[cvsnt] Update branch from head

[Bo Berglund]

Still it does not make sense to me to use one branch for *each* of your test 
machines....

If you have one checked out copy on your personal PC, then you have several 
other PC:s that you also use to work on it makes sense to me to have a sandbox 
on each of these PC:s but all on the development branch (same branch everywhere).

Then wherever you are working, you start by updating your sandbox to get the 
files from any other PC you have worked on previously. Then you do your tests 
and modifications and when you are done you *commit* from this PC.

Back at the other PC(s) you start again by doing an update and now you have 
the newly changed files in this sandbox as well (still on the devel branch).

When you finally decide that all of your tests come out OK (on all test PC:s) 
then you can merge the devel branch into TRUNK. And probably start a devel2 
branch for further work.

/Bo

-----Original Message-----
From: Andreas Tscharner [mailto:andreas.tscharner at metromec.ch]
Sent: den 28 oktober 2003 09:18
To: Bo Berglund
Subject: Re: [cvsnt] Update branch from head


Hello Bo,

> Why do you do this????

I'm doing this, because I'm working on several machines. For example: I make
a new function on my personal workstation, and then I have to test it on a
machine that is connected to a measurement system (we're making software for
measurement systems). After the test I'm sure that the function works, I can
commit it into the main trunk. The easiest way would be to have the sandbox
on a network share, but I've read on the list that this is not suggested.

Best regards
	Andreas
--
Andreas Tscharner                          andreas.tscharner at metromec.ch
------------------------------------------------------------------------
Real coders don't comment their code. It was hard to write, it should be
hard to understand.                 -- from the Linux Kernel Mailinglist