Question for Novag Obsidian owners

This forum is for general discussions and questions, including Collectors Corner and anything to do with Computer chess.

Moderators: Harvey Williamson, Steve B, Watchman

Forum rules
This textbox is used to restore diagrams posted with the fen tag before the upgrade.
User avatar
Monsieur Plastique
Senior Member
Posts: 1014
Joined: Thu Jul 03, 2008 9:53 am
Location: On top of a hill in eastern Australia

Post by Monsieur Plastique »

LWSteve wrote: Citrine and Obsidian, their playing style is completely different than the following Dave Kittinger programmed computers... (all of which I own)

Diamond
Diamond II
Star Diamond
Sapphire
Star Sapphire
Super Constellation
Super Forte 'C'
Super Forte 'B'
Constellation Forte

Just on another point, why would anyone expect an Obsidian or Citrine to necessarily play like any of those machines on your above list anyway? The Super Constellation, Constellation Forte, Super Forte, etc are much, much older programs that were written to work on much earlier hardware. Programmers will change the way their programs play based on long experience and playing strength tests, but will also adapt programs based upon new opportunities offered by newer hardware.

As for the Diamonds and Sapphires, again, I would not necessarily expect them to play the same way either. They were the top of the line programs that used far superior hardware and they also had hash tables. The programs were much bigger and priced accordingly relative to the 32K range.

The Citrine and the Obsidian are effectively only midrange Novag programs and developed from the first H8 based 32K machines (i.e Ruby, Zircon, Jade, etc). They did not evolve from the contemporary premium line of machines, nor the machines from the 80s or machines that pre-dated RISC processors.

If you were able to conclusively illustrate that, for example, Obsidian and Citrine play nothing at all like Ambers and Emerald Classic Plus once out of book, for example, then you might have a point to argue, but you are going to find that the machines produced after David left (i.e Star Ruby, Obsidian, etc) have a lot in common with the last generation of 32K machines produced whilst he was contracted to Novag.

And as mentioned before, think of the Citrine as a 32K machine, but needing more ROM because of the very large book. But if a Citrine, Amber and Emerald Classic Plus all played exactly the same opening, I would not be able to tell you blind which machine was necessarily which over the course of the game.

The Novag playing style has evolved over the years and in the 80s it was a rather aggressive, tactical and speculative style (similar to how the current 16K programs play). The 32K machines these days tend to be (in my opinion) more solid, assured and technical in relative terms.

Finally, I am not really sure why anyone would necessarily be upset that their Citrine does not play the way those old machines played. Firstly because a lot of people still own the older machines as well as the new ones anyway, and secondly because a Citrine has a higher rating than those old machines.

What I do find upsetting about the Citrine (and Obsidian) are the extremely poor opening books which are full of holes and miss fundamental lines and transpositions. This did not happen in the good old days when the books were done by pasionate experts, so I certainly agree the latest books may well have been written by the person who vacuum cleaned the office in the wee small hours.
Chess is like painting the Mona Lisa whilst walking through a minefield.
User avatar
Monsieur Plastique
Senior Member
Posts: 1014
Joined: Thu Jul 03, 2008 9:53 am
Location: On top of a hill in eastern Australia

Post by Monsieur Plastique »

To anyone still interested in the original topic, I thought of a novel answer today in relation to the Obsidian opening book.

I would get out my Star Opal and start the game against that machine until the first move out of book. I would then transfer the previously-played book moves to the Obsidian in Player versus player mode. The Obsidian would then take over the game from there.

I know it is ugly, but I would have loved a Novag computer with the Obsidian program and the Star Opal opening book (still one of the best books out there, despite having found a number of "buggy" lines where someone must have had finger trouble transcribing them).
Chess is like painting the Mona Lisa whilst walking through a minefield.
LWSteve
Member
Posts: 431
Joined: Sun Jan 06, 2013 9:26 pm
Location: WA USA

Post by LWSteve »

Star Ruby, Obsidian, Citrine and there are others...

These were all produced after Dave left the building.

These are what we call "Mongrel Mutt Novags"!

They may have some Kittinger source code in them but they are not 100% Kittinger!

As far as Citrine goes...

MP says that Citrine's opening library is all Ft' up! (I believe him)

When playing white Citrine will not vary its moves against my favorite openings and it always plays the same lines against me.
So in order to get Citrine to play variety against me when it has the black
pieces I have to constantly use "take back" on Citrine's black pieces and substitute a move of near equal value.

Does anybody else have issues with Citrine?

Post them now or forever hold your peace.

All they had to do after Dave left the building was to take a Super Forte 'C' circuit board
running at 6 MHz.... Jack it up to 30 MHz and then stick it in a Citrine box.

You would then have had one hell of a chess computer!

BUT NO! These programmers that replaced Dave Kittinger... they all had the IQ of a fence post!

LW
User avatar
IA
Member
Posts: 433
Joined: Wed Sep 10, 2008 1:12 pm
Location: Spain

Post by IA »

Interesting discussion , from my point of view last machines Novag 32K try to simulate the playing strength of computer programs , for example Novag Obsidian and Novag Citrine and play almost like the lower layers of Fritz 9 which is one of the best programs of Fritz with more aggressive play , tactical and speculative , this makes having a very solid game play , on the other hand Novag Star Ruby has the same program as the above mentioned machines but is a bit more speculative in their game and is closer to the style of play Hiarcs , are programs that Kittinger and I have no doubt but it is true that they have adapted the same 32K program to different machines , must have adjusted the engine trying to game styled slightly differently depending on machine segment and market focus .

From my point of view the engine more accomplished and better play chess 32K is the Star Ruby Novag to be highly speculative and aggressive, is a very humane and competitive engine , the Novag Citrine is very solid in the game but you lack the dynamism of the game that has the Star Ruby, the Obsidian Novag this midway of Novag Novag Star Citrine and Ruby but also has a very solid gameplay and tactical .

The programs 32K Risc of Dave Kittinger are very adaptable and with small internal adjustments seem to play a very different chess is a chess engine very chameleon.


Short so that people understand it better .



The Novag Obsidian and Novag Citrine play as Fritz 9 a game solid and very tactical and personally are two machines play a very academic style and is perfect for play and train , very good machines .

The Novag Star Ruby plays like Hiarcs 9, a very tactical gameplay and aggressive in the attacks but with a style of play more positional and better in the late game. , For me is the machine that best simulates pocket is style game Human tournament players .

The Novag Turquoise, Emerald Classic Plus .... They play like Deep Junior 9 with a very aggressive style of play and speculative.

With each layer of depth seems to play different machines , this can be seen very clearly if we compare ls Novag Novag Obsidian and Star Ruby playing almost the same but the small difference in depth of game Novag Star Ruby makes it a adversary highly speculative , if we think that Obsidian Novag longer Novag plays like Star Ruby .

All programs are the same Dave Kittinger 32K but have been adjusted to the different processors and the prevailing style of play at all times , for example in the years 90 was more prevalent tactical play romantic and machines came here as strong as Novag Emerald Classic .... After the strong presence of the PCs in the years 2000 playing style to be more tactical step , solid and this is where the heavy machines appear Novag Obsidian .... They play more like PC programs but with obvious restrictions have 32K programs .

Regards
User avatar
Theo
Member
Posts: 132
Joined: Tue Mar 05, 2013 11:34 am

Post by Theo »

LWSteve wrote: All they had to do after Dave left the building was to take a Super Forte 'C' circuit board
running at 6 MHz.... Jack it up to 30 MHz and then stick it in a Citrine box.

LW
C'mon, Star Diamond is also not that bad. Pity they didn't put that program into the Citrine.

