git@vger.kernel.org mailing list mirror (one of many)
 help / color / mirror / code / Atom feed
From: Sanjoy Mahajan <sanjoy@mrao.cam.ac.uk>
To: Linus Torvalds <torvalds@osdl.org>
Cc: David Woodhouse <dwmw2@infradead.org>,
	Junio C Hamano <junkio@cox.net>, Petr Baudis <pasky@ucw.cz>,
	git@vger.kernel.org
Subject: Re: Merge with git-pasky II.
Date: Sat, 16 Apr 2005 13:29:02 -0700	[thread overview]
Message-ID: <E1DMtuY-0002fL-3b@approximate.corpus.cam.ac.uk> (raw)
In-Reply-To: Your message of "Fri, 15 Apr 2005 08:32:46 MST." <Pine.LNX.4.58.0504150753440.7211@ppc970.osdl.org>

> And that "where did this come from" decision should be done at _search_
> time, not commit time.

I like this elegant approach, but clever pattern matching can help even
at commit time.  Suppose hello.c is simply:

  printf ("Hello %d\n", year);

And then developer A updates hello.c to:

  printf ("Hello %d\n", year);
  printf ("And   %d\n", year+1);

Meanwhile developer B updates hello.c to:

  printf ("Hello %d\n", yyyy);

How to merge these two changes?  The psychic solution is

  printf ("Hello %d\n", yyyy);
  printf ("And   %d\n", yyyy+1);

Darcs handles token renames specially, but it's not a general solution
so let's leave it aside.  The example does not have enough information
to make the psychic solution unique or reliable, but imagine that the
example were longer to solve that problem.  You'd want to describe the
delta A(hello.c) as
   
  1. duplicated message line
  2. changed 2nd line a bit

And B(hello.c) as

  1. Changed year to yyyy

In that representation, merging the two deltas becomes

  1. duplicated message line
  2. changed 2nd line a bit
  3. Changed year to yyyy in both lines

Or, by commuting the merge operations and adjusting for their
non-commutativity (in terminology like darcs's -- I'm also a physicist):

  1. Changed year to yyyy
  2. duplicated message line
  3. changed 2nd line a bit

So here some of the computation that Linus wants only at question time
(e.g. 'how did that line get here??') is also useful at merge time.
It's difficult (expensive, unreliable) to describe deltas in the form
above or, worse, to merge two such descriptions, but I hope it
illustrates the point.  And perhaps a robust and easier-to-compute
change-description language can be dreamt up, even if the general
problem of describing changes compactly is not computable -- it's almost
the same, or is the same, as finding the Kolmogorov complexity of a data
set.

Or have I missed a fundamental point?

-Sanjoy

`A society of sheep must in time beget a government of wolves.'
   - Bertrand de Jouvenal

  parent reply	other threads:[~2005-04-16 20:26 UTC|newest]

Thread overview: 148+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2005-04-14  0:29 Merge with git-pasky II Petr Baudis
2005-04-13 21:25 ` Christopher Li
2005-04-14  0:45   ` Petr Baudis
2005-04-13 22:00     ` Christopher Li
2005-04-14  3:51     ` Linus Torvalds
2005-04-14  1:23       ` Christopher Li
2005-04-14  5:03         ` Paul Jackson
2005-04-14  2:16           ` Christopher Li
2005-04-14  6:16             ` Paul Jackson
2005-04-14  7:05       ` Junio C Hamano
2005-04-14  8:06         ` Linus Torvalds
2005-04-14  8:39           ` Junio C Hamano
2005-04-14  9:10             ` Linus Torvalds
2005-04-14 11:14               ` Junio C Hamano
2005-04-14 12:16                 ` Petr Baudis
2005-04-14 18:12                   ` Junio C Hamano
2005-04-14 18:36                     ` Linus Torvalds
2005-04-14 19:59                       ` Junio C Hamano
2005-04-14 20:20                         ` Petr Baudis
2005-04-15  0:42                         ` Linus Torvalds
2005-04-15  2:33                           ` Barry Silverman
2005-04-15 10:02                           ` David Woodhouse
2005-04-15 15:32                             ` Linus Torvalds
2005-04-15 16:01                               ` David Woodhouse
2005-04-15 16:31                                 ` C. Scott Ananian
2005-04-15 17:11                                   ` Linus Torvalds
2005-04-16 15:33                                 ` Johannes Schindelin
2005-04-17 13:14                                   ` David Woodhouse
2005-04-15 19:20                               ` Paul Jackson
2005-04-16  1:44                               ` Simon Fowler
2005-04-16 12:19                                 ` David Lang
2005-04-16 15:55                                   ` Simon Fowler
2005-04-16 16:03                                     ` Petr Baudis
2005-04-16 16:26                                       ` Simon Fowler
2005-04-16 16:26                                       ` Linus Torvalds
2005-04-16 23:02                                         ` David Lang
2005-04-17 14:52                                         ` Ingo Molnar
2005-04-17 15:08                                           ` Brad Roberts
2005-04-17 15:18                                             ` Ingo Molnar
2005-04-17 15:28                                           ` Ingo Molnar
2005-04-17 17:34                                             ` Linus Torvalds
2005-04-17 22:12                                               ` Herbert Xu
2005-04-17 22:35                                                 ` Linus Torvalds
2005-04-17 23:29                                                   ` Herbert Xu
2005-04-17 23:34                                                     ` Petr Baudis
2005-04-17 23:53                                                       ` Kenneth Johansson
2005-04-18  0:49                                                       ` Herbert Xu
2005-04-18  0:55                                                         ` Petr Baudis
2005-04-17 23:50                                                     ` Linus Torvalds
2005-04-18  4:16                                               ` Sanjoy Mahajan
2005-04-18  7:42                                               ` Ingo Molnar
2005-04-16 20:29                               ` Sanjoy Mahajan [this message]
2005-04-16 20:41                                 ` Linus Torvalds
2005-04-15  2:21                       ` [Patch] ls-tree enhancements Junio C Hamano
2005-04-15 16:13                         ` Petr Baudis
2005-04-15 18:25                           ` Junio C Hamano
2005-04-15  9:14                       ` Merge with git-pasky II David Woodhouse
2005-04-15  9:36                         ` Ingo Molnar
2005-04-15 10:05                           ` David Woodhouse
2005-04-15 14:53                             ` Ingo Molnar
2005-04-15 15:09                               ` David Woodhouse
2005-04-15 12:03                         ` Johannes Schindelin
2005-04-15 10:22                           ` Theodore Ts'o
2005-04-15 14:53                         ` Linus Torvalds
2005-04-15 15:29                           ` David Woodhouse
2005-04-15 15:51                             ` Linus Torvalds
2005-04-15 15:54                           ` Paul Jackson
2005-04-15 16:30                             ` C. Scott Ananian
2005-04-15 18:29                               ` Paul Jackson
2005-04-14 18:51                     ` Christopher Li
2005-04-14 19:35                     ` Petr Baudis
2005-04-14 20:01                       ` Live Merging from remote repositories Barry Silverman
2005-04-14 23:22                         ` Junio C Hamano
2005-04-15  1:07                           ` Question about git process model Barry Silverman
2005-04-14 20:23                       ` Re: Merge with git-pasky II Erik van Konijnenburg
2005-04-14 20:24                         ` Petr Baudis
2005-04-14 23:12                       ` Junio C Hamano
2005-04-14 20:24                         ` Christopher Li
2005-04-14 23:31                         ` Petr Baudis
2005-04-14 20:30                           ` Christopher Li
2005-04-14 20:37                             ` Christopher Li
2005-04-14 20:50                               ` Christopher Li
2005-04-15  0:58                           ` Junio C Hamano
2005-04-14 22:30                             ` Christopher Li
2005-04-15  7:43                               ` Junio C Hamano
