On 29/03/2017 17:59 , Christian Beer wrote:
> If the Client/Manager was built from one of the commits above then every
> commit in master is part of the history of this version.

That's precisely my point. There's no point in thinking/guessing about
what might be part of a given release. That's approach is just bound to
fail as you will miss things - things do change over time. One way of
doing this manually now might be incomplete tomorrow.

One thing is a *fact* though: any time a release is build the *whole*
repo state will be the basis for that build, regardless of what you
think might play an active role in it. This is simply *not* a matter of
opinion.

Git is tailored for exactly this purpose, represent a coherent state of
a whole source code base with an unambiguous identifier, the SHA1. Thus
you only need to know two things in order to build a new release:

1) Decide on the state the whole repo should be in
2) Have a clean repo (no uncommitted changes)

#1 can be achieved by either using the already established release
branch or by creating a new one. That is, branch it off from master. I
described that workflow before.

#2 can be achieved by using a dedicated build tree that's not used for
development purposes. Alternatively you just (temporarily) "git stash"
your local changes away.


Cheers,
Oliver


_______________________________________________
boinc_dev mailing list
[email protected]
https://lists.ssl.berkeley.edu/mailman/listinfo/boinc_dev
To unsubscribe, visit the above URL and
(near bottom of page) enter your email address.

Reply via email to