Posts Tagged ‘Engineering Team’
Facebook makes another talent acquisition: Octazen Solutions
Facebook makes another talent acquisition: Octazen Solutions
Facebook has acquired Malaysian startup Octazen Solutions, which specializes in contact importing, GigaOm reports.
Facebook spokesperson Larry Yu called this a “talent acquisition,” as Octazen’s two employees will work as engineers for the company out of Malaysia. Facebook has made a series of design tweaks encouraging users to enlarge their social networks on the site by highlighting its automatic friend finder.
“We’ve admired the engineering team’s efforts for some time now and this is part of our ongoing effort to add experienced, accomplished technical talent to help drive the company forward in its efforts to be the central way for people to connect and share information.”
The acquisition marks the third in a series of talent acquisitions, following Facebook’s purchase of FriendFeed last August and Parakey in 2007, which brought aboard Facebook Lite lead Blake Ross and Joe Hewitt, who is responsible for designing the company’s early iPhone apps.
Facebook is in a mature, high-growth phase and has been aggressively courting engineers, both at senior levels and for entry-level positions. CEO Mark Zuckerberg has even asked for suggestions on which companies to buy on Quora, a question-and-answer site founded by the company’s first chief technology officer, Adam D’Angelo.
“The company recently received an offer to acquire most of the company’s assets and to employ those assets in a different direction. After carefully evaluating this offer, our team believes this is a wonderful opportunity of which we must take advantage.
As a result, effective immediately, Octazen will no longer accept new service contracts or renew existing service contracts, and will enter a transition period to wind down operations.”
Companies: co:Facebook, Octazen Solutions
16 month-old bug continues to crash Flash
16 month-old bug continues to crash Flash
Filed under: Multimedia
Matthew Dempsky has discovered a bug which will crash the Flash player on every supported platform. That might not seem like a huge deal, except that he discovered this bug in September of 2008 and has reported it to Adobe, which hasn’t fixed it yet.
16 months later.
If you’d like to test it for yourself, make sure there’s nothing important open in your browser window and head to http://flashcrash.dempsky.org/.
In Safari and Google Chrome, this crashes the plugin but not the browser. It took Firefox 3.6 down entirely.
Why would Matthew post such a page? Isn’t that reckless? Well, he explains on that page:
“Regarding crashing, I can tell you that we don’t ship Flash with any known crash bugs, and if there was such a widespread problem historically Flash could not have achieved its wide use today,” Lynch wrote. “Addressing crash issues is a top priority in the engineering team, and currently there are open reports we are researching in Flash Player 10.” (Source: PC Mag, “Adobe Defends Flash, Calls Apple Uncooperative”)
He goes on to say:
This page exploits a bug that I reported to Adobe in September 2008, and has affected every release of Flash on every platform since then. Despite numerous email exchanges with the Flash product manager about the bug, the bug report being hidden from the public for “security” reasons, and [although] Adobe CTO Kevin Lynch’s claims otherwise, it continues to be an issue.
…I’m not an Apple fan boy out to prove Steve Jobs right in Apple’s decision not to support Flash on the iPhone / iPad. Instead, I’m just a software engineer who at one time had to deal with Adobe’s sorry excuse for a development platform and made an earnest effort on several occasions at helping them improve it for everyone. (This issue is merely the tip of the iceberg of ridiculous bugs and random backwards and forwards incompatibilities known as Adobe’s Flash Player plug-in.) After trying to work with them to fix this issue and experiencing nothing but frustration, I just don’t give a damn anymore.
Adobe has been able to rest on its laurels with Flash, because it was a de facto standard. Now that the platform is being left behind by new mobile devices and computing metaphors, Adobe is making an appeal to the public that Flash isn’t that bad.
Adobe’s been able to do much the same with Photoshop and CS4. Even people who love the apps and use them every day have learned to live with the crashes and other problems. Adobe seemed not to be in too much of a rush to get Snow Leopard compatible versions out. Ditto for when Apple switched to Intel.
I’m amazed by people who continue to defend Flash, including those who believe that alternatives will have a chance if web developers weren’t pushed to start using newer alternatives like H.264 and HTML 5. (No, I’m not saying H.264/HTML 5 is a drop-in replacement for Flash, and I’m not even going to mention SVG.)
