[ Jason L Tibbitts III writes: ]
>
> >>>>> "DW" == Dave Wolfe <dwolfe@risc.sps.mot.com> writes:
>
> DW> After thinking about how the new command language works in Mj2, I
> DW> suspect my change would complicate the transition.
>
> Actually I don't think it would, assuming that I understand what you did
> correctly. Am I correct in assuming that your change just doesn't guess
> and always adds the default list if there is one (i.e. the command is being
> sent to the -request address)?
What I had in mind was the fact that Mj2 has the concept of defaults
(for several field types) which are set independently of the current
command being parsed. In that context, it actually means something to
not specify a command field, i.e. accept the default. In 1.9x, -l was
the only default, and as we've discovered, resulted in some ambiguous
parsing situations that persist into 2.0. (Is the ambiguity only when
parsing 1.9x-style requests, or am I misremembering things again?)
Nevertheless, the question remains, is it acceptable to change the
allowable command syntax at this late date? On the one hand, I can argue
that the ambiguity caused it to not work at all in some situations, so
changing it doesn't matter, but on the other hand, the change will break
some commands that worked just fine. Does it matter enough to put in the
additional work to refine the fix to not break obscure compatibility?
--
Dave Wolfe
Follow-Ups:
References:
|
|
