[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: [RFC] Column alignment in debconf {multi,}selects, take 2

On Thursday 06 March 2008, Jérémy Bobbio wrote:
> After creating this template to see how it could look like, I have
> noticed the alignment issue.  IMHO, having right alignment support is a
> minor improvement over no alignment support, so I would prefer to get
> alignment support in-tree first.

Yes, absolutely.

> Well… while trying to figure out where to put the needed ${!TAB} in
> partman code I have discovered several things:
> But anyway, I am not pushing changes in partman yet.  I do think there is
> a little bit of restructuring needed first.

Yes. We have talked in the past of changing to Choices-C in partman to maybe 
simplify things. Colin recently did a commit to improve preseeding in a few 
partman dialogs, but I have not yet looked at his changes.
I agree it will require quite a bit of work to get it right, but it should 
be worth the effort.

> I wonder if we should just not drop the dash completely: it just looks
> like an inverted column separator, now that alignment is done
> correctly. [3,4]

As already commented on IRC, I think the dash is useful as a kind of link 
between the English and translated names. IMO Christian should have the 
final word on this.

> I think you overlooked how the implementation works in the text or newt
> frontend: the proposed patch is not using actually sending '\t' to any
> terminals.  It is just using the character internally as the delimiter
> for columns, like the GTK+ needs.  These delimiters are replaced by the
> correct amount of spaces by strutl.c:stralign().

I probably did (or just did not understand the code with my limited C).
Thx for the explanation.

Attachment: signature.asc
Description: This is a digitally signed message part.

Reply to: