I'd prefer to avoid flame war, since I know this is an hot cold topic here. But I don't agree. On Tue, Nov 25, 2008 at 11:34 PM, Uriel wrote: > xml? Just say *no*. > > And BTW, steve's libxml is far from usable and doesn't really > implement the complete standard (but who can blame him? the libxml2 > people have been working on their implementation for years, and it is > still full of bugs and security issues, not to mention insanely slow). well... slow... may be if compared with a parser for a non-extensible language, but I'm not sure I could do better (even if I didn't try). As any other tecnology XML could be misused. But there is no reason to "hate" it so much. Depending on your application design and the (human or not) environment who required the application, it could be a bless or a mess. For what I'm thinking there is simply no other way to obtain the result. I need to encapsulate human skills to minimize the interaction between them (graphic designers and developers). With the (huge?) fixed cost of such a library development, I could minimize it. Again, just say *no* to xml, even the web 2.0 fools have given it up. > I find many approaching XML in a similar way than other use for plan 9: they are so misunderstood! Those web 2.0 guys who decided to avoid XML are often moving HTML pieces. And they fill them selves smart. Or evalutating JSON. It could seem smart, actually. But when client change (say from a browser to a flash or to a web service) your problem begins. Moreover those smart guys mix data and presentation. They are smart if their only need is to sell a web 2.0 site to someone actually need a blog. For anything more, they are chosing a dangerous road. > > uriel > Giacomo PS: I've had way to appretiate your estetical opinions about software design. Only upon this subject I deeply disagree with you.