VTS vs Nemo - excerpts from Nemo's letters.

 ──────────────────── ────────────────────
 VTS & Nemo mail
 ──────────────────── ────────────────────

 VTS'99

   I picked up letters from '97 and '98, bo-
earlier than those published in the past
number.  I think there's something there too
useful.  And I gave letters from '99 to
give it to another person (like Nemo does
from Open Letters ;). However, with them (for now)
there was a bummer ;(


 Technical characteristics -
 - not the main thing. The main thing is to do
 and along the way, to improve.

 (c) Nemo

 Everyone is inventing lo-
 it's a stalemate, but no one wants to dig.

 (c) Nemo

 date: 10/15/97.

   ...

 4.  You need to upgrade in the Spectrumand not
modernization!  Upgraded computer
cannot be sold or repaired.
The owner becomes a hostage to his company
pa... So, gradual upgrade is being replaced
one-time for the price of a new computer, and in 90% of cases
tea - this new computer is IBM...

   ...

 7.  Any graphic extension sub-
implies replacing TV-set with a display, and this
already IBM again... Spectrum is exactly that
linen that is based on the TV standard, which
unshakable.   Sp-architecture is ideal for
TV-set, and this can be proven. Any other
the graphics mode will be worse (by
to the whole set of parameters),than stan-
dart Sp-screen. Pay special attention to
reference to the word "standard". It is quite possible
wife scenario for the development of events when things-
there are severalmonsters - and the standard of destruction
tightens, i.e.   script for "Babylon"
pandemonium" - confusion of languages (stan-
darts).

   ...

 

   The actual volume of Spectrum releases in
St. Petersburg - 300-400 pcs./year. Most likely
they will buy several pieces a month - but
and that's already good. If such points in Russia
These will be dozens - this will be the rise. I
I am sincerely convinced that concentrating
production (as Mr. Zonov does) - death
true, although it is beneficial for Mr. Zonov, but
not for Spectrumists.  Most profitable
participation of the largest number of people throughout
CIS with small but effectively operating
capital, rather than the concentration of capital
la and production at one point (one line-
tse).  In the first case, the development of interest
everyone is involved (in the development of the platform and accompanying
driving, because this will be the only way
increase capital turnover); in the second
In no case - only one person - Mr. Zonov
(and even then, insofar as... under di-
twist refers to the escalation of technical
characteristics, mitigation and support, and
platform, and as a result, discredit
Spectrum, i.e. in simple words,
production will turn into a bluff. When
everything will collapse, there will be no one to askand not with
whom.  Mr. Zonov himself made excellent money from this
seems to be selling IBM to former Spectrumists.
If he is left alone, he will simply put
the public before the fact - Spectrum died - and
there will be nothing to object to - he isthe only one
ny.)

   Regarding the production of GS, XTR and
etc.  These are independent producers, and
I cannot force them - if they accept
such a solution, then the technology(s) will be
transferred.  The fact that in St. Petersburg with such difficulty
managed to create at least some kind of community
manufacturers, relies primarily on
mutual respect for rights and freedoms as self-
cost units.
   There are also intermediate options, for example
measures, selling boards at a dealer discount.

   ...

 export of technologies to ZF#7>

   ...

   Doesn't really exist now
problems with escalation of technical characteristics
ristik, but there is a problem with the software
compatibility of different platform clones
Spectrum and software support most
"advanced" (or rather sophisticated) modes
lei. You can, for example, put 16 M "brain
gov", and immediately a crowd of their lamers come running
buy.  Who will be responsible when they
will understand the benefitsinstalled 16 M
exactly as much as from the same 16 M, le-
cravings in the sideboard, - i.e. that they were cheated?
There is a good saying - "the best is the enemy"
good."

   ...

 

 date: 12/18/97

 1.  According to is-dos, support is required for
places.  We are from here, from St. Petersburg, just not in
able to work with users, -
we need specialistsright on the spot.

 2. Who will finalize the IDE-drive circuit and how
- this is not important. The scheme has no
shield. In principle, it can be soldered using snot,
- is it only worth it? Produce peripherals
more profitable than motherboards - but
only ifsomeone is doing-
there are motherboards.If you are interested -
no production (most likely this is assembly
in place from ready-made empty boards) - please -
Lusta, do it!

3.  pcb. files are the final product
electronics development systems - package
PCAD (on IBM).  Having pcb. files and production
production of printed circuit boards, can be remade
car anytime, anywhere. Own
but KAY standards are going down the drain...
pcb file, on the one hand, is an array
data in PCAD, which is operable, i.e.
change and modify in PCAD; on the other hand
sides are technological information in
in the form of a photomask (electronic form),
schematic diagram and program for checking
leniya, adapted to a specific line
for the production of printed circuit boards.
   ...

 

   ...

 3. The standard is characterized, first of all,
go, mass use.  Then the price
software (per user)
is significantly reduced - this is the first thing. Second
swarm - to “squeeze” something into the framework of Spect-
rum is very difficult, almost always catastrophic
software compatibility drops physically
ability... Third, homemade products are very expensive
gi, - if they are made en masse, -
It’s easier to make a new board;  those. second
hand, upgrade - with a human face (in
differences from IBM). In this case, care
expansion of the platform is partly taken upon themselves
users - manufacturers this now
no longer possible...

  this means the following: buying
a new board, the user (himself) sells
old (at a slightly lower price), because
its capabilities may be quite sufficient
definitely a less wealthy person. When
this low-end Spectrum is becoming more
affordable and not thrown away
car wash, but continue to work for new owners
ev.  Those.  as a result, the total number
users is increasing.>

   I already wrote -make peripherals!Almost
all viable development concepts
Spectrum architectures  can be taken beyond the
well, i.e. they are not related to the basic
architecture.
   As for the temporary hut, it’s a system
synchronization, the skeleton of the computer, its microphone
roarchitecture.  To change it, the computer
literally needs to be mutilated. You can
Is it possible to sell it after this? And repair-
what to do?

 4.  About standards in general.  This is a utopia:
collect information, agree on stan-
dartah.  There were such attempts both in our country and in
abroad.  No good (or bad)
too) they didn’t bring it. Those. the result is - well -
left.   It is necessary to develop an ideology
Spectrum.  Those. quite general provisions
niy,which would be generally known, which
can then be detailed on site in
in the form of specific developments.  Those. impossible
poor management (dictat of the standard-
mi) is impossible if there is no agreement on general,
conceptual issues... And vice versa,
a unified idea of architecture in general
may be more effective. Those. from non-
mediocre dictate of standards for management
management of development trends. This allowed
makes it possible to concentrate rather than disperse the intensification
Leah  GOST is not possible on Spectrum - this is
centralized, from above, planting stan-
darts (in the USSR it worked, because there was no
hierarchy, it cannot exist in Spectrum).

 6.  TV-set is, first of all, a TV-stan-
dart.  So, Spectrum uses it on
all 100%. Those. this is the “butt” standard Peak-
the village should be square, exactly under
this item is "failed" video processor
Spectrum. If the pixel is standard, i.e.
square, then the video signal reaches its
th upper limit on bandwidth
TV-set;  neglecting this and making vi-
deo pulses are shorter, we get mo-
brightness dilation (instead of duration) and
phase distortion (i.e. distortion type
sticking together of points, i.e. distortion-micro). B
in the image of sprites (images) this is
naturally, and in texts this is very important.
So your eyes will water. TV-set used
children are necessary inliterally, to dress
to your ears. For IBM display, the distance from
image plane to the operator's pupil
- 50-70 cm, for TV-set it is 1-1.5 m.
   The square symbol is essential
for the screen as a graphic one. 32 familiar places
- this is 256 pixels, i.e.  ideal for
8-bit processor with 16-bit ad bus
res.  This is where the phenomena come from.
nal graphics capabilities (formerly
total dynamics) Spectrum. You can also
show that Spectrum was designed
specifically for TV-set, i.e. all architecture
subordinated to this idea.
   At the same time, it might be worth creating
video card (but on the bus) and a little experimentation
rement.   Find out, for example, how
for a long time a person can do without glasses,
reading 80 characters on third generation TV
niya... Or compare the time of reading one
of the same text, a kilobyte or so
200, on TV with 80 characters and with 40-50 characters
lami. I think the result will be sobering for many.

   ...

 9.  Now multicolour.  This is healthy
idea, but no standard. If multicolor
Pentagonists get carried away, then whatever you order
do?  - set 320 lines per frame
versus 312 lines (TV standard) - the same
insanity   The problem is in different ways
computer synchronization (INT, string length,
number of lines), and finally, port #FD, - real-
lysing this port in a way like thismade in the Pentagon, significantly reduces
port space (orsophisticatedde-
encryptor of this port with processing of all 16
bit). This, in turn, puts at risk
SPb-standard (port #1FFD) and makes no sense
tire, - well, what kind of tire if there is no
What are your free port addresses? Those. arch-
Pentagon clone texture isolated and not
capable of development. On the other hand, this
and not bad - they have a standard - how
They say, although it’s ugly, it’s uniform -
but... Hardly anyone is making Pentagons now
will - give guarantees on this "plate"
spaghetti" no one will. And if there is no ha-
guarantee, then there is no production - one possibility
fraud.

 step 7. Pentagon/Scorpion.
 Pentagons on the market 40% and Scorpios
40%.  The nice thing about GMX is that
these architectures aremutually exclusive.
In order to work in the Pentagon, not
it is necessary to disable Scorpion's architecture -
and vice versa. GMX assumestwo computers
yutera in one bottle...
   In KAY-1024 (v1.5) it is supposed to find
compromise. Then further development is possible
twisting within one architecture.

 The following is done in KAY-1024:
   - no wait 3.SMhz (normal) mode, 312
lines, 69888 t./INT.
   - 6 bits #7FFD untouched for use
introduction of shortened port addressing #7FFD and
#1FFD by #FD.>

   The trouble is that it's not 100% pentagon and
not 100% scorpion. Although the capabilities of multi-
The colors are the same, the bus extension is the same.
But old demos will be buggy. At first I
was worried about GMX, but then calmed down
rested.  It's justa lotion, and not new
comp.  Who will connect them, who will service them?
live?  where can they get displays?
Those. the lotion cannot bemassive.Cro-
Moreover, the reputation of Mr. Zonov, not without my
help ;-), gave a crack. And sell those
now bullshit will be more difficult for him... In general,
strategic direction of activity
Zonova is: 1).reserve
Spectrum channels and the monopoly on them (and in-
formation).  2).taking advantage of the monopoly, quiet
transfer users to IBM all at once -
this is much more profitable, here you can use
tvovat (unlike Spectrum) large
capital. Making fetishes in such a context
text is quite justified - preserved
image of the Spectrum company (advertising again). A
that this is bullshit, and the Spectrum is faster
bends, the more it is on the market - so
it’s just good for you: the more they’ll buy
IBM... "Spectrum-2000" is even less real,
than TRANSFORMER and Peters. Without a goal -
architecture in GMX, it is impossible to
you can create a new computer - this is a joy
I need the day.  And who will buy it -
those users who bought from Zonov
IBM?...

   ...

 clause 10. Pink Floyd.
 With Pink Floyd the situation is as simple as three
kopeks. Format editor on the eve of release
first version, PF explained to the authors that
some driver... In one sense or another, in
to one degree or another, Format's editors bear
responsibility to users for
your publications.  There is nothing to criticize yet
(what if something works out for them), so
how to praise...
   In general, it takes 2-3 to create an OS.
years of labor professional group prog-
rammists.  Therefore their claims seem
unreasonable to me.  For now, talk about
that the OS is being created, and even more so that
it  was created - prematurely. In addition,
in my opinionAccording to information, the system is loose, unstable
textured, which is called "without a king"
in the head." The principles on which
it is based...
   This means that the system will only be under Pe-
ters. What about mass appeal?

   ...

 

 date: 03/25/98

  this letter was preceded by mine
a message in which there was such a big
load about MPA - multi processors arc-
hitecture, inspired by GS-like things.
And he was justly burned
burning at the stake ;-) (and now also public
nomu ;-).>

   I got acquainted with your ideas. Special
they did not arouse enthusiasm, I explain in
what.

 1.  You amuse yourself with chimeras. This is very
pleasant thought-making, something reminiscent of
to nonsense. Mental frenzy. Then you
and given the opportunitypracticalaction
me so that you do real things
mi.  There are things that work, and then there are
things that don't work. In your mind you are not
you will understand a priori (a priori - lat.). Such
large-scale projects are beyond the capabilities of anyone, not
not to mention me, and even more so about you. This
titanic ((asuric - eastern;
demonic - west)antique titan - po-
Lubog) projects such as the transfer of Siberian
rivers or electrification of the entire country (auxiliary
line DneproGES). In fact,practically,
i.e. really, they turn around Potemkins-
such villages and mass psychoses.
You can't lose your sense of reality. Best
way and the best guarantee of seeing things that way
kimi, what they are - itpractical
work.  There are quite objective
injuries that can be seen and felt
only during practical work.

 2. Architecture can only develop
evolutionarily, i.e.  continuously.  Not now
there areproblems with increasing power
themes, architecture, gadgets, etc. etc.
The problem is in the users' brains - there
brake

 Example_1:  Scorpion (available since '92
years), KAY (produced since '94) have
25bkb of memory. Where is the math support?!! She's not there
!!!  (there are already more than 50% of such computers on the market)
Where are the programmers? They amuse themselves
measures and invent their own OS (which
are not needed).Real software productabsent. So were these the same 25bkb?
There are none! And you offer such a trick-
tit...
 Everyone is inventing shovels, but
nobody wants to dig...
 We need people who share small,
modest, practical things, and do not lie on
sofa and don’t indulge yourself with chimeras.

 Example_2: Bus. Do you think the tire appeared
was in '94? You are wrong. The first con-
conceptual developments were made back in
'92 as an alternative to the meaningless
Scorpion's system connector.   Company
(c)Nemo & SKIM then produced several
hundreds of cars 48k/month. I'm like a face, you-
performing the functions of the chief engineer (i.e.
marketing, public relations, strategy
development), had everything possible for that
moment of completeness of information andhad real
experience  in Spectrum. Now let's see what
it worked.
   92-9Chgg.  - development; 94-9bgg. - pub-
Lika is skeptical and giggles,
96-98  - the advantage of the tire is realized and
the market is partly demonopolized;  98 -
the public begins to make a chimera out of a tire. For
period 92-98 about 20-50 invested in the tiremillion rubles - zero return... You are capable
for such investments (which will never come to you
not to return) for such a long time -
new deadlines? Note that any re-
the result (long-term) will only be if
Are your ideas really viable?
Moreover, technical characteristics and especially
benignities do not play any role at all in
long-term time scale.
   This is the deepest misconception that Spect-
rumlacks technical characteristics
tick. On the contrary, they are already in it excess
accurate.Redundant in the sense that they are not
holding consumer value, i.e.
practical, custom filling
I eat.  Everyone sells opportunities, not products.
you are usable. This reminds
"MMM" shares.  Moreover, Spectrum is
form in which it is (if, of course, it is
mean a mass-produced home computer, and
not a field for technocratic experiments)
fully complies with domestic realities
liams of a global plan. Ifcomplicate it
("develop"), then the machine will be torn off -
noy from reality (reality). Mo-
The goal is that we need to develop the market,
platform - rather than architecture.  And here
This is something no one wants to do. This
unprestigious, and there will be no laurel wreaths
- the work is menial and thankless. And already
only as much as it was possible to
you can develop the market inappropriate way
develop architecture.

