Changes to the Weekly Competition website

[GenTheSnail]

Whenever I check the 'disable timer delay' box, the setting resets after I log out.
This is pretty annoying, since I'll get a few moves into the solve before I realize that the timer hasn't started yet.

 

[ichcubegerne]

I would like to see the events that are not part of the weekly comp anymore to be separated from the rest on the PB sheets. This makes it a bit easier to see what events are important there.

 

[ichcubegerne]

Whenever I check the 'disable timer delay' box, the setting resets after I log out.
This is pretty annoying, since I'll get a few moves into the solve before I realize that the timer hasn't started yet.

I dont have this problem. I checked that box years ago and it never unchecked automatically^^

 

[Keroma12]

I have attempted to fix both of these now. Multi still has a number of problems, but I think I have that much working. Please let me know if you see any problems with them.

Sorry to report that I don't think anything has changed that I can see. For multi, in the weekly view tab, the only way I can compete without using the timer is by being in 'use the timer mode', typing in the number of cubes I want to scramble, seeing and applying the scrambles, and then clicking 'restart without timer' (or whatever it says exactly). While applying the scrambles, if I ever switch tabs (with Ctrl+Tab or Ctrl+1 for example), it will act as if a key (the spacebar) has been pushed but not let go. In particular, the scrambles disappear, and next time I hit any key in that tab the timer starts.

(The same thing happens in other events: I'm in weekly view, click on an event, and then do Ctrl-Shift-Tab to go to my cstimer tab, but this causes the site to think the spacebar has been pushed but not let go. Not a huge deal because when I get back to this tab I just click 'enter times manually' and it goes to manual mode, ignoring the started timer.)

If this isn't clear then I can try figuring out how to record my screen.
It's probably not a priority compared to some of the other stuff you're working on. Thanks for all the work you put into the site!

Although I do notice now that if I go to edit my multi result after having submitted it, I can see the scrambles that I used, which is very useful, thanks! (Could be useful to have this for all events, particularly blind ones to help figuring out DNF causes.)

Edit: These two issues have been fixed.

 

[Mike Hughey]

I have just implemented a new feature that I realize will be used by very few people (despite it taking a great deal of work). But those few people might provide some new capabilities that we wouldn't otherwise have, and besides, I wanted the capability myself, so I went ahead and did it anyway.

I am now supporting database exports, similar to those provided by the WCA. I am actually supporting 4 different formats, instead of the WCA's two:
1. An SQL export that basically mirrors our internal database structure, leaving out the private data such as passwords. Note I am also hiding first and last names, since there is an option to hide them, and I don't want old database exports to expose names that people might have since decided to hide. This export contains comments and FMC solutions, so it has more data per solve on average than does the WCA database.
2. A TSV version of the above export.
3. An SQL export where our data has been converted to tables to more or less match those of the WCA. It is my hope that this export will allow people with software written to work on WCA exports to be able to use our data with their software, basically unchanged. One thing I'm hoping is that this might allow Kinch ranks to be calculated for us, much as they are done elsewhere on the web for the WCA data. I basically tried to map or default any fields that we don't have an equivalent of so as to match the database table structure as closely as possible.
4. A TSV version of export 3.

Since this feature will likely have very few users, I would appreciate any feedback any user might be willing to give me. I have done some basic testing of the exports, but I would not be surprised if I have made some mistakes. Please contact me if you see anything I have done wrong, or anything I can do to improve the output (especially for the WCA format). I will be happy to work with you to help make your application of the database export work as much as possible.

The exports can be accessed here:
https://www.speedsolving.com/competitions/results/export.html

This can also be reached from the menu at the top of the statistics pages.

Again, please let me know of any problems you might see; I will attempt to fix them the best I can.

 

[ichcubegerne]

Hey, just wanted to know if a kinch score ranking is coming in close future?

 

[Mike Hughey]

Not sure how close, but I do hope to do it before too long. The database export was work done partially to better support adding more complex statistics. I certainly hope to have it done sometimein the next 6 months or so, but there still are a number of other things I need to do first.

That being said, if someone already has software to do kinch rankings for the WCA, theoretically our WCA database export version should be directly usable to generate them with the same code right now. So someone else could independently implement this themselves now for until I get to it!

(I say theoretically because I'm still not sure how well I matched the WCA export. I just realized I probably didn't fix the multiBLD results to match their format, so I probably need to fix that first. But if someone takes on this project, I will be happy to work with them to get my database export in better shape.)

 

[Shadowjockey]

Your eventids (at least in ranksaverage) are just numbers, while that makes sense, the WCA uses shorthand names like 222 and 333.

 

[Mike Hughey]

Your eventids (at least in ranksaverage) are just numbers, while that makes sense, the WCA uses shorthand names like 222 and 333.

Oops - one of those things I figured I'd miss - thank you. I'll try to take care of it in the next database export. Please continue to let me know as you find other things; these should mostly be easy to clean up.

