Re: 4.3.0.1 Possible Release and Keeping XFree86 CVS Up to Date

[Brian Ford <ford at vss dot fsi dot com>]

On Wed, 30 Jul 2003, Harold L Hunt II wrote:

>-p, -U3, and -u weren't really the question here, but thanks for the
>tips anyway.
>
Sure, sorry.

On Wed, 30 Jul 2003, Brian Ford wrote:
>
>>So:
>>
>>cvs diff -up -r xf-4_3-branch xwin > xwin.diff
>>
>>But, if your sources are already at the head of the branch (like they
>>should be to do this), the -r xf-4_3-branch is redundant.
>>
>Head of the branch as in I checked out using:
>
>cvs -z3 checkout -r xf-4_3-branch xc ?
>
Yes.

> I always do a 'cvs update' before diffing for stuff that I am sending to
> xfree86.org, specifically so that I don't keep sending them the same
> things over and over again.
>
Good.

You do know that -r tags are sticky, ie. They are remembered until you use
another one, or remove them with cvs update -A.

I just wanted to make clear the difference between updating/checking out
with a branch tag vs. a static (snapshot) tag.  A branch tag is a
dynamic moving target and always refers to the head of the branch.  Thus,
updating here does keep you up to date.  A static snapshot tag is just
that.  It is a point in time snapshot that will never change (unless
something gets retagged because of a mistake).  Updating there is
pointless.

>You are both referring to a diff against an -r xf-4_3_0_1 checkout, not
>against a -r xf-4_3-branch checkout, right?
>
Yes.

>Is the head of the branch xf-4_3-branch? I assume that is what you mean.
>
Yes.

>Who has to apply my patches by hand? This has never been brought to my
>attention. Who is "they"? Are you referring to the patches I send to
>xfree86.org or the patch sets that I distribute with Server Test Series
>releases? The latter are meant to be applied by hand and only server to
>highlight what has changed to those that are keeping track.
>

I thought I had seen in a past thread that one of your xfree86.org
developer contacts apologized for a delay in getting a patch applied
because he had to apply it by hand.  I can't seem to dig that thread up
now, so it was either an isolated incident, or I was confused.  Sorry
for jumping to conclusions.

>So my question is, can I make patches for the xf-4_3-branch based off a
>HEAD checkout without actually having to pull the xf-4_3-branch tree? If
>so, then I can just create patches to keep the xf-4_3-branch up to date
>without actually having a copy of it on my disk. This would be
>convenient.
>
Officially, no.  Depending on how much has changed, maybe.  If you are
going to try though, I'd just use regular cvs diff.  Otherwise, you might
inadvertently pull something off HEAD that you did not mean to.

It would be better to either pull the whole branch, or temporarily move
the applicable part of your HEAD tree over with cvs update -r tag and
then move it back with cvs update -A.  This is tricky, but manageable.

--
Brian Ford
Senior Realtime Software Engineer
VITAL - Visual Simulation Systems
FlightSafety International
Phone: 314-551-8460
Fax:   314-551-8444