On Wed, 6 Jul 2011, Sven Schreiber wrote: > to prove that at least some (n>=1) people do read the manual here are > some things I noticed today: > > 1) I once was taught on this list that "loop for i=1..x" is just a > tolerated syntax error, not the correct form (which is "loop i=1..x"). > However, I noticed today that this idiom is used in the manual, so I > suggest to correct that. (This is something I probably could even do > myself in cvs.)
Thanks for spotting that: it's now fixed in CVS. > 2) similarly, the "secret" short form "-q" for "--quiet" is also used in > the manual and probably never explained We need a short section in the manual on option abbreviations. I think that (at least) -q for --quiet and -v for --verbose ought to be reliable (i.e. not used for a different purpose in some commands), but I'll have to check. > 3) and a suggestion: in a foreach loop many variables which are > contiguous in the dataset can be referred to like this: > > loop foreach i var1..var99 > > This is explained in the manual, otherwise I wouldn't know it. When I > was dealing with 51 weekly dummies today I thought that it would be very > useful to have this syntax also for definition of lists, or maybe even > directly in estimation commands. For lists, yes, it would be nice -- but I think that would make it unnecessary in estimation commands. Allin
