Newer posts are loading.
You are at the newest post.
Click here to check if anything new just came in.

March 29 2012

Developer Week in Review: Google I/O's ticket window open and shuts in record time

This week, I'd like to take a moment to thank the good folks over at Parkland Medical Center, who took pity on the retching, sweat-covered soul who appeared on the doorstep of their emergency room last Friday morning. They swiftly (well, after 15 eternal minutes in the waiting room, which is pretty swift for a walk-in to an ER) got him hooked up to an IV and introduced the two God-given holy fluids of morphine and Dilaudid. On a totally unrelated note, I'd like to proudly announce the birth of a healthy 3mm kidney stone at 5PM last Friday. Donations to its college fund can be made ...

Extending the trend line doesn't look good

Google IOLast year, Google I/O sold out in under an hour. This year, it only took 20 minutes. If we extend the trend-line out a few years, the only people who will be able to get in will be those who have access to micro-second responsive stock market trading programs and hyper-tuned eBay auction sniping software.

At least, however, Google fans have some clue when the registration opens for their conference. Those of us still waiting for Apple's WWDC conference know it will have to open for registration soon, but the exact date and time is a mystery. Thankfully, the multi-thousand dollar registration fee tends to make WWDC a bit slower to fill up, but it will still be a race for those who require authorization from their management to go (some of us get authorization months in advance, specifically for this reason).

If there's a solution to this classic supply versus demand problem, I can't see it. Regional conferences reduce the benefit of getting all the developers together in one place and would have the companies sending their development staff to the four corners of the world. Maybe Apple and Google need to start renting out football stadiums instead of conference facilities.

More pigs spotted airborne

For those who have been taking a skeptical view of Microsoft's avowed embrace of the open source movement, there's more reason to believe it's genuine. This week, Microsoft released a whole crop of its .NET technology to its CodePlex open source repository, and the company did it under the hyper-liberal Apache 2.0 license rather than something proprietary and restrictive. In addition, Microsoft has started using the developer-friendly git source control system — another attempt to make itself more compatible with the open source community as a whole.

Of course, releasing portions of its proprietary environment as open source is still an attempt to get people to use Microsoft's technology as a whole, including Visual Studio, but the more it puts out there under licenses that include patent grants, the more possible it is to incorporate compatibility with Microsoft products in non-Microsoft platforms and products.

Was the cake made out of 0xDEADBEEF?

It's practically unimaginable today, but when the gcc compiler was first released 25 years ago this week, the only way to compile your code was to pay your hardware vendor for a proprietary compiler package, sometimes costing tens of thousands of dollars.

In the intervening years, the shining star of the Free Software Foundation (FSF) has become the go-to (excuse the expression) compiler for most modern compiled languages, available on and for just about every hardware platform you can think of. It doesn't have the death-grip hold on the industry it once did, with Apple among others moving to LLVM, but it was the first and for a long time the best compiler money couldn't buy. You may argue with the current philosophy of the FSF, but give it due props for opening up the world of programming to the world by making free tools available to anyone who wanted them.

Fluent Conference: JavaScript & Beyond — Explore the changing worlds of JavaScript & HTML5 at the O'Reilly Fluent Conference (May 29 - 31 in San Francisco, Calif.).

Save 20% on registration with the code RADAR20

Got news?

Please send tips and leads here.


March 22 2012

Developer Week in Review: The mysterious Google I/O machine

We're in the countdown days to the two big annual developer conferences (not counting OSCON, of course ...). Google I/O will open registration on March 27th, and if past history is any guide, WWDC should also start (and end) signups around the same week. So, get your credit cards warmed up and ready. Last year, both conferences sold out in less than a day (Google I/O in under an hour!).

And speaking of Google I/O

Google IO game

Just what is the purpose of the Rube Goldberg-esque physical puzzle that has gone up on the Google I/O website. Does it have something to do with a puzzle that potential attendees will need to solve to register? Will attendees be flung around from session to session by giant pendulums? Is it all just a cool demo of Chrome? And does it have anything to do with ancient Mayan prophecies?

In any event, it's a fun (if simple) game, worth a few moments of your time, but unlikely to absorb more than 15 minutes of your attention. Now, if they added achievements and a Zombie mode, that might be something.

So much for sandboxing

Reports of a successful exploitation against the Chrome sandbox appeared recently, and now word has broken that a new Java exploit not only breaks out of the sandbox, but manages to install itself into system memory, where it can mess around with privileged processes. Worse, unlike the Chrome exploit, which was reported to Google and not in the wild, this new Java hack is being actively distributed on popular Russian news sites.

Since the entire point of a sandbox is to keep malicious code from getting access to system resources, it is truly disheartening to see how frequently sandboxes are being penetrated these days. If there's one piece of code that needs to be rock-solid, it's the bit that keeps the bad guys from doing bad things. That it fails so often in reality either indicates that developers aren't doing a good job, or that it's a really hard problem and it may be time to rethink sandboxing as a valid security approach.

Go is almost a Go

For those who have been eagerly awaiting Google's attempt to reinvent the wheel new programing language, Go, the wait is almost over, as RC1 has just hit the street. According to the developers, this is very close to what the final 1.0 release will look like. If you've been waiting for a stable version of Go to kick the tires, now is probably the time.

As with most new programming languages, I am maintaining a healthy degree of skepticism as to the long-term viability of Go. This is not because of any inherent faults of the language, but because of the institutional inertia that new languages have to fight to gain acceptance. Whether Google's influence will be enough to get Go ensconced in the pantheon of mainstream languages is yet to be seen.

Fluent Conference: JavaScript & Beyond — Explore the changing worlds of JavaScript & HTML5 at the O'Reilly Fluent Conference (May 29 - 31 in San Francisco, Calif.).

Save 20% on registration with the code RADAR20

Got news?

Please send tips and leads here.


February 09 2011

Developer Week in Review

I am led to understand that there was some kind of sporting event on Sunday. As our brave New England lads were not involved, I foreswore it (except for watching the commercials on Hulu, of course ...) But now that the nacho-and-beer-induced stupor that many of you seem to have been suffering from has worn off, let's see what happened around the industry last week.

The 49th caller wins tickets to Google I/O

If you had your heart set on a trip to the beautiful George Moscone center for Google's annual developer love-fest, you needed to have fast fingers this year. The general registration signup opened and closed again in under an hour, after the website took a beating usually reserved for iPhone reservations.

Your weekly Oracle news

Proving once again that Oracle totally fails to understand the open source community, there's now a major todo brewing around the continuous integration project formerly known as Hudson. The basic timeline appears to be as follows:

  • Oracle takes down the source repositories for more than a week to do infrastructure changes, without warning the developer community.
  • The founder and chief committer created a GitHub project so that work could be done while Oracle got their act together.
  • Oracle got snotty and said that the GitHub project couldn't be called Hudson because Oracle owned the trademark.
  • The developer community said "Fine, we'll change the name to Jenkins."
  • Oracle said "You can call it whatever you want, but it'll be a fork."
  • Hilarious hijinks ensue!

Keyboards for real programmers

LISP MachineAfter last week's (DMCA'd) peek at the Internet, circa 1994, here's another blast from the past: a collection of keyboard photos from the early LISP machines! These keyboards had a set of modifier keys yet to be surpassed in the world of programming, sporting not only a control and alt (meta) key, but also a hyper and super key. All of them could be chorded, leading to such memorable key presses as "control-meta-hyper-yu shiang whole fish."

Regrettably, the page fails to display any keyboards from LISP Machine, Inc, which was my first corporate employer (after a six-month-stint working at the MIT AI Lab.) However, if you look at the photo, you can see what was the bane of my existence when I did field service calls for LMI, the notorious A=M ECO (Engineering Change Order). There's a series of white wires running down the center of the front wire-wrap panel. They are twisted-pair with one wire grounded on either side, and were put in because the original single-wire runs were sensitive to induced current from wires they crossed. There were 32 of them, they needed to be unwrapped on both ends from the pins, teased out with dental tools, and then the twisted pair wires added in place of them. It took hours, and woe to you if you broke another wire while removing the target ones. But tell that to the kids today ...

You young whippersnappers come back next week, and Uncle James will tell you more tales of computers new and old. Until then, get off my lawn! Suggestions are always welcome, so please send tips or news here.

