Tip:
Highlight text to annotate it
X
>> I’m here with Eye.fi. Could you please introduce yourself?
>> PRAJ: Hi, I’m Praj, I’m with Eye.fi. We’re a small startup based in Mountain
View, California. >> So, what do you guys do? What kind of parts
you guys make? >> PRAJ: So, what we do at Eye.fi is amazing.
We make this Eye.fi card and it automatically transfers pictures from your camera to your
computer or web--web- sharing sites such as Picasa or YouTube.
>> Well, have you implemented any of the Google APIs?
>> PRAJ: Yes, yes. We’ve been integrated with Picasa Web. We’ve used the authentication
of APIs extensively and, at the same time, creating albums and uploading pictures. And
I should say, the same with YouTube. We’ve accomplished a great deal with YouTube as
well transferring movies from your camera directly to your YouTube accounts.
>> Great. I mean, can you tell us a little bit about why you chose Informatix (ph), Picasa,
and YouTube APIs over other APIs like, let’s say, Flicker? Or have you--you’ve already
implemented Flicker as well? >> PRAJ: Yes, we’ve implemented Flicker.
And we do have a lot of users, you know, there’s a lot of demand for it, and the APIs are very
simple. They can--they can get a little tricky in terms of the data we get back, but they’re
very simple and, you know, that the clients--, the clients libraries that are there right
now are very robustly written, and we actually in fact use the GData API client library to
upload to YouTube. >> Do you have any anecdotes that you want
to share… >> PRAJ: Yeah, those are…
>> For everybody? >> PRAJ: Yeah, absolutely. There was a time
where we were trying to--this was when we were trying to upload videos for the first
time from an Eye.fi card. And, at some point, we were just like, “Where did the upload
go?” ‘cause it was just, you know, lost in the system. And we figured that there was
an issue with the memory management inside those classes, so what we did to remedy, to
rectify that was basically do chunk uploads. So, we were, instead of like, say, pushing
in a gig file in to the class to YouTube be where--you know…
>> Splitting it up? >> PRAJ: Splitting it up into 4 kilobytes
each and then sending it through. So, that was pretty, pretty neat. I mean, you know,
basically you’re doing a socket connection and kind of using that technique to it to
do it. >> And then there was actually a bug report
and we’re not aware of the bug and we’re kind of resolving it.
>> PRAJ: Well... >> Well, that’s a good plan. It’s a very
interesting stuff. >> PRAJ: Thank you.
>> You have anything else that you’d like to add?
>> PRAJ: Yeah, as far as the APIs are concerned, you know, the Google data APIs are awesome
and--in fact we extensively use the Maps API as well. So, what the Eye.fi card is going
to do is it actually geo tags the pictures and the videos that you take.
>> : Cool. >> PRAJ: So, we can show it very, you know--
in a real-time basis as far as the users are concerned to--they can see where they took
the picture. And we actually plotted down on, you know, the Maps to show them.
>> That is--that’s a really nice feature, I think.
>> PRAJ: Yeah. >> It could be released well in the future,
especially on, like, vacations wherein you want to see where you took the picture.
>> PRAJ: Exactly, right? You don’t know where you are at and, you know, it’s (INDISTINCT).
>> Great, thank you for coming at the I/O and this interview.
>> PRAJ: Thanks for inviting me.