Git development
 help / color / mirror / Atom feed
From: "Philip Oakley" <philipoakley@iee.org>
To: "Thomas Gummerer" <t.gummerer@gmail.com>, <git@vger.kernel.org>
Cc: <trast@student.ethz.ch>, <gitster@pobox.com>,
	<mhagger@alum.mit.edu>, <peff@peff.net>, <spearce@spearce.org>,
	<davidbarr@google.com>
Subject: Re: Index format v5
Date: Fri, 4 May 2012 14:25:05 +0100	[thread overview]
Message-ID: <2A6D903A618147EEBBFEF99234104EE6@PhilipOakley> (raw)
In-Reply-To: CALgYhfMKdbv8TiT4ALDSvD3pSXHEPLWHM09DxYnRmRdBWRjh8Q@mail.gmail.com

From: "Thomas Gummerer" <t.gummerer@gmail.com> Sent: Thursday, May 03, 2012 
6:25 PM
>I have been drafting the Version 5 of the index format over the past
> few days with the help of Thomas Rast, Michael Haggerty, cmn and
> barrbrain on IRC. It will save with prefix compression on the path, and
> using a crc32 over the stat data, instead of the full data, since it is 
> only
> used for checking if the file is changed. (Thanks Michael Haggerty for
> this hint. Unless we are missing something this will save another
> ~4 MB on the Webkit index.
>
>
>
> GIT index format
> ================
>
> = The git index file has the following format
>
xxx
>
> == Directory entry
>
>  Directory entries are sorted in lexicographic order by the name
>  of their path starting with the root.
>
>  Path names (variable length) relative to top level directory (without the
>    leading slash). '/' is used as path separator. '.' indicates the root
>    directory. The special patch components ".." and ".git" (without 
> quotes)
>    are disallowed. Trailing slash is also disallowed.
>

Does the prohibition of ".git" prevent the potential for versioning of the 
.git directory itself (e.g. .gitignore the pack & objects themselves)?

It should be possible for one's current repo status to be tracked and 
summarised by a single sha1, say as an indepenedent branch, as you would the 
code itself.

[my use case is in a managed environment where one has plenty of networked 
project storage, but no real availability of a network server, so anybody 
and his mate could corrupt ones network repo with badly thought through 
tweaks. The lack of a server is the root cause, but that's not something the 
project can fix, so thinks ... 'version the meta data...']

I can see that the lack of the leading "/", and the ".." are a policy 
statement about paths being relative to top level directory, with direct 
path referencing. The typical avoidance of ".git" should be a note about how 
regular git works, rather than an absolute prohibition.

Philip 

  parent reply	other threads:[~2012-05-04 13:25 UTC|newest]

Thread overview: 49+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2012-05-03 17:25 Index format v5 Thomas Gummerer
2012-05-03 18:16 ` Thomas Rast
2012-05-03 19:03   ` Junio C Hamano
2012-05-04  7:12   ` Michael Haggerty
2012-05-07 22:18     ` Robin Rosenberg
2012-05-03 18:21 ` Ronan Keryell
2012-05-03 20:36   ` Thomas Gummerer
2012-05-03 18:54 ` Junio C Hamano
2012-05-03 19:11   ` Thomas Rast
2012-05-03 19:31   ` Thomas Rast
2012-05-03 19:32     ` Thomas Rast
2012-05-03 20:32       ` Junio C Hamano
2012-05-03 21:38   ` Thomas Gummerer
2012-05-07 18:57     ` Robin Rosenberg
2012-05-03 19:38 ` solo-git
2012-05-04 13:20 ` Nguyen Thai Ngoc Duy
2012-05-04 15:44   ` Thomas Gummerer
2012-05-04 13:25 ` Philip Oakley [this message]
2012-05-04 15:46   ` Junio C Hamano
2012-05-06 10:23 ` Nguyen Thai Ngoc Duy
2012-05-07 13:44   ` Thomas Gummerer
2012-05-06 16:49 ` Phil Hord
2012-05-07 13:08   ` Thomas Gummerer
2012-05-07 15:15 ` Michael Haggerty
2012-05-08 14:11   ` Thomas Gummerer
2012-05-08 14:25     ` Nguyen Thai Ngoc Duy
2012-05-08 14:34       ` Nguyen Thai Ngoc Duy
2012-05-10  6:53         ` Thomas Gummerer
2012-05-10 11:06           ` Nguyen Thai Ngoc Duy
2012-05-09  8:37     ` Michael Haggerty
2012-05-10 12:19       ` Thomas Gummerer
2012-05-10 18:17         ` Michael Haggerty
2012-05-11 17:12           ` Thomas Gummerer
2012-05-13 19:50             ` Michael Haggerty
2012-05-14 15:01               ` Thomas Gummerer
2012-05-14 21:08                 ` Michael Haggerty
2012-05-14 22:10                   ` Thomas Rast
2012-05-15  6:43                     ` Michael Haggerty
2012-05-15 13:49                   ` Thomas Gummerer
2012-05-15 15:02                     ` Michael Haggerty
2012-05-18 15:38                       ` Thomas Gummerer
2012-05-19 13:00                         ` Michael Haggerty
2012-05-21  7:45                           ` Thomas Gummerer
2012-05-16  5:01                     ` Michael Haggerty
2012-05-16 21:54                       ` Thomas Gummerer
2012-05-19  5:40                         ` Michael Haggerty
2012-05-21 20:30                           ` Thomas Gummerer
2012-05-13 21:01 ` Philip Oakley
2012-05-14 14:54   ` Thomas Gummerer

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

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

  git send-email \
    --in-reply-to=2A6D903A618147EEBBFEF99234104EE6@PhilipOakley \
    --to=philipoakley@iee.org \
    --cc=davidbarr@google.com \
    --cc=git@vger.kernel.org \
    --cc=gitster@pobox.com \
    --cc=mhagger@alum.mit.edu \
    --cc=peff@peff.net \
    --cc=spearce@spearce.org \
    --cc=t.gummerer@gmail.com \
    --cc=trast@student.ethz.ch \
    /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.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox