On Tue, 2 Mar 2004, Reinhold May wrote: > > Please don't encourage people to dig about in the internals. > > We want to be able to change them freely, but if you start > > publicizing "solutions" that use the internals, that then > > becomes part of the public api. > > Dave, I see your point here, but on the other hand I realise as > well that I want something quite special, that, if Flavio or > anyone else incorporated it into DateTime::SpanSet, would > probably clutter the interface, especially if other people would > want to have their own requests in as well. And it's exactly > the clean interface, I really like about the DateTime packages.
Well, in this case what you want can be provided with a perfectly generic method, as Flavio demonstrated in his followup post. > So, if Flavio offers me some help using internals, I'm > grateful and I'm aware that that part could be invalid using > later builds. If want you wanted was so specific that you were the only person who ever wanted to do it, you might be right. But in this case, your post highlighted a weakness in the DateTime::SpanSet API, and that's where the problem should be addressed. -dave /*======================= House Absolute Consulting www.houseabsolute.com =======================*/
