Rybka Cluster Rental Program

[Jeremy Bernstein]

Maybe you ought to stop being angry, get over it and move on. Vas is out of enduser computer chess as far as I can tell, driven out by his experience.

I'm not angry, just vaguely annoyed, and I won't be buying any further Rybka products in the future, unless there's a big change in customer service. Of course, my 65 euro won't make a difference to Vasik, and that's fine. If enough people vote with their wallets, maybe we'll see some changes. But if the future of Rybkachess.com is selling server time to people with nothing better to do with their money (such as buying a pimped-out multicore system and putting a strong, free engine on it for about the cost of 4 runs on Vas' cluster), it's all moot.

As I'm certain that this venture is doomed, though (returning to the topic of this thread), I guess the enduser customers will eventually regain their importance to Rybkachess.com, if Vas wants to continue to live from programming chess engines.

[tomgdrums]

In this case, it's the world view of a member of the "producing class". From a purely legal perspective, Vas has absolutely no obligation to his licensees -- we can agree on that. However, if Rybka is to be a functioning business, Vas has to provide the additional "service" of fixing bugs, in order to provide the software that the user is licensing. If you buy a table from a carpenter and it turns out that one of the joints is a little funny and squeaks when weight is applied, and the carpenter hopes to have your business in the future, you can be 100% certain that a fix will be found within a reasonable amount of time and applied at a reasonable cost (it was, after all, not your fault). I expect the same from software engineers, myself included. Maybe I'm a sucker for listening to my licensees and fixing the bugs they find, absolutely gratis. I'm convinced, and experience bears me out, that goodwill sells more software. Leaving aside all questions of pride in good craftsmanship, etc.

In your description, Vas is a tormented genius/delicate flower whose feelings have been grievously assaulted by an angry mob of non-paying non-customers. Sorry, but Vas entered the business world with open eyes, and I don't give a rat's ass about his feelings -- I am a paying customer and expect the software experience I believed to be licensing when Vas took my money, and I expect that to be provided in a timely manner. As there is no demo version of Rybka with which one can test for gross errors in craftsmanship (MPV stalling, for instance), as one must rely on the "goodwill" and "trustworthiness" of the developer, I feel fully justified being ticked off.

Falling back on the straw man of "the angry mob" as any sort of rational justification for Vas' attitude toward his paying customers is a lazy and arrogant position. If you see your customers (real or potential) as a bunch of inconvenient whiners standing between you and your peace and quiet, you should probably stop selling whatever it is you're selling.

Jeremy

Exactly my point, he has stopped development for the consumer base of endusers. Development now appears (to me anyway) to be more concentrated on dealing with distributors, other manufacturors, other businesses and quite possibly investors. This way he gets the fun of being a business guy without the horrors of having to deal with the internet phenomenon of the masses allowing at least the vocal part of themselves to become a mob led by thugs? Wasn't that also partly the reason *you* left talkchess?

Well, we're talking about different thugs. I don't consider Vas' angry customers to be a mob led by thugs. They are dissatisfied consumers and I completely identify with that dissatisfaction. If some portion, even a large, loud portion, of those angry voices are non-paying non-customers, well, that's the price of doing business, and the price of being at the top -- you think that Bill Gates and Eric Schmidt cry themselves to sleep at night?

The thugs that catalyzed my decision to leave talkchess are the pure result of that internet phenomenon you describe -- a relatively unremarkable clique of self-proclaimed experts and their suggestible hangers-on who feel entitled telling everyone else how to live their lives. But there's virtually no overlap between this "mob of thugs" and the one you're complaining about.

Jeremy

If there are valid complaints, well fine, but the whole mess over at talkchess seems to me to be a mass assault on all fronts where perhaps validly complaining users are just mixed in with the haters. How to begin to even tell them apart, and why bother, there's no satisfying the hater community. Maybe you ought to stop being angry, get over it and move on. Vas is out of enduser computer chess as far as I can tell, driven out by his experience. btw, when a successful commercial programmer brings out a new release there are armies of hangers on of other programs who try to sabotage the release with lies of bugs that don't exist, bad reports and so on. There's only one defense against that; render the forum where it happens as unimportant by the simple expedient of ignoring it. Which is what ALL commercial programmers do, is it not?

Actually ALL commercial programmers do NOT do that. Hiarcs just released a bug fix that some customers brought to Mark's attention. AND the Junior programmers (through the Hiarcs site) released a new version (with a discount for previous customers) that addressed something that I wanted in Junior. And I was more than happy to pay for the new release of Junior because it wasn't a fix. It was an addition. BUT the difference is that the Hiarcs/Junior people listen to their customers and fix what is broken and add things that are asked for. (The Junior team has also quickly fixed bugs for their releases on the Hiarcs site)

The relationship is sound and a two way street.

And have you noticed that the Hiarcs board is for the most part civil? (not always...this is the internet) I think the civility at that board is due to the fact the customers know that Mark and the Hiarcs team DO listen and WILL fix what is broken and WILL OFTEN add things that are asked for. In other words the Hiarcs team APPRECIATES their customers and works to strengthen the relationship. I am a Hiarcs customer for as long as they are around because they release quality products and appreciate their customers.

The Rybka team is often dismissive of it's customers (not just in actions, sometimes in words as well) which then leads to the customers who helped make Rybka so successful feeling unappreciated and stuck with software that does not always perform as advertised. That is not a good relationship.

It is really rather simple.

[notyetagm]

Maybe you ought to stop being angry, get over it and move on. Vas is out of enduser computer chess as far as I can tell, driven out by his experience.

I'm not angry, just vaguely annoyed, and I won't be buying any further Rybka products in the future, unless there's a big change in customer service. Of course, my 65 euro won't make a difference to Vasik, and that's fine. If enough people vote with their wallets, maybe we'll see some changes. But if the future of Rybkachess.com is selling server time to people with nothing better to do with their money (such as buying a pimped-out multicore system and putting a strong, free engine on it for about the cost of 4 runs on Vas' cluster), it's all moot.

As I'm certain that this venture is doomed, though (returning to the topic of this thread), I guess the enduser customers will eventually regain their importance to Rybkachess.com, if Vas wants to continue to live from programming chess engines.

Yes, I cannot wait to get my hands on a new Sandy Bridge mutilcore system and run Houdini 1.5 on it.

[notyetagm]

...

Actually ALL commercial programmers do NOT do that. Hiarcs just released a bug fix that some customers brought to Mark's attention. AND the Junior programmers (through the Hiarcs site) released a new version (with a discount for previous customers) that addressed something that I wanted in Junior. And I was more than happy to pay for the new release of Junior because it wasn't a fix. It was an addition. BUT the difference is that the Hiarcs/Junior people listen to their customers and fix what is broken and add things that are asked for. (The Junior team has also quickly fixed bugs for their releases on the Hiarcs site)

The relationship is sound and a two way street.

And have you noticed that the Hiarcs board is for the most part civil? (not always...this is the internet) I think the civility at that board is due to the fact the customers know that Mark and the Hiarcs team DO listen and WILL fix what is broken and WILL OFTEN add things that are asked for. In other words the Hiarcs team APPRECIATES their customers and works to strengthen the relationship. I am a Hiarcs customer for as long as they are around because they release quality products and appreciate their customers.

The Rybka team is often dismissive of it's customers (not just in actions, sometimes in words as well) which then leads to the customers who helped make Rybka so successful feeling unappreciated and stuck with software that does not always perform as advertised. That is not a good relationship.

It is really rather simple.

Another reason why I am such a huge fan of the Houdini engine.

Is any chess programmer more responsive to problems with his software than Houdart is? He is the best!

[Uly]

If enough people vote with their wallets, maybe we'll see some changes.

And this is the problem, that's not happening. My understanding is that Rybka 3 was selling just as good as months after release, days before Rybka 4 arrived. I think Vas understood that the engine will sell well no matter what, that he could have made Rybka 3 40 elo stronger than 2.3.2a, instead of 100, and it would have sold as well, it only caused him to give 60 elo more for free to reverse engineers.

I've talked with Vas and he seems happy and calm, which means Rybka 4 is selling well regardless of the bugs and service, that's why the 4.1 version doesn't seem like a high priority (but at least it's in the "plans, not promises") and why he decided Rybka 3+ wasn't necessary after all (the people complaining about it bought Rybka 4 anyway.)

So in the end it seems the problem is the customers that will buy anything regardless.

[BB+]

I'm not angry, just vaguely annoyed, and I won't be buying any further Rybka products in the future, unless there's a big change in customer service.

I can say that I felt the same way about the blow-over with Rybka 3 bugs (the problems with the specifically-advertised "permanent hash" for example). I agree with Ovyron that there's just simply little incentive for him to output any bug-fixes (I think he did fix a number of the bugs reported with R3, but releasing an actual update proved too steep an obstacle). Certainly the merger with ChessBase for R3 has been one of the facets in place here.

he was on the talkchess forum and left due to a mob hostility and unpleasantness (a not uncommon reason),

I'm not sure what the reasons were. Peter Skinner was a moderator then (12 Feb 2006), and VR opened a poll involving whether people would be willing to "subscribe" to an engine for a version-period, and PS found this to step across the bounds of "#4: No blatant commercial exhortations" -- when CCC moved in March 2006, it seems that VR simply didn't make the trip. [PS was generally depressed that few on CCC backed him up on this-- there seemed to be a distinct division between those hungrily testing every 1.0betaX release, and those who thought this spammed the forum. He also noted that VR was using a Yahoo! group for a similar purpose, against the stated rules, and that Gothic and Fruit had previously been stifled from over-promotion on CCC. He then "resigned" as moderator and CCT8 director, but things seemed to have cooled down. As one respondent put it, if the community would prefer to have such "commercial" posts, perhaps the charter should be amended...].

Here is the post of VR in question:

Subject: continual releases in computer chess
From: Vasik Rajlich
Message Number: 485975
Date: February 12, 2006 at 04:52:00

Hello,

my experience with various releases over the past two months has been quite positive, and one of the things that I have tossing over in my mind is the idea of continuing a sort of "rolling release schedule" for versions of Rybka after 1.2. Customers would pay for a subscription, rather than for a specific release.

Probably, the release schedule would be on the order of once per one or two months.

If this happens, of course existing customers will have their order turned into a subscription which starts when Rybka 1.2 is ready. In fact, I'd probably add some sort of a bonus for the early customers.

There are three problematic issues:

1) How to combine it in the future with CD sales.
2) How to convince someone whose subscription period just expired to renew immediately, when the next update will be only marginally better than the previous one. Basically, in that scenario, you are selling the future.
3) This whole approach depends on continual progress, and loses its point when progress comes to a crawl.

This last point is probably the most important. Note that it's not only a question of engine strength - for example, once the UCI-KNEX (knowledge extensions) are out, then a faster release schedule will again make more sense.

Naturally, this isn't strictly a computer chess issue. For a bit more eloquent general justification, please see here (scroll down to the section on "Releases"): http://www.paulgraham.com/road.html

In the case of computer chess, here is why I think customers win:
1) The cutting-edge functionality is delivered immediately, rather than after a two to twelve month window of waiting for the next release
2) The final product is better, because
a) it is better-tested
b) it is possible for the developer to try more radical features and get feedback
3) It's more fun

There is also one additional nice point:
4) Piracy requires more work

Vas

[Uly]

(I think he did fix a number of the bugs reported with R3, but releasing an actual update proved too steep an obstacle).

Delivering the update would have been simple (as it shows with the ease he decreased the strength of current development Rybka when compared with Rybka 4, presumably), but he just didn't have enough time.

I figure Vas's has his current plan in a stack of priorities, with the urgent ones at the top, and he neglects the ones below them until it's their turn. When he runs out of time, he releases what he has and creates a new stack, the priorities at the top of the old one may now be at the bottom or disappear.

But when something is called to his attention he may reorder one priority to the top, or shuffle the stack, so proposing new ideas to him can be dangerous. Recently I proposed him a new feature and he implemented in the next version, which I thought was great (it was pushed at the top), only later I realized that that time used could have been used for something more important, now elements at the bottom are in danger of being left for Rybka 5's stack (including Persistent Hash ).

[Chris Whittington]

I'm not angry, just vaguely annoyed, and I won't be buying any further Rybka products in the future, unless there's a big change in customer service.

I can say that I felt the same way about the blow-over with Rybka 3 bugs (the problems with the specifically-advertised "permanent hash" for example). I agree with Ovyron that there's just simply little incentive for him to output any bug-fixes (I think he did fix a number of the bugs reported with R3, but releasing an actual update proved too steep an obstacle). Certainly the merger with ChessBase for R3 has been one of the facets in place here.

he was on the talkchess forum and left due to a mob hostility and unpleasantness (a not uncommon reason),

I'm not sure what the reasons were. Peter Skinner was a moderator then (12 Feb 2006), and VR opened a poll involving whether people would be willing to "subscribe" to an engine for a version-period, and PS found this to step across the bounds of "#4: No blatant commercial exhortations" -- when CCC moved in March 2006, it seems that VR simply didn't make the trip. [PS was generally depressed that few on CCC backed him up on this-- there seemed to be a distinct division between those hungrily testing every 1.0betaX release, and those who thought this spammed the forum. He also noted that VR was using a Yahoo! group for a similar purpose, against the stated rules, and that Gothic and Fruit had previously been stifled from over-promotion on CCC. He then "resigned" as moderator and CCT8 director, but things seemed to have cooled down. As one respondent put it, if the community would prefer to have such "commercial" posts, perhaps the charter should be amended...].

