Return to gitsofar.mdwn CVS log

Up to [NetBSD Developer Wiki] / wikisrc

web commit by mspo

    1: ## NetBSD with git so far
    2: 
    3: [core statement on vcs](http://mail-index.netbsd.org/tech-repository/2015/01/04/msg000497.html)
    4: 
    5: ### Low memory hosts:
    6: 
    7: * [tuning for git on low memory](http://mail-index.netbsd.org/tech-repository/2015/01/08/msg000520.html)
    8: 
    9: git appears to have slightly different memory characteristics depending on the
   10: protocol used.  Over http I am able to get a full clone with all history on a
   11: 256 + 256 raspberri pi.  If you bump up the memory to 512 + 256 it makes ssh
   12: possible, which means writes are possible.
   13: 
   14: The link above has some tuning I used to get memory requirements way down.
   15: 
   16: It should be noted that git support a "shallow" clone (--depth 1) which ignores
   17: most history but allows commits and full development.
   18: A shallow clone works on very small systems; I would guess 128MB + a little swap
   19: is enough.
   20: 
   21: git is slow during 'status' by default since it searches the entire tree for a
   22: change.  It will produce a warning with tunable options if the command runs
   23: slowly.
   24: 
   25: *Update*
   26: 
   27: After some complaining on the git@ mailing list a patch has been produced which
   28: drops the memory requirements down quite a bit.  I can now, without much tuning,
   29: work on my 512 system.
   30: 
   31: ### CVS in parallel
   32: 
   33: I do not think this is a good idea and do not plan to advocate for it.
   34: Git does have a cvs server built-in but I have not taken the time to set it up
   35: for testing because it is slightly involved and I don't see the purpose.
   36: 
   37: ### Conversion
   38: 
   39: One-shot to create the new True Source.  I don't think there will be many cvs
   40: hold-outs.
   41: 
   42: See above for CVS server provided if ongoing conversion is really desired.
   43: 
   44: ### existing cvs dependencies
   45: 
   46: is there a list of these?  build systems?
   47: The entire build infrastructure of NetBSD should (even without giti) change into a "jobs"-oriented workflow instead of a "server"-oriented workflow.
   48: 
   49: Very recent (summer 2017) events have shown that the ability to move things around is very important.
   50: 
   51: 
   52: ### How should NetBSD be setup
   53: 
   54: High level- private box for write master using ssh, any number of additional
   55: systems with read-only mirrors over http:// and git://
   56: 
   57: Also see a great description of how DragonflyBSD is setup:<br>
   58: [dfbsd server setup](http://lists.dragonflybsd.org/pipermail/users/2015-January/207421.html)<br>
   59: [dfbsd workflows](http://lists.dragonflybsd.org/pipermail/users/2015-January/207422.html)<br>
   60: [dfbsd config](http://lists.dragonflybsd.org/pipermail/users/2015-January/207424.html)<br>
   61: 
   62: ### how to install
   63: 
   64: git should fit into NetBSD src/tools easily.  I have not personally tested
   65: cross compilation.
   66: 
   67: ### workflows
   68: 
   69: See DragonflyBSD examples given above
   70: 
   71: There are many many workflows supported in git.  For the most part I think
   72: NetBSD developers would follow the "feature branch" workflow from the main repo
   73: (or private/semi-private clones before merge).
   74: 
   75: Public collaboration is a big feature of git since it can format patches into
   76: at least two different email formats and they can be submitted to a bug report
   77: or to a mailing list, which should allow clean apply.
   78: 
   79: A non-developer could also post a pull request to github or host his git repo
   80: for a friendly developer to add as an origin and pull his branch.
   81: 
   82: (git origin add future-developer http://example.com/~greatguy/src.git)
   83: 
   84: 
   85: ### log message formats
   86: 
   87: Try to references named branches/tags instead of sha-1's
   88: Also using the dates for commits instead of commit id's
   89: 
   90: ### how to convert
   91: 
   92: https://github.com/netbsd/
   93: 
   94: ### No lock-in
   95: 
   96: I am unable to anticipate the next generation of SCM.
   97: Don't do anything weird like change history and we should be fine.
   98: 
   99: Maybe when we have 30 years of project history it will be time to consider
  100: restructuring the project.  :)
  101: 
  102: git is the most widely used VCS ever so it has the best chance of conversion tools existing.
  103: No future tool will be able to exist without a git-conversion script.
  104: 
  105: ---
  106: 
  107: I think this is less a function of the tool and more a function of the project not
  108: allowing non-"standard" actions.
  109: 
  110: ### Who, When, and How Long?
  111: 
  112: * ESR/IIJ/Joerg - convert
  113: * sometime, eventually, maybe
  114: * assumptions/proposal:
  115: 
  116: Assuming conversion starting from date(x) to freeze(y) is relatively easy, the
  117: refinements of Joerg/ESR conversion can continue to run in read-only mode as they
  118: do today.  This means the "switch" is a few hours only for:
  119: 
  120: 1. cvs goes read only
  121: 2. history from last git conversion pull until now is appended
  122: 3. cvs is turned off
  123: 4. git is made available over ssh
  124: