I was very impressed by this talk. Matthew showed us how to use the Apigility tool to streamline the API development process by doing the tool setup itself, then following up with a very brave live-coding demonstration. Although the "gifomatic" API might seem trivial to some, I felt it was concrete enough to communicate to lesser experienced devs who might not yet have written their own APIs (e.g., yours truly) how they should work. I left with a curiosity about exploring the Apigility tool further, and a reminder to play around with API development in my own side-projects.
Chris's talk covered a wide range of command line tools available to a Linux server administrator, much of which was quite new to me in my young career. Given the amount of information therein, it's simply not possible to dive down in too great of detail on any particular tool or utility, but I thought he provided enough information on each to make the value each tool provides clear, and to help the attendee determine whether it's something they'll need to pursue further independently. I took copious notes, and I am glad to also have access to the slides to review anything I might have missed.
Very good data.
I would have loved to have some time on setup process and configuration. Maybe you should introduce the slide about the window.performance before you go into the T seconds stuff next time since it seemed that some people assumed you were talking about something more like time to first byte.
I was a little confused if the main product tracks the time it takes users to respond to your page and make a selection. You talked about it a bit but after showing the apdex it seemed that it wasn't.
I liked citing of the claims like the 80% benefit. The pace was good. The apdex T seconds calculation stuff I think would benefit by being a little more concise with just your warning about setting t too high.
Very informative and good. Thanks.
Good talk overall, slides could have been more readable, but I was in the back.
Examples were a good way to show the concepts applied to real world cases, would have liked to see more emphasis on them and less on the history (although I know that's likely a hard balance to strike).
Super helpful tools. Even good for doing things other than apis. Unfortunately some don't apply to soap as well. Maybe you should loook into a soap->ui converter. It appears apigee has something like such. That would be another huge addition
MWOP: Can you let me know what code bugs there were? I've had people point out a few in the past but I'm pretty sure I cleared all of them out by now; certainly things like defining a function on a slide and not using it. :-)
I agree generators play really nicely with functional style; unfortunately there's only so much that can fit into an hour long talk, or into people's brains at once.
If Vesna hadn't disclosed so in her opening slides, I would not have guessed that this was her speaking debut. Her presentation was well-structured and she kept good pace with the material, stopping at well-placed jokes to liven things up.
A few of the slides were cluttered -- mainly those with embedded charts and graphics, sometimes positioned alongside bullet points. Still, this was an amazing effort, considering Vesna picked up deck.js two days before the conference to recreate her slides after some ill-timed data loss of the originals.
Overall, I was most impressed with the discussion and Q&A around and after the presentation. Of the few talks I attended, this by far had the most audience involvement and conversation. Vesna did a great job fielding questions and even dropped an anecdote about an email conversation she had with a former Amazon employee (Greg Linden?) regarding some oft-cited statistics about website response times. That was a subtle hint at the amount of preparation Vesna had done on the topic, and it showed.
I wish I could offer some constructive comments, but I thought it was perfect. Well-presented information about various DI-related topics with a concise description/demonstration(i.e., code examples) of the differences. Thanks Jeff!
I thought this presentation was a good companion piece to Ben Ramsey's "Debugging Zen" talk earlier in the day. Thinking like a computer scientist involves taking a step back to examine the problem from a different perspective, and using your intuition to help determine whether you're considering all of the possibilities.
All that said, clowns are forever horrifying. Nice work!