Boost logo

Boost :

Subject: Re: [boost] [git] Line endings in bootstrap.bat on Windows
From: Dave Abrahams (dave_at_[hidden])
Date: 2013-11-27 12:31:57


Steven Watanabe <watanabesj_at_[hidden]> writes:

> AMDG
>
> On 11/27/2013 08:03 AM, Dave Abrahams wrote:
>>
>> What do you consider to be "correct" handling?
>>
>> If you can figure out what's really desired, we can easily implement it.
>> Or, maybe Niall's patch already implements what you want. So far,
>> though, there have been so many emphatic demands from the community that
>> no history be "lost," that we have been extremely reluctant to process
>> the file contents as part of the conversion.
>>
>
> That is exactly the problem. You /are/ losing part of
> the history.

Which part?

If you mean the svn:eol-style attribute, then it is correct that we're
not representing it in the converted repository. We could represent
that in various ways in Git without actually altering the bytes of the
file. Or we could alter the bytes of the file, but that would itself be
lossy.

> There are really two options:
> a) Correctly preserve svn:eol-style, however git represents it,

Git doesn't represent it.

>
> so that the files produced by a git checkout are always binary
> identical to those produced by svn checkout.

How things are checked out by Git depends on a user's local settings
(e.g. autocrlf), so there is no way we can represent things so that a
git checkout is *always* identical to an SVN checkout.

(Is the same true of SVN?)

> b) Change the line endings to match whatever the rules are in the git
> repository.

This is imprecisely stated. What rules?

> I would strongly favor option (a), unless git is
> too broken to handle it.

Again, the problem is that there's no consensus around what "correct"
means, here, and nobody has been able to articulate precisely what is
desired.


Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk