Re: RFC: Clean up "show remote"

[Daniel Jacobowitz <drow@false.org>]

On Tue, Jan 24, 2006 at 12:44:28AM +0200, Eli Zaretskii wrote:
> > Date: Mon, 23 Jan 2006 00:15:44 -0500
> > From: Daniel Jacobowitz <drow@false.org>
> > 
> > > With your change, the part in parentheses will not be displayed, so
> > > how would the user know that Z0 is a packet that sets software
> > > breakpoints?
> > 
> > Instead, you'll get something that looks like the current output of
> > "show print":
> > 
> > software-breakpoint-packet: Support for the `Z0' packet is auto-detected, currently unknown
> 
> Right, I didn't catch the part before the colon, since it's in another
> place in the code.
> 
> But doesn't the above looks backwards, English-wise? we first announce
> the full packet name, and then say that support for its short alias is
> <whatever>.  The "foo: Bar" syntax is usually reserved for messages
> that come from `foo', which is not really the case here.

Well, it doesn't look at all unnatural to me.  It would if there were
only one of them, but there's a whole list.

To recap (from memory, forgive typos):
(gdb) show remote software-breakpoint-backet
Support for the `Z0' packet is auto-detected, currently unknown
(gdb) show remote
software-breakpoint-packet: Support for the `Z0' packet is auto-detected, currently unknown
hardware-breakpoint-packet: Support for the `Z1' packet is auto-detected, currently unknown
[more...]

> > I'm trying to cut down on how badly this wraps, since we can't
> > word-wrap in translated messages (is there any way to do that?
> 
> The only good way is to break it into two separate messages and mark
> them in i18n comments as belonging to the same phrase.  Translators
> will then break the translated message as appropriate for their
> language (or even produce a single line, by translating the second
> part as an empty string).
> 
> > GCC seems to wrap using spaces even in i18n output; maybe we
> > could do the same, if I am interpreting that right?).
> 
> If that's what GCC does, GCC is not doing TRT: wrapping on spaces
> works for Latin and other similar languages, but can be dead wrong in
> other scripts.

This would be a terrible shame.  I asked for opinions from a couple of
GCC developers, and Paolo Bonzini suggested:

> if a translator is anal about where to break, he should use non-breaking spaces
> (such as Unicode 160) and those answer false for iswspace

> iow, doing mbstowcs, breaking at iswspace, and converting back, should work.
>  but it's sort of a mess

> ask eli if that's correct.  maybe it's not.

Is that right?  If not, is there some forum where I could get a
definitive answer?  No one's ever complained about the GCC wrapping
(well, people have complained about a related problem in the English
wrapping output, but not about bad wrapping in translations as far as I
can see); and it's very convenient.  I'd love to be able to make
GDB word-break output.

Right now we just write overlength messages to the screen.  I don't see
how that can be an improvement over breaking them at whitespace instead
of at the margins, but maybe I'm missing something?

-- 
Daniel Jacobowitz
CodeSourcery