I was just looking at my "old queue" list and there are a bunch of single characters in there. I rarely, if ever, add single kanji to my studying, but the list is comprised almost completely of these single characters. Why would it be like this?

Yes, I have the same problem. I'm guessing that when the list was generated, the "add individual characters" feature was on. I remember seeing this as an option before, but I never use it and I really don't want all these single characters in my list. Can you guys get rid of them please?

Ah, good eye. After some investigating this looks to be a system wide problem. The system was a little overzealous when it came to finding everything that was in the queue. Basically, when you added a word to the queue, it would also add the related character vocab items so that they could be tracked along with everything else (ie, how many characters do you know, given that you know these words?). But these character items were also marked as added to the queue.

I'm going to try to redo the queue lists so that they only include the 'top' level vocabs that were added, and not these individual characters unless they were added explicitly. This will probably mean most of the old queue lists will be pared down a bit. In the meantime, don't add from the old queue list, or those individual characters you add during your studies will stick around even after I run this script. Unfortunately due to limited information in the system, I can't differentiate those that you add now from ones you added before.

There's also something off with dates. I've got a tiny sprinkling of words in dated sections from over the year+ I've been using skritter, but most are in undated sections of 100s.

Choose one or the other, surely?

Yeah, the data on the items added from the queue is spotty at best. It seems that it's less helpful than I imagined it would be. Turning the queue into a list turned out to be trickier than I anticipated.

From what I understand, the main problems with the old queue lists are that:

- the ordering isn't very helpful, in that many or most are undated

- the not-yet-added queue items are sprinkled around in the list

- there are characters and components in the lists that shouldn't be in there

I may scratch the attempt to organize the already-added queue items and just let them be fairly disorganized, and instead focus on making sure that the added and the not-yet-added words are clearly separated, and try to filter out the components and characters that shouldn't be in the list. Would that strategy cover everyone's needs for working with the old queue?

That would be good for me, thanks!

I have the same problem. Right now the old queue is set on adding and there are all those single characters that I never added outside of words (like 经 or 然) and have no intention of studing on their own.

To prevent them from being added I now have to manually remove all unstudied characters from that list, correct?
Since if I just say I don't want to add from that list, all other words that I DO want to study would be removed too?

Or can I just pause the queue (that's still an option right?) until the bug gets fixed?

Yes, you could stop adding from the list (i.e. pause) until it gets fixed.

Yes, everyone should pause these lists until we fix it (hopefully later today or tomorrow--we're working hard on it). We are going to automatically pause them all soon, but haven't got that done yet.

I'm scripting through them now and turning off the Old Queue lists from being added. Now that that's set, I'll focus on the few remaining bugs and Nick will start working on redoing the Old Queue lists so they're better organized.

I've bled my brain out wrangling with the logic to adjust the data just so, and I think it's going to solve the problems. It should be done in a couple hours. For users with extra large queues, and potentially a few other users whose usernames are nearby, I might need to do it manually tomorrow morning. Let me know if your Old Queue list doesn't get changed to eliminate unwanted single characters and the randomly ordered sections by date.

Thanks for your patience with this one!

Great - looks like it worked!

Looks much better, but there still seems to be some minor issues.

Eg 册 cè: book; (mw for books)

is in my old queue done sections, but not actually studied.

The data aren't going to be totally exact, wispfrog. If you want, you can go through and ban anything you see that you don't want in there. It should be pretty close, though.

I'd probably want it added anyway, it was just a surprise to find it in the section that wasn't going to be studied. I guess I can reset the list to add right from the beginning to pick it up.

Hmm - so I reset my old queue to the beginning again, but skritter is /really/ /really/ slow at going through it to add the one new word near the end - its going to take many minutes. And its now brought that incarnation of the flash from which I pressed add to a dead halt!

I can actually watch it making progress by opening that list up in another window, and watching the % complete change when hitting refresh.

Maybe this is one cause of overloading the servers!

Ok, something is still broken with the old queue list. When it did eventually get all the way through the Old Words sections, (having found about 3 extra words doing that) it has completely skipped the 'Not Yet Added' section and marked the whole list as finished. Study settings say no sections are being skipped.

Oh that's a bug. Looks like we forgot to link the old sections with the new properly. I've fixed yours and will soon run a script to fix any other lists like this.

As for making it go faster... I'm not sure that's possible. We can experiment with ways to make it go faster, but it's going to take a while to chug through hundreds of words in a list no matter what. We'll see what we can do!