On not claiming bugs
====================
A small point, but I'll finish what I started.
You said
"Is this a bug in Stata for mac or does everyone get the same result
(the title for the t-axis "year" remains visible)."
The advice I quoted from the FAQ (which is advice, and not a rule;
there are no rules in the FAQ) is very well meant and based on the
total experience of this list, for several reasons. The most
obvious, but one not applicable here, is that it is not very
diplomatic to ask questions about user-written programs
implying that problems are really the programmer's fault!
The tactful pose is always that the user has missed something.
Of course, all conscientious Stata programmers want to know
about their bugs and to fix them as quickly as possible.
Even with StataCorp, presumed to have thicker skins, I divide
it up this way. As people begin learning Stata, they may be
tempted to blame Stata for what almost always turn out to be
their own errors or misunderstandings. In fact the same
temptation remains after 10 or 20 years of use. Conversely,
very experienced people tend to go straight to StataCorp
in private emails with what they think are bugs. Either way,
there is little reason to claim a bug in public.
-color(none)-
=============
You have a very good point about "color(none)".
I thought at first that Stata was just ignoring
that as irrelevant, but that is I think not the
case.
Again, I don't think this is a bug, but an exception. The
standard way of blanking out text is to blank it
out with "". Blanking it out with -color(none)- would make
definite sense in the way you argue, but I guess it is not
allowed for the following reason. In general, if you just change
the colour of something, in this case text, it is still there
with the same size and shape, just a different colour, even if that
is "none". In particular, whatever would be next to it remains
in the same position as before. But that is not what
you expect if you blank out text. If you destroy an inner
title, you expect an outer title to move closer, and
so forth. So I guess that allowing
-color(none)- to destroy things would create other difficulties.
This therefore raises a question of consistency across different
graphics elements given the principle that the graphics
code is written in as general a way as possible.
There is no doubt much more to be said by Stata's developers.
should they choose to correct or amplify this guess.
Nick
[email protected]
> -----Original Message-----
> From: [email protected]
> [mailto:[email protected]]On Behalf Of Fredrik
> Wallenberg
> Sent: 19 May 2004 23:53
> To: [email protected]
> Subject: Re: st: Re: Transparent background in graphics
>
>
> I am indeed using Acrobat (or sometimes Keynote) for my presentations,
> but the problem would be the same if I was printing transparencies...
> I still cannot overlay the Stata graphic any background image and have
> that image show through.
>
> As for "Is this a bug?" I did *NOT* say "This is a bug" and thus I
> think my comment is allowed. Now please tell me what color(none) is
> supposed to do. I quote from the help file regarding colors:
>
> " none no color; invisible; draws nothing"
>
> Well, I am going to change my statement and say that YES, IT
> IS A BUG! ;-)
>
> Fredrik
> *
> * For searches and help try:
> * http://www.stata.com/support/faqs/res/findit.html
> * http://www.stata.com/support/statalist/faq
> * http://www.ats.ucla.edu/stat/stata/
>
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
