I thought this was the best of the ruby session. I though the presentation was very good and easy to follow.
I was a little dissappointed. I heard about MariaDB last year at UTOS. I was very excited to hear what had happened in the last year with MariaDB but instead got a bit of a history lesson.
Also, the razzing of Callaghan who joined the audience, left the rest of us feeling quite awkward. Such a prominent member of the open source community, who has contributed so much should not have been subjected to being openly called out for having not contributed to MariaDB directly. As I understand it MariaDB pulls from MySQL main and as such enjoys many benefits from his work if not directly. I am certain that if MariaDB were to become the choice at FB he would contribute directly.
Anyhow, what substance there was, was good and I very much enjoyed hearing about the future direction of MariaDB.
I thought it was a good introduction into why we should do test driven development and some methods of how to implement this approach.
Thanks for taking the time to come visit and share with us. It is very interesting to hear how Facebook has taken MySQL well beyond where any RDBMS has gone. And to think that some people are still skeptical about using MySQL or any open source database.
Also thank you for answering my questions about the influence of NoSQL within Facebook, at least as well as you could without sharing company secrets.
Thanks for taking the time to share what you have done with D3. About to start a project that will use D3 for some visualization. Had a hard time following a bit of your demo, I am sure the network problems didn't help.
Thanks for taking the time to share Wordpress with us. I have recently started hosting a Wordpress blog and really enjoyed seeing some of the standard features.
The presenter was totally unprepared. He spent most of his time trying to setup a new rails project. I wish the presenters were required to submit slides before the conference. This was a total waste of time.
The printed schedule said this talk was called "power up your css with preprocessing" or something like that, so imagine my disappointment when it was all about the box model and floats. I won't blame Justin for that. However, even the online schedule says "common pitfalls" and implies it's kind of an intermediate/advanced talk, but the box model is something you should have learned the first week. I think the talk should have been called "CSS Layout 101" and he could have fleshed it out with that goal in mind.
I also didn't think the float examples made the concept as clear as it could be. This page has some clearer pictures (in my humble opinion): http://css-tricks.com/all-about-floats/
JLW - thanks for the feedback. I didn't intend to suggest that "other people's code was rarely, if ever, to be trusted". I stated that it is often more trustworthy. In fact, the point I was trying to make was that there are times when it is a very bad thing to invent your own code rather than using already proven code.
However, there are also times when it makes sense to use your own code rather than "some 13-year-old's code" (which is what I said in the talk). I was referring to random WordPress plugins and such that people like to install with no knowledge of how security-conscious the author is.
I apologize for not making this clear. I'll try to make that more clear in the future, along with some more details on access control. Access control is somewhat of an inexact science, however, as it really depends on your framework.
Wish I could have been in the room, I listened from the hall. It would seem obvious that when a Keynote speaker gives another presentation, it should be held in a larger room.
Anyhow, thanks Phil, it was great to hear a little more concrete presentation of what you have been talking about for so long with personal clouds and online identity. It is very obvious now why such infrastructure must be the way of doing things in a trillion node network. The pain it is now when my personal info changes, and having to change it in dozens of online silos. How would I ever manage it 1, 2 or 3 orders of magnitude larger, is unfathomable.
I hope soon that the data silo creators start to see the light and move towards this model sooner than later. I can see some moves in that direction with things like Google and FB signin, but even those are just silos. The open framework that eliminates the branded signin a replaces it with a link to your personal cloud will be amazing.