If we all went with the “de facto standard” we’d be using Internet Explorer 6 on Windows. Actually, we’d probably be using Internet Explorer 4.
No doubt that Flash has done some great things. At one time, it was cutting edge stuff. Now it’s a dull butter knife.
And I’d be remiss if I didn’t remind you about ClickToFlash which I’ve reviewed previously.
(Hat tip to Craig Hockenberry and Mike Damm for bringing this story to our attention.)
TUAW16 month-old bug continues to crash Flash originally appeared on The Unofficial Apple Weblog (TUAW) on Sat, 06 Feb 2010 16:30:00 EST. Please see our terms for use of feeds.
Read | Permalink | Email this | Comments
Read the whole story…
Former Apple Software Chief Joins His Pals At Elevation Partners
Former Apple Software Chief Joins His Pals At Elevation Partners

Avadis “Avie” Tevanian, a former chief software technology officer at Apple, has jumped ship to Elevation Partners, where he will be a Managing Director.
Tevanian joins Fred Anderson, Apple’s former CFO who is a managing director at Elevation. Elevation is also the largest investor in Palm, and brought in Apple exec Jon Rubinstein as CEO of the mobile company. Tevanian will help Elevation discover new investment opportunities and develop strategy for the firm.
Tevanian spent nearly 10 years at Apple. Prior to serving as Chief Software Technology Officer, he was Senior Vice President of Software Engineering, where he led the software engineering team that developed Apple’s operating system OS X. Earlier in his career, Tevanian was Vice President of Software Engineering at NeXT Computer. Tevanian holds a B.A. in mathematics from University of Rochester and earned his M.S. and Ph.D. degrees in computer science from Carnegie Mellon University. While at Carnegie Mellon, Tevanian designed and engineered the Mach operating system, which OS X is based upon.
Elevation also recently pulled in Palm director and former eBay exec Rajiv Dutta late last year to help the private equity firm raise money for a second fund.
Crunch Network: CrunchBoard because it’s time for you to find a new Job2.0
Search, Monetize and Fact Check YouTube Transcripts with Speakertext
Search, Monetize and Fact Check YouTube Transcripts with Speakertext
You’ve probably never heard of Matt Mireles and Bjorn Liljequist but with a $4000 dollar budget and an engineering team paid in iPhones, the two already have Meebo founder Seth Sternberg as their advisor and praise from VC Fred Wilson. The duo’s filtering service Speakertext will launch at tomorrow’s New York Tech Meetup and the concept is a simple one – to make video interesting.
Like Tubechop, Speakertext allows users to omit the boring parts of a video; however, the service’s transcription component offers a new and important twist. Says CEO Mireles, “At some point, longer videos become useless. It’s the metadata and the fact that we’re allowing it to be indexed that make this a great tool.”
The service uses the YouTube API and replaces the YouTube player with a Speakertext player. Users can search video text for relevant quotes and embed the linked quote or the Speakertext player and video into their blogs. To index your own video with the system, you can either transcribe it yourself or opt into a Mechanical Turk package. For $20 dollars per hour of video, you can have speeches, events and podcasts transcribed. The company plans on creating a premium service and launching on additional video platforms in the months ahead. For now, Speakertext offers the following benefits:
Attribution & Monetization: Says Mireles, “Video piracy often happens because users can’t get the clips they want from the original publisher. We’re building a monetizable search strategy for video publishers.” Because Speakertext just sets the original video to a new start point, there is no need for a new file to be uploaded by a secondary publisher. In this way, the secondary publisher highlights the content they want and the original publisher maintains credit, links from search and potential revenue channels.
Accuracy: Stop for a moment and think about the quagmire of misinformation swimming across the web right now. From religious figures to political candidates to CEOs, there is no shortage of misquotes in the media today. If you want to be a transparent company, government or influential entity, video transcription is a great way to dispel myth from fact. The company will also crowdsource transcription accuracy in a method similar to that of the Worldwide Lexicon Project.
