Dutch PHP Conference 2012
6 - 9 Jun 2012 at Amsterdam RAI

Interesting to see so many negative remarks on this talk. It made me fell uneasy too, but I acctually think he is more or less spot-on in his prediction.

But on the other hand I beleive that most of us "traditional web developers" allready have made the shift. html and forms still get used but most of our time gets spent with client side js or in the backend. That won't change in the near future.

To get rid of the mess with iOS-apps, Android apps, Windows Phone apps etc. etc. I really do hope for solutions like PhoneGap will mature enough to make it easy to make cross platform native apps from a single point of entry.

It was so obvious this was intended for a US audience only. You just don't shock any European by saying that Asia has countries with bigger economies than that of Canada! Some statistics did say something but there were just too much. It's nice you did add something extra to your message with every statistic or else it would have been really boring. Though the keynote ended sooner than planned it really shouldn't have lasted any longer. I didn't completely agree with the message either: I don't think Mobile will completely replace the desktop market although I guess it will probably be bigger. At the end it was said that we can probably solve the problem of having 4 bilion extra internet users and contributors. I think this is a wrong statement: I think THEY will solve the problem themselves much quicker, because they don't care about all the old technologies we have to still deal with every day. Google, Apple, Facebook: In the future they will have larger Asian or even African counterparts and companies that will fill completely new gaps. Not a bad speaker though.

Good talk

RT +1 @Jan_Pieper: "Cool talk, cool presentation and a very nice tool :-)"

Heavy on code but i actually liked that and a geat talk to explain the code

@Onno:

Not sure its the correct place to start the discussion, but anyway: It isn't really bogus if you will take a look how most of the puppet modules are build. Yes, *eventually* you tell the system which package it needs to install but from your point of view, this is irrelevant. You only need to tell it to install "apache", not httpd or apache2. A good example can be found in the "example42" modules: https://github.com/example42/puppet-apache
Mostly this is due to the fact that distributions tend to use different names of packages, or even use different layouts (/etc/httpd/ vs /etc/apache2 for instance).

The difference between "what" instead of the "how" is mostly visible in the resources and arguments themselves: when I tell a service that it needs to be enabled, I don't have to specify that is needs to do this through update-rc.d, chkconfig, manual installation in rc.x directories or another way. I let the agent take care of that.

Very good talk, and well prepared. Frisby.js looks to be a bit too complex with too much nesting of if-statements all over the place.

But overall this was one of the most interesting presentations at the conference.

Good talk, but like in previous comments i would have liked to see some examples in action

I agree with the previous comment that i would like to see different examples for differnt cooupling types even when you give a talk around them it makes the types look the same.

Very interesting insights into code quality expressed in a way that I haven't heard before.

It spoke to my heart and gave me new ways of expressing my QA desires and gave me guidelines and new ideas to reevaluate my views on specific parts of coding standards.