The wiki for save.GetSolutions has been updated with an example showing the table format.
The solutions table consists of keyword-values pairs instead of subscripts. For a solution table entry sol, you can reference sol.name, sol.score, and sol.filename.
Having the filename is interesting, but I don't think there's a way to use it in a Foldit recipe.
There doesn't seem to be a way to retrieve the comment associated with a shared solution.
Ok, so the 'hash' names refer to the names one sees in the directory listing on disk.
For example: cmp-binary-47b3e462d77cde5897b37de9ae27765f
This would be the latest devprev version.
It does however still tell me there is a newer version after which it downloads another copy of itself. Is that intended behavior ?
The filename is mostly there for uniqueness, since you could save a solution with the same name (and even score). The filename is the only thing that is guaranteed to be unique.
For players, you can probably just ignore it and never be bothered. But when you ask to load a solution, Foldit needs to know precisely which one, so we include it in the table.
Are these released? You did a release, then a rollback, and you said you were going to re-release but I didn't see if you did or not.
My "filter" namespace is empty.
20180525-f7af35c587-win_x86-devprev
Sorry, I may have miscommunicated.
The update was rolled back entirely. Both main and devprev should be as they were before this devprev was released.
Devprev has been re-released, 20180725-d14dae34c3-win_x86-devprev is now current on Windows.
The scoring bug that affected previous devprev releases seems to have been corrected.
Looking at the current aflatoxin puzzle, a solution with added segments opens at the correct score in the newest devprev. In previous versions, the credit best score failed to deduct the penalty for the added segments, leading to an invalid score.
The new Lua functions described above are now available in devprev.
So far, no new outright bugs, but there may be an issue with "objectives". In a couple of cases, I've seen the score marked as invalid (crossed out) when I didn't expect it to be. Once this happened while changing tracks. I have been able to reproduce these problems. I suspect the unexpected behavior may relate to the scoring bug fix, but we'll see. The problems so far cleared up when I restarted the client or opened a shared solution.