Education and Research: Video has been a boon for visual learners. Now, rather than simply searching titles and descriptions, transcription will allow users to find valuable educational resources. The company is already speaking to open courseware proponents to increase access to quality materials.
Mireles points out that Google is currently in the process of automating captions for YouTube. Says the startup founder, “We have no doubt whether searchable video like this will be the standard in 5 years. We just want it to be us that does it.”
Speakertext will launch to the public at 5pm PST tomorrow evening at speakertext.com.
Thanks to David Cohn for the tip!
Phone Modders, Take Note: HTC Releases Hero Source Code
Phone Modders, Take Note: HTC Releases Hero Source Code
Following weeks of requests from open-source developers, HTC has released the Hero Android source code on their developer site.
In response to the mobile devs, HTC has previously stated it was waiting for its own developers to provide the source before releasing it publicly. As late as last week, HTC representatives had emailed developers saying, “At the moment we do not know when the kernel source for the Hero will be released,” and “We are still pushing our developers to provide us with the source code and for the links to be added.” Since the Android kernel is licensed under the GPL, this delay was creating both dissatisfaction and controversy in the community.
However, just as a few developers were beginning to talk about enforcement actions, the company posted the code, and everyone lived happily ever after.
Or something along those lines. GPL non-compliance and hints of internal process and delivery issues don’t mode well for the mobile manufacturer. After unfavorable coverage of the company’s “foot-dragging” on Slashdot and long threads of modev complaints, we do hope that HTC’s future Android projects will be more swiftly opened.
The Hero, as a device, is significant in itself, hence the enhanced perception of cruelty in HTC’s not releasing its source code sooner. It’s created huge waves in geek circles, beating out the iPhone for Gadget of the Year at the prestigious T3 awards and generating enough gadget-geek slavering to power a small city.
So, will the gadget geeks and modevs have to push for open sourcing every time a cool, Android-powered device is released? Where was the major malfunction that led to these delays? Were the HTC engineers thrown under the bus to allow leadership to save face, or do the HTC powers that be simply need to get their engineering team under control? Let us know your thoughts in the comments.
A busy week for real-time search — here’s a list to keep tabs
A busy week for real-time search — here’s a list to keep tabs
After scarcely a peep for much of the summer, a handful of the real-time search startups we profiled earlier this year have ramped up their offerings this week. They’re part of a wave of companies that are mining the increasing amount of data shared on sites like Twitter and Flickr to offer search results based on what’s relevant now.
In general, we’re seeing more traction from companies that are trying to distribute their search and data collection technology rather than centralize it in one destination site. One of the older companies, OneRiot, turned on its revenue model this month by selling sponsored search results. (It has a distributed approach, partnering with at least 40 other companies to feed its results into other sites.) Tweetmeme, which has a retweet button that’s seen at least 50 million times a day in addition to a search engine, launched analytics for companies that want to track the viral spread of their content through Twitter.
(Here’s a basic primer on all of the companies for background.)
And then here’s what’s new:
Topsy, which raised $15 million over the summer, released two plug-ins: one for Wordpress and another one for your browser. When people tweet about a blog post, the plug-in will find it and include it as a “native” Wordpress comment at the bottom of the post. The browser plug-in adds a Topsy search bar to the top right-hand corner of a browser so users don’t have to navigate away from the page to search.

Crowdeye, founded by former members of Microsoft’s search engineering team, now lets users tweet directly from the site. Visitors can also track the most popular content from specific domain names like VentureBeat.com (see the snapshot below or click here to test it out).

Scoopler, founded by two recent college graduates who met in the U.K., cut the load time on its pages and added a “Discovery bar” at the top of its page to show trending topics and recent searches. The company also added channels of content, making it easier for visitors to keep track of popular links in topics like technology and sports.

To help publishers and brands figure out how much additional traffic Twitter is driving them and optimize it, U.K.-based Tweetmeme launched sophisticated analytics features. Tweetmeme will break out retweets by geography and show a publisher who their most influential readers are based on how far they drive a piece of content through their social network.

