git.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Shawn Pearce <spearce@spearce.org>
To: Junio C Hamano <gitster@pobox.com>
Cc: git@vger.kernel.org
Subject: Re: [PATCH 09/11] streaming: read non-delta incrementally from a pack
Date: Sun, 15 May 2011 17:58:46 -0700	[thread overview]
Message-ID: <BANLkTin7DALU8sfFxzJo62DGtNE2fsZxXw@mail.gmail.com> (raw)
In-Reply-To: <1305505831-31587-10-git-send-email-gitster@pobox.com>

On Sun, May 15, 2011 at 17:30, Junio C Hamano <gitster@pobox.com> wrote:
> +static read_method_decl(pack_non_delta)

I am not a huge fan of these decl macros... but I can see how writing
out the same function prototype 3 times is annoying.

> +       switch (in_pack_type) {
> +       default:
> +               return -1; /* we do not do deltas for now */

Haha. Deltas are going to be painful. Very, very painful.

We actually try to stream deltas in JGit. Our implementation isn't
useful. It only works if the base object is only accessed in
sequential order by the delta instruction sequences.

I had plans to add an external delta base cache to
$GIT_DIR/objects/delta-base-cache using a block file format that is
random accessible, but has CRC-32 checksums on each block to still
ensure there isn't silent data corruption on reads. You can read more
about it here http://egit.eclipse.org/r/1724 but the patch is probably
stalled and will get abandoned.

I think the better strategy is to avoid delta compression altogether
for objects that are so big we cannot materialize them as a contiguous
buffer. What a reasonable limit is, I don't know... but its probably
got to be around 25-50 MB. The Android project (as an example) has 6+
MB XML documents in their source code repository that are very delta
compressible.

-- 
Shawn.

  reply	other threads:[~2011-05-16  0:59 UTC|newest]

Thread overview: 49+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2011-05-16  0:30 [PATCH 00/11] writing out a huge blob to working tree Junio C Hamano
2011-05-16  0:30 ` [PATCH 01/11] packed_object_info_detail(): do not return a string Junio C Hamano
2011-05-17  0:45   ` Thiago Farina
2011-05-17  2:36     ` Junio C Hamano
2011-05-16  0:30 ` [PATCH 02/11] sha1_object_info_extended(): expose a bit more info Junio C Hamano
2011-05-16  0:30 ` [PATCH 03/11] sha1_object_info_extended(): hint about objects in delta-base cache Junio C Hamano
2011-05-16  0:40   ` Shawn Pearce
2011-05-16  0:30 ` [PATCH 04/11] unpack_object_header(): make it public Junio C Hamano
2011-05-16  0:30 ` [PATCH 05/11] write_entry(): separate two helper functions out Junio C Hamano
2011-05-16  0:30 ` [PATCH 06/11] streaming: a new API to read from the object store Junio C Hamano
2011-05-18  8:09   ` Jeff King
2011-05-19  1:52     ` Junio C Hamano
2011-05-16  0:30 ` [PATCH 07/11] streaming_write_entry(): use streaming API in write_entry() Junio C Hamano
2011-05-16  0:30 ` [PATCH 08/11] streaming_write_entry(): support files with holes Junio C Hamano
2011-05-16 10:53   ` Nguyen Thai Ngoc Duy
2011-05-16 14:39     ` Junio C Hamano
2011-05-17  1:18       ` Nguyen Thai Ngoc Duy
2011-05-17  5:23         ` Junio C Hamano
2011-05-16 13:03   ` Thiago Farina
2011-05-16  0:30 ` [PATCH 09/11] streaming: read non-delta incrementally from a pack Junio C Hamano
2011-05-16  0:58   ` Shawn Pearce [this message]
2011-05-16  5:00     ` Junio C Hamano
2011-05-16  0:30 ` [PATCH 10/11] sha1_file.c: expose helpers to read loose objects Junio C Hamano
2011-05-16  0:30 ` [PATCH 11/11] streaming: read loose objects incrementally Junio C Hamano
2011-05-16  0:47 ` [PATCH 00/11] writing out a huge blob to working tree Shawn Pearce
2011-05-18  6:41 ` Jeff King
2011-05-18  7:08   ` Jeff King
2011-05-18  7:50     ` Jeff King
2011-05-18 15:12       ` Junio C Hamano
2011-05-18  8:17 ` Jeff King
2011-05-19 21:33 ` [PATCH v2 " Junio C Hamano
2011-05-19 21:33   ` [PATCH v2 01/11] packed_object_info_detail(): do not return a string Junio C Hamano
2011-05-19 21:33   ` [PATCH v2 02/11] sha1_object_info_extended(): expose a bit more info Junio C Hamano
2011-05-19 21:33   ` [PATCH v2 03/11] sha1_object_info_extended(): hint about objects in delta-base cache Junio C Hamano
2011-05-20 23:05     ` René Scharfe
2011-05-21  1:49       ` Junio C Hamano
2011-05-19 21:33   ` [PATCH v2 04/11] unpack_object_header(): make it public Junio C Hamano
2011-05-19 21:33   ` [PATCH v2 05/11] write_entry(): separate two helper functions out Junio C Hamano
2011-05-19 21:33   ` [PATCH v2 06/11] streaming: a new API to read from the object store Junio C Hamano
2011-05-20 23:05     ` René Scharfe
2011-05-21  1:49       ` Junio C Hamano
2011-05-19 21:33   ` [PATCH v2 07/11] streaming_write_entry(): use streaming API in write_entry() Junio C Hamano
2011-05-20 22:52     ` Junio C Hamano
2011-05-19 21:33   ` [PATCH v2 08/11] streaming_write_entry(): support files with holes Junio C Hamano
2011-05-19 21:33   ` [PATCH v2 09/11] streaming: read non-delta incrementally from a pack Junio C Hamano
2011-05-19 21:33   ` [PATCH v2 10/11] sha1_file.c: expose helpers to read loose objects Junio C Hamano
2011-05-19 21:33   ` [PATCH v2 11/11] streaming: read loose objects incrementally Junio C Hamano
2011-05-19 21:44   ` [Not A PATCH v2 02/11] interdiff Junio C Hamano
2011-05-19 22:21   ` [PATCH v2 00/11] writing out a huge blob to working tree Jeff King

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=BANLkTin7DALU8sfFxzJo62DGtNE2fsZxXw@mail.gmail.com \
    --to=spearce@spearce.org \
    --cc=git@vger.kernel.org \
    --cc=gitster@pobox.com \
    /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;
as well as URLs for NNTP newsgroup(s).