 Now a little about architectural development
tii. Let us have a set of parameters
architecture (i.e. speed, volume of pa-
memory, port allocation, etc.). This
vector, i.e. point corresponding to the con-
vintage architecture (with a set of specific
parameters).  Architecture change is
difference between two specific points in
space of architectures, i.e.
 _ _
 A(to) - A(to+t)

Every point in the space of architectures
(possible) corresponds to the volume of software: V.
Then the change in consumer satisfaction
tion: _        _
              V(Ato) - V(Ato+t)

Заметьте, что

  ┌ _     t   _
  │ A(to)───> A(to+t)
  <   _    t     _
  │ V(Ato)───> V(Ato+t)
  └
               _
т.е. изменение A, V происходит во времени,
т.е. t.

   Реальным  развитием будут такие измене-
ния A и V:
    _       _        _
   dA(dt) = A(to) -  A(to+t)
       _      _        _
   dV(dA) = V(Ato) - V(Ato+t) > 0 !!!

   Далее,  предположим, что существует та-
кая функция:

   u = f(V,t,быстродейс., объем памяти...)
            └────────────/─────────────┘
_
 A

 1. Zone formulation of the problem (Scorpi-
on). Find and implement dA such that when
dt->0 dV=const>0.

 2. Statement of the problem from (c)Nemo (KAY).
Find and implement dA such that when
dt=const dV=max.

 Comment:  1-no inputs at all-
to give money, but to receive it... The most expensive
go - transformation, market development is not
required.  The parasite's approach: tear off someone else's
device and...
   2-development itself.


 Now about software compatibility.

 P.P.S.  - direct software compatibility
bridge.
 Let there exist n E(belongs to) N
programs for Spectrum (any kind in general).
We will check the programs in the following way:
zom:
   - First, let's check on the proprietary Spect-
rum'e (standard) - if the program is running, then
p E N, p=p+1 (of course, pThen there is a limit:

i
 P.P.S. = lim - < 1
 p->oo p ^
 └─(now)

 T.P.S.  - transitive program syn-
capacity.
 We simply check any program from
n E N on clone (KAY, Scorp...). If it goes
then i=i+1. Then there is a limit:

i
 T.P.S. = lim - < P.P.S.< 1
 n->oo n ^ ^
 ├───────┘
 └─(now)

 Then the real development of architecture will be
children in caseexistence of such A that
T.P.S.> P.P.S.

   So, now there is no real development.
Still, the hardware bells and whistles are a bluff.

 

 date: 30-05-98

 1. KAY bus (Nemo-bus) is not designed
for the number of slots more than 4. Do more than 4
slots is to make a barn for idiots.
A lot of things come together here. And speed-
yours, and design, and price, and practicality
probability, probability of successful docking, possibility
possibility of standardization, etc.
 Output: 1).  make multimaps (stan-
perfect size!).
 2).  Enable Cent peripheral port
ronics: 70% of devices have nothing to do on the wireless
no. In this inclusion it works as a "system"
dark connector"

 2.
 ...
   Specifications are noodles
on ears for USERS. There are no technical
obstacles to making a shed
(monster). Look at Peters - I'm selling
there are more boards and computers than them. Because,
what I do works. That's not the point -
le that does not burn out when turned on and does not
buggy - ahreally works, i.e. full
not consumer meaning, - they are also de-
devices bark into the sideboard (and after a month or two
- on the mezzanine, six months later - in the trash -
ku...), sort of “pink elephants”. Why do they
do they do this?  Are they fools or what? Not really. Im
advertising is needed, just like Zonov, - for
PC sales. What is this like?
In fact, they are not interested.

   ...

 

               ***