Anyone in naclports-owners (https://chromium-review.googlesource.com/#/admin/groups/5210,members, binji@, bradnelson@, sbc@) has permission to create new branches in this repo.

I don't have that permission, although I can grant it to myself. Would you prefer that I use my break-glass permission and handle this myself, or delegate to the actual repo owners?

i chatted with bradnelson@ via hangouts at the time and he said he was fine with the idea.  can't comment wrt their preferences for admin requests like this though :).

Sad that we are stuck there :(    Any chance that will ever change?

But given that we are.... sgtm.  Done.

when i say "we're stuck on v35", what i mean is:
* we've tried to upgrade to newer versions like v47+, but the resulting builds of openssh are unreliable/flaky and crash at weird places.  reverting back to v35 and all the problems go away.
* considering the future of (P)NaCl (i.e. everything is going to web assembly and (P)NaCl is going away), it isn't worth anyone's time to investigate these problems.
* i'm pretty sure there are no advantages to us using a newer toolchain wrt security/stability/optimization here.  we've got a great sandbox already regardless of the version, and the performance appears to be sufficient (all the bottlenecks are in the HTML/JS side of things with rendering).
* SecureShell only cares about 3 packages: glibc-compat, zlib, openssl.  i can handle updates of those myself since i'm already a maintainer for glibc/zlib/openssl in CrOS & Gentoo (and have worked with upstream for all three).

so once webasm becomes a reality we can use, we'll most likely look to jump ship from (P)NaCl entirely.

the branch is created, but the ACLs aren't set up to allow us to push CLs through for this branch.  the idea as to let us self-serve on this w/out hassling you, or do you guys want us to send these old ones to you too ?

I'll let binji/bradnelson/sbc answer that question.

What permissions do you need exactly?  Do you want to be able to push directly to that branch?

If the changes are few I think going through code review is still probably a good idea?  At least perhaps at first?

push would be nice ;), but just the ability to Code-Review/Commit-Queue/Submit CLs ourselves is sufficient.  we were planning on doing code review, but against ourselves.  if you want, i can add you guys to my CLs.

reviewing yourselves sounds fine.  Not sure if that means you need to be added naclports-committers, it might just work as it stands.

Added vapier@ and rginda@ to naclports-committers

adding me/rginda@ to naclports-committers would work as well.  i just figured adding an exception for refs/heads/secureshell would keep us from stomping on master by accident.  if you aren't worried about that, then sure, let's go that route :).

Looks like that gives to "push" and "Submit" (if i'm reading the ACL stuff correctly).

The CQ has not been used in the long time, and never on a branch, so you might not want to use that :(

it does indeed ;).  but my plan was to go the gerrit/review route (assuming it works out).