Just a quick note. What was Stata thinking (by investing in a Windows
version)? I know, I know, lots of people use Windows. But, that the
machine crashes probably has nothing to do with Stata. It is quite likely
that some or many of those people writing Stata about the machine crashing
when Stata does something have the same problem whenever they use *any*
non-MS software--netscape, WordPerfect, anything. That has been my
experience with *every* windows operating system since MS "superceded"
Windows 3.1--the last one where you could easily switch out to DOS.
Windows crashes at the slightest wind. Every release is promised to be
stable, and every release fails. A major reason for this is MS *never*
tells non-MS programmers all the rules of Windows, so that non-MS
programmers will necessarily write software that will at times clash with
some arcane Windows rule (and crash or hang the system). And, MS
incompetence means even their own non-OS software, hastily written, poorly
tested, and user-unfriendly, also crashes constantly.
MS used the former mentioned predatory technique and others at the outset,
and continues to do so. I hope Stata will not end up diverting more and
more energy and investment to the impossible task of writing stable
software for an OS for which Stata will never receive complete
documentation. Instead, I hope Stata pays equal or more attention to
writing for operating systems that are designed for users to get work done
(e.g., linux) as opposed to operating systems designed to maintain the
dominance of one northwestern U.S. corporation. It is only by this
approach that Stata will avoid getting bogged down into the quagmire of
spending more and more effort just trying to get the menus right on the
screen, and will be able instead to steadily introduce truly useful
advances (e.g., more models, more options, more presentation quality
stuff--more more more!).
Not to start an MS-war, but something *had* to be said re: the thread
about windows crashing when something is changed.
Respectfully,
Sam
On Sun, 5 Oct 2003, Nick Cox wrote:
> Fred Wolfe
>
> > This is to report a graph box problem, Stata se 8.1, up to
> > date, windows 2000
> >
> > This line works
> > graph box peuro eq5d5 eq5d5sf sfindex euroqol
> >
> > Addition of the second line crashes Stata and throws me out
> > to operating
> > system.
> >
> > graph box peuro eq5d5 eq5d5sf sfindex euroqol ///
> > legend( label(1 "HAQ") label(2 "EQ5D5") label(3 "EQ5D5-SF") label(4
> > "SF-6D") label(5 "EuroQol"))
> >
> > The legend line worked perfectly in other non-box graphs.
>
> There is nothing wrong with the -legend()- call.
> However, you need a comma before the option, i.e.
> just before or just after the "///".
>
> Crashing may seem a little disproportionate
> here as a response to this small syntax error.
>
> However, as far as Stata is concerned
>
> legend( label(1 "HAQ") label(2 "EQ5D5") label(3 "EQ5D5-SF") label(4
> "SF-6D") label(5 "EuroQol"))
>
> is all part of your varlist, and for other reasons parentheses ( )
> are allowed as part of a varlist specification.
>
> My wild guess is that Stata is thrown by the _nesting_ of
> parentheses -- in, as said, what it thinks is a varlist.
>
> Nick
> [email protected]
>
> *
> * 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/
