Tip:
Highlight text to annotate it
X
Jenny Donnelly: To kick off day two we've got one of my favorite speakers, the infinitely
enthusiastic and passionate and prolific engineer, Dav Glass, who's been the heart and soul of
YUI for a long time and has now spread his wings beyond our team to work on even more
influential and key open source projects for Yahoo. I'm really excited for this talk because
it's going to be a bit of a stroll down memory lane. We've gone through a lot together and
there's a lot of detail in there that he's managed to recall, so I'm very excited to
welcome Dav Glass for our opening keynote. Thank you.
[applause]
Dav Glass: Alright. How are you all *** doing this morning?
[applause]
Alright, so that goes out to Tripp, wherever the hell he's at, because I told him in the
first 20 seconds I'd drop the F bomb for him.
I've got a nice little shirt here for you. Kind of figured that fit in.
Are we having fun yet? Are you awake? You're not loud enough. Are you awake?
[cheers]
Alright, damn it.
This was an amazing little talk for me to try to put together because I went and dug
really deep in a lot of places, and I tried to pick out some of the key things that have
changed in YUI over the years and how that's influenced open source at Yahoo, and how Yahoo's
initial open source has really helped YUI.
Because as you can see here, this is Amotz and he is our chief architect, and he is absolutely
100 per cent behind anything that we want to do open source. I've had words with him
before, and it's like my goal at Yahoo is to open source any damn thing that doesn't
have a patent on it. If there's a piece of code they've got sitting there that's useful
for anyone and it doesn't have a patent on it, it's going. I'm going to toss it up on
GitHub, we're going to support it, we're going to do everything that we do. As YUI has done,
Yahoo's going to do that with everything else.
My focus right now is on Node. We have open sourced I think 15 of our core modules that
were internal. I've given them a goal of open sourcing two modules a month, so we're just
going to keep open sourcing and open sourcing and open sourcing. I believe we actually have
our first pull request going into Node next week, and it will actually land in master
which is huge for us to finally have a real Node core fix.
It's a simple one but nobody had ever thought of it. When you fire up Node you can hit it
with a sig user 1 to turn it into debugging mode. Well we just said if you hit it again
it turns it off. I mean it's dead *** simple, but nobody'd ever thought of actually
doing that before. Until you jump into an application the size that we run and you need
to get that application into a certain state, and then you want to turn debugging on so
you can debug whatever the hell you're looking at, and then you turn it off and you get your
application into another state, turn the debugging back on and look at what you're doing. Nobody
thought of that because they're all still building these little bitty tiny things and
we're not.
It was great to actually have that landing in core. I've had several pull requests that
I've issued for documentation and little stuff, but this is our first real chunk of code and
I can't wait for the guys to keep doing it more and more.
Like I said, with this talk I'm actually going to try to go back in time a little bit. I
know this is not history for a lot of people because they think of history as being like
1960s, but in the web development world we move so fast.
In 1996... Does anybody know any of Yahoo history of what Yahoo was built with back
in 1996? No, it was not PHP. Filo actually wrote his own web server called the Filo Web
Server and we used that to serve pages called the Filo Server Pages, and they were actually
written by David Filo himself. They were actually at Yahoo for a long time, because when I started
in 2006 I was editing Filo Server Pages on some of our very, very old systems.
But in 1996 that all changed when we actually adopted Apache. That was our first really
big step into open source and we haven't stopped since then. After we adopted Apache we started
fixing it and fixing all the bugs and making it even stronger because we were growing faster
and faster and we were pushing that thing harder than anyone else had ever done. So
we started helping out with that.
Anybody know who the little red guy is over there? Come on.
Audience member: BSD.
Dav: Yeah, alright, BSD. BSD and Perl were the other things we started playing around
with. Yahoo actually has a significant amount of code inside of FreeBSD because that was
the number one serving architecture that they used for years. We just gave it all back,
we didn't keep any of it. Every change that we made we gave right back to FreeBSD like
we do everything else.
So that's kind of where it started at, but I'm going to jump a little bit further ahead
into 2002. Anybody know what happened in 2002? That's when PHP actually showed up. So there's
a big difference between 1996 and 2002. We started helping with PHP in 2002. I believe
that's around the time we hired Rasmus. Rasmus worked at Yahoo and we worked with that. Most
of our engineers were actually committing code back into PHP. Everything that we did,
anything that we modified we were helping to strengthen that language.
It's really good to look at that because if you compare that to now, we're doing the same
thing with JavaScript and Node. That's why this kind of leads into YUI itself, because
we know what we like, we found something that suited what we needed to do at the time, and
we fixed it and we made it better, we made it stronger, we made it hold up to the stress
that Yahoo can actually put on something.
After that, I'm going to jump to 2006. We're going to skip up because I'm trying to get
to the actual YUI stuff. Anybody know what happened in 2006? What version of YUI? It
went up on SourceForge and it was actually 0.9.0 was the first public release of YUI.
That's old. We actually scraped this off of Web Archive. That is the first landing page
for YUI in February of 2006.
It was pretty crazy because it actually got picked up by Slashdot, so there's the actual
Slashdot article from February of 2006. Commander Taco actually posted in, check out the category
it's in: Hey We Should Check This Out department. So it was pretty big at the time, I mean there
were not very many yet. There was Prototype, Scriptaculous. I'm not exactly sure of the
date that jQuery came out, but this was February 14 2006. The comments are pretty decent to
read too. When it came out, it was actually launched on 2/13 so Slashdot picked it up
an extra day later.
But these were the core pieces that were actually in the library in 0.9.0. Now the really cool
part is that a lot of this stuff... I'm going to turn my presenter's notes on, I forgot
them. There were actually two versions of YUI at this time. There was actually a 1.0.
Nobody saw that because 1.0 was our internal version, and there was nothing different between
the two. They were exactly the damn same. But for some reason they decided to publish
the public one at 0.9.0, and the internal one was 1.0. That was a bit confusing for
people when suddenly we shipped it to 2, YUI 2 came out after this.
During 2006 they changed quite a few things. Why are my slides not moving to the next one?
There we go. In May of 2006 we had 0.10, and that's where all the CSS tools came in. Those
are the things that people were baffled by because we had fonts and we had reset, and
that was the first version of Grids. That was our first actual step into doing more
than just JavaScript, which was different than everybody else.
In 2006 they actually kept kind of busy. You can see that there were a lot of releases
here in 2006. There was a whole boatload of new features that came in and someone decided
we would open this little mailing list called YDN JavaScript. How many of you guys were
ever on that mailing list?
Alright, so this might have been before some time, but one of the other things that happened
to show up in 2006 was this guy. This is my first post to YDN JavaScript in 2006. July
13 2006 was the first post. By the time we closed that mailing list I had 4475 posts
on that mailing list helping people out, and the majority of them was a one line link to
an example that I wrote to help them instead of actually explaining the damn issue. At
that point I think I had 250 plus examples on my site, and the YUI team itself only had
like 90.
The other one that showed up there was Mr. Caridy back there. Mr. Caridy showed up out
of the blue in December of 2006, helping people the same way I did, but at the time he was
in Cuba. We would only talk by IM, because at that time Skype didn't exist. We had never
actually met but we talked like every day.
And then we shifted off into 2007. In 2007 we actually got a real website. The website
looked a little prettier, we added some more colors here and there. At that point is when
we officially released 2.2.0. This was a pretty big release because there was a whole bunch
of new stuff that we added in. We were changing all kinds of things because of DataTable.
DataTable was like the first of its kind, it was a really big widget and everybody was
like oh my God, look at everything this does! We actually released it at Yahoo like six
months before, that's why there was no 2.0.0 and 2.1.0, because those were internal ones,
and we came out with 2.2.0 as our first really big YUI 2.
What really made this one important was, back to Caridy again, Caridy had this thing called
the Bubbling Library. Have you guys ever heard of this one? The Bubbling Library actually
was really interesting because a lot of the fundamentals from YUI 3, Caridy was exploring
a lot of those in the Bubbling Library. He had this really firm grasp on the event bubbling,
of being able to put widgets inside of widgets inside of widgets and having them propagate
up and down. It was really ahead of its time at that point.
There was another one that came out, another extension. Have you guys ever heard of YUI-Ext?
I almost guarantee you you have heard of that. You guys heard of ExtJS? That used to be YUI-Ext.
It originally started as an extension for YUI and it was a plugin layer that sat on
top of YUI 2, and then eventually that ExtJS became Sencha. The last time I checked you
can see some of our old event system and stuff that were in several of the releases up until
a few years ago. You could look in there and see that their drag and drop was actually
powered by YUI 2's drag and drop, and their event was powered by YUI 2's event system.
We made a big shift in 2007 and 2008.
The other thing that happened in 2007 was I actually joined the team at that point.
A lot of people like to think that I was the first community member to join YUI and I actually
wasn't. I was not the first one because I was a Yahoo before that, so I worked at Yahoo
in 2006 and I worked in a different team. But I was getting to the point where Eric
Miraglia was actually throwing me out of classes. I would show up to take the YUI classes and
he was like you, just leave, you can teach the class, I'm just going to write it down
that you were here, and just go away.
At that point I took a class with Thomas Shaw and we were doing an IO class and everybody
in the class was just trying to make a simple IO request. He walks back to my machine and
I've got this animated thing bouncing around, showing the weather. I was bored. He's like
you need to come talk to the YUI team. I officially joined the team in 2007.
In 2007 we had a big chunk of modules come out. All of these releases came out. We went
from a little 19 modules to 31 modules in 2007.
In 2008 we got an even better website. They realized hey, this thing is starting to take
off, we're going to give you a little bit more, so we started redesigning a few things.
We had a whole list of really big modules come in, but the biggest stuff was in 2.3.0.
I really believe that 2.3.0 was the turning point for YUI. Everything that happened after
that point was a direct result of this thing.
The first thing is the loader. I mean how many of you guys use YUI 3's loader? Alright.
So that thing was actually built in 2007 and 2008 in YUI 2, and we prototyped it at that
time, and it was revolutionary. Even now it kicks ***, but back then nobody had ever heard
of it before. The editor was my first contribution to it. The reason I put editor up there is
because it was the first YUI widget that just used everything in the entire system. The
dependency tree was huge. That was when we started thinking what the *** are we going
to do with all this stuff?
The final thing that actually showed up is the combo loader. That was what actually showed
up and changed the *** planet when it comes to what we do. I mean how many of you
guys use the combo loader? Isn't that *** thing cool? That thing completely changed
the way that we do frontend engineering. I just happened when I was doing my slides,
I had a bag of Combos there and I just thought that was a really cool picture to put up for
it.
I truly believe 2.3.0 changed the way YUI was.... Everything that YUI had done before
it completely changed, because at that point we had to have a build system, we had to register
our modules, loader had to figure out all this intricate stuff. Then we had to figure
out how to load it, and then we realized that's a lot of *** files so we had to develop
our own technology to do this stuff, to slam these things together so that we could make
one single request or two single requests and just make it faster and easier.
The other thing that came out of 2008 was the Grids. I mean how many of you guys use
Pure? Okay you guys know that that's technically the YUI Grid system? Shake your heads if you
understand that, okay. The Grids actually came out in 2008, the first version. I went
home one night and I hacked up this little thing called a Grid Builder. You guys ever
remember using that thing?
That thing, it took off because Mr. Jeremy over there happened to post about it, and
it actually fried my server. I had it hosted on blog.davglass.com, and I had a phone call
from a buddy of mine that was hosting my little bitty box and he said smoke is literally pouring
out of the back of this thing. That's when Eric moved it to developer.yahoo.com where
it is today, and I had to throw up another machine and put one little redirect that threw
all of that traffic over there because it literally just ate up my entire machine.
