From: Brian G. <br...@qu...> - 2003-07-23 19:49:51
|
> > There's a difference between deprecating them and getting rid of the > > ones we have, breaking everyone's templates in the process. > > Those templates will break at some point anyway (if tools do go away). > Does it really make a difference if it happens now or later? It sounds > like we're deciding if it will happen or not, regardless of when. Of course it matters! If we say "Context tools are going away in two years", people have ample opportunity to gradually work the CTs out of their templates. This is why facilities get deprecated instead of removed -- as a warning that they may eventually go away. It takes a long time for deprecated facilities to go away, and for good reason -- it takes a long time to wean people off them. If we rip things out and people's WM apps stop working, peoeple will eventually stop using WM because its not stable. That doesn't help us. But in any case, its putting the cart before the horse -- there has to be something that does the same thing before we can deprecate the functionality we have now. > How about exploring a way to expand this capability to regular > $Variables too. Sure, lets talk about that -- that might be cool. > > Just like directives. Lets get rid of them too then! > > Come on man, gimme a break. :) Me too! > > Then do so, deprecate the context tools we have, and then, in a > > version or two, we can talk about getting rid of CTs. > > So you're saying, w/o any further discussion, that we should deprecate > all our existing tools? Hmm. No, what I'm saying is that without any further discussion, we should do nothing, but that if you want to do something, getting rid of them is not a candidate, although deprecating them might be. > Spending the time talking about a better mechanism is a waste if you > can't be convinced what we currently have needs to be changed (or even > dropped completely). That's backwards. I'll be more inclined to deprecate them if a better mechanism exists! |