The cranked up Super Expert C exists, also a cranked up Diablo 68000 (with hashtables!) ---> take a look at the "Revelation"

Kind Regards,
Theo
User avatar
Steve B
Site Admin
Posts: 10140
Joined: Sun Jul 29, 2007 10:02 am
Location: New York City USofA
Contact:

Post by Steve B »

Kittinger left Novag about 10 years ago

The Forum contains A Wealth Of Information
we had a discussion about this here 4 years ago
we came to the conclusion then that Kittingers engine is still in all of Novags computers although probably somewhat modified

here is the discussion which contains a link to an interview that Bryan had with Dave Kittinger in late 2008 which sparked our discussion
sadly the link to Bryans interview with Dave Kittinger no longer works

http://hiarcs.net/forums/viewtopic.php? ... 62d0d9f989

here is a different link to the Interview

http://kittinger.yolasite.com/

Deja Vu All Over Again Regards
Steve
IanO
Member
Posts: 162
Joined: Sat Aug 15, 2009 11:43 pm
Location: Portland, OR

Post by IanO »

Monsieur Plastique wrote:Programmers in this sort of class - as history now tells us, are very few and far between. The programmers in the early days (of which Kittinger is one) had to have considerable expertise in working with marginal hardware, severe space limitations and needed to be able to work directly with primitive I/O interfaces. You just don't throw out those years of expertise when someone packs their bags, because the chances of economically improving on the base product they provided to you are next to nil. This is why you can still buy a Morsch program in a Mad Catz plastic toy 20 years after it was written.
And more's the pity! Because it is not worth new development, the dedicated computer market is like a time-capsule of the 1990s, both in terms of opening trends and the techniques used. Correct me if I'm wrong, but I don't think any dedicated computer ever sold even uses the null-move heuristic, not to mention late-move-reductions, razoring, automated tuning, and the host of other techniques discovered since 1993. Heck, they even dropped hash-tables in this age of super-cheap memory and 32-bit embedded cpus!

I've considered sponsoring an embedded chess computer competition, designed to run on the type of $5 controller you would choose to use in a modern clean-sheet chess computer design. Something too small for Linux, so that you couldn't just copy-paste from existing engines. An Arduino board, for example. There are even cheap ARM kits, with enough SRAM for a small hash table. It would be interesting to see how much stronger such engines would be compared to the dedicated pinnacles of 1993.

Ian
User avatar
Steve B
Site Admin
Posts: 10140
Joined: Sun Jul 29, 2007 10:02 am
Location: New York City USofA
Contact:

Post by Steve B »

IanO wrote:
Correct me if I'm wrong, but I don't think any dedicated computer ever sold even uses the null-move heuristic, not to mention late-move-reductions, razoring, automated tuning, and the host of other techniques discovered since 1993.
Ian
Actually you are wrong
Dedicated chess computers have become very sophisticated
one has only to look at the Phoenix line of dedicated chess computers
which run many modern PC engines including Shredder and Hiarcs...
(in addition to running emulations of the old classic engines)
of course most of the computers discussed here are the older dedicated chess computers but that does not mean that more up to date models dont exist

Not Exactly Regards
Steve
IanO
Member
Posts: 162
Joined: Sat Aug 15, 2009 11:43 pm
Location: Portland, OR

Post by IanO »

Steve B wrote:
IanO wrote:
Correct me if I'm wrong, but I don't think any dedicated computer ever sold even uses the null-move heuristic, not to mention late-move-reductions, razoring, automated tuning, and the host of other techniques discovered since 1993.
Ian
Actually you are wrong
Dedicated chess computers have become very sophisticated
one has only to look at the Phoenix line of dedicated chess computers
which run many modern PC engines including Shredder and Hiarcs...
(in addition to running emulations of the old classic engines)
of course most of the computers discussed here are the older dedicated chess computers but that does not mean that more up to date models dont exist

Not Exactly Regards
Steve
That is the only one. And only by being very commodious (800mhz, 16mb) so that engines designed for PCs may be run.
User avatar
Monsieur Plastique
Senior Member
Posts: 1014
Joined: Thu Jul 03, 2008 9:53 am
Location: On top of a hill in eastern Australia

Post by Monsieur Plastique »

IanO wrote:I've considered sponsoring an embedded chess computer competition, designed to run on the type of $5 controller you would choose to use in a modern clean-sheet chess computer design. Something too small for Linux, so that you couldn't just copy-paste from existing engines. An Arduino board, for example. There are even cheap ARM kits, with enough SRAM for a small hash table. It would be interesting to see how much stronger such engines would be compared to the dedicated pinnacles of 1993.

