Tip:
Highlight text to annotate it
X
VAN RIPER: Hi, I'm Van Riper, and I work with our Google
Developer Groups all around the world.
Today we're starting our first ever Google Developers Tech
Roundup, a monthly show bringing to you the latest and
greatest in what's happening for Google Developers.
I'd also like to introduce my co-host, Stephanie Liu.
STEPHANIE LIU: Hi, everybody.
My name is Stephanie Liu.
I lead the Global Programs team for developer relations.
I work very closely with Van.
VAN RIPER: And we also have Adriana Cerundolo.
ADRIANA CERUNDOLO: Hi, I'm Adriana, and I am the
Community Manager for the Google Developer Groups, and I
have the pleasure of working with Steph and Van also here
in Mountainview.
VAN RIPER: So today, we're going to talk a little bit
about I/O, since that was recently some of our favorite
moments and some of the top news.
We'd also love to hear from you what you found
interesting as well.
But I'll start with my favorite
moment from the keynote.
It was actually before the keynote even started.
Because of my work with the community, that really kind of
gave me chills when I saw 250 plus organizers all stand up
before the keynote even began, waving their local banners and
chanting GDG, GDG.
I love the passion of our developer community, and it
was great to see that in action at the keynote.
How about you, Stephanie?
STEPHANIE LIU: I think my favorite moment from the
keynote was a mid section.
He was talking about his sci-fi-- you know, how we're
living in the future, his whole "Star
Trek," sci-fi thing.
So I thought the Google Now demo was awesome.
Being able to see the thinking, the principles
behind how we're moving forward in that space with
answer, converse, anticipate, with anticipate being the
really futuristic, we're giving information before you
even know you want it.
I thought that was awesome.
And as a bonus, when Amit got up on stage, he said he grew
up in India.
And all of our Indian GDGs, you could hear them, they were
near the front row, just went crazy.
So I thought that was special.
ADRIANA CERUNDOLO: Yeah, I'll have to say for me my favorite
part was the Maps update.
Because I've always loved maps.
Ever since I was a kid, I would have them all over my
walls and just loved geography.
So I thought it was great to see that now we have
customizable maps.
And it's also that integration was [INAUDIBLE].
And you can see reviews and so on all in your map.
And they actually went and did, and they panned out to
using Google Earth and up into the universe
view during the keynote.
And that was the backdrop for the rest of the segment, and
that was really awesome.
Loved that.
VAN RIPER: Yeah, but actually turning now to more about the
substance of the keynote, the things that I personally--
and for developers in particular--
for myself, as a long time Java developer, I was really
excited by the announcement of Android Studio.
I can remember when I was a young developer using early
access releases [INAUDIBLE]
first IntelliJ IDEA for Java developers.
And now it's kind of a new world with Android development
having the early access preview releases of Android
Studio, which is based on the very solid community edition
of IntelliJ IDEA.
So now there's a free environment that's based on
this environment that is a pleasure to use.
And even though it's in early access, I found Android Studio
already to be a pleasure to use.
So I encourage you Java developers to check that out.
What about you, Stephanie?
STEPHANIE LIU: Well, I thought that some of the kind of news
that may have been lost in the shuffle was some of the
sessions that we did on Google's two languages, which
are Go and Dart.
And there was a theme, which was performance.
So for Dart, if I had to choose a session, I'd
recommend Web Languages and VMs, Fast
Code Always in Fashion.
It's Always in Fashion with Lars Bak and Kasper Lund.
So there's a cool demo in there, so look for that, where
they're zombies.
So you have that.
Just like the performance enhancements using that single
instruction multiple data, which Dart supports.
And for Go, Go 1.1, there was also a bunch of optimization
that came out with the .1 release with garbage collector
stuff, [? mathematician ?]
and some tweaks to standard library.
It's compatible with 1.0, and you'll note, if you upgrade,
you'll probably notice your code running a little faster.
And if I had to recommend a talk to watch for that, it
would be Rob Pike's talk about advanced Go [INAUDIBLE]
currency patterns.
So if you've been hesitant because the languages are
still pretty nascent, I'd recommend you check them out
because it's fast moving.
It's getting more mature.
The teams for Dart and Go are actually both very responsive.
