[uecker]

A lot of the ideas around git were known at this time. People mentioned monotone already. Still, Linus got the initial design wrong by computing the hash of the compressed content (which is a performance issue and also would make it difficult to replace the compression algorithm). Something I had pointed out early [1] and he later changed it.

I think the reason git then was successful was because it is a small, practical, and very efficient no-nonsense tool written in C. This made it much more appealing to many than the alternatives written in C++ or Python.

[1]: https://marc.info/?l=git&m=111366245411304&w=2

[oblio]

And because of Linux offering free PR for git (especially since it was backed by the main Linux dev).

Human factors matter, as much as programmers like to pretend they don't.

[dspillett]

> since it was backed by the main Linux dev

For “backed by” read “initially written by”.

I don't particularly remember Linus making any push for git to be generally popular. While he was more than happy for other projects to use it and be his testing resource, his main concern was making something that matched his requirements for Linux maintenance. BitKeeper was tried and worked well¹, but there were significant licensing issues that caused heated discussion amongst some of the big kernel contributors (which boiled over into flame-wars more than once or twice), and those were getting worse rather than going away².

A key reason for Linus trying what he did with Git, rather than using one of the other open options that started around the same time or slightly before, was that branching and merging source trees as large as Linux could be rather inefficient in the others — this was important for the way Linux development was being managed.

Of course most other projects don't have the same needs as Linux, but git usually wasn't bad for them either and being used by the kernel's management did give it momentum from two directions: those working on the kernel also working on other projects and using it there too (spreading it out from within), and people further out thinking “well, if they use it, it must be worth trying (or trying first)” so it “won” some headspace by being the first DVCS people tried³, and they didn't try others like mercurial or fossil because git worked well (or well enough) so they just didn't get around to trying the others⁴ that would have worked just as well for them.

----

[1] Most people looking back seem to think/imply that BK was a flash in the pan, but Linus used it for a full couple of years.

[2] A significant problem that caused the separation, rather than it being because BK was technically deficient in some way for the Linux project, was people reverse engineering the protocol to get access to certain metadata that would otherwise have required using the paid version to see, which the BK owners were not at all happy about.

[3] So yes, human factors, but less directly related to one particular human that is Linus, more the project he was famous for.

[4] That sounds a lot more dismissive than I intended. Of course many did try multiple and found they preferred git, as well as those who did the same but went with one of the others because they were a better match for the needs of that person/project.

[oblio]

> I don't particularly remember Linus making any push for git to be generally popular.

Outside of giving one of the highest visibility tech talks in history, at Google (back when Google was the mega hip FAANG), declaring Subversion (the then leading SCM) brain dead?

Marketing works in many different ways, as does signaling. Geeks wear suits, too, their suits just aren't composed of suit jackets and suit pants, they're composed of t-shirts and jeans.

[dspillett]

> declaring Subversion (the then leading SCM) brain dead?

I remember that more as berating the incumbent leader in non-distributed VCSs, than promoting a specific DVCS, and that git wasn't mature at that point (the move from BK had not happened). Though maybe my remembered timeline is muddled, do you have further reference to that talk so I can verify details?

[oblio]

This is trivially google-able.

Well, I guess I might as well play ChatGPT :-)

https://en.wikipedia.org/wiki/BitKeeper

https://en.wikipedia.org/wiki/Git

https://sandeep.ramgolam.com/blog/linus-torvalds-talks-about... -> https://www.youtube.com/watch?v=4XpnKHJAok8