On-Line #58 13 октября 1996

Интервью

Interview - Interview with Sergei and Andrei Zonov Larchenko (developers of computer "Scorpion")

                INTERVIEWS



   Interview developers computer Scorpio Sergei Zonova and
Andrew Larchenko newspaper ON-LINE.



   OldMan> Moscow Machine SCORPION is not particularly popular. 
In your opinion, why? 


   Serge Zonov> I think that's the same reason she was in St. 
Petersburg Today the most popular of the Spectrum-compatible 
machines. I I mean, first of all, the presence (or absence in 
Moscow) more or less complete infrastructure support and 
maintenance users as well as a firm that continually promotes a 
specific model of computer on the market. Moscow hackers can 
with I disagree, believing that Scorpiona have many 
shortcomings and that is why he is not as popular in Moscow. 


   Typically, the arguments are those or other examples
inoperability and incompatibility of the computer with some 
demo programs written in the Pentagon and "tuned" by its 
technical specifications: the port addresses, a period of 
staffing scan, etc. And the Pentagon is just and has been 
developed in Moscow actively promoting, as in the city to do it 
a lot easier. Although any more or less versed in circuitry and 
programming experts agree that the Pentagon is not the best 
with these views develop. So it turns out that what I said at 
the beginning of the answer to your question, and is the main 
reason for the popularity of a computer in a particular region. 


   Of course, all other factors are also important, but
they are, even in combination, remain in the background. But 
despite this, in Moscow, many owners Scorpionov. There is a 
club Scorpion, etc., can not be said about Peter and the 
Pentagon. 

OM> The main problem SCORPIONa - incomplete decoding port # FD.
In St. Petersburg organized a real war with the command OUT (# 
FD), A. Could you possibly not the first time to talk about

this trouble: do not perceive your car at all the command or
perceives, but under certain conditions. Moreover, not
Remember to mention the older versions SCORPION ZS.

ZS> hold my opinion that not "incomplete decoding port
# FD ", a more complete (than in the simple ZX128, but not in 
the Spectrum 128 +2, +2 A, +3) decoding ports 1FFDh, 7FFDh and 
ports AY-8910/12 is the cause of improper conduct computer

when using OUT (# FD), A. In developing the computer was 
originally laid, the additional system port having the same 
address as in the Spectrum 128 +2, +2 A, +3, and precisely 
address 1FFDh. The distribution of digits in this port chosen 
as close as possible to the distribution of digits of the Port 
Spectrum 128 +2, +2 A, +3. 


   In the derivation of the port to a shortened at # FDh, at 
junior byte address bus arrives code FDh, and the high byte 
address bus - the contents of the battery. As in Scorpio in the 
upper part of the address decoding involves discharges A15, 
A14, and in old versions, and A12 (which was odd), then if you 
do not keep track contents of the two most significant bits 
accumulator, the command OUT (# FD), A, you can instead of the 
desired destination port 7FFDh, fall into either port 1FFDh, or 
in ports AY and vice versa. All this will tend to lock up the 
computer. Much reliably use the output from the full 16-bit 
addressing. And if you just can not do without the commands 
with a short addressing what should monitor the contents of the 
two bits are the battery so you fell "at". 


   For some reason, when referring to a shortened AY at 
programmers (and mostly all this concerns the Moscow and 
Ukrainian programmers) and all, but when accessing a system port

address 7FFDh track only one significant bit of the battery. 
The reason for this, in my opinion, lies in the fact that most 
of them work at the Pentagon, where there are no additional 
system ports, and they do not check how their works on other 
computers. All of the above information on sampling conditions 
for all I / O port is in the "Guide Users of the Scorpion, in 
the "Quick Reference Guide for programmers, in addition, long 
walks and sale scheme Scorpio on which all this can be easily 
seen. 


   I want to note once again that there is absolutely no 
problem with Port # FD because of his "incomplete" decoding, it 
is only necessary to write programs correctly (or use the full 
address decoding, or monitor the contents of two bits are the 
battery). Additionally, it should be noted that during the 
discussions on this topic held in the Spectrum EnLight-96, 
agreed with me are recognized in the Spectrum environment 
authorities as Stalker and Max Iwamoto.


OM> Admit it, the Scorpion has other problems?

ZS> Absolutely. As in any other computer problems are and
in Scorpio. But they can be divided into two groups:


   1. Problems such as: why "the computer crashes, or why he
improperly executed this or that team, and does not work
drive controller.


   2. Problems such as: why did he display only 256 points, not
512.


   That is, the problem of qualitative characteristics of the 
computer as compared with similar models, and compared with 
other classes of computers. In my view the serious problems and 
shortcomings of the first type in Scorpio there, and the 
problems of the second type of course is, and always will, 
since here the estimate Users are always subjectively and 
assessment criteria are different. 


   Among the shortcomings and problems of the first type of 
which I have heard or read called the following: abnormal 
organization of the port FFh, inability to connect additional 
devices such as modem, programming and even some "nonsense". 


   In about port FFh. On the Scorpio it is implemented 
completely as well as in ZX 48/128, namely: when referring to 
any unrealized physically (ie, does not exist on the parent 
board, nor the additional peripheral boards) input port 
accesses the port that received "the people" is the name - Port 
# FF. This port received the current attributes of the screen 
when scan is in the working field (current paper and

ink), or code of FFh, when there is a map of the curb.


   The question remains how the computer learns about the 
connection of additional ports of entry located on the 
peripheral boards? This question for many remains so 
misunderstood. Although I have nothing here and did not invent, 
and repeated all one to one, as was done in firm ZX Spectrum.



   In fact, a very interesting picture is obtained: dozens of 
people go on about this or other compatible PC with a brand 
ZXSpectrum, often written on this subject the whole didactic 
tracts, as then it turns out that the Spectrum they are in the 
eye is not seen neither alive nor in the form of the scheme and 
therefore do not know, for example, clicking on its bus system 
derived signal IORQGE, in addition to signal IORQ. Namely, in 
the proper use of these signals and the "crux of the matter" 
correct connection of additional peripheral boards as a 
corporate ZX-Spectrum. And Scorpio, because in this matter in 
Scorpio is repeated all the way thought of Mr. Sinclair, and 
nothing new here was not invented yet. 


   Talking about the correct connection of peripherals, I
I mean, first of all, the uselessness of any pererezok and
modifications of the motherboard circuitry for connecting 
peripherals - just put in and everything works without any 
conflict. The bottom line is that it is a signal IORQGE and 
serves as a signal by which selects all ports on the 
motherboard, including the port FF. Signal IORQGE repeats 
signal IORQ until not selected any peripheral device. As soon 
as you select some peripheral device, it takes control signal 
IORQGE, but it puts on this wire logic 1 thereby blocking the 
sample of all ports on the motherboard including port and FFh. 
Signal IORQGE formed not just as very very simple from the 
signal IORQ. And you can do it perfectly on any 
Spectrum-compatible computer (except, perhaps, those who 
gathered in the main to 155 series), namely: need to break the 
wires suitable for 20-pin Z80 (this conclusion IORQ) to include 
a resistor 470-560 ohms. After this signal output resistor 
connected to a 20-pin Z80 will be named IORQ, a signal to the 
opposite conclusion resistor - IORQGE. In addition to the 
peripheral devices must be implemented cut-off device on the 
wire IORQGE, when the device is not selected, and the filing 
logic 1 on the wire IORQGE, when the device is selected. 


   If we are to discuss the problems of Scorpio, it should be 
said about the following errors: incomplete decoding ports 
TRDOS - is not on the signal A1 (fixed in the board Turbo +), 
more stringent decoding ports AY, as a consequence - not 
running digital audio (fixed in the board Turbo +), the correct 
period of human clock (20 ms) and therefore , the signal INT - 
hence the slow run some programs written for the Pentagon, with 
its extended period of the signal INT. Interrupt handler does 
not have time to be completed by early next signal INT, and as 
a consequence, the program runs two times slower. This defect 
is also corrected in Turbo +, although this is wrong. Most 
Scorpio is a significant defect and lack of 100% compatibility 
with the company ZX-Spectrum on the timing diagram TV-scan, 
because of this program that paint any inscription on the curb, 
the Scorpio print correctly, these labels do not be. The 
elimination of this "defect" results in a significant 
complication and appreciation of the scheme, none of the common 
variants of Spectrum-compatible PCs is not implemented in 
Ckorpione in this regard will not change anything. 

OM> And yet, abbreviation of ZS, it is that - delusions of 
grandeur? Or you it wanted to emphasize that SCORPION ZS - it 
is not SPECTRUM ZX? I even saw somewhere about such a statement 
- "... Enjoyed the Spectrum and the scorpion ...", coming true,

not from you ...

ZS> In the unwinding of any commercial project in terms of 
output on the market are very important by many factors, 
including the positive experience of previous cases. Since my 
first development of Leningrad-1 was the most common model of a 
family of protozoa Spectrum-compatible computers (with all its 
deficiencies), then highlight the fact that Scorpion ZS 256 is 
its continuation, development, designed by the same author, I 
considered important. "Promotion" Scorpius began absolutely 
from scratch without any initial capital, the only thing I had 
- it's kind of no fame in the Spectrum environment, and do not 
use this fact would, in my view, wrong. 

OM> What models of the Spectrum you consider the most 
"dangerous" competitors SCORPION'u or at least worthy? 

ZS> "dangerous" - no, because the more models, the more
fans at the Spectrum in Russia, the more involved in this
the scope of people, the more the market itself Spectrum theme. 
A the appearance of some models exceeding the Scorpion, makes

think about eliminating its shortcomings and eliminate this 
advantage. In short - I want to say that healthy competition - 
engine of progress. Among the worthy competitors Ckorpiona 
should be noted of course "Profi", although it is very 
different computers. The tasks set by developers of these 
computers, as far as I know, were also different. In Profi 
focus placed on professional work in the CP / M, at a 
high-quality screen, etc. In Scorpio, most important to me at 
development scheme was simple, compatible co Spectrum-ohm focus 
primarily on system TRDOS (despite its numerous deficiencies) 
and, most importantly, the ability to continuously improve 
capabilities of the computer in general for through the 
development of software embedded in computer primarily in the 
informal service monitor. 

OM> From ZX-FORMATa Moscow learned about the existence of 
another perspective (at least, since it was presented) version 
Spectrum - KAY 256. What do you think about this? 

ZS> I find it difficult to talk about the prospects of this 
computer, because my opinion is very subjectively. But 
nevertheless, I did Say it all.



   Certainly "abstruse" philosophy of equity of "Nemo" may give 
the impression that KAY-256 is the height of design thought. 
Knowing how and by whom designed this computer, I can judge its 
capabilities and future prospects. Computer KAY-256 (the name 
comes from the name of Spectrum 128k + AY, when it added more 
128kBayt and got 256) was born in the throes of creativity 
Slava Skutinym (country should know his characters), after he 
worked for a year tuner cards in our company, and learned all 
the nuances of Scorpio. 


   Took from him all the best in terms of circuitry, namely: 
the scheme of distribution of port addresses, an organization 
of 256 Kbytes to IMS 565ru7, the scheme of quartz at 14 MHz for

the computer and the floppy drive controller, the location of 
the main chip (thus ensuring the same density wiring as in 
Scorpio), system connector, making the board for three size 
sockets (actually there are 2, because one is reserved for the 
controller FDD, missing on the board) and calling it not a 
system connector, a system bus, and changing only the schema TV 
scan, reducing the to two the number of shells used in this 
site, due to the use of more advanced ICs, it has become to 
promote your computer on the market, saying its incomparable 
advantages. 