A little over a week ago, OneRiot, unveiled what it believes will be its primary revenue model. It’s selling text ads that will appear alongside relevant search results. The layout is similar to Google’s but OneRiot is selling sponsored placement for content, not commercial goods and services. This is because the company believes that when visitors search for real-time results, they aren’t necessarily in the state of mind to buy products. They’re looking for context or news, unlike in traditional search. Therefore, ads for goods and services won’t necessarily work. Instead, publishers will pay to have their relevant content promoted when visitors are looking for recently published items.

A Look Behind The Curtain At Facebook’s Optimization Efforts
A Look Behind The Curtain At Facebook’s Optimization Efforts
Facebook is big. Really big. So it comes as little surprise that every tweak made to the site (like the subtle change to the header a few days ago) can have a pretty substantial impact on the way people use the social network. Earlier this week Facebook’s Engineering team posted a note written by intern Zizhuang Yang, who has spent the summer researching how changes in things like load time can affect users. Yang writes about three main experiments he conducted over the last few months, including one involving overall site speed and two in the way pages load, and the results are quite interesting.
The first experiment examined how Facebook users would respond to a general slowdown. Yang found that regardless of site speed, users spend around the same amount of time on Facebook. That might sound like good news (at least they don’t get frustrated and leave immediately), but it means that if the site is running slowly users are going to be seeing fewer pages in the same amount of time, which Facebook obviously doesn’t want. So — no surprise here — Facebook is striving to make the site as fast as possible.
The second experiment involved the order in which items on the page should load. Yang writes that Facebook has been internally debating whether the page should display everything as quickly as possible, even before some necessary scripts to actually interact with the site have loaded, or to show a white page until everything is good to go and then render it all at once. Yang writes, “In all groups of users, keeping the page blank resulted in lower usage statistics. Thus the debate was resolved.” So if you’re ever on Facebook and you find that a certain button isn’t working for the first second or two after a page loads, this would explain it.
The third experiment involved loading stories in the News Feed. Regular Facebook users have likely noticed that the site will automatically fetch more News Feed stories as you scroll down the page. This feels like a nifty new feature, but it was actually designed by Facebook to cut back on load times — News Feed used to show 30 stories at once; now it loads 15 at first and only shows the next 15 if you scroll down the page. What Yang found, however, is that when people do scroll beyond the initial 15 stories they’re shown, they’re happy to wait the extra second or two for 30 new stories to load, which results in a signifiant boost in engagement. This makes perfect sense — if I’m actively reading through the News Feed (as opposed to just seeing it because it’s Facebook’s home page) it’s because I’m killing time or trying to catch up on my friends’ past posts. The more stories shown during this ‘catch-up’ time, the better.
Also interesting to note is that Facebook seems keen to put its internship program in the public eye — just last week the site’s blog included a post from an intern who build the Facebook Pages to Twitter syndicator, and now we’re seeing the fruit’s of another intern’s summer experiments. This may well be part of the company’s plan to attract new talent during its recent hiring spree.
Crunch Network: MobileCrunch Mobile Gadgets and Applications, Delivered Daily.
TokBox Adds Document Collaboration Powered By EtherPad
TokBox Adds Document Collaboration Powered By EtherPad
TokBox, the web-based video chat application, has announced that it has partnered with EtherPad to bring document collaboration straight into TokBox. Now, all Tokbox users can simultaneously collaborate on one text document or “pad”, while chatting in a video call. What’s cool is that when multiple people edit the same document at the same time, all changes are instantly reflected on everyone’s screen who is in the document. Once your done with your document, you can save it for later use.
Essentially, you start a call on TokBox and invite up to 20 friends or coworkers to the call. You then put in your URL for EtherPad in an existing URL pad.
After speaking with TokBox CEO Ian Small, TokBox wants to focus more on collaboration, and they have started a great relationship with EtherPad. Small also mentioned that EtherPad will be integrating TokBox into their service in the next couple of weeks. According to Small, EtherPad will be able to collaborate on documents with the addition of voice and video chatting powered by TokBox.
It’s still unclear what TokBox’s business model is, but they’re possibly positioned well because they are not limiting their service to an operating system, but just to the browser. TokBox also recently laid off 50% of their engineering team, and all the companies founders have left the company.
Crunch Network: MobileCrunch Mobile Gadgets and Applications, Delivered Daily.



