[illwieckz]

This was part of:

• NaCl process creation improvements #2001

Especially that commit:

• efb8fa9

It had been temporarily put aside when testing and merging:

• Rebasing slipher's work on my CMake and VirtualMachine code #2018

In #2018 I temporarily removed what was not touching cmake as it was not needed at the time, and this was conflicting with branches I used to test #2018 (2018 includes preparatory work so I had to test it worked).

I now rebased all the branches that would have conflicted with that.

The commit message is the sentence from the original commit that described that specific part of the patch.

This is already reviewed by me and approved, so @slipher can merge anytime he wants to.

This changes the way a rebuilt deps 11 version is done, and then such rebuilt deps would not fit to rebuild a last-release engine, but it works with the master branch, and rebuilding a last-release archive can use the already built deps.

The alternative would be to bump first but then some bumping commit may precede what will change again. As long as releases can build and master can build, it is fine.

[slipher]

The commit message is the sentence from the original commit that described that specific part of the patch.

I think with that sentence I was describing both the build.sh change, and the CMake changes to stop copying that file. With the remaining part I might describe it as "Deps: don't package amd64 nacl_bootstrap"

This changes the way a rebuilt deps 11 version is done, and then such rebuilt deps would not fit to rebuild a last-release engine, but it works with the master branch, and rebuilding a last-release archive can use the already built deps.

Right, that's how I see it. The two guarantees should be that the engine works with the version that it auto-downloads, and that it works with building the current version of the script from scratch. Being able to overwrite the published version of the deps is a non-goal.