Ian
Well, if I were made of money, I have often fantasised about getting the rights to the original source code for numerous "landmark" Novag machines of the past, hiring some gun software / hardware engineers and bringing back the Novag brand name. But this time round, it would actually mean something unlike what the brand has represented over the last 15 years.

First up, I would ditch all these silly 4K / 16K programs, because quite frankly, they are complete and utter crap. All the fixed and variable levels are as buggy as heck and the only blunder free levels are the ply and extended ply levels. If you could see the way I can absolutely decimate these 16K programs - even on a 6 ply level that can take over 5 hours for 40 moves - in a way that makes me look like a GM and them like a beginner - you would surely agree.

Instead, my entry level Novag model would be based on the "proper" 16K program - Primo / VIP, but upgraded hardware. I would then have two mid-range models - one based on the very best of the "old" hardware - so one of the Super Forte thingies whatever and another more budget version based on the earlier 32K programs - probably Emerald Classic Plus. There would be no Citrine - I would just bring out a new wood and portable version of the Sapphire II, again updated as I saw fit.

All models would have custom opening books lovingly crafted, thoroughly checked and fuly debugged using extremely strong chess engines for sanity checking.

Whilst I have to disagree with LWSteve about the Kittinger thing, I do tend to feel that the latest Novag machines (that are nevertheless fairly strong), seem to play a more sterile game for whatever reason as compared to classic Novags. I recently played the Citrine and was quite astonished (apart from getting it out of book by move 2) to find it was more interested in closing the position and simplifying. And of course, after it did that, it did not have a clue what to do and no hope whatsoever of scoring a win.

OK, here endeth the dream.

But wait..there is more...proof of what a joke the so-called "ELO 1900" 16K programs are that have dominated the Novag midrange landscape for the best part of two decades now. Here is a game I played yesterday when I gave the darned thing 6 whole hours to play 40 moves, compared to my "walk by every so often, stare at the board for minute and make a move" approach. Except, I have to admit, the bishop sac on move 20 where I did have to take about 20 minutes to go through all the variations in my head (apart from anything else, I really had to be certain it was a forced mate). But in any case, my clock was set to 40 in 2 and I had only used about 25 minutes up to that point.

These programs have been sold as "Elo 1880" and later "Elo 1900" programs for two decades or so now. In my opinion, they play around the high 1300s Elo - low 1400s Elo at the absolute best, being generous, very strong tailwind, operator half drunk.

I fear I am going to go to my grave clueless as to what motivated Novag to ditch an excellent 16K program already running on cheap hardware for a buggy and extremely weak low level hobby program, leaving as it did a massive gap in their range between the 16K models and the next tier upwards that as remained to this very day. I would never be able to beat those 16K programs like this.


Event "Lvl BE6 - 6 ply - around 240 minutes for 40 moves"]
[Site "?"]
[Date "2013.09.05"]
[Round "?"]
[White "Parle, Jonathan"]
[Black "Novag Star Opal"]
[Result "1-0"]
[ECO "B15"]
[PlyCount "55"]

1. e4 c6 2. d4 d5 3. Nc3 b5 4. a3 dxe4 5. Nxe4 Nf6 6. Nxf6+ exf6 7. Bf4 Be7 8.
Nf3 O-O 9. Bd3 g5 10. Be3 Re8 11. c3 Bg4 12. Qc2 h6 13. O-O-O Qd5 14. h4 Bxf3
15. gxf3 Qxf3 16. Be4 Qg4 17. Rdg1 Qd7 18. hxg5 fxg5 19. Rxh6 Kf8 20. Bxc6 Nxc6
21. Rh8+ Kg7 22. Qh7+ Kf6 23. Bxg5+ Ke6 24. Qe4+ Kd6 25. Bf4+ Ne5 26. Bxe5+ Ke6
27. Bf4+ Kf6 28. Qe5# 1-0
Chess is like painting the Mona Lisa whilst walking through a minefield.
Reinfeld
Member
Posts: 486
Joined: Thu Feb 17, 2011 3:54 am
Location: Tacoma, WA

Post by Reinfeld »

Resurrecting this question from a year back, because I'm interested:

Earlier in the thread, Monsieur Plastique said this:
The Citrine and the Obsidian are effectively only midrange Novag programs and developed from the first H8 based 32K machines (i.e Ruby, Zircon, Jade, etc). They did not evolve from the contemporary premium line of machines, nor the machines from the 80s or machines that pre-dated RISC processors.
Comment: Obsidian boasts of its RISC processing right on the housing.

MP also said this, which is absolutely right:
What I do find upsetting about the Citrine (and Obsidian) are the extremely poor opening books which are full of holes and miss fundamental lines and transpositions. This did not happen in the good old days when the books were done by pasionate experts, so I certainly agree the latest books may well have been written by the person who vacuum cleaned the office in the wee small hours.


Well, yes. The two machines have classy housings and lousy opening books. Obsidian is my travel choice, because I like the extra touch of the wooden pieces, the option to play with a friend at the beach. Yeah, it's cheesy, but look at the competition. Fidelitys are OK, I suppose. I actively dislike the button configurations, generic pieces and touch aspects of the Mephisto/Saitek line, right up to the current models (exception: the Explorer design.)

I'll throw this point to the mercy of the forum, but it seems to me that Novag dominates the design aspect of dedicateds, and it isn't close. Their machines are the most fun to USE, to carry around.

That said, if you've got the best housing figured out, why skimp on the features? I would LOVE an Obsidian board with Star Diamond strength and features. I would buy it in an instant.

Fernando said this in a related thread five years back:

Source: http://hiarcs.net/forums/viewtopic.php?p=28016
As we have said many times: strenght is not the issue they care, but housng and frills to hipnotyze average customer.
- R.
"You have, let us say, a promising politician, a rising artist that you wish to destroy. Dagger or bomb are archaic and unreliable - but teach him, inoculate him with chess."
– H.G. Wells
User avatar
Fernando
Admiral of the Fleet
Posts: 3059
Joined: Tue Jul 31, 2007 4:35 pm
Location: Santiago de Chile

Post by Fernando »

Monsieur Plastique wrote:
LWSteve wrote:When Dave left the building he took his ball and bat with him and his source code also!
As someone who was a programmer for a living and worked around people who were contracted programmers and IT professionals, if that is the case (he took his work with him) then that the most unique and bizarre commercial arrangement I have ever heard of.

I have never, ever heard of any software-related contract where the company a given contractor works for does not have rights to the source code during the contracting period as well as after the contracting period has ended. It is a completely unworkable commercial arrangement, which is why it does not happen in my experience.

If this was the way the commercial IT world operated, then companies would fall into a heap left right and centre and the world would literally plunge into chaos. At worst, rights would still remain with the company but the programmer may have a shared right in certain situations and / or after a certain period of time expires. I think this sort of arrangement might have been the case with Richard Lang, for example, since he has (so I understand) licenced source code for the Revelation machines.

The whole point behind being a programmer is that you are providing an intellectual service to a company. The company is paying for that service. If after you leave, the company has to cease using every bit of intellectual property you created for them, then they have gotten nothing in return from you.

Kittinger's programs were developed over many years and they are the product of a vast amount of experience, ingenuity and sheer hard work. I do not believe for even one moment that Novag said to themselves, "let's start from afresh now Dave has moved on". Programmers in this sort of class - as history now tells us, are very few and far between. The programmers in the early days (of which Kittinger is one) had to have considerable expertise in working with marginal hardware, severe space limitations and needed to be able to work directly with primitive I/O interfaces. You just don't throw out those years of expertise when someone packs their bags, because the chances of economically improving on the base product they provided to you are next to nil. This is why you can still buy a Morsch program in a Mad Catz plastic toy 20 years after it was written.

So far as playing style is concerned, you do not need to make many changes to the source code to change the style of play of a chess program. Some very small changes can dramatically change the way a program plays, yet you might be changing the values in a few tables here and there for example.

I agree that the latest Novags do not play like the old ones, but in my experience, none of the changes in playing style are evidence that a completely new program was written, because there is a much simpler (and cheaper) reason that instead explains it.

Perhaps we all could follow a line of thought that accept that K program were still used as a base for the reasons of company interest, etc, but we also can agree that a "style change" is in fact not other thing that a program change due to some alterations, bugs. new hardware etc, so it can be also said that it is a different program. If something play differently, then is another program as a result of anything done to the original stuff.

Fern
Festina Lente
User avatar
fourthirty
Full Member
Posts: 763
Joined: Fri Dec 06, 2013 8:46 pm
Location: San Francisco

Post by fourthirty »

Reinfeld wrote:Well, yes. The two machines have classy housings and lousy opening books. Obsidian is my travel choice, because I like the extra touch of the wooden pieces, the option to play with a friend at the beach.
- R.
Reinfeld - I agree 100% I really enjoy the design and gameplay, AFTER the opening sequence. I know this has been discussed before on this forum, but it certainly appears the Obsidian has no variety in it's opening play.

Greg
User avatar
scandien
Member
Posts: 206
Joined: Mon Sep 12, 2011 1:15 pm
Contact:

Post by scandien »

hello,

i have resurrect this thread, because i have just finish the Khmelnitsky test for Novag Obsidian and Novag SCORPIO.
I am using this set to check the playing profile of the machine.
I think the results are interesting, and shows clearly that those machine are related ( even if different ) each other.


here are the comparison of the profiles.
Image

and my comment below:

SCORPIO 68000

The Scopio/Diablo was a very strong and well balanced machine. It is as good in endgame than in middlegame. Surprisingly the SCORPIO doesn't play in counter attack a most other program, but it has a good skill in attack. It is able to identify an opponent weakness and to mobilize his piece to reach a far objective. This is, in my view, his main strength. His tactical skill is good but it's positional play is at the same level ( strategy skill). Once again this is quite rare , because machine tend to rely on their tactical skill.
The Scorpio can consider very agressive/active move , and is really a tough/untertaining opponent for human ( the resuts versus human are very good as indicated in DIAMOND's chapter).
The Scorpio is really a unusal machine.




DiamondII :


The Diamond is a tremendous opponent, as indicated by his result and level of play. It is rather balanced , but his strength rely mostly on his tactical skills ( threat detection, calculation). If it plays quite well the endgame , it is not so efficient with theorical endgame. The Strategy ( and positional play) is his weakness, and it can weaken his position by following tactical purpose.
If it is strong in counter attack , it is rather weaker in attack , and it can be in trouble if it has to identify a weakness in his opponent's position, and if it has to mobilize his piece to attack this objective. This weakness is related to his strategic skill.
The profile indicates that , if Diamond is related to the DIABLO/SCORPIO, Kittinger have neglected the endgame and strategic skills, to concentrate on tactical ones. this seems quite efficient for games versus machines as the DIAMOND is better than DIABLO in those encounter.(2091 for DiamondII versus 2010 for SCORPIO*). This is less efficient in Man-Machine game where the too machine seems to be nearly of the same level (rating 2136 for DiamondII versus 2126 for SCORPIO*). Results on FICS are even better for SCORPIO (2180 FICS versus 2120 FICS)


*source selective search N°117
Novag Obsidian :

the tactical skill are reduced to impove positional play
The Novag Obsidian seems wery close to the SCORPIO , except for the Edgame and counter attack skills. The Counter attack capabilities have been greatly improved , this imply that the obsidian will be more dangerous when we try to attack it, but in counter part the endgame skills dropped (all have a price i suppose). The attacking skill have been reduced a little too, and the tests show that the Obsidian will not consider some active/agressive lines if his position will degraded.



best regards

Nicolas
User avatar
spacious_mind
Senior Member
Posts: 4000
Joined: Wed Aug 01, 2007 10:20 pm
Location: Alabama
Contact:

Post by spacious_mind »

Hi Nicolas,

Interesting report. Thanks for sharing it. I am assuming that the further they reach the outer circle the stronger the program and the more a program makes a perfect circle then it is most balanced in its abilities?

Best regards
Nick
Post Reply