Quote:
Originally Posted by cyana
again, thanks a lot for your efforts! Unfortunately, IE 10 crashes when I try to create an account at calibre2opds.myjetbrains.com so I cannot add comments there. However, your implementation of author links would be really great and is exactly what I had in mind. I would not drop the x-refs for tags in the book's page, they are very useful to switch to more books of the given genre.
|
I was just thinking of discontinuing a separate cross reference section on the page as the authors, tags and summaries shown in the book details would become links that are functionally equivalent to the current cross-reference links as long as you want cross-reference links in the first place.
Quote:
Meanwhile, the catalog generation is about 30%  - my flight to London leaves on Sunday at 1 PM. So I don't think it will make it. What happens to the cache when I cancel the catalog generation? Will it continue at the point where it was interrupted, or will it restart from the ground? If so, a break/continue option would be very helpful - considering that an upload under such conditions might be very likely for win7 users and can take a week, so an unexpected interruption might happen from time to time - however when it is not necessary anymore to create an opds catalog, the upload time will drop significantly. The time consuming part is the sheer number of the files, not their size.
|
No - it is not possible to interrupt a generation and resume. There s lots of info cached in RAM so stopping the program loses this.
I must admit that I do not understand why it takes so long to run the generate? I can only assume that the way your system operates is not like most cloud services where the files are held locally - and a background process does the sync to the cloud. Calibre2opds then normally just generates the local copy and lets the sync service run separately at whatever speed it can achieve. It sounds as if you might be trying to write directly to the cloud in real time?
The cache file that is used to help optimise copying is currently only written out at end-of-run. I need to think if there would be any advantage to writing out a partial cache file on an interrupted run. I would have to think about what the effect would be on the next generate run to see if it is likely to speed things up or slow them down.