GNU bug report logs - #69454
Not possible to insert an empty vtable

Previous Next

Full log

View this message in rfc822 format

From: Joost Kremers <joostkremers <at> fastmail.fm>
To: Adam Porter <adam <at> alphapapa.net>
Cc: 69454 <at> debbugs.gnu.org, Eli Zaretskii <eliz <at> gnu.org>, larsi <at> gnus.org, eric.marsden <at> risk-engineering.org
Subject: bug#69454: Not possible to insert an empty vtable
Date: Thu, 02 May 2024 09:31:01 +0200

Hi Adam,

On Tue, Apr 30 2024, Adam Porter wrote:
> Well, I'm not sure how useful it would be to calculate the widths of columns and
> show them if there are no data objects to render in them; especially, if the
> columns have to be recalculated and the table has to be re-rendered as soon as
> objects are shown in it.

Yes, that's what I meant when I said the code using vtable.el would have to take
that into consideration. In my case, I do that by leaving only the last column's
width undefined.

>  It would seem to merely serve as confirmation of which
> columns are defined in the vtable.

Yup, that is exactly right, but I think there are circumstances in which that
may be useful. Specifically, cases where the data being displayed is generated
by the user. (Think database front-end, even though I realise vtable.el isn't
suitable for displaying extremely large amounts of data. In my use-case, the
amount of data is manageable, though.)

> IOW, it seems that, if there are no objects to show, it might be just as good to
> short-circuit the columns and just insert some kind of "[no objects]" string.
>
> I don't have a strong opinion on this, though.  I think I would generally favor
> whichever approach required the smallest change to the code.

In that case, go with your suggestion, because mine certainly would require
bigger changes. 😀 


-- 
Joost Kremers
Life has its moments

Previous Next

GNU bug tracking system
Copyright (C) 1999 Darren O. Benham, 1997,2003 nCipher Corporation Ltd, 1994-97 Ian Jackson.