Keeping up with Web Technology
After getting into Spring Roo I had a look at JBoss Seam, which does some of the same RAD stuff. JBoss Forge is actually eerily similar to Spring Roo. I don't know enough about it yet, but it doesn't look different enough. While looking at videos I found this interesting video that summarizes JVM Web Frameworks (my niche). Then one thing struck me...
I was agreeing with that video as though
"Yes, that is the tech I am familiar with. You are accurately describing my Eco-system."
only to find out that video is more than 2 years old. Since I only really started making an effort to "get with it" about a year ago, I am not surprised I have some catching up to do.
I went looking for this guy's web presence and found a very relevant blog post about this topic. Burr Sutter's (the guy from the webinar video) personal account of how he's running out of time to keep up with new developments would apply to a lot of developers, including me. He is in a better position to "keep up" since his frequenting of conferences and knowing people "in the know", is something the majority of developers can't afford/do. That is NOT to say we can't make more of an effort (slight dig to myself).
I have a Twitter account that I use to try and follow people that post things in my blind spot that I find interesting: front-end tech, mostly. If it were not for some Mozilla trend a few weeks ago, I would not have found out about asm.js. My imagination started running wild with the ideas of what high performance code delivered through a browser could actually DO! I'm sorry to say the first thing I thought of was a Bitcoin generator client distributed discretely by a high-traffic website. Shame on me.
As I review the state of the art of JVM Web Frameworks, not ignoring stuff like Compojure/Ring, I get the feeling that there are two opposing forces in the world of Web Applications.
- That of the full-stack, fully integrated, development pipeline
- Opposed by the multi-tooled, custom tailored, cherry picked solution
I have to say that before writing it down like that I was cursing the lack of even MORE integration. However, I have been shouting for the longest time that "enterprise" is a dirty word, and that a quality solution built to the specific needs of a company will have better adoption and lower total cost of ownership than an out of the box, one size fits all application needing years of integration work. Now this is a gut feeling based on personal experience and I am not allowed to name names. Anecdotal evidence tells me that IT service providing/consultancy companies will feed the fears of clients; telling them
"your continuity is too precious to risk on something you have so little experience with". "Let US take care of it! We're experts!"
...they shout. They're not wrong, they're just assholes. They bill hours based on the whims of an account-manager that has very little (if any) idea of what realization would entail. They invent SLA's that bleed companies dry. Well, not all the time. Some companies have a LOT of blood.
I'm going off on a tangent here, but it sort of ties together. Watch me tie it all together. Keeping up with the state of the art in any fast developing field is hard. If it's hard for domain experts, of course it's hard for laymen. You can't fault a business-person for not wanting to take risks where he/she sees them, even if they're not really there. That's why a good consultant would be worth so much, but too much of that and you end up with a big consultancy bureau with its claws in you all over again.