UPDATE: Here is the resulting stats from all the Hacker News Traffic What happens when you are the #1 article on Hacker News
Over my nearly 20 years in software I've seen thousands of resumes and the biggest chore is weeding out who seems like a good potential match. Unfortunately you have to trim down the list of resumes to people you want to phone screen with, then people you actually want to have come in person. Most of the time you'll see bullet points such as these:
Rockwell Collins 2000-2009
Technologies used: Java, C++, PHP, Python, Ruby, MySQL
Ok great, we use Java and PHP and Python but how competent are they in those languages? They most likely weren't developing the same amount of time in each language. From paper you can't really tell anything so it's a maybe pile because they just list too many things. If I knew they spent 70% of their time on Java and they have around 5% familiarity with Ruby now that's telling me something of value. I know how to better read this person and adjust my interview questions accordingly.
Today on Forrst.com I came across what I think should be the future of the resume. It's in Infographic form created by Jamie Kite (http://ravenousfig.com/). At a glance I get exactly the information I need to make an informed decision on this person. I can see exactly how much of each technology they worked with at each respective position. If you were to pair this with a data output that is machine readable, google's problem of 75,000 resume's per week could be significantly made easier to wade through. HR people in enterprise companies look for keywords, bottom line. They see PHP on a resume and you're gonna be interviewed for a PHP job. If I could pull from a service where I said give me a list of people available who are currently doing 40% or more with PHP then that starts to get fun, and when you start to think about linking your public github repositories so you resume starts to become a living thing that sits on your site or is hosted by a resume service. While obviously nothing is perfect I think the idea of adding weighted cues for skill sets is something missing from the current resume model.
**** UPDATE ****
The main thing I take from the resume is the ability to see percentage of use across skillsets. When I see someone has MongoDB on their resume for instance I want to prepare questions for them that tend to get into the nitty gritty of how Mongo operates, why it was chosen, what weaknesses you found, what strenghs, how do you deal with the document size limits, what do you do when your index outgrows memory,replications strategies etc... Turns out most people spend 2% of their time writing queries against Mongo and slap it on the resume. Since I like to prepare a custom interview for each person I'd like to know they probably know how to query MongoDB but don't know the low level details and I can shift my questions into other areas they're more likely to provide competent answers in. Google does similar things with sending out a questionaire to fill out with your level of expertise among a variety of subjects. They then tailor questions based on that, however thats more back and fourth than should be required.
A rough xml sketch could be
Here is a link to the full resume image: http://forrst-production.s3.amazonaws.com/posts/snaps/57782/original.png?1296924317