search

seongjaelee / nvatom

nvatom (unpublished from atom.io)
MIT License
75 stars 14 forks source link

Performance issue or not? #22

Closed alex-kovac closed 9 years ago

alex-kovac commented 9 years ago

Hi! seongjaelee XL thanks for working on this package.

I would like to ask all of you how is notational-velocity package working for you? Especially how many notes are you having in the nv folder and how is your Atom performing.

I have a folder of about 2000 notes that used to be nvAlt notes folder and an issue with this package. When notational-velocity in Atom is summoned, Atom basically locks down for 10 minutes or more. After that, it becomes responsive... kind of... but as long as this package is activated, Atom will block/freeze when nv shortcut is activated. I am guessing nv is indexing the notes (?)

seongjaelee commented 9 years ago

Thanks for your e-mail.

Yes, unfortunately it loads all the notes in the directory every time you start Atom. This is inevitable since we need to search the contents.

If we want to improve the performance, we can do 1) Aynchronously read files so that you can still use Atom while reading files, and/or 2) Cache the read information so that we don't have to scan it every time we start Atom.

Unfortunately, I'm not working on the project for now, so you'd rather implement it by yourself.

On Sun, Aug 9, 2015 at 12:21 PM, Aleksandar Kovač notifications@github.com wrote:

Hi! seongjaelee XL thanks for working on this package.

I would like to ask all of you how is notational-velocity package working for you? Especially how many notes are you having in the nv folder and how is your Atom performing.

I have a folder of about 2000 notes that used to be nvAlt notes folder and an issue with this package. When notational-velocity in Atom is summoned, Atom basically locks down for 10 minutes or more. After that, it becomes responsive... kind of... but as long as this package is activated, Atom will block/freeze when nv shortcut is activated. I am guessing nv is indexing the notes (?)

— Reply to this email directly or view it on GitHub https://github.com/seongjaelee/notational-velocity/issues/22.

jonmagic commented 9 years ago

I have a branch I was working on a couple of months ago that switches search to use lunr.js via an npm package I built called docquery.

I mention it because I was working towards a solution where the search index could be cached and reloaded from cache, rather than re-reading every file on the filesystem each time.

I'm not sure when I'll pick that up again, but it might be a viable solution for situations like this.

alex-kovac commented 9 years ago

Thanks for such quick replies!

I am not sure if I will be able to do anything either, unfortunately. But who knows...

seongjaelee commented 9 years ago

24 is now merged and in 0.2.0. Thanks Jonathan, and thanks Aleksandar.