That was a big turning point for us because then we realized it wasn't just JavaScript
we wanted to do, it was also CSS. It was very, very important to us, so we've always maintained
that we're not just a JavaScript library, we're a JavaScript and a CSS library. The
CSS is very important.
The other thing that we got in 2008 was our first logo. That was our first official logo.
It was so weird that all of us sat there and stared at our damn keyboard for how many years
and didn't realize that YUI was right there on the *** keyboard side by side.
[laughter]
Why did we not notice this? We're a bunch of geeks staring at our *** laptops and
we never noticed that those keys were side by side until someone at one of the frontend
engineering summits took a picture of a laptop really closely and posted it, and we're like
well duh, and we made it a *** logo.
The next thing in 2008 that was pretty cool was that this guy shows up. The funny thing
is that he showed up in Florida from Cuba. I didn't ask.
[laughter]
I have no *** clue what the hell happened in between the time that he disappeared until
the time he showed the hell up. I'm not saying that somebody dropped a big purple raft anywhere,
but he showed up and nobody asked. I will not ask.
But it was great to see that suddenly he just kind of disappeared for a little while, and
I used to talk to him every day and I'm like I don't know where the hell Caridy went and
all of a sudden he pings me and he's like dude I'm in Florida. I'm like what the ***
are you doing in Florida? He was like don't ask, I'm like okay, and I never have. All
these years I have never asked and I do not want to know, but he made it to Florida and
I'm really happy about that.
The actual final thing that happened in 2008 was we went on GitHub. This is the fundamental
total changing point for everything that Yahoo and YUI has done in open source, is GitHub.
When they launched that thing you can actually see that Yahoo joined it six months before
YUI did. So Yahoo actually had an account there and they were pushing source code to
GitHub before YUI had gotten their account. I didn't know that until I actually went to
make these screenshots, so I actually had to look up and find out what was there beforehand
because I didn't realize Yahoo had the account before YUI did. I mean that was pretty interesting,
that one surprised me. I was in shock on that.
The advent of GitHub, like I said, changed everything the way that we did. We had to
actually step back and look at what we were doing. I can remember having the conversations
with Adam back there when we first learned about this thing called Git. He came to me,
he's like dude you've got to check this *** out, it's totally cool. I was sold within
the first two minutes of looking at it, I'm like I see this is what we want to do. Eric
Miraglia actually pointed me to GitHub because it was in private beta, and he was like you
need to see this.
I started playing around with that, I started talking to the owners of the place, and then
Adam and I, without talking to anyone else on the team, decided that we're going to move
everybody from CVS to Git. We didn't take an opinion, we didn't ask anybody about it,
we just said *** it, we're going to do it. I spent several months exporting everything
from our CVS history into the two different GitHub repos.
I'm proud to say that in 2009 that not only were we doing stuff but Yahoo open sourced
our traffic server. Have you guys ever heard of the Apache Traffic Server? The Apache Traffic
Server probably serves a shitload of content on the internet that nobody even realizes,
and that thing used to be called the Yahoo Traffic Server. We open sourced it in 2009
and gave it to the world and said here, you guys can have it.
One of the other things we did in 2009 was we open sourced our version of Hadoop and
merged it in with the main codeline, and we've been a very big supporter of Hadoop after
that.
But for my personal part of it, 2009 was a very important year. I still believe it was
the best thing that I have done in the six and a half years that I was on the YUI team,
was pushing that *** to GitHub. Those are the exact tweets from the middle of January
of 2009 when I pushed the source code for YUI 2 and YUI 3 to GitHub. Before that we
had never released our source code. We had always done a build and we had put it in a
zip file and we gave it away that way, and people gave us *** for it. I gave us ***
for it, I was pissed.
