New mylist file flag: corrupted [DONE]

[egg]

PetriW will probably get me for this, but...

[:twisted:] Why not get AOM just get the hashes and request a patch??? [/:twisted:]

Seriously, though, if the hashes were stored in AniDB (it shouldn't be too hard to get AOM to populate them). This would make it easy to AOM to see if any of the chunks/parts are the same (rather than just depending on the file size) and be able to authenticate which files are corruptions of each other (although unless one is marked CRC valid it will not know which is "right"). The CGI would not be able to make these correlations from just an ed2k link, but AOM could do it easily with the files...

[wahaha]

This would make it easy to AOM to see if any of the chunks/parts are the same

And that's where it'd break
Neither does AniDB store a hashset nor is it possible to derive a chunk-hash from the existing (overall) hashes. All AoM sees is some file that doesn't match any file in AniDB. The size might be a good guess, of course, but that wouldn't work for truncated downloads or divfixed files for example.

Well, if all AoM-clients would form some kind of network themselves*, they could organize all this without the website/server...

*) Somehow, I sometimes tend to mention pretty megalomaniac approaches ^^;

[egg]

Neither does AniDB store a hashset nor is it possible to derive a chunk-hash from the existing (overall) hashes.

I know, another hashset would have to be added... That's what I meant, and AOM could populate it fairly easily if it was created...

*) Somehow, I sometimes tend to mention pretty megalomaniac approaches ^^;

Hey, we all get to dream sometimes...

[kidan]

Well, my Idea was something like an advanced ed2k-watch (AOM-driven).

That way you should be able to tell, if a file is unique (due to local corruption) or if it is spread and deserves permanent addition to the database.