Re: [Gnu-arch-users] Encoding handling proposal

[Charles Duffy]

On Mon, 2004-08-30 at 17:12, Marcus Sundman wrote:
> On Tuesday 31 August 2004 00:24, Charles Duffy wrote:
> > "Move the method with *this* name to be above the method with *this*
> > name" is more robust, and could be applied (or rejected with an
> > intelligent explanation about why it won't apply) -- but doing it means
> > a file-format-aware diff/patch tool pair.
> 
> I agree that such an approach would be more robust, but I don't think the 
> benefits outweigh the problems. I strongly believe in having only one patch 
> file format, but I wouldn't mind having it support some "fancy" 
> transformations. That way features could be added to future diff tools 
> without having to update all existing systems.

I disagree that a "generic patch with extra primitives" is a sufficient
general solution. In particular, it's every bit as useless as GNU diff
in cases where complex document structure transforms are implied by a
patch, or where human-intelligible merge failure messages are needed, or
where blindly doing "best-guess" application would result in a
syntactically invalid (and thus unreadable/uneditable/unfixable)
document -- cases where a document-type-specific diff and patch would
shine.

One of the unfortunate things about arbitrary file types is that many of
them are opaque to the point where "these changes wouldn't merge" output
that's useful to a human simply can't be produced without having the
diff represent high-level changes and having the patch tool apply them
and report in a high-level way what failed.


Anyhow, you ignored the whole thing in my initial post in this thread
about metadata. Is that because you agree with it, or I'm only echoing
sentiments that are already adequately fleshed out, or you merely never
got around to addressing it, or...?