[brian_m]

This will affect a vanishingly small number of FD users but for them and their opponents, it's a very significant problem. The bug was reported to Fred about 18 months ago, but it's still there, as one of my opponents complained a couple of days ago.

In order to be affected, you need to have defined a system in FD where one or more of the openings varies by seat and/or vulnerability.

In my case, my regular partners and I play a system where an opening bid has one meaning in 1st and 2nd seat except at adverse vulnerability, and a different meaning otherwise. In order to cover this combination I found I had to define the bid four times (once 1st/2nd NV, second time 1st/2nd all vul, 3rd time 1st/2nd adverse. 4th time 3rd/4th, I think I have the combinations correct).

If your opponents are using the downloadable client, then everything works just fine. However, if someone is using the web client to play, then the web client always shows the second meaning of the bid. I suspect, but haven't proved, that this is because the web client always uses the last meaning of the sequence as defined in the FD file, irrespective of seat and vulnerability.

What is beyond doubt is that someone using the web client always sees the 1st/2nd adverse and 3rd/4th seat meaning of our openers, even in 1st/2nd and not adverse. i.e. when the other meaning should apply.

I've checked this when we have had one opponent using the downloadable client and the other on the web client - they were both adamant about the descriptions they could see.

Like I said at the start, this is a very obscure one, and probably affects a vanishingly small number of players - but be warned that if you should happen to play a system that varies as mine does, you need to check with your opponents whether they are using the web client or not, as they will receive incorrect explanations if they are.

[mgoetze]

Hm, I seem to recall it worked fine with just different seats, I never had anything with different vulnerabilities. Anyway it seems like FD has been all but dumped for the web client, a lot of the info doesn't show up anymore.

[brian_m]

mgoetze, on 2011-June-22, 01:36, said:

Hm, I seem to recall it worked fine with just different seats, I never had anything with different vulnerabilities. Anyway it seems like FD has been all but dumped for the web client, a lot of the info doesn't show up anymore.

Yes, you're right, what I see would be explained solely by the web client ignoring the vulnerability.

If a lot of the info doesn't show up then I think we're getting to the stage where we need some sort of flag in the profile information to show how a user is connecting. It could cause an awful lot of bad feeling otherwise.

We had one earlier today, I asked him twice whether he was using the web client and got no answer (and he was showing a Canadian flag so it shouldn't have been a language problem). I just gave up at that point, if he got wrong explanations then nuts to him. I'm certainly not going to key everything in just because he can't be bothered replying.

It's very sad that FD has been dumped, if what you say is correct. It really is a good system - and if Fred had managed to modify it so that it had an "auto-learn" mode, where you could tell it to remember an explanation that you'd keyed in, then it would have been even better.

I think Mike Mardesich was talking about modifying OKScript to work with BBO at one point. I'll have to take a look for his website and see whether he ever did it. I had substantially the same system keyed in for use when I played on OKBridge.