Respect of the no-history flag in gclient sync for submodules

[uazo]

Hi devs,
I would like to point out a small patch that would like to fix the failure to respect --depth=1 during the fetch of the submodules in gclient sync: it would improve the fetch times for those who, like me, need only the sources (without history) for compile chromium to a specific tag, minimizing unnecessary disk space. Probably it would also be an advantage for your infrastructure as it could lead to a minimization of network traffic.

reading the code, I realize that probably that choice is due to a specific will linked to
lack of support of fetch of a specific tag if not present in the main branch, but, using it for years now, it seems to be a choice today anachronistic: personally I have never encountered any problems, although I do not have the tools to fully test it.

The patch is available here.

I was wondering if the chromium team was interested in a more in-depth verification, I won't hide from you that instead for me it would like to be the way to better understand the procedure necessary to propose patches to (my) upstream.

Thanks for the attention.

Carmelo

[Bruce Dawson]

I'm not qualified to review that patch, but in general we are very receptive to useful patches. The process is explained here:

https://chromium.googlesource.com/chromium/tools/depot_tools.git

Create a branch, create a commit, use "git cl upload" to upload it, then find reviewers and send it for review (the instructions suggest finding reviewers before upload, I find it easier after upload, either works). The current diff lacks any explanation of what the issue is or how your patch fixes it but your commit message will give you an opportunity to supply that information.