So I spent my entire Christmas break importing all the CVS history, mapping all the user
names, rewriting all the history in CVS into Git, to formalize these Git repos and push
them right after everyone came back from break. A lot of you guys out here know that at the
end of December nobody does ***. We can't push anything because we don't want to break
production servers and stuff in the middle of the holidays. So I waited until after the
holidays were over and pushed this thing up, and I still think that is the greatest thing
I've ever done.
That is my pride and joy, was pushing that up, because it changed everything. Everything
that that team had ever known how to do, we had to learn again because we were now no
longer in private, we were now no longer doing this on our own, the people were weighing
in. That's you guys, you guys are telling us what you want to do, starting with this.
The other thing that happened then too was that we actually launched our first website.
We moved from developer.yahoo.com to YUILibrary. That's the original design that Nate Koechley
and I put together. Again, I didn't ask for permission, I just built the damn thing. I
build it and I put it up. The majority of that website was built during my paternity
leave when my daughter was born in 2009. I sat while she was asleep, I mean I was at
home with the kids but I mean, damn, it was a newborn, she sleeps all the *** time.
While she was sleeping I just pulled my laptop out and I built the site, I put the forms
together, I built the bug tracker by hand because we needed one and I couldn't find
one that worked.
This was our first attempt to rebrand ourselves. We tried to do that several times.
The other thing that came out in 2009 was YUI 3. 2008 we were doing preview releases,
but 2009 was the first official beta release of YUI 3.
Anybody else know what comes right after this, what else came out in 2009? Because I did
it the first YUIConf. Notice it now? The gallery. We launched that thing at the first YUIConf.
2009 I stood up in front of the entire room and just amazed them all because nobody had
ever done that before. We were opening up our CDN to anyone and allowing them to host
their modules.
We had a good sense of humor about it because I actually built the gallery in less than
a month. I had three and a half weeks to build the entire thing. I did half of it and Adam
back there did the other half. When I got up on stage to tell everybody the whole room
was like oh my God, this is cool, and I was walking you all through the whole thing of
how you would actually publish, and I went to submit my CDN request and the *** back
there denied me.
[laughter]
He actually denied my pull request in front of the room of everybody. I had to like change
my entire damn presentation to fit in and do it all over again so he could submit it,
and then he actually did a CDN push right there from inside the room.
I don't know how many of you guys were here for that first one, but we actually had a
CDN push because there was a member in the audience, Greg Hinch, he posted the form module
during my talk. As I was doing the talk he was following along, submitting the pull request
and everything that he needed to do and that was the first official one that went in the
library. How *** cool is that? It's totally cool that somebody does something like that.
Like I said, the gallery changed everything because now you guys were able to build modules,
and you guys were able to put them inside of your code exactly like we do, and you were
able to leverage our CDN, you were able to leverage all of our infrastructure to get
what you needed to get done. Like I said, nobody had done that before. It had never
been done before, especially on our scale, saying that we're just going to give you our
entire edge network, you get it all.
I didn't get a chance to do the stats on it but I know that we do one gallery build per
week, and that's roughly 52 a year, and that's 5 years. You guys do the *** math, I'm
not going to do it this morning. But that's a lot of *** builds, a lot of builds.
That was, I believe, toward October or November of 2009.
We step into 2010 and we started to make that new journey as a new team. I mean we were
learning things. The first thing that I did was of course I shook the whole *** planet
up and I put YUI on Node. Nobody had ever done that *** before. It had never happened.
A full JavaScript library running inside of Node, and it was like 0.0.30 or something
like that. Not only did I have YUI 3 running on Node but I had... I have no idea why that
threw over one. But I had YUI 2 running on Node, and it was actually rendering server
side HTML and ***. It was a complete prototype, it was a total *** hack.
I think that also changed the way that we looked at things because now we were no longer
just working with web browsers. This technology showed us that we could be using our tools
in all of these different places. It opened us up. I mean now we've got Arduinos, we've
got NodeCopters, and we've got that new Tessel that came out. Have you guys seen that little
device? It's got Node built in to the damn hardware. I mean how could we have seen that
in 2010? I knew that I loved this thing so I wanted to keep pushing it.
