From: everything-list@googlegroups.com
[mailto:everything-list@googlegroups.com] On Behalf Of LizR
Sent: Saturday, March 29, 2014 6:37 PM
To: everything-list@googlegroups.com
Subject: Re: [foar] Amoeba's Secret now available in paperback

 

If my joke works at all, it needs you to take that quoted line out of
context. (If I understand correctly, committing in a version control system
is booking in your changes so they are accessible to others...?)

 

That is one aspect of it for sure.

Looked at from another angle it is also the process of merging in the change
deltas (for CVS type repositories; GIT does it differently, but conceptually
it is similar) However; if you think of the case of a main trunk branch (and
there can be multiple such branches); committing a branched change set, back
into main is the act of merging these changes into this trunk-line of main. 

 

On 30 March 2014 13:30, Russell Standish <li...@hpcoders.com.au> wrote:

On Sun, Mar 30, 2014 at 12:46:48PM +1300, LizR wrote:
> On 28 March 2014 20:03, Russell Standish <li...@hpcoders.com.au> wrote:
>
> > I used to get everything to the commit stage, then go home.
> >
> Typical guy :-)
>

I don't know about the "guy" bit, but certainly typical for someone
with a spouse/significant other, and life outside of work :).

And as I mentioned, if I knew I was going to have a quiet evening at
home (as opposed to going out to theatre, say), and I thought the
commit was not likely to be problematic, then I would sometimes
commit later in the day on the understanding that I would log in again
remote at say 8:30 or 9 pm - just to check things, and fix any
unpredicted problems, or back out if things went completely pear
shaped.

The point was that the repository system (which is very common - the
only exception I know of is Aegis) forced this sort of behaviour.

Incidently, in Aegis, the start of a commit would lock the
repository. If the commit builds and passes its regression tests, the
code is added to the repository, otherwise its is failed, and the next
person attempting a commit is processed.

At no stage is it possible for a commit to break the build.

Trouble is Aegis is not popular, mainly because it doesn't play nicely
with the Windows operating system. I have tried to come up with a way
of implementing this protocol with the other popular SCMs used -
mainly subversion, but also perforce, but haven't succeeded. Git comes
close though - people commit to their local repo, then post a pull
request. The owner of the master repository then does a pull, and
either passes or fails the commit. If the master repository owner is
automated, then you get pretty much the Aegis protocol.

Cheers


--

----------------------------------------------------------------------------
Prof Russell Standish                  Phone 0425 253119 (mobile)
Principal, High Performance Coders
Visiting Professor of Mathematics      hpco...@hpcoders.com.au
University of New South Wales          http://www.hpcoders.com.au

 Latest project: The Amoeba's Secret
         (http://www.hpcoders.com.au/AmoebasSecret.html)
----------------------------------------------------------------------------

--
You received this message because you are subscribed to the Google Groups
"Everything List" group.
To unsubscribe from this group and stop receiving emails from it, send an
email to everything-list+unsubscr...@googlegroups.com
<mailto:everything-list%2bunsubscr...@googlegroups.com> .
To post to this group, send email to everything-list@googlegroups.com.
Visit this group at http://groups.google.com/group/everything-list.
For more options, visit https://groups.google.com/d/optout.

 

-- 
You received this message because you are subscribed to the Google Groups
"Everything List" group.
To unsubscribe from this group and stop receiving emails from it, send an
email to everything-list+unsubscr...@googlegroups.com.
To post to this group, send email to everything-list@googlegroups.com.
Visit this group at http://groups.google.com/group/everything-list.
For more options, visit https://groups.google.com/d/optout.

-- 
You received this message because you are subscribed to the Google Groups 
"Everything List" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to everything-list+unsubscr...@googlegroups.com.
To post to this group, send email to everything-list@googlegroups.com.
Visit this group at http://groups.google.com/group/everything-list.
For more options, visit https://groups.google.com/d/optout.

Reply via email to