2005-04-15  6:28                                 ` Christopher Li
2005-04-15 11:11                                   ` Junio C Hamano
     [not found]                                     ` <7vaco0i3t9.fsf_-_@assigned-by-dhcp.cox.net>
2005-04-15 18:44                                       ` write-tree is pasky-0.4 Linus Torvalds
2005-04-15 18:56                                         ` Petr Baudis
2005-04-15 20:13                                           ` Linus Torvalds
2005-04-15 22:36                                             ` Petr Baudis
2005-04-16  0:22                                               ` Linus Torvalds
2005-04-16  1:13                                                 ` Daniel Barkalow
2005-04-16  2:18                                                   ` Linus Torvalds
2005-04-16  2:49                                                     ` Daniel Barkalow
2005-04-16  3:13                                                       ` Linus Torvalds
2005-04-16  3:56                                                         ` Daniel Barkalow
2005-04-16  6:59                                                         ` Paul Jackson
2005-04-16 15:34                                                 ` Re: Re: " Petr Baudis
2005-04-15 20:10                                         ` Junio C Hamano
2005-04-15 20:58                                           ` C. Scott Ananian
2005-04-15 21:22                                             ` Petr Baudis
2005-04-15 23:16                                             ` Junio C Hamano
2005-04-15 21:48                                           ` [PATCH 1/2] merge-trees script for Linus git Junio C Hamano
2005-04-15 21:54                                             ` [PATCH 2/2] " Junio C Hamano
2005-04-15 23:33                                             ` [PATCH 3/2] " Junio C Hamano
2005-04-16  1:02                                               ` Linus Torvalds
2005-04-16  4:10                                                 ` Junio C Hamano
2005-04-16  5:02                                                   ` Linus Torvalds
2005-04-16  6:26                                                     ` Linus Torvalds
2005-04-16  8:12                                                     ` Junio C Hamano
2005-04-16  9:27                                                       ` [PATCH] Byteorder fix for read-tree, new -m semantics version Junio C Hamano
2005-04-16 10:35                                                       ` [PATCH 1/2] Add --stage to show-files for new stage dircache Junio C Hamano
2005-04-16 10:42                                                         ` [PATCH 2/2] " Junio C Hamano
2005-04-16 14:03                                                       ` Issues with higher-order stages in dircache Junio C Hamano
2005-04-17  5:11                                                         ` Junio C Hamano
2005-04-17  5:31                                                           ` Linus Torvalds
2005-04-17  6:01                                                             ` Junio C Hamano
2005-04-17 10:00                                                         ` Summary of "read-tree -m O A B" mechanism Junio C Hamano
2005-04-16 15:28                                                       ` [PATCH 3/2] merge-trees script for Linus git Linus Torvalds
2005-04-16 16:36                                                         ` Linus Torvalds
2005-04-16 17:14                                                           ` Junio C Hamano
2005-04-15 19:54                             ` Re: Merge with git-pasky II Petr Baudis
2005-04-15 10:22                           ` Junio C Hamano
2005-04-15 20:40                             ` Petr Baudis
2005-04-15 22:41                               ` Junio C Hamano
2005-04-15 19:57           ` Junio C Hamano
2005-04-15 20:45             ` Linus Torvalds
2005-04-14  0:30 ` Petr Baudis
2005-04-14 22:11 ` git merge Petr Baudis
     [not found] <000d01c541ed$32241fd0$6400a8c0@gandalf>
2005-04-15 20:31 ` Merge with git-pasky II Linus Torvalds
2005-04-15 23:00   ` Barry Silverman
2005-04-16  0:32     ` Linus Torvalds
  -- strict thread matches above, loose matches on Subject: below --
2005-04-26 18:55 Bram Cohen
2005-04-26 19:58 ` Linus Torvalds
2005-04-26 20:30   ` Tom Lord
2005-04-26 20:31   ` Bram Cohen
2005-04-26 20:39     ` Tom Lord
2005-04-26 20:58     ` Linus Torvalds
2005-04-26 21:25       ` Linus Torvalds
2005-04-26 21:28       ` Bram Cohen
2005-04-26 21:36         ` Fabian Franz
2005-04-26 22:30           ` Linus Torvalds
2005-04-26 22:25         ` Linus Torvalds
2005-04-28  0:42           ` Petr Baudis
2005-04-26 21:26     ` Diego Calleja
2005-04-26 20:31   ` Daniel Barkalow
2005-04-26 20:44     ` Tom Lord

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

  List information: http://vger.kernel.org/majordomo-info.html

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=E1DMtuY-0002fL-3b@approximate.corpus.cam.ac.uk \
    --to=sanjoy@mrao.cam.ac.uk \
    --cc=dwmw2@infradead.org \
    --cc=git@vger.kernel.org \
    --cc=junkio@cox.net \
    --cc=pasky@ucw.cz \
    --cc=torvalds@osdl.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
Code repositories for project(s) associated with this public inbox

	https://80x24.org/mirrors/git.git

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).