Wait and NoWait — ZXNet «hardware.zx»

From Vyacheslav Kalinin To All 18 February 2006

Hello, Jukov Juk> Kay; Leningrad 1,2,3; - wait Juk> Pentagon, Pro, ATM, Quorum - wait-free Juk> The number of ticks in a frame is the same for everyone except the Pentagon. Juk> Therefore, it’s a big deal, whether with or without wait. But this affects Juk> about how multicolors and border effects will look. What about the odd numbered teams in yellow scorpion?

From Konstantin Zhukov To All 18 February 2006

Hello, Soplik Kay; Leningrad 1,2,3; - waiters Pentagon, Profi, ATM, Quorum - wait-free The number of ticks in a frame is the same for everyone except Pentagon. Therefore one figs, what with wait, what without it. But this affects how they will look multicolors and border effects.

From Victor Ivanov To All 18 February 2006

Hello, All The question is essentially lame, but I can, I’m not a coder. Which of the current clones are NoWait and which are not, and whether they slow down much. Thank you.

From Andrey Alexandrovich Titov To All 19 February 2006

Hello, Jukov Juk> Kay; Leningrad 1,2,3; - wait Juk> Pentagon, Pro, ATM, Quorum - wait-free Juk> The number of ticks in a frame is the same for everyone except the Pentagon. Juk> Therefore, it’s a big deal, whether with or without wait. But this affects Juk> about how multicolors and border effects will look. How are these the same thing??? Pentagon 128 (Classic) - 71680 Pentagon 128 (Grandromax) - 73000 (or something like that) Profi - 69888 (or the same, something like that) Scorp - 70908 etc. etc.

From Victor Ivanov To All 19 February 2006

Hello, Soplik That's why I ask. It just turns out that if you sculpt a toy in multicolor, then scorpions can be delete it immediately :( Yes?

From Vyacheslav Kalinin To All 19 February 2006

Hello, Soplik Sop> That's why I ask. Sop> It just turns out that if you sculpt a toy in multicolor, then Sop> scorpions can be crossed out immediately Sop> Yes? No, but there will be a lot of izvarats, with odd bars.

From Chunin Roman To All 19 February 2006

Hello, Jukov Juk> As for the Pentagon, I already said that it has a different number of clock cycles Juk> frame, and therefore performance. And Scorp (I don’t know how) Juk> old boards, but on green) 312 lines for 224 clock cycles, i.e. like Juk> Pro 69888. This is a terrible misconception about the speed of the Pentagon. Performance The Pentagon is the same, it just has fewer frames per second (i.e. extremely non-standard video signal).

From Chunin Roman To All 19 February 2006

Hello Titus Tit> We are talking about the same thing. Coder when writing one-frame effect Tit> thinks only about the number of clock cycles in the frame, and is not interested Tit> absolute speed of the machine. If I write a one-frame effect Tit> taking 71000 clock cycles, then not on any machine with less Tit> number of ticks in a frame, there will be no one-frame effect. Now it’s correct - the number of clock cycles in the frame, not the performance (local or global) :). With such inaccuracies in wording I will fight mercilessly ;) because they introduce beginners and not only into very large misconceptions (such as the Pentagon is the fastest). Performance is still measured in the number of operations per second.

From Konstantin Zhukov To All 19 February 2006

Hello Titus Tit> How are these the same thing??? Tit> Tit> Pentagon 128 (Classic) - 71680 Tit> Pentagon 128 (Grandromax) - 73000 (or something like that) Tit> Profi - 69888 (or something like that) Tit> Scorp - 70908 Tit>, etc. etc. As for the Pentagon, I already said that it has a different number of cycles in the frame, and hence the speed. And Scorp (I don’t know how it is on old boards, but on green) 312 lines by 224 measures, i.e. like Pro 69888.

From Andrey Alexandrovich Titov To All 19 February 2006

Hello Conan Con> Logical error. Frames (interrupt periods) are different for and Con> other clones, so measuring performance with them is the same thing Con> what to come to the competition with your own chronometer that runs Con> slower than others and, using it to measure the race time of your favorite athlete Con> say: . Con> Con> As for dems, it is not that is critical for them, Con> and the number of clock cycles between interrupts. No more, no less, but exactly that Con> for which the demo was written. Turn on turbo mode on Con> (7MHz) and make sure that the speed is (absolute or local) Con> many demos are very . We are talking about the same thing. When writing a one-frame effect, the coder only thinks about the number of clock cycles in the frame, and he is not interested in absolute performance cars. If I write a one-frame effect that takes 71,000 clock cycles, then not a single to a machine with fewer clock cycles per frame, there is no one-frame effect there will be.

From Andrey Alexandrovich Titov To All 19 February 2006

Hello, CHRV CHR> This is a terrible misconception about the speed of the Pentagon. CHR> The performance of the Pentagon is the same, it just has frames per second CHR> less (i.e. extremely non-standard video signal). Absolute performance - yes, but local performance is important for dem (in flow of one frame), and it is higher ;-)

From Konstantin Sviridov To All 19 February 2006

Hello Titus Tit> Absolute performance - yes, but local performance is important for us Tit> performance (within one frame), but it is higher ;-) Logical error. Frames (interrupt periods) are different for and others clones, so measuring performance with them is the same as coming to competitions with your own chronometer, which runs slower than others and, Having measured the race time of your favorite athlete, say: “At what time is he racing?” fastest! Record holder!>. As for demos, what is critical for them is not , but the number cycles between interrupts. No more, no less, namely the one under which the deme and was written. Turn on turbo mode (7 MHz) on the and make sure that performance (absolute or local) is not even suitable for many demos taste>.

From lvd To All 20 February 2006

Hello, CHRV CHR> Now it’s correct - the number of clock cycles in the frame, not CHR> performance (local or global) :). With similar CHR> I will fight mercilessly with inaccuracies in wording;) because they introduce CHR> beginners and not only into very big misconceptions (such as the Pentagon CHR> fastest). CHR> Performance is still measured in the number of operations per second. And this, sorry, is called fighting windmills. Gives a chance to different people finger with a different font to show off once again =) Normal people immediately understood what was going on. Beginners themselves also do everything with over time they will understand everything and learn everything... or they will not understand and will not learn, in this In case, they don’t even care that the Pentagon has different things there and why =)))

From Vyacheslav Kalinin To All 20 February 2006

Hello, Jukov Juk> And Scorp (I don’t know how on old boards, but on the green one) has 312 lines per Juk> 224 clock cycles, i.e. like Pro 69888. I'm talking about something else! Teams occupying an odd number of clock cycles are complemented to an even number. Accordingly, if the program contains many odd-clock instructions, then the number of clock cycles can essentially drop to ~65000.

From lvd To All 20 February 2006

Hello, CHRV CHR> Vadik is not constructive. CHR> Then we ourselves, first of all, suffer from these misconceptions. CHR> Thanks to such formulations, the Pentagons are incorrect CHR> become the standard. Well, I’m saying that all sorts of correct formulations are for pedants =) A Pentagon... it’s good that at least it’s a standard, otherwise there would be a pipe at all.

From Chunin Roman To All 20 February 2006

Hello, lvd lvd> And this, sorry, is called fighting windmills. Gives a chance lvd> different fingers with a different font to show off once again =) Vadik is not constructive. We ourselves are the first to suffer from these misconceptions. Thanks in such incorrect formulations then become the standard for the Pentagons.

From Andreas Kaiser To All 20 February 2006

Hello Titus Tit> '71680' was taken into account by demomakers at least since '94. Yes, I don’t argue, but the Pentagon-128 appeared earlier than 94, having at the same time 128K, BDI and AY on board.

From Andreas Kaiser To All 20 February 2006

Hello Titus Tit> IMHO, Pentagon 128 has become a standard due to its prevalence in Tit> the times of the birth of the domestic demoscene - once, unchanged Tit> there are two basic parts of its circuit diagram (if you remember the others Tit> clones, then almost each of them had several variants), and, Tit> of course, thanks to the magic number '71680' ;) IMHO, just because of BDI and 128K. To the people who have an idea about this "71680" there wasn't much at that time.

From Andrey Alexandrovich Titov To All 20 February 2006

Hello, CHRV CHR> Vadik is not constructive. CHR> Then we ourselves, first of all, suffer from these misconceptions. CHR> Thanks to such formulations, the Pentagons are incorrect CHR> become the standard. IMHO, Pentagon 128 became a standard due to its prevalence in times the emergence of the domestic demoscene - once, the immutability of the basic part of it circuit diagram - two (if you remember other clones, then almost each of them there were several options), and, of course, thanks to the magic number '71680' ;)

From Andrey Alexandrovich Titov To All 20 February 2006

Hello icebear ice> IMHO, just because of BDI and 128K. To the people who have an idea about this ice> "71680" was not very much at that time. Demomakers have been taking '71680' into account since at least '94.

From Chunin Roman To All 21 February 2006

Hello icebear ice> Yes, I don’t argue, but the Pentagon-128 appeared earlier than 94, having ice> with 128K, BDI and AY on board. The AU was a slap on the processor's seat. Only one appeared on board Solon model. Pentagon 128 has been actively appearing since 1990-1991. I showed it to Conan, it seems my board still has errors and a trace in which the top layer was the bottom and the bottom was the top. And then the boards with the mark came “96” (this is the symbol) 1992. Now they have already become standard.