   Although in my opinion this computer is the obvious 
shortcomings: lack of port FF, use to work with the printer 
port AY8910 (in principle, it is not so bad, but if you do not 
is AY, then there is no printer), the lack of on-board FDD, and 
the main drawback - lack of service-monitor which stops the 
further improvement of the capabilities of the computer without 
changing the circuitry. But it all becomes clear to the user 
KAY-256 only after he becomes them. Mr Skutin, advertising 
their computer for some reason did not say on the same 
page-ZX-Formata (other publications known to me) that he did 
not have enough money to split half the costs of developing the 
informal service monitor, and deliver it to your computer. In 
my opinion, to do so would be to more useful than writing 
treatises on community development and apofeticheskih methods 
turbirovanie processor, which few people understood. 


   Personally, I'm in several times to read some articles, but
did not understand. Pleading to those whose opinions I
respect as the opinion of experts, explain to me what it says. 
I understood that and they put it mildly, not everything is 
clear. 


OM> What are you most proud of what you think is ingenious idea
in your development?

Andrew MOA> I beg vstryat in the conversation, since talking 
about Sergei somehow immodest. At one time I really Bad scheme 
generator, power supply and processor and disk controller, and 
a large field of metallized holes on board for technical 
creativity. Convenient, in my opinion, and come up with 
nothing. 

OM> Let's talk about the shadow screen ... You saw the 
monitor-debugger STS? My opinion - STS order of magnitude 
steeper than TM as the interface, so and according to their 
functional abilities, in particular, it is much quickly trace. 
Why do not you try to "outrun" STS? 

AM> However, the view was not confirmed by specific examples 
... A Seriously, let's first understand who, what and why

needed.


  From the outset, an integrated debugger (and its volume is 
relatively all embedded software is now

somewhere around 20-30%) thought to debug their own
programs and to depart from this principle, we do not want to. 
Other case - the interface. Here I have to agree. The new user 
interface has already been invented and is compatible with the 
current, but All is not enough time.



   Oh, and everything else, why take away someone else's bread? 
If you professional capabilities and built-in debugger (by the 
way, bought with your computer or (as the count?), free annex), 
then you can buy the STS and use it. Who forbids? By the way, I 
plan to implement something like "remote" or the hardware 
debugging - it is a partial answer to the following question. 


   Specifically about the "fast lane" STS - all the time, to 
the debugger, again, in my opinion, you need some compiler with 
anything. After all, nice to debug their programs as source 
code ... 

ZS> I want to note also that moment that any changes, and
the more drastic (such as to make the Shadow Monitor)
in new ways, with a powerful user interface, steeper and STS
etc. - All of these changes require a significant investment of 
time and therefore money. So always have to make a choice: what 
is most important to date on what is necessary to work in the 
first place? Moreover, to rely not only the opinion of 
well-trained hackers, but also listen to the tens and hundreds 
of end users from different regions of our country. 


   So, on the basis of this integrated view, we believe in
Today our primary task as quickly as possible
progress of work on the controller HDD, for writing various
service programs, the shells (yes so much more you can think 
of). In my view the emergence of HDD, that works well in 
TR-DOS, iS-DOS and other systems will give new impetus to the 
development Spectrum in Russia.



   At present, work on the HDD in the process of preparation of 
detailed documentation on working with hard drive and a new 
version ProfPZU. This will take 1-2 weeks plus 2 weeks to pay 
the debts that have accumulated since the summer to sell 
motherboards Controllers HDD. And then - please, can be freely

order controller either with us or through Inforkom.


   Moreover, we have decided to change our marketing policy and
do everything possible to ensure that our Prof.PZU (from version
4.0) could work at all Scorpion-ah as the Turbo and non Turbo, 
a "left" and "right". 


   Will work Prof.PZU - can be used in HDD
TR-DOS, etc. In addition, we plan to give a Inforkom
Article on how to remake other Spectrum-s, so that at
They could work with TM ProfPZU. Again, these can be
connect the controller HDD. Controller itself is 
multifunctional. It is IBM slot that allows you to connect 
various IBM peripherals, such as the HAYES-modems, 
SoundBlaster-s etc. In addition to this socket is provided for 
setting the clock real-time, set IC NVRAM for storing custom 
settings on your computer, etc. 

OM> Shadow monitor, as you say - the best tool for
professional. In this case, I'm not a programmer, and
a hacker-cracker - to protect its program of TM is not easy, but
very simple. MAGIC BUTTON even more powerful means of hacking.
Fully protect yourself from MAGIC can only ever holding a stack 
of at the beginning of RAM, that not every maker on the forces, 
and some machines, particularly the Pentagon, it is easy to 
cope with this perversion. Do you, as I understood completely 
ruled out the possibility of using the function MAGIC. Do you 
have the desire to build ROM in a completely independent 
procedure "will dump on drive? Just do not say you respect the

Copyright and hack someone else's program is not good ... In 
anticipating that, firstly, if the defense does not break, they 
develop, and no one will (not want to look like so that the 
Spectrum poorly in this issue). And secondly, in my opinion, 
hacking and analysis of professional programs - the best 
teaching method for beginners. 

AM> The best tool for professionals - bright minds burdened 
with deep expertise in his field of interest. As these 
knowledge extracted? Question whose answer is ambiguous and can 
vary from dispositions and capacities of the individual. I 
think that now, when so much is available and special 
okolospetsialnoy literature when you can get access to the 
algorithms and source code, just anything, "hacking" gradually 
fades into the background - is the analysis. 


   Breaking the protection is probably necessary, but do not 
forget about "Direct" problem - the programming, but it will 
have nothing to defend. Not without reason, an increasing 
number of global companies refuse to "Direct" to protect their 
programs, shifting those functions to other parts of the 
project. By the way, we really respect the copyright. A small 
digression - usually, "neglect" lasts as long as the fact that 
most of these thefts Copyright has not touched you personally, 
then everything falls into place.


OM> close within the meaning of the question ... Why are you so 
persistently refuse the opportunity at any time to change the 
configuration of your computer? Simply put, why not put a 
button press which "turns" Scorpion ZS in the standard Spectrum 
ZX, benefit they are very similar ... It is no secret - many on 
their own, make such revision, which actually does not matter 
... 

ZS> First of all, we do not reject this possibility. In the new
developments and try to embody this idea. Just one of the
problems of any production is also a certain awkwardness, 
which, in turn, caused by the need ensure the stability of the 
company. In our plans for new models of the Scorpion to make an 
additional position in the menu to switch off additional ports 
without disabling the shadow of the monitor. For users of older 
versions of the computer, we plan to produce custom branded LSI 
installation where the user of his old Scorpio gets modern with 
all the new additional features. 

OM> I think the use of RST8 - not a good idea (it mildly) ... 
Prove the converse 

AM> better to offer an alternative. Actually, it is not our 
idea, and the idea of ​​Sir Sinclair. So complaining to him.

It is through the RST 8 calls the system resource
Intreface One. I would like to organize a full emulation of the 
device (at the level of file calls), but our hacker so 
assiduously set about adapting, that did not want to waste 
time. 


   We are aware of, and by the way, warn everyone that using 
RST 8 programmer "binds" its program to your computer. But, for 
professional programmers is not a problem - release various 
versions are not difficult (if not, then perhaps talking about 
professionalism is too early). For starters, the same users, it 
seems to me, it's not bad Support - something, such as custom 
libraries supplied to systems programming. 


   To date, under the RST 8 introduced several new challenges
concerning the management of computer equipment - HDD controller
eg. How to implement it on another? In addition,
expanded range of semiconductors to support operating systems.

OM> Scorpion ZS - one of the cars with extended memory. Why the 
same so humiliatingly little programs that use this advantage

(Adaptation of games under 256 KB, in my opinion - nonsense), 
do not you interested in this?


