Hi John, Walter, I agree with your comments. However, let me make the following observation. Any human project can only reach perfection by following an asymptotic behaviour. So the choice is: - wait for ever for the perfect documentation - meanwhile provide access to the source. It is our experience that exposing the source has the big advantage to show how classes are really used in the real life. I am not just talking of the root source here. John, it would be really constructive if you could send a list of the functions for which you have found the documentation missing. Rene Brun On Wed, 13 Oct 1999, John Zweizig wrote: > > Walter F.J. Mueller wrote: > > >It worries me a little that the `source browser' is considered one of the > >more important features in the ROOT documentation. It should be superfluous > >for most, applications should be written against an interface specification, > >not against an implementation. > > I agree entirely. The fact that the root source code provides useful > documentation is a sad comment on the state of the documentation of the root > classes. I have found many methods with parameter lists of single character > variable names where there is no mention of how the parameters are used, what > units are assumed, etc. Perhaps the most useful feature a documentation tool can > have is to encourage the author to provide all the necessary information by e.g. > requiring a description of all non-void parameters and return values. >
This archive was generated by hypermail 2b29 : Tue Jan 04 2000 - 00:43:40 MET