Statalist


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: st: RE: egen & sum()


From   "Nick Cox" <n.j.cox@durham.ac.uk>
To   <statalist@hsphsun2.harvard.edu>
Subject   RE: st: RE: egen & sum()
Date   Wed, 26 Nov 2008 12:48:48 -0000

It is crucial to distinguish between 

1. Command -foobar- changes in behaviour. Either the old behaviour is no
longer obtainable (clear cases: correction of bugs, addition of new
options) or you need to invoke -version- to get the old behaviour. 

2. Command -foobar- is superseded or made redundant by something else. 
For the most part, the old stuff lies around indefinitely. Anyone
remember -fit- for example? In Stata 10.1 -fit- is still there,
unchanged since 1998. -for- is still there, etc. If something has gone
completely, the odds are you won't miss it (e.g. the very old -menu-).
Whether it is documented is a different matter and a delicate one. I
have kept all my old manuals from version 2 onwards, but many are pretty
dusty and I think the idea that everything should be documented for ever
more is a little Utopian. 

Nick 
n.j.cox@durham.ac.uk 

Richard Williams

At 05:29 AM 11/26/2008, Nick Cox wrote:
>Although it's StataCorp call, there is no obvious reason why any of the
>old names should ever stop working. The most obvious reason for change
>is that StataCorp might want to reclaim some of them, but as said they
>are now considered objectionable.

I don't know about these particular functions, but Statacorp's policy 
in the past has been that names can stop working once you get to the 
next version, e.g. if a name was changed in Stata 9 the old name will 
continue to work in Stata 9, but once you get to Stata 10 it is only 
guaranteed to work under version control.  Personally, I would prefer 
that old names continue to work even without version control for all 
eternity.  The now undocumented -index- function was brought back to 
life when several baffled and bewildered people who couldn't figure 
out what the new name was requested it.

It probably is better to break down and learn and use the new names, 
since anyone else who is reading your code will be baffled and 
bewildered by your use of functions that are no longer documented.

Given that, thanks to version control, very few commands ever die 
completely, I continue to wish that there was some easy way to get 
help for all these things that now go undocumented.

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   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