ZS> Knowing how in our country "valued" copyright software
ensuring we are not entertained (and still do not harbor) no 
illusions about the emergence of a large number of programs 
using 256 KB. The main purpose of the additional memory - meet 
the needs of the informal service monitor without using the 
main 128 KB. 

OM> Will there be a Scorpion 512 (> = 1024)

SZ> Work in this direction are underway. Since there is not 100%
certainty than they run, we did about it and do not tell. You 
can make anything, but how much will it cost for the end user 
and want whether it is available - here the main question. So, 
what we're working, will be of a board with Memory capacity up 
to 2.5 MB graphics 640 * 400/200, built-in timer and interrupt 
controller and other "twists". But perhaps none of this and 
will not be released because it is not known yet exactly how it 
is economically feasible. And, in our opinion, is far more 
important to support new developments of existing computers 
than develop new ones. 

OM> You already have implemented the hard drive on the 
Scorpion. Clyshal business for small - software support. Do you 
have this difficulty? If yes, contact via ON-Line with a call 
for help ... Specifically, tell me what you need and what you 
can offer for this work. 

SZ> You imagine wrong problem. In your opinion is it that we 
finally have a controller board HDD, and now how do I write 
under this board software. Everything is just different. 
Software support for HDD worked for over two years and only now 
have a working options. Had previously been tried in practice 
and several schemes and support MFM-HDD, ie those options, 
which only now are conversations, in particular on pages 
ZX-Revue. No us in this work, unfortunately, will not be able 
to help. Today we are like I have said, during the preparation 
of documentation, verify the functionality of our controller 
with different models of IDE-drives, operating time service 
program supports HDD, etc. Do not really want to go to market 
with the raw, undocumented, product, and without detailed 
documentation work will be virtually impossible. Who waited for 
two years, I hope wait a little longer.


AM> Unfortunately, "software" industry on the ZX-Spectrum
does not develop as we would like, and not as accustomed
operate the system and application programmers on other 
platforms. I could not get a decent compiler with assembler 
(when I started, and now the texts of the Shadow Monitor in any 
of the existing simply vlezut) was not a library program: all 
had (and have) written from scratch, starting with "independent 
poll keyboard "and ending with emulators and simulation of the 
system). 

OM> I have long ago there was a crazy idea - to realize the 
hard drive hardware ban on writing to disk. This significantly 
facilitate the fight against software viruses, which I'm sure 
there will be a proliferation of "dope." What do you think 
about this? 

SZ> Pure hardware - for IDE-HDD is impossible, or rather a very 
cumbersome, and software and hardware is quite real. Since our 
version of HDD controller goes through all the management of 
TM, then This problem is easily solved by adding the line in 
the menu on the monitor. To the extent necessary - time will 
tell, but certainly over the it is worth considering.


AM> And in general, noted that less talk and wait for viruses
so they do less ...

OM> Make advertising Scorpio briefly but impressively Explain
users why they need a Scorpion?

SZ> If we turn to those potential users who want to work in the 
Spectrum-compatible environment, then I told them I say the 
following, in my opinion the most important: you should opt for 
the Scorpio, if only because it is the only computer on today, 
which on the one hand standard released, and the other - 
constantly evolving and improving. This applies to both 
hardware (the introduction of Turbo mode, the increase in 
volume up to 512kB ROM, the emergence of new controllers: 
IBM-keyboard, mouse, HDD, and integrated software security: 
increasing the service functions built into the ROM computer). 
If you trace the unfolding of various Spectrum-compatible 
computers for the last 2-3 years, you do not can not agree with 
me. 

OM> If you do not mind, we conclude the topic Scorpion ZS. 
Unexpected Question: What do you think about electronic media 
on the Spectrum that you've seen? 

AM> In general - not bad, just a diva at times give much of 
their already there! Or divorced?


OM> Can you say anything about the network in Moscow ZXNET?

AM> Unfortunately, no. I know it only by hearsay, which
sometimes reach me at FIDO networks and the Internet.

OM> That's on a sad note, we shall conclude the interview ....