And it would be like a cool time to get into it.
My own personal story about Go is, well, my fiance and I got
engaged like seven months ago or six months ago.
The first concession, the first wedding planning
concession I made, was to do our wedding website on App
Engine Go instead of App Engine Python.
Because he really loves Go, and he thinks
it's a great language.
And so I could see it because I'm being a good future wife,
compromises.
But also I thought it would be a great to do Go.
And even though I haven't made any significant patches yet, I
still got three months.
So I'm like [INAUDIBLE] digging into it.
VAN RIPER: And don't forget the PHP announcement as well.
It may not be a new sexy language like Dart or Go, but
I think there's a lot of developers out there that--
STEPHANIE LIU: Stately?
VAN RIPER: There's a lot of developers out there using PHP
that were probably very happy to see that
announcement as well.
ADRIANA CERUNDOLO: It's awesome.
I just want to say I hope that's the only concession you
have to make, at least before your wedding
starts with the planning.
STEPHANIE LIU: Oh, yeah.
[INAUDIBLE] hot pink.
Hot pink and purple as my colors.
I'm just kidding.
It's not.
ADRIANA CERUNDOLO: I love it.
I love it.
It's going to be great.
You're going to look great in that color.
I can just see it.
Well, I just want to say my personal favorite update was
with Apps Script.
I think this is an important update.
So we already know that you can use Apps Scripts to extend
into Google Spreadsheets, and now you can extend into Docs,
which I think is really important because now you can
use custom menus, dialogs, and triggers.
This is something that personally is going to make my
life much easier, especially the update with
Apps Scripts and forms.
So it's going to make those interactions with form
submissions that much richer.
And also there's going to be much more that you can do with
the form submissions to get information from when they
submit and so on.
So it's just going to make all the work that we do with forms
and the events just much more efficient and easier.
So I'm excited about that.
VAN RIPER: So that was just a few of our favorite
announcements and technical components from I/O. I know
there's a lot more--
there's way more that was covered there on Google+ and
Chrome, and you need to really check out all the videos from
I/O. And we'd love to hear from you what your favorite
pieces of what you thought the most significant developer
announcements were.
Let us know in the comments.
For our next segment, we have Kim Cameron from the BigQuery
team joining us.
Kim and Amy Unruh gave a great talk about the sensors used to
collect data at Google I/O this year and how the Google
Cloud Platform was used to do cool stuff with the data.
So take it away, Kim.
KIM CAMERON: Hey, everyone.
So you may have heard about the project called the Data
Sensing Lab at Google I/O. And I'm here to talk to you a
little bit about that.
So first of all, before we get into what it actually was, how
did it get started?
So our colleague Michael Manoochehri on the BigQuery
team went to Strata in February, and he was really
impressed by a project that the O'Reilly Data Sensing Lab
team had set up.
It basically took about 50 sensors, put them all over the
conference, collected environmental data, and then
did cool stuff with it, visualized it for people.
It was a pretty big success.
And Michael thought, why can't we do that even bigger.
So he decided to do this at Google I/O. Instead of 50
sensors, we did 500 sensors.
And like at Strata, we set them up all
over the Moscone Center.
And we wanted to figure out cool stuff about the data, and
we wanted to show it to you in really interesting ways.
So I really encourage all of you to go to the YouTube talk
that will be shown at the end to get full
details about the project.
There was a lot involved.
I'm going to summarize it a bit for you right now.
And there were two main components, the hardware end
and the software end.
So first off, for the hardware, we took 500 sensors,
as I've mentioned.
And these were made up of two main pieces.
First of all, there's an Arduino board that
makes up the base.
Arduinos you've probably heard of.
They're open source electronic boards.
They're really cheap.
You can get them for $20 at any Radio
Shack or other store.
To that, we added a custom sensor shield.
And the O'Reilly Data Sensing Lab team ended
up doing this work.
We partner with them on this project.
Each sensor shield has some built-in sensors,
environmental sensors, for temperature, pressure,
humidity, and also light.
And then they have the capability to include three
additional sensors, specifically air quality,
pressure, which is-- we had these mats around Moscone that
if people stepped on them, they registered--
and also RF noise, which we had tuned to GSM.
