Using the MacBook, I'm worried that if the hard drive is on the road and it gets spoiled then the 3-day code can be lost to everyone.
So I actually have a TMP repository on our main server computer, and I can do it hg push The dilemma is that unless I'm committed to the first time, And with the previous experience, we can not work until we are ready to prepare a central server (to share the code with peer, merge, etc.) (The reason is, we have selected files Can not push - all of us Do not push committed files or do nothing). So how to solve it?
Is there any way to say, "copy all modified files (and added files)" on the central server / user / peter / 2010-06-18 or not committed anyway Get it on the server?
In general, simply commit is urged with a DVCS that "Commit, often less Do "locally often submit, then your SMP repo gets pushed on the server for backup purposes. When you are happy with your work, you push your TMP repo to your shared "central" server.
There is no need to ensure that after every change everything is compiled, the ideal is to make sure that the changes you make back are 'tip' in each combination situation. drop offs. Generally triggers a continuous integration build server to turn on a build after each change group (change group) arrives, and it only considers the 'tip'.
If you can not really stand the changes, then you can not stand on your own, then one step supporter will work on it by pushing a untouched patch for a different line of repository. Will use it.
Out of the three options presented:
- Fix your workflow so that the changes in the central repos can handle changes that do not compile, when They are part of a change group that compiles
- Use the mercurial queue to keep patches in an approached patches repository
I think that First is most done, and usually "right" way.