Join Bridge Winners
All comments by Hans van Staveren
1 2 3
You are ignoring the author of this comment. Click to temporarily show the comment.
Check the Orlando FAQ, second item.
http://championships.worldbridge.org/orlandows18/faq

I doubt this is frequently asked, but the info is there, although not totally understandable. Anyhow it states: Since the format of some events is not yet known (the number of entries is not final) the organisation produced the codes for the Rosenblum Cup, the Open Pairs, the Mixed Teams and the Mixed Pairs.

And the .sqd files are reproduced there.
I encourage anyone with interest in this subject to make a copy. As stated in the documentation at this point you cannot do anything with it, but it is important that copies are kept. It keeps the WBF honest(they would be anyhow).

The page talks about a championship secret, but that was in an old version of the proposal, and says something about an ash procedure, where they probably mean SHA-256. Even that you only need if you both distrust the WBF and me, since the software does it automatically.

Anyhow, during these championships we will iron out the starting wrinkles. I plan some info dissemination, perhaps something in the press conference and/or a daily bulletin article and/or a Q&A session about it. We will see.
Sept. 11, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
Just in case this was tongue in cheek and not just a remark, the documentation is in the same ZIP file as the programs.
Sept. 11, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
I understand that most people will be not so strict. Good security does not necessitate it. But for example the WBF and EBL try to keep the machine that generates the hands not connected. I respect that, and for sure do not want to force them to change the policy.

Luckily I do not need to do that, since the current design of the software keeps the public and private data separated, which means that organizations that want to keep this strict policy can just make sure they never put the private data on a connected computer, and can shuffle the public data around on a USB stick or so.
June 20, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
I will probably add this as an automatic option. However, apart from explaining this to the bridge world, my major issue is that it needs Internet connectivity. The computers I expect the software to run on should not be on the network.

My current thought is to write the procedures such that the key file will never be connected to the net, but the description file(which contains the salt) can temporarily be.
June 20, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
All looks fine. Leaving me with this burning question:
given that I am trying out some SciFi code on pretty reactionary guys, would they understand this?

My post crossed your post. Sorry. I understand your explanation.
June 19, 2018
Hans van Staveren edited this comment June 19, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
Ok. This looks indeed fine for the DI. I did not know about this.
Three questions:
1) Would using these for the session keys be safe?
2) Is there Perl sample code to automatically query this?
3) How can I say in advance which of the broadcasts I am going to use? I should not have a choice.
June 19, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
You are ignoring the author of this comment. Click to temporarily show the comment.
The hands for this software actually come out of a slightly modified version of BigDeal. If you can crack BigDeal you can crack this. I do not think you can crack BigDeal.
June 19, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
A first draft of the documentation appears here:
https://www.dropbox.com/sh/guv798tyju3avv5/AADNMwCrKv0QvXa-KLUVDLkKa?dl=0

Any comments are welcome. Technical details will follow, software will follow. You are welcome to Email me at sater@xs4all.nl if you are dying for more info.
June 19, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
I would be very careful with the “rigged for sure”. As far as I know, most Tournament Organizers are honest blokes. If only because being honest is the most lazy option.

Anyhow, with a bit of luck I will publish software and documentation in a couple of weeks.
June 18, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
I just heard from Maurizio that the first three hands in the final session of EBL Oostende were (potential) slams. Again, I have no doubt that Maurizio is honest, but if players complained what is his defence at the moment?
June 18, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
Indeed, the software is there to show that the organizer itself was honest, and did not do anything with the hands.

Still nothing prevents the organizer, or the duplication room, from selling the hand records to certain players in advance.

I start with the *easy* problems…
June 18, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
Anyhow. I think we are about to get a reasonable software.
I do not want to bother the rest of the group, but anyone that would like to be informed about developments, please send an Email to sater@xs4all.nl

I will let you know about developments, and if there is enough interest I might setup an Email alias or so.

Be the first to know….
April 30, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
People see star patterns in the sky. That can never be helped.
April 30, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
Ok. I might be convinced. What about the following proposal:

We drop the Tournament Secret. Instead the software administrates all sessions, and the Delayed Information in a master file, and a key, or perhaps pair of keys(reserve set!) per session in a separate key-file.
The master file contains the SHA-1 signature of the key-file.
The software will still make hands for one or more sessions, the players that want to check still need to check only one SHA-1.

The TO can now publish the master file before the tournament, and the key-file after the tournament.
April 30, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
This would be possible with the software. However, if you really want people to check on you it is perhaps asking a bit much for them to check 40 or so SHA-1 hashes when one would do.
I really do not see a serious threat for a tournament when breaking it requires rolling back two different crypto-hashes when you have two weeks.

This is a trade-off. We can see where it will go. Any TO can split his tournaments up as far as he wishes. I think the current proposal, with one seed for the whole Bermuda Bowl is strong enough.
April 29, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
I do *not* want the TO to reuse seeds across tournaments.
The software will generate a seed per tournament, which will be disclosed after the tournament.
However, think Bermuda Bowl. This runs for two weeks. After the first round robin there is a set of hands known. Now someone could go and start trying to find the original seed. If he finds it within two weeks it is in time for the final.
This must not happen, that is why I am going to be generous with the bits in the seed. They are very inexpensive anyhow…
April 29, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
As described in the documentation the seed used for hand generation would follow directly from the triple of tournament secret, delayed information and session name.

This will allow players to check on the tournament organizer after the tournament, when the secret is revealed. My current plan is to make the tournament secret pretty big, some 320 bits+. I have to make sure that during the two weeks or so of the tournament even some pretty powerful hardware cannot crack the secret.
April 28, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
Ok. The simple answers about restraints is this: absolutely none.
There are 53,644,737,765,488,792,839,237,440,000 possible bridge hands, and as far as I can prove BigDeal will produce each of those with equal probability.

So in theory you could get a set of 26 boards which are all passed out by the majority of players.

Your comment about more competitive auctions arises from the fact that manual shuffled boards, by a combination of lazy shuffling and weak elderly hands, tend to be more balanced, or boring.

But if you are used to these manually shuffled hands, you will find the first time with properly shuffled hands interesting…

Hans
April 28, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
By the way, especially for new readers on this subject.
If you think directing the dealing program sounds fishy let me try to explain in simple terms.
When BigDeal starts dealing hands it needs some sort of starting value. Normally this is a random 160 bit number, generated by keyboard input.
However, any 160 bit number will do, as long as it is unique and unpredictable.
Whatever number is chosen, the resulting set of hands will conform to expected statistics. Of course statistics on any set of hands do not mean a lot.
But(and I am willing to explain this in a lot more detail to those interested), this proposal will *not* lead to different sorts of hands.

Hans
April 28, 2018
Hans van Staveren edited this comment April 28, 2018
1 2 3
.

Bottom Home Top