Hey Gilles, That would be awesome thank you! I'm also going to start searching and including all the search lists for the algs for others to search for ones they like too.
Here is the biggest problem. We all have different styles. I vastly prefer RULM to anything else, and will often choose a sub-optimal RULM alg over an optimal length one. Also, some people are better at using the optimal algs and getting them to full speed. So here is the idea. I am a self proclaimed RULM user. I prefer RULM to anything else, although I do have lots of RUF and RULFD types algs that I like a lot. So once someone finds a list of potential algs for a case I could search for the RULM alg, and someone who is better at making the optimal alg fast could search for that. We can group ourselves accorind go out styles. That way, on the central site where we post these algs, each case will have upwards of hopefully 10 different algs that are optimized for different styles. If 10 isn't possible, at least a few to choose from. I don't know the best way to set up this group interaction, perhaps we can all vote for our favorite and vastly preferred styles to start with. I am very much a RULM solver though, and prefer it to all other types by far. If too many of us are the same style then we can branch out. I also like RUF a lot and could search for those. Basically we need people to just search for algs in all types of metrics, and people to sift through algs once they are found. Imay even be able to get Richard Patterson to help us for the sifting part. I am willing to run my processor 24-7 to search for algs, as well as look for good ones among those found. We should coordinate as a team and work on these together. Let's serioulsly turn team [zb] into a group effort to improve the ZB method. I do think with a truly optimized ZBLL, that this method can be one of the best, if not the best. But it will take a flawless ZBLL. Let me know what you guys think. I want to start with my T-orientations case 1 algs first. Most are sub-3 already, but I want to optimize those that aren't and also research and provide algs for different styles as well. What do you guys think? Should we make a database and divy up all the COLL cases? What would work best, I want as much input from everyone as we can get so that everyone can enjoy what they are working on, and do what they want to work on. Chris definitely still team [zb] as well --- In zbmethod@yahoogroups.com, Gilles van den Peereboom <[EMAIL PROTECTED]> wrote: > > Hello, > > Even though I stopped learning ZB a few weeks ago, I'm still > interested in that extension. > I'm kind of fed up with learning an huge amount of algorithms without > always understanding what I do. > But I'm really interested in finding new ones either by myself or with > the help of a computer. > > If you want some help, I could help for exemple on one of the > orientation case (maybe split it with someone if anybody is > interested). > > I'm rather fond of the Double-Sune orientation so if there is any case > you prefer to have first (since I'm probably not gonna use this in > competition), just tell me which one :-) > > Gilles. > > 2005/12/11, cmhardw <[EMAIL PROTECTED]>: > > Hey everyone, > > > > I've been talking to Richard a lot lately, and he really inspired me > > about one thing. > > > > I think ZBLL can be extremely fast, but we have to set a higher > > standard for it. We have to set the standard of sub-3 average for > > every alg (just straight execution) and that would make the LL have a > > sub-4 solve time with recognition also (in theory). > > > > Well I don't think even a few people could do this in a short amount > > of time, so here is the suggestion. ZB doesn't have much popularity > > yet, since it isn't fast. Sadly it will take that to get many to > > notice. Also, why learn ZBLL quickly now and redo 300 algs later when > > it does gain some popularity after people have learned it, when we can > > all work on generating algs (with the sub-3 requirement) and make it > > fast now? > > > > It would work like this, learn ZBLL in any way you want, and generate > > and save all algs for that path. If you are learning algs that have > > already been generated, then learn the algs you want from a webpage or > > from someone, but generate other cases only to make them fast, learn > > them later. > > > > We could each take a COLL case and generate algs and spend a few weeks > > or maybe a month on optimizing each case to sub-3. > > > > After that we can meet here and post the best ones on a central > > website (perhaps Ron will lend us a space in the Algs section). > > > > I think if we make ZBLL fast right from the start, then we can learn > > the fastest ZBLL right from the start and have a chance for the method > > to be fast. > > > > I know a lot of you guys are very serious about the method, so I ask: > > who is willing to help me make a sub-3 requirement for ZBLL a reality? > > I can start with my slowest T algs (some are already easily sub-3) > > and work from there. Again you don't have to learn the algs you > > generate, learn however you want, but spend time each day helping > > optimize the ZBLL. > > > > I think with all of us it will still not be much man power, but it > > will be enough to make a start. > > > > I will now require sub-3 for all algs, and try to spend a few weeks to > > 1 month on each COLL case to really get this right. I will stop > > learning until my T and U algs are fully sub-3 optimized. > > > > Also if sub-3 is just not possible, let's find the true limit, and get > > under that. > > > > ZB can still have a change, but we have to have a higher standard than > > sub-4. I'm already convinced that a sub-4 ZBLL stands no change > > whatsoever against Fridrich. > > > > I'm going to start on my worst T cases and go from there. Is anyone > > else interested? Let me know, I would really love for all of us to > > really take this method seriously (we already do, but I think a new > > level of dedication will be required for ZB to be fast) and move this > > method into being a top level method? > > > > Chris > > > > > > > > > > > > > > ________________________________ > > YAHOO! GROUPS LINKS > > > > > > Visit your group "zbmethod" on the web. > > > > To unsubscribe from this group, send an email to: > > [EMAIL PROTECTED] > > > > Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service. > > > > ________________________________ > > > ------------------------ Yahoo! Groups Sponsor --------------------~--> Get fast access to your favorite Yahoo! Groups. Make Yahoo! your home page http://us.click.yahoo.com/dpRU5A/wUILAA/yQLSAA/MXMplB/TM --------------------------------------------------------------------~-> Yahoo! Groups Links <*> To visit your group on the web, go to: http://groups.yahoo.com/group/zbmethod/ <*> To unsubscribe from this group, send an email to: [EMAIL PROTECTED] <*> Your use of Yahoo! Groups is subject to: http://docs.yahoo.com/info/terms/