One of the other things that happened in 2010 which was that slide there, that was our first
official pull request to YUI 3 and it was for *** editor of course. But that one
was a big thing because we had never had someone contribute code before. Besides at Yahoo,
we had never had someone contribute code to YUI before that point because they couldn't,
we didn't give them the damn source code. Internally Yahoos could do it but no one in
the public could, and that guy was in the public, he wasn't a Yahoo. So that's the first
official pull request for YUI.
How *** cool is that? I mean come on. We went all those years being completely isolated
and hidden behind. Nobody knew what we were doing. We were taking out there, we were like
here, this is our stuff, if you like it, *** help us. So far it's going really, really
well. If you guys can't read that, there it is. 9/10 of 2010 at like 4am. I don't know
what the *** he was doing up at 4am besides... I'm the only one that's up at that time.
But at that point everything just completely changed for us. We started hiring more people
into the team that had the experience in open source, because at the time not a lot of people
knew how open source worked. GitHub changed that even more because it made it even easier,
because back in the old days it was a *** nightmare to get code into somebody else's
stuff. It would take months and months and months to get a patch, it would take sometimes
years to get new features. Now we're able to do it with a push of a damn button, it's
like oh my God.
We switch around to 2011, and this one right here, this is one of my personal favorites.
Where the hell is Eric? Okay, I wasn't looking. There's Eric. Eric shows up in 2011. Now you
know why this one's important? I just get crickets man, come on. Somebody. He is our
first contributor to YUI that was hired to the YUI team from his work in the public.
He had never been at Yahoo before, he was not an internal transfer, he was a community
member that was actually working with the public and was so good at what he did we wanted
him to come work with us. That's *** cool, right? I mean come on, wake the hell up!
[laughter]
Jeez! I know it's early but damn, it's 10 o'clock, it's noon for me! I've had like two
coffees already, I've been here since 5am. Wake the *** up!
But anyway, this one is awesome because it was a first for our entire team. No one had
ever done that before. All of the team members up until that point had been other Yahoos,
they had come from other departments, they had been other people at Yahoo that had used
this, but we had never officially hired a community member.
I was insanely happy when Eric joined us because it was another shifting point in the way that
YUI was going, is that we were becoming less and less of a corporate team and more and
more of a community team. It just keeps going downhill from there.
The other thing that happened right after Eric did was we actually got a new logo, and
this time it was different, it was newer, it was shinier, it was prettier.
The other thing that comes along with that, Ryan and I spent how long? Four months? Five
months?
Ryan Grove: Six years.
[laughter]
Dav: We spent a lot of time doing the new website, so that was another really big turning
point for us too because we were on our own. Ryan and I were the ones that accept pages
for the site going down. If it didn't work it was our fault, we didn't have an entire
ops team, we didn't have 100 servers to sit there. These were ours, we built them, we
maintained them. The night that it launched I was up all night and I think I deployed
35 times that night while everybody was using it.
It was a big shift because not only was it prettier and it was more well laid out, Ryan
did a great job with Selleck which is our documentation tool. We were able to actually
combine YUIDoc and Selleck in a very simple way that we could dump it into Mongo. We had
really super *** simple search, we didn't have to hardly do anything to get a really
good set of search out of it. And the search is awesome. Ryan is the one that wrote it
so everybody has to personally walk up and thank Ryan for the search because it's totally
*** cool.
How many of you guys have used that to find some *** information? Alright, so that's
his fault. He's right over there, he's in the hat with the... See? [laughs] So you have
to personally thank him for that because he worked very, very hard.
The other really, really cool part of this is that most of it's written in Node. That
was our first really big chunk to write one of our own sites in Node. But what's even
more interesting is the fact that it was written in two different languages. There's PHP on
one side and there's Node on the other. We blended them together and y'all can't tell
the difference, that's what's totally awesome. You can't tell when you jump from one part
of the site to the other which one of that is Node and which one of it is PHP until you
look under the hood.