And we had about 100 each that had one of each of the
optional components, and then all 500 had the base ones,
which were, like I mentioned, temperature and pressure and
humidity and light.
So we set those up all over Moscone.
And the next issue was, how would they communicate with
each other.
So we partnered with Digi International, who also were
involved at Strata.
And they hooked up a mesh network.
And this allowed the sensors to communicate with each other
over a much greater distance than if we had set up, say, a
Wi-Fi network.
So then the software piece comes into play.
And that's where our team built the app and figured out
how we were going to work with this data.
And we used many parts of the Cloud platform, specifically
App Engine, BigQuery, Compute Engine.
We used App Engine Data Store.
Lots of pieces involved.
We wanted to be able to build an infrastructure without
doing a lot of the core work to put the pieces together.
We wanted to be able to take in the data, analyze it and
visualize it very simply.
So just to give you a brief rundown of how we did that.
First of all, we have all these streams of data coming
at us from this mesh network.
And we need to be able to put that somewhere quickly so we
can keep getting the data.
So that, we used App Engine.
And we specifically used App Engine task queues to go off,
grab the data asynchronously, and put it into the App Engine
Data Store, where we could use Data Store's system to be able
to query things like the most recent 100 readings for heat
from all of the sensors, things like that.
We also used the task queues to get the data into BigQuery.
And we wanted to do that to be able to enable more robust
queries for things like aggregates, so the sum
function, for example.
So we had the data in both of these data sources.
We also used Compute Engine to do some analysis on the data.
And moving forward, we're hoping to be able to do even
more types of analysis.
And we also used some visualization scripts to be
able to output things that were going on in the kiosks
that were around Moscone.
Specifically, I think we had one that was showing where the
most recent pressure was around the conference.
You could see in front of the doors there was a lot of
traffic, at the base of the escalators.
We even put some sensors in some interesting places, like
there was a blimp that was floating around.
We had a few of the sensors up there.
As well as in, there was a car that you could sit in.
We had a pressure mat in there so you could see how often
people were getting in and out of the car.
And so we did this to really help see more about what was
going on at Moscone than you might normally be able to see.
And we think that this opens the door for doing really cool
analysis on things like, if people walk around in a
certain area, does the air quality go up or down?
And actually, we had one visualization for that, with
the help of Tableau, who ended up hooking up some
visualizations through their connector to BigQuery.
And we did see a correlation between number of steps that
people were taking in an area, and the air quality did
actually go down, which was kind of funny.
So this project really helped us do cool things like that.
And we think in the future we can do even more.
And to wrap it up, we do plan on continuing to
work on this project.
We're hoping to do further installations at other
conferences or events.
And we are also planning on open sourcing the project in
the future, as well as the dataset is available, the
entire BigQuery dataset.
You can go look at it yourself.
And all this information is listed at datasensinglab.com,
where we're going to have all of that stuff in the future.
VAN RIPER: So Kim, we have some of the sensor hardware
used at Google I/O here in the studio.
I have one in my hand right here.
How would developers get this hardware
for their own projects?
KIM CAMERON: So that's a good question, and actually, at the
conference, that was one of the main things people were
asking at the booth.
How do I get this sensor shield?
Because obviously you can roll your own, but they wanted the
ability to have it now.
And you can actually pre-order those at the
datasensinglab.com site.
The O'Reilly Data Sensing Lab team is planning on making
those available in the future.
So you can certainly access those now.
VAN RIPER: Great.
Well thanks, Kim for that brief overview
of your full talk.
I hope people go and see the full talk by Kim and Amy after
seeing this introduction.
And there will be links to that video as well as links
where you can give us feedback in the notes to this show.
Well, that was awesome.
ADRIANA CERUNDOLO: Man, I really think we better wrap
this up before the air quality continues to decrease.
VAN RIPER: Sounds good.
Well, I hope you enjoyed this first edition of the Google
Developers Tech Roundup.
We hope to see you a month from now for our second show.
In the meantime, if you have something you'd like us to
share around Google Tech, use the hashtag #gdtech.
We're going to close with a little montage of our photos
of Google Developers in action from I/O. And again, use the
#gdtech hashtag to share photos with us that you'd like
to see on the next month.
So until next time, take care.
[MUSIC PLAYING]