Skip to Content.
Sympa Menu

sm-discuss - Re: [SM-Discuss] patches

sm-discuss AT lists.ibiblio.org

Subject: Public SourceMage Discussion List

List archive

Chronological Thread  
  • From: Ismael Luceno <ismael.luceno AT gmail.com>
  • To: sm-discuss AT lists.ibiblio.org
  • Subject: Re: [SM-Discuss] patches
  • Date: Mon, 30 Apr 2012 19:45:27 -0300

On Mon, 23 Apr 2012 13:00:29 +0900
flux <flux AT sourcemage.org> wrote:
...
> Regardless of whether git compresses the actual file or not, this has
> no bearing on whether the file will be compressed for any instance of
> the grimoire not coming from a git clone. This means all "regular"
> users of the grimoire (scribe add stable, scribe add test, etc.) will
> only have the uncompressed file.
>
> Additionally, even in git clones, the file is only compressed by git
> in the git index. Any working tree version of the file is not
> compressed by git, but is as-is.
>
> The question should not be "does git compress this?". The question
> should be "do we want a 10kb file showing up in all (working)
> instances of the grimoire?".
>
> Personally, I'm fine either way in this case, as a single instance of
> 10kb added to a grimoire that sits at over 100mb as it is won't matter
> much. But consider this: if we have 1000 patch files, and each patch
> is 10kb, that will add 10mb to the total size. Continue ad nauseum
> for the full implication. Whether this becomes a practical reality is
> up to the collective vision, but I believe there was some policy
> about a minimum size for a patch file to be compressed.
>

I think it's worth considering compressing the whole grimoire, as I
proposed earlier...

The point is that, without tail-merging, the most waste comes from
file sizes not being multiple of the filesystem block size.

Attachment: signature.asc
Description: PGP signature




Archive powered by MHonArc 2.6.24.

Top of Page