Comment Getting the job done quick is all that counts. (Score 5, Interesting) 323
Being in Webdev for 15 years I can say that getting the job done quick is all that counts. Most of the web is run by the bizarest of contraptions in software you can imagine - but they get the job done. Take for instance Wordpress: It's a prime example for bad software architecture and the inner platform antipattern.
But it works. It delivers, Any idiot can download and install WP, pop in a theme and start fiddling. The webev gets called in when the system is all gummed up and feature x,y or z has to be added with magic programming trick (i.e. dirty hacks) quickly.
Same goes for PHP as a PL. Strange, bizar and hilarious, but it get's the job done.
That's what counts.
All that been said, it's precisely because of this that your skills as a webdev determine wether you'll have some freedom to pick your job and a fair salary or if you'll be treaded badly. I've been through so many projects that I can tell you even the crappy devs don't mean it. If there's a crew of 5 coding without versioning, that's because their to dumb to know any better and they won't listen to you if you're not ready to walk out of a job that only pays you a McDs salary.
If however, you've got the skills and the tools, most people will think you're a demi-god. Use whatever technology you want, but be able to deliver. I've started building my own toolkit a while ago - it involves bash-cli snippets and PHP code - and dive into any mess my client/boss requires me to work with, be it Wordpress, Drupal, Joomla or whatever. I've since become good enough that I can make some demands, but I have no illusions about my outlook in the webdev world. It is a volatile occupation and unless you move into Java/Oralce, SAP or MS territory, it will stay that way.
The upside is the freedom we have. We get to use FOSS most of the time as primary tools of trade and get to try out new things 5 times a week - neat. You can't have it both ways.
In a nutshell: If you want to stand your ground, you have to be good at both: Overall problem solving experience and proficient expert knowledge in the current tools of your trade. If you stick to building those mostly from tried-and-true FOSS technologies, you'll keep pointless learning to a minimum. For instance, I make a point of using grep to search for snippets of code in a project. My IDE may be dead 3 years from now, as may be the system I'm using. grep will be around until I die.
My 2 cents.