3 ms·
I'm really starting to think it would be a good PhD dissertation project to (semi?) automatically reduce Chromium to the bare minimum needed for a given Electro
by azhenley 8y ago
I'm really starting to think it would be a good PhD dissertation project to (semi?) automatically reduce Chromium to the bare minimum needed for a given Electron app. The techniques could probably be applied to reducing any library and would be of interest to researchers in SE and PL, along with industry.
- DonHopkins 8y agoThat's not just tree shaking, that's more like rainforest shaking!
- cat199 8y ago> good PhD dissertation project why? would not existing static analysis/dependency pruning take care of it if someone had the time? I don't see what would be 'novel' here from a theoretical point of view.. still a good project possibly though
- azhenley 8y agoThat is the research of it, we don't know if existing techniques would work! And surely there are novel improvements that could be made, particularly for this specific application. Also, I imagine there are many cases where static analysis would not be enough for the JS app.
- barbecue_sauce 8y agoSounds more like an undergrad thesis.
- andai 8y agoAn old man once said, "write the thesis you wish to see in the world."
- nickpsecurity 8y agoThat's exactly what I thought. You just include what you use. Either developer does it on a confijguration screen, a tool does it with analysis on what code reaches what other code, or both. It would also help if Chrome is designed to make this easier. eCos is prior work on OS side that made me always think of designing systems for easy stripping. http://ecos.sourceware.org/about.html http://ecos.sourceware.org/about.html
- Andrex 8y agoHm, would headless Chromium be a good starting point for this?