git "philosophy" understanding problems - no central server?
 

Hi,

I was reading Git User's Manual and I was wondering about one passage.

I don't see these advantages. OK, you can still work even if your Server is down, but if you have *no* central server to share the development, then what?
How to apply all these changes? Don't tell me, that, for example, Linus Torvalds sits days and nights by filtering and applying new patches he received?
I mean, there are several Kernel git repos available.

Can somebody explain this please?


Thanks

Why at all applying changes requires a central server ?

You have a tree and you apply changes to that tree, thus producing an updated tree. What does a server have to do with this ?

I'll try to better explain myself.

The fundamental issue is not GIT, it is personal responsibility and traceability.

I.e. despite the fact that the wide-spread approach is that every programmer on a project can apply changes and then nobody knows who screwed up it's a horrendously wrong approach.

The right IMO approach is that there is personal responsibility. This means that there is a person responsible for the integrity of project tree (and I used to be such a person on a number of VLSI projects). The responsibility implies that either the responsible person makes all the changes or he/she approves of all the submitted changes.

I.e. there is a simple technical part of applying the change and there is a part requiring human intelligence, and that part requires human understanding of the proposed changes.

So, again, no server is necessary and in this case Linus and other key Linux people are the ones who apply changes after analyzing them.

Honestly, it sounds like a lot of work. But okay; you have good arguments.
So http://git.kernel.org/ for example, exists only to distribute the branches after Linus decided which patches come into it's trees?


Thanks