zim-wiki team mailing list archive

[Jaap Karssenberg <jaap.karssenberg@xxxxxxxxx>]

On Tue, Aug 20, 2013 at 11:29 AM, Klaus-Dieter Bauer <
bauer.klaus.dieter@xxxxxxxxx> wrote:

> 2013/8/20 Jaap Karssenberg <jaap.karssenberg@xxxxxxxxx>:
> > [...]
> >
> > The suffix, or the "-attachments" folders solve a problem for indexing.
> This
> > is a problem I will solve differently when refactoring the index. The
> idea
> > is that zim will inspect the content of text files to determine whether
> or
> > not it are pages. I don't mind allowing alternative suffix as well if
> that
> > makes interoperability easier, but for default will stick to ".txt".
> >
> > [...]
> >
> > Regards,
> >
> > Jaap
>
> Just a quick list of issues that are raised by the ambigous file
> extension and would require major (probably OS-dependent) pains to fix
> without changing the extension:
>
>   + No distinction from plain text files by icon (inconvenient e.g.
>     when viewing results of desktop search).
>   + No double-click-to-open in file browsers / search results without
>     breaking the behaviour for non-zim .txt files.
>   + ZIM must check .txt files for ZIM page format,
>       - when indexing
>       - when listing attachments
>       - ... ?
>   + When processing notebooks through an interactive shell,
>     distinguishing normal text files from zim page files is unviable
>     and inconvenient when processing them with shell scripts.
>
> Is there some reason, why you want to keep the ".txt" extension? I
> can't see much of an advantage.


I'm sorry to be a bit blunt, but the point in the previous mail was not to
hijack the previous topic with this discussion - so see my changed topic
line for this mail.

The topic on the file extension has been discussed several times on this
list. I do not deny that there are issues with the current format, but I
have not yet seen a workable solution that I could apply without much work
on my side.

In short, when you change the extension you will still need to support
existing notebooks with .txt files. This means you have still have to solve
detection of txt files in zim + you need to deal with multiple extensions.
How do you deal with conflicts where there are two "pages" with the same
name but different file extension - which one is the real page ?

I do have thinks on my roadmap to put things in place to ready the way to
allow multiple extensions. Ones we can support multiple, we can support
switching.

Admitted, this is a slow process, especially as I have not been able to do
much coding in the last few months. If you want it faster, please help out
with implementing an upgrade path for existing notebooks.

Regards,

Jaap