From kde-core-devel Wed May 30 15:08:54 2001 From: Stefan Westerfeld Date: Wed, 30 May 2001 15:08:54 +0000 To: kde-core-devel Subject: Re: PLEASE don't ! (Re: kdesupport: let's get rid of it, now.) X-MARC-Message: https://marc.info/?l=kde-core-devel&m=99123553622455 Hi! On Wed, May 30, 2001 at 04:22:16PM +0200, Rob Kaper wrote: > On Wed, May 30, 2001 at 04:02:38PM +0200, Guillaume Laurent wrote: > > ... It doesn't look like a hassle for you because you never really had to > > do it. Take my word for it, it *IS* one big fscking hassle. > > I am using bare Slackware and don't check out kdesupport to compile KDE. > It's still no hassle. > > > I was already annoyed to discover that all of a sudden I had to get libxml2. > > Given that it's only for help files, I didn't even bother. > > Which is why there's a difference between "requirements" and > "recommendations". If you don't have a requirement, you cannot compile > because configure _fails_. > > Others are enhancements and you can use KDE just fine, you will just miss > certain things such as real-time docbook parsing, sound, advanced Javascript > support.. This is exactly what I don't like. The idea of recommendations. As developer I will have to deal with the fact that some people don't feel like they need libaudiofile, and then artsd will have to contain all that of "and if you don't have libaudiofile, fail gracefully, report the error, hint the user what is wrong, continue working nicely" code. It makes be wonder if writing a wav loader wouldn't solve the problem more elegantly, which on the other hand means generalized that KDE developers should replicate functionality of "recommended" libraries, as you can't rely on it anyway, but don't want to bother the user with that. So... I think the strategy of recommending things is bad, and removing kdesupport will make more users run into problems at runtime. Cu... Stefan -- -* Stefan Westerfeld, stefan@space.twc.de (PGP!), Hamburg/Germany KDE Developer, project infos at http://space.twc.de/~stefan/kde *-