cvs commit: LFS/newxml/chapter06 iana-etc.xml mktemp.xml
matthew at linuxfromscratch.org
Mon Mar 22 13:41:08 PST 2004
On Mon, 22 Mar 2004 14:18:31 -0700
Gerard Beekmans <gerard at linuxfromscratch.org> wrote:
> On Mon, 2004-03-22 at 13:56, matthew at linuxfromscratch.org wrote:
> > * Added the 2 files that Zack's previous patch created and noted
> > to self
> > that CVS is a computer program and therefore has the right to
> > treat me like the error-prone human I am. It should note that
> > there are files not under version control and prompt me for what
> > to do with them *before* making the commit.
> You mean like subversion? Unless I misunderstood the docs.
Your telepathy skills are honed tonight aren't they Gerard...I was
just writing an email regarding that very tool :)
I've just tested subversion out here, and it's `svn commit` usage
follows that of CVS, as in it doesn't warn you that you might have
forgotten to add a file.
Looking at http://svnbook.red-bean.com/svnbook/ch03s05.html it seems
as if I might be using a insufficient "work cycle" in that I should
be doing an `svn status` (or `cvs status` in our case) *before*
making the commit. Having said that even without the "-v" flag, the
output of `cvs status` is far too verbose to tell at a glance what was
added. It'd be useful if it had a `cvs
status--only-new-removed-or-changed-files` or`cvs status--brief` option,
but alas it insists on printing 8 lines for each and every file, and
hits the server to do it as well! `svn status` by default shows *1
line* for only those files that were modified, added or removed from the
working copy, and doesn't need to communicate with the server in order
to do it - i.e. it has the functionality/usability *I want* from cvs by
default! Can't ask for more than that.
More information about the lfs-book