Optics would need line of sight. The earth's curvature puts a very hard limit to how far you could communicate. The accepted standard in underwater is VLF, but as the name implies you are only capable of very low data rates.
The name implies only the frequency, which doesnt have anything to do with data rates. The VLF band is tiny though - only 27kHz. The small bandwidth is the main reason it would be slow.
True but if your emitter is not underwater like the receiver then you will have to account for air/water refraction not sure how that would work in the real world. Also I suspect it would be fairly easy for a rogue state to infer submarine locations from emitter beam orientation in low orbit space.
Sure! As far as normal DVCS features go, I'd say Gitlab and Github are almost at feature parity(I happen to use both extensively). The UI may just come down to a matter of preference.
Github indeed has somewhat superior code-review tools. But I've found Gitlab to be perfectly adequate.
When it comes to CI/CD, Gitlab CI blows Github Actions out of the water by a HUGE margin.
So, if I just wanted a tool that got things done for me, instead of fragmenting myself with Github + Github Actions + TravisCI / CircleCI, etc, I'd just get one Gitlab subscription and call it a day.
[0]: https://www.usgs.gov/media/images/triangulation-locate-earth...