I first started turning angle brackets into dollar signs on a full time basis in the summer of 2002, a little over 6 years ago (I did contract work, internships, and worked part time before then).

Allow me to tell you a little bit about the ol’ interweb job back then.

The company I worked with was on the cutting edge… of backwards compatibility, that meant that in 2002 we were developing for 4.x browsers circa 1997, Today that would be like developing for IE 6 and Mozilla 1 (as a focus!) so it wasn’t really a big deal it just wasn’t exciting. Tables and spacer GIFs were the name of the game (GIFs were still patented back then you know!) We used Flash too, we tried to publish for Flash 3 or 4, and of course we offered non-flash, non-javascript, non-cookie versions of pretty much everything thing because a surprising number of users back then thought they were bad. Some clients (especially in the the government) were in offices where they used dial-up to connect to the Interwebs from their Windows 98 computers with 640×480 256 color screens, used IE 4 with no plug-ins, no scripts, no cookies… o and they still wanted nice websites. Boy howdy did we deliver! It’s amazing what you can do when you have to.

If you’re wondering just how much things have changed in the way of rich content access I just checked the stats for a couple of popular wide audiences websites I have access to and browsers with no-flash access make up only 1.5% of the the hits, including those from mobile devices. Just for comparison about 3.3% of people in North America are visually impaired, 0.4% are legally blind, 1% are functionally illiterate. How often do you develop with non-flash users in mind? How about the blind or even illiterate ones?

So where is the future?

I see the bulk of the skilled development work shifting from a) making pretty designs, b) marking up content the client wrote, and c) hammering out oodles of non-interactive sites to developing really interactive rich applications.

This might actually mean fewer web designers, and developers over all; but let’s face it, at one time a lot of people made a lot of money by “thinking in fortran” and punching cards, those people didn’t die when the technology changed. We just need to be ready for the changes. A lot of the really dull repetitive, wheel re-inventing jobs will be eliminated, and the people who aren’t capable of moving on will certainly be forced to find new jobs, some of the jobs may move to emerging markets, but don’t expect “basic” web jobs to be sticking around here all that much longer. At-least not very many of them.

CMSs are getting pretty good, rich editors are almost decent and if IE gets it’s act together the lumps with that will smooth out pretty quickly. I’ve even seen a fairly convincing demo of a product that cuts up PSDs into XHTML and CSS automatically.

By 2019 I see development of the typical small business website going a bit like this (and I’m being very conservative here). A designer makes a layout (probably based on something from a library of wireframe designs), perhaps even in a tool which transparently works directly with XML, CSS and SVG (Or whatever code we’re using by then!) bypassing Photoshop (or whatever the big tool is by then) all together (in most cases). That design is point and click marked up for use with a template system and is then plugged into a CMS. The client will fill out the content, which I’m sure will still be difficult to cause to happen, even in 2019.

It’s already pretty close to being the case, all we’re really losing is the skilled work of turning designs into code and plugging them into a CMS, and don’t expect to be doing a lot of work writing CMS software either. How many people are writing new word processors these days? At one time that was the product of the future, Now you have three, Office, OpenOffice and Pages, and you can almost ignore the last two, actually most businesses do. I’m expecting an end to end Adobe solution that actually works, right now they don’t have that, but Adobe isn’t stupid, they’re just slow.

That means Joey McCodemonkey isn’t going to be able to eek out a living turning PSDs into HTML anymore; so what will he do? Well that means Joey is going to have to do what all those programmers who were writing desktop publishing software back in the ’90s had to do when that dried up. He’ll have to get a job doing something else, and for a lot that’s going to be outside of programming. Technology careers have a tendency to chew people up and spit them out. How many programmers do you know? How many who have been doing it for more than 5,10,15,20 years? Those numbers shrink pretty fast don’t they. It’s not like people were not programmers in 1989, sure there might have been fewer but they did exist, and probably in greater number than you’d think.

So what will the Joeys who can adapt be doing?

Well they’re going to be writing apps, web apps but we’re not really going to consider them that because it’s all going to be “web” apps, and you can drop the web because that’s not going to be important. You can call them, internet hosted applications or something if you like. The line between desktop and web apps is already under attack, heck Microsoft tried it with the whole HTA fiasco back in the day. More recently Microsoft, Apple and Google have all been playing around with web based office suites.

So what does it come down to?

Rich, unskilled-human friendly interfaces to important and interesting data, just like desktop apps today, just like desktop apps in 1999. Our idea of what good usable software means changes all the time. In 1989 color terminal based software was the bee’s knees in usable computer software. In 1999 it was rich GUI desktop applications. In 2009 it’s rich web apps. In 2019 it’s going to be rich interfaces delivered over the Internet. What will rich mean? Well I suspect it’s going to be ever more interactive and ever more social. Indeed John Gage was a genius when he came up with the one time Sun Microsystem’s slogan of “The Network Is The Computer”. It’s been lauded as an idea before it’s time, and it only gets more true with time.

So if you want to get picky about re-sizing your images for better quality and watching for artifacts you should shoot for 604 pixels either wide or tall.

As for the actual file size, I did a lot of tests and it seems (as of July 28, 3008) that Facebook re-compresses your images regardless of what their original size might be, this equates to about a 50% quality setting in most apps but there is no size it won’t re-compress, even if it makes the file bigger.

Mike makes a good point, you can find out more in his Facebook Gallery