RE: [Groff] behaviour of \fP

[Nicola Bernardini]

Sorry for being late in posting.
It seems to me that if a font was not found in the current device,
it would remap the current font to itself anyway, so that reverting
back would give the previous font (which was always the same, but
at least it is not the wrong one).
Being really old and a user of the original DWB I sometimes happen to
remember how the original troff used to work :)

Nicola

------------------------------------------------------------------------
Nicola Bernardini
E-mail: address@hidden

> On 21-Jan-00 Werner LEMBERG wrote:
> > 
> > Consider the following with the ascii device, using tmac.an:
> > 
> > 
> > This is
> > .B bold
> > text.
> > And this is \fCtypewriter\fP text.
> > 
> > 
> > As you can see, the last `text.' appears in bold face, which is of
> > course unwanted.
> > 
> > Question: Shall we change the behaviour of `remembering the previous
> > font'?  I can imagine the following:
> > 
> >   If a font is selected which doesn't exist for the current device,
> >   ignore it but update the `previous font' so that \fP gives a
> >   reasonable result.
> > 
> > What's the behaviour with the original troff?  Do you think that this
> > change would cause a lot of inconsistencies?

The way the original troff used to work was that '.B' was working
for only that word, then the current font would change back to whatever
it was set to, then switch to 'C' font (and if not ignoring that font
> 
> I don't know how "original troff" behaves in this respect. However,
> I believe that groff should do the decent thing in these circumstances.
> If absence of font 'C' means that there's no font change, than
> \fP should be ignored as well. If I understand Werner's proposal
> correctly, this would be achieved by "copying" the current font
> to the previous font if the font change request fails.
> 
> This would (or should) make the phenomenon device-independent:
> following \fP, the text would be in the intended font whether
> or not a preceding \f[anything] failed because the requested font did
> not exist (which you have to accept anyway).
> 
> I've tried to think of circumstances in which Werner's suggestion
> would do something unwanted, but haven't seen anything so far.
> 
> Ted.
> 
> --------------------------------------------------------------------
> E-Mail: (Ted Harding) <address@hidden>
> Date: 24-Jan-00                                       Time: 04:09:35
> ------------------------------ XFMail ------------------------------
> 
> _______________________________________________
> Groff maillist  -  address@hidden
> http://ffii.org/mailman/listinfo/groff
>