Here is the post of VR in question:

Subject: continual releases in computer chess
From: Vasik Rajlich
Message Number: 485975
Date: February 12, 2006 at 04:52:00

Hello,

my experience with various releases over the past two months has been quite positive, and one of the things that I have tossing over in my mind is the idea of continuing a sort of "rolling release schedule" for versions of Rybka after 1.2. Customers would pay for a subscription, rather than for a specific release.

Probably, the release schedule would be on the order of once per one or two months.

If this happens, of course existing customers will have their order turned into a subscription which starts when Rybka 1.2 is ready. In fact, I'd probably add some sort of a bonus for the early customers.

There are three problematic issues:

1) How to combine it in the future with CD sales.
2) How to convince someone whose subscription period just expired to renew immediately, when the next update will be only marginally better than the previous one. Basically, in that scenario, you are selling the future.
3) This whole approach depends on continual progress, and loses its point when progress comes to a crawl.

This last point is probably the most important. Note that it's not only a question of engine strength - for example, once the UCI-KNEX (knowledge extensions) are out, then a faster release schedule will again make more sense.

Naturally, this isn't strictly a computer chess issue. For a bit more eloquent general justification, please see here (scroll down to the section on "Releases"): http://www.paulgraham.com/road.html

In the case of computer chess, here is why I think customers win:
1) The cutting-edge functionality is delivered immediately, rather than after a two to twelve month window of waiting for the next release
2) The final product is better, because
a) it is better-tested
b) it is possible for the developer to try more radical features and get feedback
3) It's more fun

There is also one additional nice point:
4) Piracy requires more work

Vas

When I wrote the Charter (yes, yes, existentialist issues here!) the exact text was "flagrant commercial exhortations", the thinking behind which (they only have to ask if they are not sure, as with the rest of the Charter) was to allow information and to prevent commercial spam. Hence the use of those specific words; exhortations (plural) = once is ok. Exhortation to distinguish between strong 'advertising' and informative text. Commercial = the spam is effectively for financial reasons. Flagrant = add extra strength to the requirement.

Basically, the idea being to allow just about everything apart from relatively mindless repeated commercial spamming. The text of Vas quoted comes nowhere near the requirements of the Charter term and described the action of PS renders him guilty of abuse of the Charter.

[BB+]

Multiplying Vas rental prices by some dreamed up number of users is not the way to value the thing. It's the spin-off of having developed a working example of a new technology, having a software engineer team in place with developed skills, being first with a new idea, showing that one can make things happen. The opportunities that will arise are unpredictable but real as others with funds to spare see possibilities, need a quality team for some related development etc. Vas is building a software development business in which the chess turns out to be just a means to an end. It's the business that will have the value, not the loss-leader showpiece product.

I think this is one of the most over-looked aspects of "the Rybka project", that is, general consciousness of Rybka as a brand name. For instance, there was a series of matches of Rybka against GMs (some at odds, and I think Milov at 2700 was the highest rated), which helped make the name known [while incidentally putting an end to most of the debate about humans versus computers]. There was the attention-grabbing $100K Convekta challenge (which got whittled down to $10K), competing in every tournament possible for about 3 years, even the Rybka Forum itself is a "branding" exercise.

VR himself was on top of this from the start, for instance (emphasis added):

Subject: Rybka 1.0 Announcement
From: Vasik Rajlich
Message Number: 466792
Date: December 05, 2005 at 04:44:36
[...]
If any of you are interested in helping the Rybka project succeed, the following are all useful areas of contribution:

1) Comments, feedback, and CPU time for beta versions.
2) Purchase the full version of Rybka 1.0.
3) Get Rybka, and computer chess in general, "out there" into the world of chess - articles, clubs, into the general chess consciousness.

This last point is for me the biggest. The computer chess community has tremendous expertise and knowledge, and computer chess is interesting and fun. As programmers, we struggle with the question of what chess knowledge really is
in a much deeper and more interesting way than chess players do. A chess player will learn something obvious about positional play, and never really stop to inspect it - because as a human, he doesn't need to. On the other hand, when your program is constantly rebeling against everything you taught it, or plays worse with those last few bits of what you thought were knowledge, you end up asking much tougher questions. So - for those with the ability and interest, let's get out there and spread the word. [...]

I might add that, for all the above things I mention, it could well be that merging with ChessBase boosted Rybka more than anything else. In 2005 at least, VR seemed to think that he could target a broader spectrum of users than just the "aficionado" market serviced by ChessBase.

[pulern]

Hi,

Are there any published main line of a given test position, where you can see the strength of the 100 Cluster at say 3min and 30min?

Would be much easier to estimate the value of the product then, as I could compare it with my own hardware.

Paul