From my point of view, though, we had noticed even last year that the Velocity agenda was leaving us with fewer tangible takeaways. What I’d like is something holistically useful. Front end performance is important, and there was a big “revolution” in that space with Souders etc.’s work a couple years ago, but it’s a small piece of the entire puzzle.

Even just the performance track needs to be about more than just front end performance and it needs to focus on specific takeaways that a Web ops kind of person would implement. There’s other venues for pure Web designers to go to IMO – that’s not this crowd; this crowd is a specific mix of Web generalists – part UNIX admin/open source wonk, part developer, small part designer. The more you go toward stuff only a pure Web designer who does nothing but CSS all day would care about, the more people like us will say “Hmmm, I think this year I’ll pass.”

Also some of the stuff was very speculative (browsers, the weird proposed google protocol, quickling) that have no practical takeaways for someone. It also seemed like some of it was about highly esoteric fixes that would be a) for a super Web designer JS/CSS hacker and b) only relevant to the “top 5 guys”. It’s somewhat interesting to see the things that Google/Yahoo/Twitter/Facebook have had to do to operate at their scale, but 99% of the audience is smaller scale and won’t be doing that. “We spent 6 months optimizing those last two pixels!” Remember that the conference shouldn’t be for the top 1%, if it wants to make $ it should appeal to Joe Internet Startup/Jane Enterprise Web IT Person.

Anyway, here’s what I’d propose as a mix.

Performance – covering front end, back end, caching, memcached, hadoop, distributed computing, etc. (I liked some of the non-FE stuff in these veins at last Velocity)
Operations – config management, monitoring, control, logging, security (OWASP type), etc.

Focus on the core. There’s already cons for developers. There’s cons for designers. There’s cons for Web marketeers. But this is the first venue that brings together Web technologists specifically, especially devs/ops/devops.

I’ve heard a few folks say that they want this year’s Velocity 2010 to include more on the traditional backend Ops problem space. I completely agree, and we’re addressing that for Velocity 2010 & the Velocity Online Conferences (OLCs)

Thanks!

-Jesse Robbins

I’m part of the conference program committee, and I’m knee-deep in reviewing proposals for 2010′s conference, and there are really great non-frontend proposals. Having said that, given that client-side response time can be improved by >70% just by changing client-side code, I’ve always enjoyed the mixture of the two crowds.

Is there a view that >50% of the conference is geared towards client-side performance?