Yeah, Ryan's over there shaking his head. Most people can't. You can because you know
what parts are.
But it was interesting, it was an interesting challenge because we had to actually make
three totally separate systems work together flawlessly. Ryan and I spent a lot of our
extra time working on this. I mean we were still full on in the team so we had all of
our normal things that we had to do, so we took this on as a way for both of us because
we knew we wanted to get it down and it had to get done, so we did it ourselves.
I loved it, I think that was the first time I'd ever worked with Ryan full time and I'm
not going to say any bad things about him. But no, Ryan's awesome and I love him to death.
The other thing that came out in 2011 was official support in YUI for Node. There were
no more hacks. We actually launched it into our CI system and we were testing over 7000
unit tests every time we committed on Node on the current version, the previous version,
and the next version, so we wanted to make sure that YUI always runs on Node. Because
we knew that that was going to be a partnership and it was a partnership that we wanted to
keep.
Moving on to 2012, the next thing that happened, which you can kind of see where it's going
here, is that Mr. Tony... Where is he at? Over there, there he is. Everybody look over
in that corner and embarrass Tony. Tony showed up in 2012 and he made the second official
external contributor to join the team. I think it's fantastic. I mean that's two people that
we had so far. These were community members, these were people that were in the IRC channel
helping other people. They were dedicating their own free time to helping others, they
were teaching people, and not only were they doing that but they were also bringing code
back in. They were submitting pull requests and they were fixing bugs. That was really,
really awesome.
It's this giant explosion of cool *** that just keeps happening because the next thing
is I'm going to embarrass Ryan again, him and Luke, when I announced at last year's
YUIConf and I physically stood on stage and added them with commit access. That was the
first time in history for I think any Yahoo open source product that an external user
had full commit access to our source. Tell me that's not *** cool, right?
[applause]
I know that it's going to keep going, it's just going to keep going and going, because
how many committers do we have now? 14 committers. We've got two, three now... Four? We have
four now that are not Yahoo employees. No?
Audience member: I don't know.
Dav: Yeah he doesn't *** remember. I was too damn tired last night to look it up.
But the whole point is that we had never done that before, so we were stepping in even more
and throwing more and more of the community. Like I said, we started learning early that
we had to make this change, so we just kept slowly and slowly and slowly making this change.
Then we come to what we've been doing in 2013. I've been following it, I mean even though
I wasn't around for half of 2013 because I was laying on my back with a busted back,
2013 had a brand new series of things come in.
We've now got Pure. I was just completely flabbergasted, I was speechless, period, when
I saw Pure, because I had no idea it was coming. I was gone the entire time they worked on
it and then come to find out the thing's got 6000 stars on GitHub. It actually has more
stars than YUI 3 does. Most people don't realize that's all YUI 3 stuff, all that CSS is YUI
3 stuff anyway. But I love it, and I think it's great.
One of the final things that actually happened in 2013 is Caridy is an official YUI team
member. From 2006 to 2013 he was just cool, but now he's super cool because he's a member
of the actual team. It was awesome to get that notice too because I think it happened
when I was out as well and I didn't know about it until I got back.
It's cool because that technically makes him, what, number three, number four, community
member that worked their way into the YUI team? It took him a while. He had to earn
it. But he made it, and we're really happy about that.
One of the other ones is Juan. You guys know Juan made it, right? Where he's at? There
he is. Embarrass him.
[applause]
That one was insanely happy because I didn't even know it happened. I saw a tweet from
him that says I'm at Yahoo and I'm like what, where the *** did that come from? It actually
had Sunnyvale on the geolocation at the bottom. Like what the ***? Why did nobody tell me
this ***?
I mean you can see that they're moving, and it's constantly moving to all the new stuff
and everything. It's fantastic to see. Even considering that I came in in 2006, seeing
all of this stuff and looking back you can see the key points, you can see where it's
going and where it's been. I love it. I mean the YUI project is such a fantastic project
and it's even better when the community is involved.
