I hate web front ends

[PAH]

Problem is, as you suggest, we have to make a living and if the client insists on a 'web' front end, even after I have demonstrated that it would not 'work too well' etc... I would still build it and bank the cash.

That's in the rule of contractor survival. Give them what they want. Anything else and you're rocking the boat, unless you happen to be in a position where they genuinely believe you're capable of directing the future of their operations, such as a system architect.

Often I've encountered situations where the team I'm working with have proposed some outstanding solutions, some that would actually be easier to implement. But there's always someone who has to ok it (e.g. project manager) that either has other agendas or a lack of initiative and courage to settle for anything other than mediocre.

At the end of the day I know I'm pocketing the same amount of cash, it's just my morale that will suffer, knowing I'm stuck with a solution that I could do in my sleep, hoping that maybe in the next contract I will have the chance to push the boundaries and maybe even try something genuinely new.

[sorry for the long post, I must have dreamt of Denny last night]

[miffy]

[sorry for the long post, I must have dreamt of Denny last night]

I thought only Wilmslow did that!

[DieScum]

At the end of the day I know I'm pocketing the same amount of cash, it's just my morale that will suffer, knowing I'm stuck with a solution that I could do in my sleep, hoping that maybe in the next contract I will have the chance to push the boundaries and maybe even try something genuinely new.

It's also your reputation. I had a project implementing a fancy new technology instead of a solid old one. It was never going to work and when you drilled done the 'decision making' had been done by 'none too bright permies' eyes wide with a 'vision' of using the latest technology and commission chasing sales folk.

[r0bly0ns]

That's a browser issue; for example, Safari doesn't show the same behaviour.

Internet Explorer is, unfortunately, too broken to be of any use as a web browser.

Unforunately so many users use it that you still have to have 2 versions of every script, one that is W3C compliant and one that works in IE.

[NickFitz]

Well that's fine and dandy if the client wants to play ball and implement a new infrastructure that draws on the above eight.

That's what their existing infrastructure is already like. It's what all distributed infrastructures, from the Internet down to your home-office LAN, are like. The point is that you have to know that the eight things listed aren't true for any distributed environment whatsoever, as the starting point for making something that will work reliably in a distributed environment.

[NickFitz]

Unforunately so many users use it that you still have to have 2 versions of every script, one that is W3C compliant and one that works in IE.

Or you use a library that has already abstracted all of the cross-browser problems away for you.

To be honest, since version 6 (released seven years ago), IE's support for the W3C DOM, though imperfect, has been sufficiently standards-compliant that cross-browser scripting isn't really a problem. Also, with an understanding of the mysterious world of the "hasLayout" problem, most CSS issues can be dealt with quite easily, and the remaining few are generally well understood and have straightforward fixes - setting "display:inline" on a floated element having a margin on the side it's floated to eliminate the 3-pixel text jog, for example.

[Xenophon]

Or you use a library that has already abstracted all of the cross-browser problems away for you.

To be honest, since version 6 (released seven years ago), IE's support for the W3C DOM, though imperfect, has been sufficiently standards-compliant that cross-browser scripting isn't really a problem. Also, with an understanding of the mysterious world of the "hasLayout" problem, most CSS issues can be dealt with quite easily, and the remaining few are generally well understood and have straightforward fixes - setting "display:inline" on a floated element having a margin on the side it's floated to eliminate the 3-pixel text jog, for example.

NickFitz is awarded +5 Xeno Geek Points.

[r0bly0ns]

Or you use a library that has already abstracted all of the cross-browser problems away for you.

That library still has to contain 2 versions of a fair few things though, even more if you have to be compatible pre IE6.

To be honest, since version 6 (released seven years ago), IE's support for the W3C DOM, though imperfect, has been sufficiently standards-compliant that cross-browser scripting isn't really a problem. Also, with an understanding of the mysterious world of the "hasLayout" problem, most CSS issues can be dealt with quite easily, and the remaining few are generally well understood and have straightforward fixes - setting "display:inline" on a floated element having a margin on the side it's floated to eliminate the 3-pixel text jog, for example.

Yeah, things are much better these days, however there are some things that really grind my goat, like IE not supporting the border-spacing CSS tag, meaning that you have to have a mix of attributes & CSS for tables.

[NickFitz]

NickFitz is awarded +5 Xeno Geek Points.

[Xenophon]