speeding up rescans

[aliasname]

likely Simon knows the answer but perhaps someone else also knows.
I recently was in need to do some rescans more often due to fiddling with minim settings. Then I came across the Logging level and tried "Trace" getting a quite good idea about what happens.
Now I see lots of lines in the form:
main: adding file tag XXXX=yyyy

And there's plenty I know Minimserver nor a control point nor a renderer must be aware of. Since so the question springs to mind if there's a way to skip some of these? Would that speed up rescans?

And what's the best way to make minimserver skip them?

To prevent just giving the hint to delete these information from the embedded tags, no I need that data for other software whocj also utilises the same collection.

[simoncn]

Are the tags that you don't want MinimServer to read listed in the indexTags or itemTags properties?

[aliasname]

(08-01-2019 19:54)simoncn Wrote:  Are the tags that you don't want MinimServer to read listed in the indexTags or itemTags properties?

no, not at all. But some have tagvalues sized up to 1024bytes. And since the log did show they got detected I did wonder if minimserver perhaps processes them for no benefit. I tend not to browse my library by tradenumbers for example, nor by audiofingerprints and such like.

[Wilhoite]

(08-01-2019 20:27)aliasname Wrote:  

(08-01-2019 19:54)simoncn Wrote:  Are the tags that you don't want MinimServer to read listed in the indexTags or itemTags properties?

no, not at all. But some have still recommend these quality fat burners for belly fat and tagvalues sized up to 1024bytes. And since the log did show they got detected I did wonder if minimserver perhaps processes them for no benefit. I tend not to browse my library by tradenumbers for example, nor by audiofingerprints and such like.

I'd love an option speed up the scans as well. Could a list of things to check and uncheck for scan be possible in the future, or some feature like that?

[simoncn]

I posted this reply to the OP's question to the wrong thread by mistake, so I am reposting it here.

The trace messages are intended for debugging and might not have the meaning you expect.

In this case, the tag value is not added to anything if the tag doesn't appear in indexTags or itemTags or is one of a small number of tags that MinimServer always reads (for example, Album and Title). For other "unused" tags, the tag name is written to the cache file and nothing is saved in memory.

These "unused" tag names are saved in the cache file to speed up rescanning when new tags are added to indexTags or itemTags. If a new tag is added to either of these, MinimServer checks the cache entry for each audio file to see whether the new tag is present in the audio file. If the new tag is present in the audio file, MinimServer reads the file to get the tag value. If the new tag is not present in the audio file, MinimServer does not need to read the audio file but can use the data in the cache entry.

[simoncn]

(18-01-2019 12:32)Wilhoite Wrote:  I'd love an option speed up the scans as well. Could a list of things to check and uncheck for scan be possible in the future, or some feature like that?

I am aware that rescanning time is an issue for large libraries and I would like to improve this in a future version of MinimServer. This would require a siginificant redesign of the rescanning code to reduce the cost of rescanning files that haven't changed.