 

[One Wheel]

Would it be possible to change the scrambles for 2BLD so that the DBR corner isn't necessarily "solved"? I throw the cube up in the air after scrambling, but it seems like it would be better to throw in an L, D, or B move in the scramble somewhere.

 

[Mike Hughey]

Would it be possible to change the scrambles for 2BLD so that the DBR corner isn't necessarily "solved"? I throw the cube up in the air after scrambling, but it seems like it would be better to throw in an L, D, or B move in the scramble somewhere.

Interesting. I never really thought about that being an issue before, but it does make sense. When I do 2BLD, I focus on a different piece, so I never noticed the DBL piece (it's actually DBL, not DBR, that is "solved") is always in the same place.

I intend to move to official WCA scrambles for the official WCA events sometime this year. When I do, I will try to address this issue as well. I will essentially try to treat it the same as the 3BLD scrambles, using L, D, or B moves in place of the wide moves, just like you suggest. Thanks for pointing it out. But it will probably be quite a few months before I get to it.

 

[One Wheel]

Interesting. I never really thought about that being an issue before, but it does make sense. When I do 2BLD, I focus on a different piece, so I never noticed the DBL piece (it's actually DBL, not DBR, that is "solved") is always in the same place.

I intend to move to official WCA scrambles for the official WCA events sometime this year. When I do, I will try to address this issue as well. I will essentially try to treat it the same as the 3BLD scrambles, using L, D, or B moves in place of the wide moves, just like you suggest. Thanks for pointing it out. But it will probably be quite a few months before I get to it.

Great! I didn’t figure it would be a high priority, if it’s somewhere on the list of things to do that’s great.

 

[xyzzy]

I will essentially try to treat it the same as the 3BLD scrambles, using L, D, or B moves in place of the wide moves, just like you suggest. Thanks for pointing it out. But it will probably be quite a few months before I get to it.

Note that you can't just append L/D/B moves to the end of a normal scramble, because that still guarantees that the DBL corner will never end up in URF. To ensure a truly random orientation, you would sometimes need to use stuff like D2 R or L' F2.

Well, I'm sure you'll figure it out when you get to it.

 

[Jacck]

Curios: when I look at this weeks results in FMC there is at scramble #1 an oakley pugmire (MaSteR PugGleS) with exact the same 26-solution as Tommy. He isn't listed in the summary and imo it isn't a valid entry anyway, more a problem in the database?

 

[Mike Hughey]

Curios: when I look at this weeks results in FMC there is at scramble #1 an oakley pugmire (MaSteR PugGleS) with exact the same 26-solution as Tommy. He isn't listed in the summary and imo it isn't a valid entry anyway, more a problem in the database?

I've added new software to help deal with bogus entries, but it's still not complete. This user's results were clearly invalid, so his results were "quarantined". That solution will be removed when I finish the changes. While this problem still shows up, all the numerical results have been (at least temporarily) removed from the tables, so his entry does not affect any of the rankings.

 

[Mike Hughey]

Note that you can't just append L/D/B moves to the end of a normal scramble, because that still guarantees that the DBL corner will never end up in URF. To ensure a truly random orientation, you would sometimes need to use stuff like D2 R or L' F2.

My intention is to mirror the sort of thing that's done with 3BLD scrambles, based on the code that the WCA uses in TNoodle for those.

 

[ichcubegerne]

Switching the view on the records page doesnt do anything (eg clicking mixed or slim)

 

[Mike Hughey]

Switching the view on the records page doesnt do anything (eg clicking mixed or slim)

Correct - I haven't implemented it yet. That's why I haven't announced that new feature here yet.

I just stuck it in partially implemented because that much was pretty easy to do after I put in necessary changes for the database export. I'll need to finish the other things I'm doing so I can have time to get back to implementing all the other options on the records page.

Too much to do - I can't keep up. The biggest problem, and the thing I'm focusing on now, is actually the quarantine system for people entering bogus results, because they've become such a problem lately. If it weren't for that sort of thing, I could get a lot more useful things done.

 

[ichcubegerne]

In cubingtime you can connect your profile to your WCA-ID and there is a small icon displayed in the rankings which shows if you did that. It helps people identifying fake results to report and verifying really good results which are reasonable because they are from fast people.

 

[Mike Hughey]

Thanks; I have already been considering allowing the WCA_ID to be attached to your profile. It would have the added advantage of allowing merges of my WCA-style database export with the official one, for those willing to expose their WCAIDs.

I do hope to put out a new database export tonight with a bunch of fixes: the eventids fixed for the ranks tables, as well as corrections for DNF/DNS/multibld results. I'm still not sure I will have everything right yet with the WCA database export, but it should be much closer than the previous one.

Edit: the new database export is available now. Please let me know if you see more mistakes in it that should be fixed.