Improve initial load time and cache management of logseq

Initial loads on logseq are quite slow. If I have a fleeting thought that I want to document quickly, I’m forced to use notion for now, as (surprisingly) it loads a lot quicker than Logseq. I believe this is due to the fact that we prefetch and parse all pages in the entire graph? Where notion probably fetches more intelligently (Maybe main page or most used pages), so the load up time feels a lot faster.

If we gave the user the journal very very fast, and allowed parsing and loading of distant pages in the background, that would be a huge upgrade and allow me to use logseq as my only notes app

Logseq doesn’t parse all pages in the entire graph when refreshing the browser. It’ll restore the database indices to memory, it depends on how many notes you have. It takes about 2s for me. The last time I checked it takes about 4s for 10k notes. Make sure you don’t check the “Disable cache” checkbox in the chrome devtools.

I actually notice this more on mobile, it takes quite a bit longer on mobile (I use it via PWA on iPhone). I often come across problems on mobile that require a reindex or sign out + sign in (width error causing uneditable area or manual git errors where it’s easier to just reclone), and that takes a really long time on mobile.

I was comparing opening time to notion, but honestly notion might just feel faster because of their nice opening animation.

1 Like

Yes, mobile could be a bit slower on IndexedDB ops compared to desktop browsers.

notion might just feel faster because of their nice opening animation.
Now we have something to do :smile_cat:

I have closed this as it it quite dated and there have been a lot of improvements since the post was made.