Yeh, if I see the script, it should be understandable 
Still a description with some information should probably be somewhere
but script works for me.
Still a description with some information should probably be somewhere
but script works for me.Yeah, I'll add a description to our FAQ.
Actually, the reason it was done server-side is that it actually calls the live roll-generation code with the parameters supplied by the verification page. I opted for this method in order to make 100% that there are no divergences between actual betting and verification code.
So given the above, I'd prefer to actually keep it server-side. Do you see a problem with this as long as I supply the exact method used to calculate the bet outcome so you can verify that it's correct and (optionally) implement this verification on your site?
Please don't get me wrong. I have no problem doing this client side, but I think there's a certain charm (code-wise) when the verifier calls the actual code which is used during live-betting.
In theory if you are rigging it (of course you are not
but example) then you would easily return the server-side results as verified too. By making it client-side, this trust would be less. Of course in theory we still have to trust that your JS script is legit, but at least we can check that (if you have the knowledge.) Actually that's why an external script (even if it's a site-made JSFiddle!) is much better.That being said. I still disagree with some fundamentals as discussed a long time ago: https://bitcointalksearch.org/topic/m.12015013 (basically: needs to generate the clientseed between every bet in the browser with cryptographically secure method.) I don't think you guys changed that yet (to for example RHavar's solution.) I also discussed it a bit at Rollin thread - they did change it after few days: https://bitcointalksearch.org/topic/--687571 IMO the provably fair implementation is barely provably fair at this moment. So TBH I think that verification script isn't a high priority compared to that.
Obviously I don't think you are doing anything malicious or whatever, but the provably fair method can have some improvements. I also realize you have been busy and I can see the amount of time the Rubies addition must have cost, so I am glad you guys keep improving. Still I think that provably fair implementation should be "fixed" too
Anyway, thanks for adding the FAQ text, I will have a closer look at the Rubies bet algorithm if I got a time
Thanks for your input. Can we discuss technical details on skype (or some other realtime chat tool) sometime?
Nico, still waiting for you to contact me on skype so we can discuss how to change this to everybody's satisfaction.
