calibre friday preview releases

[BetterRed]

Originally Posted by chaley

This is intentional. The purpose of the check box is to prevent writing very large template results to an OPF. I didn't see the point if the template results are tags. One doesn't have comma-separated images.
Your second image is not the definition of a composite column. It is a standard text column.Yes, and specifically only non-tags composite columns.

Got it.

Suggest making it clear in release notes that it only applies to plain "Column built from other columns" columns.

I mistakenly assumed the OPF file being referenced was the one that gets embedded in format files (i.e. content.opf in an EPUB), rather that the metadata.opf in the book folders. But I see from the tooltip that it only gets excluded when a format file is Sent to a device, but not when its Polished, Saved to disk, Converted etc.

BR

[Comfy.n]

Originally Posted by BetterRed

But I see from the tooltip that it only gets excluded when a format file is Sent to a device, but not when its Polished, Saved to disk, Converted etc.

BR

I did some tests today on EPUB formats, and found that Polishing (option: Update Metadata in the book files) does exclude the composite template results, reducing the OPF size up to 60% in some cases, depending on how large were the images contained in template results.

[BetterRed]

Originally Posted by Comfy.n

I did some tests today on EPUB formats, and found that Polishing (option: Update Metadata in the book files) does exclude the composite template results, reducing the OPF size up to 60% in some cases, depending on how large were the images contained in template results.

The tooltip needs to reflect that, this is what's there now:

If you check this box then the result of evaluating this column's template will be stored in the backup OPF stored in the library. The same is true when sending to a device, assuming the format has an OPF. One reason to uncheck this box is that the column contains large images.

I was a little surprised to read it was limited to Send to device, but then I assumed it was cut down on comms traffic.

I don't embed uuencoded data in custom columns, my interest was to remove clutter rather than reduce size.

[Comfy.n]

When I did those tests, I was actually minding clutter not size, as in "the less clutter, the less delay in processing metadata changes", as ownedbycats speculated.

But it seems chaley was right in his reply, as I have all functioning well and fast at the moment.

(I had experienced some delay in metadata processing recently, regarding author notes, but after deleting my author Bio column and recreating it, issue was gone)