Add Typescript increasses the bundle size

[selankon]

Add Typescript increasses the bundle size

[rschristian]

To compare, you should look directly at the bundles (rather than directory). This means bundle.<hash>.js w/ JS vs bundle.<hash>.js w/ TS, and so on. First narrow down what specifically has changed; as you're only looking at raw directory size, for all we know, you might've made a typo in an import statement and now there's an extra copy of an image or something.

For analyzing bundles, there is the --analyze flag on preact build, which you can use to get an idea of the makeup of your bundles. If a JS bundle did increase after the conversion, you can use that to figure out what's changed (besides doing a direct text comparison, which can obviously be a tedious process in large bundles).

[rschristian]

Please don't use raw directory size as a source of comparison; that is not a sound way to compare JS bundles.

Creating a new JS & TS Preact-CLI app shows a difference of ≈0.003% between the output bundles, which is to be expected. But to test anything besides the bare templates, I need a minimal reproduction.

I'll be honest here: asking maintainers to look through a full app containing dozens of libraries and a +3308, -2030 diff for a perceived bug is being incredibly disrespectful of our time. You need to invest some time narrowing the (perceived) bug down before you ask other people to investigate it for you. I'd be more than happy to help at that point.

I can reopen if you provide a minimal reproduction.

[selankon]

Ok thanks for your comment, I will investigate further before post anything. Could you recommend me a better way to analyze the production build size?

Thanks for your job

[selankon]

Just FYI, thi bundle size increased from 424K to 568K. I have to research why, but probably not related to preact. See