Re: Timestamping Tags?

[Greg A. Woods]

[ On Sunday, March 3, 2002 at 11:53:34 (-0800), Claude Johnson wrote: ]
> Subject: Timestamping Tags?
>
> And I realize that tags are all
> about corroborating the state of the code at an exact moment
> in time.

That's not quite right.  In fact in some situations it's not even close.

Normal milestone tags as well as branch tags (and branch base point
tags) are usually representative of the state of a given working
directory at a given point in time.  A working directory does not have
to be updated to be tagged (though of course it should normally not have
any uncommitted changes).

Normally "cvs rtag" (which tags a module independently of any working
directory) is run with another tag (-r EXISTING_TAG) specified, and it's
normally only supposed to be used to create branch tags from branch base
point tags or to rename an existing tag.

About the only time a tag is specifically representative of the state of
the code (on a given branch) at a given point of time is when you
created the tag using "cvs rtag -D".  Given the the inaccuracies date
specifications can lead to this is not normally recommended practice.

-- 
                                                                Greg A. Woods

+1 416 218-0098;  <address@hidden>;  <address@hidden>;  <address@hidden>
Planix, Inc. <address@hidden>; VE3TCP; Secrets of the Weird <address@hidden>