I have always been a really big guy on being the community. I love the community aspect
of it, that's why I was answering questions. I don't think I've been on the forum in nine
months but I still have the number one posts. Satyam hasn't beaten me yet. I had number
one posts on the mailing list. It's really heartwarming to see the difference that comes
through. It's even nicer now that I haven't been on the team in the last six to nine months
and still watching what's been happening, because I see it, seeing that they're still
pulling in community members and they're still making it a community project and not just
a corporate project.
The super, super cool stuff about all this is that Yahoo does this not just with YUI.
We do it with almost every other product that we have, and most people don't realize it
because you guys aren't in that industry. I'm stepping into that industry now and I'm
seeing all these other projects that Yahoo is giving away and all of this massive source
code. They're going to learn the same thing, so that's what I've been trying to do is go
into these other projects and teaching them the same things and the same things that we've
learned, the things to do and the things not to do and the right ways to be doing this
stuff.
So yeah, I ran a little quick on that one so that's all I've got for you because we
started a little late. But thanks. This is the first YUIConf that I've spoken at that
I was not on the team. It was different.
The gravatar that I've always used with my hat on, that was taken in 2007 at what actually
became YUIConf. We used to have these things at Yahoo called the F2E summit, and the F2E
summit then turned into the YUIConf. In 2007 I spoke at the first one. It was my first
technical speak at a YUIConf, and I wasn't on the team either. It's kind of funny that
in 2007 I did my first talk and then here I am in 2013, I'm doing this one and I'm not
officially on the team.
But like I said, I want to say thanks for the six years that I was on the team. I really
loved it. I'm still technically involved, but they just take their time to ask me a
question. They wait until they've got me cornered somewhere and then ask me a question. But
again, thanks a lot everybody, and I hope you guys have a good day, alright?
[applause]
Jenny: Thanks Dav. We have time for questions. If anyone has a question please raise your
hand.
Audience member: Thank you for sharing this, it's totally awesome. I would like to know,
every single year of YUI is, I mean, different surprise in a positive way to me. I would
like to know what's going to be in the next coming future, like for 2014.
Dav: That's Eric.
Audience member: Oh yeah, sorry, I didn't come yesterday morning, so...
Dav: That would be Eric, that's his department now.
Eric: Some of the stuff that I talked about yesterday was really around looking at how
we can revamp the module system with ES6 modules and kind of align with some of these new emerging
technologies. Also yesterday Tilo talked a lot about the event gesture support that he's
been revamping. Clarence and also Caridy were talking about some of this awesome stuff we're
doing with Node and ExpressJS.
Jenny: All these videos will be available on the YUI Theater video channel so you can
check them out after the conference.
Audience member: Now that you are not part of the team anymore but moving to other things,
how do you see Yahoo adopting Node.js as the backend or middleware technology all across
the board?
Dav: Well right now we're not going all across the board. What we're actually doing is anything
new that is written, they get to choose which frontend framework they want to use, and most
of them are actually switching over to Node.
Several of our latest deployments, as in Flickr, the new view a photo page, have you guys seen
the new page where you actually look at a photo? That's actually built on Node. You're
actually hitting our Node server for that page, and that's the number one page on Flickr.
All of their traffic, that's the one that gets the *** kicked out of it the most and
they're like well, *** it, turn it all the way up, so they just turned it all the way
up to 100 and they let it run. That's pretty cool.
There are parts of my.yahoo.com and there are parts of our media sites, so some of the
sports sites, are all running on Node. Parts of search are. Yeah, so we've got parts of
search, parts of the media properties. They're just making the switch slowly but surely.
We can't rewrite everything now, but what we do is as we're moving forward and writing
new things, we're trying to apply that as the technology we use when we build it.
Andrew: Any more questions?
Dav: Anything else?
Andrew: Come on, it's Dav.
Dav: Yeah, usually I get like 100 billion questions. Alright.
Jenny: Thanks Dav.
[applause]