Bas said:figuerres said:*snip*
^ This, with a vengeance. I hate writing web applications. It's my job, but I hate it. I hate having to ditch smooth development of rich desktop applications in favor of having to write unreliable hacks in a language that's intended to display pages about cats rather than serve as an application framework. In order to work around the many disadvantages web applications have over rich client applications, we keep stacking more and more stuff on it, like AJAX and whatnot, and it just becomes more and more of an unworkable mess.
So because we've all collectively decided that we should just keep adding to these awful outdated technologies to bring applications to the web, my job consists of pushing a dead horse over a finish line, using a wobbly wheelbarrow. Every. Single. Day.
Lately, I've opted to to try and automate our development process rather than write more web stuff. I'm actually preferring having to wrestle with the EnvDTE namespaces, rather than writing for the web. It's that bad.
I think a lot of the problem is how ill-conceived the basic/fundamental existing stuff is. That's certainly what seems to waste much of my time whenever I design a webpage. We don't so much need new stuff as better stuff. Most web standards should be taken outside and set fire to, then replaced with something better, except it's too late now and we're stuck with the stupid things.
I mean, who the heck creates a layout system that doesn't allow for one item's x and y positions, widths, heights, etc. to be relative to other items'? (Sure, CSS can do it in a few cases, but they are exceptions and not the rule. You cannot say that one div should have the same height and half the width, and be placed relative to, another arbitrary div. You definitely cannot take properties from multiple other divs. What an absolute crock of failure.)
Then again, I could say the same about most desktop UI frameworks.
Basic XML itself is okay and useful but the technologies (and specs, documentation/books/etc.) layered on top of XML are pretty awful, on the whole. (Pet hate: Specs/docs which talk about things they haven't actually defined yet, without even telling you they haven't defined them yet, for great confusion, re-reading and wondering if you accidentally skipped a page somewhere.) I mean, if XSLT didn't suck so much it could've done what CSS does so much better, IMO. As it is, I'd sooner convert XML content into HTML via C# or C++ code than mess with XSLT, and that can't be right.