Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: st: Version control question


From   khigbee@stata.com
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: Version control question
Date   Mon, 03 Oct 2005 13:36:26 -0500

"Mark Schaffer" <M.E.Schaffer@hw.ac.uk> asks:

> Under Stata 9.0 and earlier, up to 20 estimation results could be stored
> using -estimates store-.
> 
> Under Stata 9.1, this limit has been increased to 300 estimation results.
> 
> If I have a program that starts with, say, -version 8.2-, and I run it
> under Stata 9.1, will the maximum number of estimation results that the
> program can store be 20 or 300?

Your old program (with a -version- statement less than 9.1) run
in a modern Stata version 9.1 will allow the 300 estimation
results to be stored.  Setting the -version- to a smaller number
does not turn off this new feature.

Having an older -version #- statement does not turn off the new
features introduced in later versions.  It only changes the
behavior of commands for those limited cases where version
controlling is really needed.  Take a look at

    -help version-

and scroll down to the section titled "Summary of version
changes" about a third of the way down the file.  This section
lists what behavior differs based on the -version- you set in
your program.

Here are some guidelines on what setting -version #- does and
does not do:

    1) When the syntax of a command changes so that the old and
       new syntaxes are NO LONGER COMPATIBLE, then setting
       -version #- to the older version makes the command
       understand the older syntax.  Otherwise the new syntax is
       understood.

       For many of the commands that we alter, the old and new
       syntaxes do not conflict and both continue to be
       understood (but only the new syntax continues to be
       documented).

    2) If you place an older version number at the top of a
       command or do file, it does not keep you from using more
       modern features (except in those few cases where version
       control had to be imposed, see #1).

       Sometimes people think that if they are running Stata 9.1
       and they place a -version 8- at the top of a program or do
       file, get it running in their modern Stata, and then give
       it to someone who only has Stata 8 (has not purchased
       Stata 9) that the person using Stata 8 will be able to run
       the program or do file.  The Stata 8 user will only be
       able to run it if you avoid all the new commands and
       language features found in your more modern Stata.

    3) We do not version control buggy behavior.

       I don't think anyone would want us to restore a bug when
       -version- is set to an older number -- and we do not.

    4) We USUALLY do not version control output style
       modifications.  For instance, if we make the output of a
       command look nicer (use smcl, clickable links in the
       output giving additional help of information, slight
       changes in the spacing of the columns, ...), the nicer
       look happens regardless of the -version #- setting.

Ken Higbee    khigbee@stata.com
StataCorp     1-800-STATAPC

*
*   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/



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index