vacp2p / vac.dev

https://vac.dev
7 stars 12 forks source link

Nescience blog post #127

Closed moudyellaz closed 1 year ago

moudyellaz commented 1 year ago

Added semantic line breaks

moudyellaz commented 1 year ago

Thank you very much for this post! Will continue my feedback tomorrow.

It would be nice to have a section on the current state, including the benchmarks and linking to the zk exploration repo at that point.

I already linked the zk exploration page, and we can add a small paragraph at the end about the status.

moudyellaz commented 1 year ago

Thank you for the article!

I left some in-line comment.

Some more global comments:

  • a lot of the complexity and volume of the article stems from having a public state. A reader might question the whole point of having a separated state. I guess more attention should be addressed to the question of motivation of this decision and possible use cases that show the superiority of this approach compared, to just having a unified private state.

( It would be nice if you can do this and we can review it together)

  • I am strongly against the notion of limited recursion depth, until proven necessary. It significantly limits development, and in general makes the developer's experience miserable. Since the article already introduces the notion of gas, that should be ok. Either way, introduction of such limits is a fairly simple addition.

(again, it is just an option to consider, this is why i put it there)

  • I guess it would be more beneficial to simplify the feature set a bit, since a lot of technology (sidechains, sharding, privacy transactions etc) are complex enough of themselves, and their discussion would divert our and users' attention from the main ideas of the project.

Again, thanks a lot for your great work!

moudyellaz commented 1 year ago

Again thanks a lot for this detailed article! ❇️ I left comments in-line.

Some general comments:

  • We should add links to related work rather than just mention it without citing (i marked some of the spots where we should add links)

(done)

  • In the description of the current state you are working on, you could also mention that we will publish a further post in the future going into more detail on the current state.

    Reading the article really poses the question what the current state is. Might also look like we are just planning this and not actually doing them. You can also explicitly mention in the current state section that we are actually working on Nescience.

  • Also in the current state section, we should mention our focus areas. I suggest to mention 3 and 2 as our foci.

(done)

  • Generally, we should be careful not to try to build a full L1 and focus on the zkVM. Logos is already building a L1 with Nomos. Of course we need to design things that the zkVM expects to exist in a very specific way; but more general concepts outside the zkVM can be left as "out of scope". (Let's discuss this last point more in an upcoming meeting.)

Again, thank you for this article 🙏

moudyellaz commented 1 year ago

Again thanks a lot for this detailed article! ❇️ I left comments in-line.

Some general comments:

  • We should add links to related work rather than just mention it without citing (i marked some of the spots where we should add links)

(done)

  • In the description of the current state you are working on, you could also mention that we will publish a further post in the future going into more detail on the current state. Reading the article really poses the question what the current state is. Might also look like we are just planning this and not actually doing them. You can also explicitly mention in the current state section that we are actually working on Nescience.

(done)

  • Also in the current state section, we should mention our focus areas. I suggest to mention 3 and 2 as our foci. (done)
  • Generally, we should be careful not to try to build a full L1 and focus on the zkVM. Logos is already building a L1 with Nomos. Of course we need to design things that the zkVM expects to exist in a very specific way; but more general concepts outside the zkVM can be left as "out of scope". (Let's discuss this last point more in an upcoming meeting.)

Again, thank you for this article 🙏