Anime Library - Support Thread

[Ommina]

Hey,
I just got around to checking out your plugin... I have not used it yet, other than the configuration parts. the scan completed, but there still seems to be a large number of requests to anidb waiting.
I scanned 978 files, and when the scan was over, there were still 650+ anidb requests waiting. I have no connection problems, do you have to throttle your requests heavily to comply with their API? There is about 30 seconds between requests.

Once I use the rest of your plugin I will let you know what I think about it.

Your guess about throttling the requests is exactly correct.

It's a cost of using the API -- part of the requirements are that I am not allowed to issue more than one API request every 30 seconds "for an extended period of time". Which makes, as you're discovering, the initial run of the scanner quite painful. Or at least quite long -- it took me three days.

Once the initial scan is done and you're just adding new items, it isn't too bad, but the first trip through. Ugh.

On the other hand, though, at least it is an actual API. I don't rely on screen scraping anywhere, they actively work to ensure that API changes will not break existing clients, and the authors of the API are willing to add additional commands and features if the current specifications don't supply the data a client needs.

So as far as trade-offs go, I think it's a good one. I get a fairly rich feed of data in a consistent format and I don't need to worry about a web page change breaking the plugin.

But yeah, that first run sure takes a long time. I've made it a bit less painful in the next version, but only a bit.

[Ted the Penguin]

now here is a better question... my machine went to sleep before it finished with the anidb requests... when I woke it up I got "GENERAL FAILURE" messages, so I just quit it... what do I need to do to recover? rescan?

[Ommina]

Wow, I know the lookups are boring, but they're putting machines to sleep?

Restarting the scanner should be sufficient, no need to rehash. On startup, it will go through and requeue any requests that need to be done.

[Ted the Penguin]

hmm...that didnt seem to work...

[Ommina]

Hm.

Well, you can try the scan button, but I (expect) what will happen is it will happily tell you that there are no new files to hash - hash results are written to the database as soon as the are complete. Still, worth pushing I suppose.

I'm away from the code right now, but I'll take a look at it later this evening. What -might- be happening is that it wants to wait until tomorrow to requeue the requests. I do a limited number of updating of old data on a daily basis, and I'm wondering if the date check is also blocking the requests adds that shouldn't be waiting. I'll get back to you!

[Ted the Penguin]

well, with that analysis, I can wait until tomorrow . but I did take a look at the plugin, and it looked really nice, I now just need to skin it for sassari.

I will let you know how it goes tomorrow (or whenever I get to it, damned work!).

thanks for the help, and the sweet plugin!!

[Ommina]

OK, so that wasn't it.

When you start the scanner, does it just sit there happily and smile, claiming no requests are waiting, or does it return to the "GENERAL FAILURE" message?

As for skinning for Sassari, I wouldn't rush there just yet. I have version 1.7 sitting on the horizon, and it, of course, has a number of skin changes. (Somewhere between "a few" and "a lot".) While I've gone through Fatman-do's skinning tutorials and have tried to create the skins to be as friendly as possible to the themes in which they find themselves, I've still a ways to go.

And I'd hate to see your effort be for nought with the release of a new version.

Thanks for your kind words, I look forward to hearing what you think of it as you find the opportunity to work with it some more!

[Ted the Penguin]

I did not get any general failure messages, it just sat there dumb and happy saying it didn't have to do anything.

Thanks for the heads up on the skinning!

I look forward to using your plugin more (and having the time to do that... but who knows when work will stop being crazy)

[Ommina]

Well! I think I found out what's going on.

In order to keep files originally not found in the database from being resubmitted every time the scanner starts, I put in some extra logic to ensure they check only once every three days.

On the original insert, though, it (was) setting the 'last updated' date to the current (insert) date, which means it won't requeue the requests until, well, Wednesday or Thursday.

So, beyond running a SQL statement against the database to move those updated dates back, it's pretty much just a matter of waiting it out. And since you don't seem to be in THAT much of a hurry, I'll embrace the "wait and see if the problem goes away" solution. That said, if they don't auto-requeue on Thursday at the latest, let me know.

I've already change the insert times for the next version. I'm not sure that it is a big enough problem, though, to worry about trying to get this somewhat flakey source-control to cough up the 1.0.6 code to post a 1.0.7 release.

[Ted the Penguin]

ok, thanks for looking into that... it looks like you were right, if I waited it would try getting them again. its (slowly) draining the (large) queue now.

as you said.. I wasnt in that much of a hurry I rarely get time to mess around with my toys anymore


thanks again for this plugin, its pretty awesome... now I just want to see if it works on my new popcorn hour.