[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]
Re: st: RE: mulitple comments using // and #d;
On Sep 1, 2006, at 5:07 PM, SamL wrote:
Please don't do follow Mike's advice. Elegant though it may be to
-delimit-, several factors necessitate maintaining the command. I'll
mention only two.
One need not use delimit if one does not want to do so. But retiring
Sam (and others):
command? That seems unwarranted.
On Fri, 1 Sep 2006, Michael S. Hanson wrote:
At best, I would suggest, it is a relic of times gone by, and should
be depreciated in favor of ///.
Notice I never suggested that the -delimit- command be eliminated.
It obviously would be undesirable for legacy code. As noted above, I
suggested it should be depreciated. Type -help fit- for one example
of depreciated code.
You could still use -fit- to estimate a linear regression -- indeed, I
have a colleague who uses it frequently -- but it is depreciated:
there are no examples of -fit- in the manuals (that I could find), and
it is not recommended (or even mentioned) to most users. Due to the
occasional unusual behavior of -delimit- (evidenced both by the
exchange earlier in this thread and my own experience teaching Stata in
my courses), I would suggest that it too be depreciated.
Substantively, I would like to encourage those who provide code for
learning Stata (manuals, on-line resources, etc.) eschew the use of
-delimit- and * comments, and adopt the more modern use of // comments
and /// line continuations.
Obviously, individual users are free to use whatever conventions they
prefer -- just as people can still use -fit- if they choose. But when
there exist alternatives that are better for most users, IMO those
should be promoted.
* For searches and help try: