17 декабря 2015

        ZX Spectrum demoscene: DiHalt Lite 2016 thoughts
                        by Alone Coder

Sadly this DiHalt Lite hasn't become "united and main" Spectrum
demoparty despite the hopes. A year ago they united with Chaos
Constructions that had its own circle of authors, but this 
winter it was skipped. Because of huge number of Spectrum
democompos in 2015 the authors couldn't accumulate material for
serious releases. As a result, the first places were taken not
by Spectrum works but by a BK-0011 demo and a C-64 picture.

An idea emerged from somewhere that "to make a cow eat less and
give more milk, feed it less and milk it more". This resulted in
multiplying parties and falling the average quality of prods
below the last place at Enlight'96.

It seems that unless the number of demoparties returns to norm,
the queue of poor works, discrediting ZX Spectrum, won't stop.
Strong works will be rare and won't compete at a single
demoparty.

Demoparties today repeat each other, even the voting is
frequently made with one "zapilyator". And each of them doesn't
want to close. The best compromise for Spectrum in this
situation would be transforming one of them into "Spectrum only,
any Spectrum, demo only, realtime only" party, maybe changing
the "show" format to "seminar" format.

The first point - "Spectrum only" - will make the party more
aimed (something like X for C-64, where the strongest works
appear). Mixing all the 8-bit computers together might look
fashionable - but there are a lot parties like this, including
the biggest European ones.
Attempts to dilute ZX Spectrum prods with emulator stuff also
gave nothing to the platform. Although the desire of some
organizers is understandable, to snatch a great deal of works
for their one party of dozen in existence - works of any
quality, and to make friends even with devil for that.

The second point - "any Spectrum" - will improve originality in
works. Historically inauthentic limitation "128К or Pentagon
128+AY", forced by certain people in some of recent parties,
only made the situation worse.

The inadequacy of this limitation (it is not self-generated - 
its promoters are known by name) can be seen at least because: 
- there are 6 years between these configurations, different 
generations of data medium, and Pentagon 128+AY is not a 
definite machine but a home-made thing with arbitrary filling, 
including at least one add-on; 
- neither of these configurations was a "top one" (Timex 
Sinclair 2068 and its CP/M compatible disk system appeared 
before the light Spectrum 128K - extra memory was initially a 
virtual disk; they were released in different countries; as for 
AY interface for Pentagon 128, there are no traces of it before 
the appearance of Profi); 
- circulation of all the 128K models combined comprises a tiny 
part of all Spectrums (less than 48K by order of magnitude, down 
to zero in some countries); 
- even in circles where 48K were excluded (ZX-Net nodes), the 
share of 128K models according to polls never exceeded 74% - 
even locally and even ignoring the add-ons - the maximum was in 
1997 with quick decline in favour of >=256K by 1999: 
Moscow (ZX-Net), April 1996 - 79 (61%) 128K, 39 (30%) >=256K 
http://www.zxpress.ru/article.php?id=2073 
mostly Moscow (ZX-Net), May 1997 - 49 (74%) 128K, 17 (26%) 
>=256K http://www.zxpress.ru/article.php?id=1322 
Saint-Petersburg (ZX-Format purchasers), July 1997 - 31% 128K, 
49% >=256K http://www.zxpress.ru/article.php?id=9885 
ZX.SPECTRUM (FidoNet), July 1997 - 9 (25%) 128K, 24 (67%) >=256K 
http://www.zxpress.ru/article.php?id=4745
Gomel, December 1997 - 21 (70%) 128K, 6 (20%) >=256K 
http://www.zxpress.ru/article.php?id=3258
Krasnodar, June 1998 - 9 (64%) 128K, 2 (14%) >=256K 
http://www.zxpress.ru/article.php?id=9119
ZX-Net, January 1999 - 42% 128K, 48% >=256K 
http://www.zxpress.ru/article.php?id=7181
Omsk, January 1999 - 3 (25%) 128K, 6 (50%) >=512K 
http://www.zxpress.ru/article.php?id=11366
CIS or xUSSR, July 2000 - 9 (15%) 128K, 52 (85%) >=256K 
http://www.zxpress.ru/article.php?id=4417
Rostov-on-Don, July 2001 - 50% 128K, the rest >=256K (with wrong 
sum) 
http://www.zxpress.ru/article.php?id=7322
Izhevsk, July 2001 - 100% >=512K 
http://www.zxpress.ru/article.php?id=9429
- the share of 128K only software is not big on the whole. As 
for demos, 128K demos prevail no earlier than August 1998 (this 
is easily traced in Chip ZX collection, with software sorted by 
date of arrival and with labelled hardware requirements in the 
catalog). Moreover, since Demobit'1996, Doxycon 1998, 
ArtComp'98, Paradox'99, Cafe'999, and CC'999, compo machines 
included add-ons, unfortunately different ones; 
- despite the compatibility and circulation problems, different 
authors wrote up to one thousand software titles for various 
devices above this limitation (it's hard to say "extra devices", 
as 128К was "extra memory" in our youth, too). This includes 
about 200 demos and intros. 

