All the best. Dmitry aka -=LD=- / X-TRADE GROUP. -+- SMM version 1.08

Hello wandering ALL [skip] I'm a little tired of coding, so I decided write a letter.. [skip] Spectophone number 17: Jemini & Metallica (in general): On most machines there are 5 and 7 jars of honey blocked, the processor conflicts with them and it is better to allocate them to data (text, graphics) fika, tables - no interruptions of course). Coders don't put code there, much less do not direct vector "I"; the computer will be slow down a lot. And by the way, here’s what: to pen- The above does not apply to tagon. First Association preview: FA from the Czech Republic: Never (!) install vector I into slow memory. [skip] Based on what the FA wrote, you can conclude: water: all odd-numbered banks are slow. Is this so? - I don’t presume to judge, I’ll say only: all programs on Scorpio, where register I = #40-#7f (bank 5) fairly med- lyat. If you didn't know about this, then know. [skip..] Frame scrolling... 18.. well 19 lines.. 1 bank down, 1 up. 20 lines won't work - music... it's necessary Still preparing the line for printing... This was calculated based on transactions with stack: POP HL: LD (xx),HL: LD SP,xx Interrupt (1/50 second): 70,000 ticks with kopecks. About frame scrolling, where is LDI, and even more so There is no need to say further LD (IX+n),R. I thought and thought - what to do, what not to do... Still, he looks great... [skip] Driver 3900 baud. We need instructions: where, what, where... 1. Functions performed2. I/O addresses of the driver itself 3. Maximum memory capacity 4. The source is highly desirable And it will be maybe some 3873.5 baud. [skip] 8 images for one arrow is not bold? For her, two will be enough. Along the edges; left and right sprite and mask there must be at least 7 free bits. You can move the sprite directly in the buffer when yes, he moved further than the edge, just pe- insert the first image there and add- those or subtract 1 to the sprite's output address. [skip] Now let's deviate from the computer topic. Sometimes I think that in nature it is There are no teacups, but there are patterns. And if so, then everything is predetermined leno. Some believe that there is knowledge of our universe, galaxy, planetary systems, life on our planet no more, than an accident. Is this so? p.S. Why do people need knowledge? -To speak- rit: We won't need them. And the knowledge will not be useful to them or theirs children or their grandchildren. This is not in the Bible... Best wishes, PHOENIX -+- ZXASM 3.0

Hello, computer maniaks! I’m sitting quietly reading, not touching anyone, and suddenly Kirill Frolov scratched on the monitor to Yaroslav Kozlov... YK>> Coders, don’t put code there, especially not YK>> do not direct vector "I"; the computer will be YK>> slow down a lot. And by the way, here's what: to pen- KF> KF> 1. There won’t be much delay. (Dig deeper into the ancients KF> program for 48 specs, In the 48 spec everything is arranged differently, in addresses #C000-#FFFF you can throw vector I, but not in 128 KF> there is #4000..#7fff slow memory). KF> 2. You can get the code there! I don’t know about the vector. Maybe someone KF> will tell? And the code can’t go there, and I can’t - everything will slow down by 25%. [slow memory gobbled up] YK>> FA from the Czech Republic: Never (!) install YK>> vector I to slow memory. KF> KF> Will there be a glitch? Not a glitch, but an excessive brake :( YK>> Based on what FA wrote, you can do - YK>> water: all odd-numbered banks are slow. KF> KF> IMHO it depends on the hardware. Maybe only screen banks.This is the umpteenth time this sign will appear: slow memory: 4,5,6,7 banks, fast memory 0,1,2,3. And yet, during disk operations, register I must be set to stable, i.e. #3f, and then restore, an example of such illiteracy X-REVERSY on my 2-fielder this gamma does not load :( Was it really that difficult to execute the 2nd commands? ld a,#3f ld i,a and then restore, with this the authors protected themselves to some extent degrees from gentle kraks, by the way, this is also present in Black White glitch :((( YK>> Is this so? - I don’t presume to judge, I’ll say YK>> only: all programs on Scorpio, where YK>> register I = #40-#7f (bank 5) pretty honey- YK>> lat. [slow memory gobbled up] I really didn’t know that scorpions slow down :))))) [slow memory gobbled up] KF> some especially crippled and others, but not on the scorpion KF> !!! You yourself are especially crippled :(( [slow memory gobbled up] With best wishes, Sir DENIS /Ascendancy Creative Labs.

Hi Paul! KD>> X-REVERSY on my 2-fielder this gamma does not load :( KD>> was it really so difficult to execute the 2nd command KD>> ld a,#3f KD>> ld i,a PF> don't understand? what does he mean? Before disk operations, it would be necessary to set IM1 in reg. I write #3f. otherwise, on some computers the machine will not be able to load the file(s). All the best. Dmitry aka -=LD=- / X-TRADE GROUP. -+- SMM version 1.08

Hi, Kirill! KF> 2DL: why??? Where are the brakes from? What if MI1? if IM1, but the AND register = slow memory, it will not work. This is not due to interruptions, but to regeneration, because address regeneration = I*256+R. Best wishes. Dmitry aka -=LD=- / X-TRADE GROUP. -+- SMM version 1.08

Hello, Denis! 15-04-98 at 13:33:20 Denis Kitel => Oleg Grigoriev: DK> Hello, computes maniaks! Nothing like that. I may be a maniac, I won’t even deny it. I know, but certainly not a computer one. ;)) OG>> What about reading/writing _sectors_? Yes via OG>> #3d2f, yes OG>> on im2? DK> Reading will slow down for a very long time (if there is any DK> error control, otherwise it will be bent): , and the same thing with recording DK> bad - when trying to read there will be troubles, if it writes at all. DK> But if skillfully, you can combine IM2 and disk operations DK> but there will still be a brake. :-( What's causing the brake? 8-o... Look: IM2, the vector is #BE, the plate is at 257, on the interrupt- only EI hangs: RET, I climb into trdos via #3d2f pre- strictly prohibiting interruptions... So what? On my copier it's like this... Don't offer IM1 - this bastard makes INC (IY+64). Best wishes, Oleg aka Ruiner. -+- Shoot-Edit v6.06

Hello, Oleg! Viktor Kuzmin answering to Oleg Grigoriev [cutting-mutting] DK> Reading will slow down for a very long time (if there is of course DK> yes DK> error control, otherwise it will be bent): , and the same thing with recording DK> bad - when trying to read there will be troubles, if it writes at all. DK> But if you skillfully, you can combine IM2 and disk DK> operations, DK> but there will still be a brake. :-( OG> Why the brake? 8-o... OG> Look: IM2, vector equal to #BE, plate at 257, at OG> interrupt- OG> only EI hangs: RET, I'm getting into trdos via #3d2f OG>pre- OG> strictly prohibiting interruptions... So what? OG> In my copier it’s like this... OG> Don't offer IM1 - this bastard makes INC (IY+64). If there is no EI in those procedures (in TR-DOS), then everything is OK.OB. If there is and you do not use BASIC ROM, then you can use interruptions hang not EI:RET, but: push hl ld hl,#3d2f ex(sp),hl ei ret or ei jp #3d2f or Well, in short, something in this style (helps ;) - personally tested (though I have a Pentagon ;)))With best wishes, KVA of E-mage