[MPlayer-dev-eng] x11_common.c: it still doesn't work in fullscreen

Diego Biurrun diego at biurrun.de
Thu Mar 18 01:49:43 CET 2004


Gabucino writes:
 > Diego Biurrun wrote:
 > > Which proves that you are wrong AGAIN.
 > Huh? This doesn't prove anything. It says nothing about merging deleted
 > revisions. BTW I told you that what I did, worked, so it's quite amusing
 > you're still on the topic.
 > 
 > > We all make mistakes (and I usually make more than the next guy), there is
 > > no shame in admitting them.
 > I admit my mistakes, but so far you couldn't prove it was mine (btw there
 > was no mistake on my part, as it worked).

Again, apologies to all innocent bystanders for continuing this...

OK, I'll take a deep breath, (try to) stay calm and spell this out
slowly for you.

The text I posted does say something about merging deleted revisions.
Here is where:

    rev

       Delete the revision rev. For example, `-o 1.3' is equivalent to
       `-o 1.2::1.4'.

This is to be read as 'cvs admin -o1.164 <filename>' is equivalent to
typing 'cvs admin -o1.163::1.165 <filename>'.

    rev1::rev2

       Collapse all revisions between rev1 and rev2, so that CVS only
       stores the differences associated with going from rev1 to rev2,
       not intermediate steps. For example, after `-o 1.3::1.5' one
       can retrieve revision 1.3, revision 1.5, or the differences to
       get from 1.3 to 1.5, but not the revision 1.4, or the
       differences between 1.3 and 1.4.

The important part is the first sentence.  It means that instead of
storing all the small steps necessary to go from one revision to the
next, it stores one big step.  No changes are reverted, undone or
whatever you wish to call it, they are all merged into one big change.

So far for the background, now what was your mistake?  You typed

  cvs admin -o1.164 x11_common.c

in the hopes of undoing the changes made to the file in the commit
number 164 (Such an attempt is not entirely without merit as there are
SCM systems that support reverting changesets, but CVS just isn't one
of them.).  Lather, rinse, repeat with 165 a bit afterwards (I think
the numbers are correct, but that is entirely secondary).  What you
did _not_ do is change revision 1.166, which was the latest revision
of the file.  You only changed the history of the file.

As an undesirable side effect this created a discontinuity (jump,
hole, whatever) in the revision history.  I noticed this and fixed the
discontinuity by removing revision 1.166 and recommitting it as
revision 1.164.  No changes were done by me, I carefully checked that
I recommitted the same file.  So far so bad.  None of this was a good
idea, both things (more or less) mess up the CVS history, either by
creating discontinuities or by committing under a different name and
date, which makes it hard to track commits that go together.  I should
have realized much earlier what the real extent of your error was (not
just creating discontinuities, but failing to make any changes) and I
ordered myself some Coke for that.

What happened then?  You see the diff of my commit on mplayer-cvslog
and flame me to hell and back for "recommitting reversed shit" without
realizing that the lion's share of the Cola belongs to you.  Naturally
my commit did contain the changes you believed to have "reversed",
since they were never really "reversed" in the first place.  Reverting
changes simply is NOT a feature of CVS, no matter how often you claim
the opposite.

You say you checked and it worked and whatnot.  I believe you.  But
whatever tricked you into believing that your attempt at "reversing"
changes at some arbitrary point of the CVS revision history was really
some weird tricky side effect.  Feel free to investigate and find out
what the heck was happening.  It has happened to you before, don't you
remember the stray man page in /usr/local/ incident between you and
Arpi?

I've said it once, I'll say it again, erring is human, we all do (and
I more than most others).  However throwing accusations around without
being able to prove them is _NOT_ acceptable behavior.  Nor is
claiming that the earth is flat.  We do not seem to share the same
reality, but I can assure you that 'cvs admin' does not revert changes
in either reality as we DO share the same CVS program and also the
same laws of physics.

Friendly regards

Diego
(Who should definitely spend his time on more productive avenues.)




More information about the MPlayer-dev-eng mailing list