The returning interest to Timex and Profi is thereupon very
promising. As well as the long-going movement at Pentagon 1024 
and ATM Turbo that are far from exhausting the potentialities.
Today, unlike the 90's, there is no problem to run any software.

All this hardware is known, so the audience won't have troubles 
in judging how efficiently it's used. For example, everyone 
understands that a demo for the original Speccy has 3 times less 
of available RAM, as compared to 128K, and the only screen area 
(so there are beam problems and limitations on multicolor 
types). If we consider higher class in the form of Pentagon 
128+AY, we take into account its fast memory (which means speed 
up and tempting ease of sync), high border resolution, and the 
disk drive from where you can load in background and almost 
instantly. It's obvious that demos like Eye Ache 2, Refresh, or 
Stellar Contour can't run from tape - with loading parts three 
times as big as in 48K demos where they are already long. Demos 
supporting turbo mode allow smoother 3D (maybe even ideal for 
those Profi users who soldered 20 MHz, and this is hardly the 
limit) but they are limited in some of coding techniques, 
multicolors, digital sounds, and border effects. On the other 
hand, demos for Timex, upgraded Pentagons, ATM, Profi etc can 
avoid multicolor by using hardware videomodes. General Sound 
returns digital sound for demos supporting turbo mode, and also 
saves RAM and t-states for music, with richer sound. And so on. 
The only question here is about inclusion of SAM Coupe' that had 
separate community for a long time. Of the old sceners, Y... 
supports this inclusion. Possible arguments, apart from current 
decoding of SAM as "Spectrum Advanced Machine", might be seeming 
succession to Timex (so unusual frequencies aren't strange, 
especially as Pentagon 48/128 also has unusual frequencies and 
even lacks AY, too), and the memory mapping blows memories 
about very old 80K extension  to stock 48K with two 32K banks. 

The third point - "demo only" - will concentrate the authors'
efforts. This generally applies to graphicians and composers
that usually have no time for demo because they need solo
releases at every party :)

The fourth point - "realtime only" - is intended to return the
former quality of demos (based on definition "Realtime demo" at
European PC demoparties or on formula "pure animation is
forbidden" in rules of old Russian demoparties), after many
years of degradation.

One notable observation is that degradation is glorified as a 
form of art by the same circles that support multiplying 
parties, castration of ZX Spectrum, and hooliganism on scene. 
Maybe they don't so want bad for the platform (they might never 
mind about its future), if they are defending their mode of 
life. Anyway the sequences are seen - Russian-language 
IRC-channels turned into streams of mud, zx.pk.ru forum became 
desolate after mass attacks and trolling, somebody hesitates to 
release the already written software (S...), and some people 
even decide to quit the Spectrum scene with wording like "it's 
no more interesting to me" (N...) and "it became unpleasant to 
communicate" (S...). However the heart of the scene in its best 
years was not between internet loudmouths, but in more moderate 
private correspondence and informational press.