June 02 2010

Four short links: 2 June 2010

  1. Wikileaks Launched on Stolen Documents (Wired) -- Wired claims the first set of documents was obtained by running a Tor node that users connected to ("exit node") and saving the plaintext that was sent to the users, without their knowledge. Reminds me of the adage that nothing big in Silicon Valley starts without being some degree of evil first: YouTube turning a blind eye to copyright infringement, Facebook games and spam, etc.
  2. VC Investments in Education -- Cleantech investors are chasing a 3x larger market than Education and yet are putting 50-60x the money to work chasing those returns.
  3. Cells: A Massively Multi-Agent Python Programming Game -- a sweet-looking update on the old Core War game.
  4. Google IO 2010 Session Videos Online -- I'm keen to learn more about BigData and Prediction APIs, which seem to me an eminently sensible move by Google to play to their strengths.

May 24 2010

Google vs Apple: Google doesn't need to win

So now we're apparently in a Google versus Apple fight to the death. Google open-sources VP8 (now WebM), and Steve Jobs immediately throws cold water on it. Apple got their share of scorn at Google's I/O conference. Google thinks they have mobile/cloud/desktop integration nailed--and if what they demo'd last week actually works in FroYo (Android 2.2), they probably do.

But the notion that this is a "fight to the death" is a bit bizarre, even though it's been portrayed that way, and by none other than Steve Jobs, who earlier this year said that Google is out to kill the iPhone. If it is a battle, the terms are uneven.

My own disclosures: I'm definitely a Google fan. And I'm also an iPad-equipped Apple fan, though I am also very unhappy with the closedness of the Apple platform, and the way they treat their developers. But they make beautiful hardware, and they really understand "it just works."

Vic Gundotra nailed it in his second keynote at Google I/O. When he was starting at Google, he asked Andy Rubin why Android was important; why did the world need one more mobile platform? Andy's answer was that Google had a very dismal future if they didn't address
mobile; we'd end up with one platform, controlled by one vendor, and one carrier. It was a wise and prophetic answer. If I've pieced the chronology together correctly, this would be about the time the iPhone was coming online. And the iPhone is a great device--great, but ultimately closed.

Apple makes hardware, and the more hardware they sell, the more money they make. So Apple clearly wins if they sell iPhones to everyone--the more iPhones (and iPads), the more they win. There would be nothing better for them than driving the other smartphone manufacturers out of the market. (They don't seem to be interested in low-end, low profit margin phones, but that's another story.) So what it takes for Apple to win is clear: dominance of the smartphone market.

Google's stakes are different. They don't make money from selling phones, and they even abandoned their retail NexusOne store with very little pain. They don't make money from licensing software either, as far as I know. Google makes money from selling ads. And the more ads they sell, the happier they are. Apple is fighting for market share in cellphones; Google is fighting for market share in ad placement.

This asymmetry is very important. Google does not have to dominate the smartphone business; they just have to make sure that there's an environment in which the business of selling ads thrives. While Apple wants to dominate smartphones, Google undeniably dominates online ad sales--and they clearly see ad placement on mobile as a huge opportunity. Conversely, failure to dominate mobile ad sales would be disastrous. At best, it would limit their potential; at worst, if we're heading for the end of the "desktop/laptop era", it could seriously threaten their core business.

Making money selling mobile ads requires that Google keep the smartphone market open, plural, competitive. As long as there are multiple smartphones in the market, content developers will be driven towards open standards like HTML5. Developers will build richer and richer HTML content for the phones--and Google will thrive in its core business, placing ads on HTML pages. Google doesn't need to "win"; they just need to "not lose", to keep the game open, and to drive open technologies to the next level where they can compete successfully. In the long run, a closed system can only thrive if it's the only player in the game. If we've learned one thing from the growth of the Internet, it's that open standards that can be implemented by many vendors trumps closed systems, and enables the kind of competition that drives out monopolies.

Just as an athlete will inevitably perform better when he's relaxed and not worried about losing, Google's big advantage in the smartphone wars may well be precisely that they don't need to win. Googlers are justifiably proud that US Android sales have snuck ahead of iPhone sales. Of course, that's 50-odd phones available for all US carriers, versus two iPhone models available only from AT&T. And when the iPhone 4 comes out, Apple will certainly see a big burst of sales. But that's not what's really important to Google; all they need to do is keep the game open, for themselves, Palm/HP, RIM, and the other smartphone vendors--and to establish the kinds of standards that enable a competitive market to thrive.

There is a real threat to Apple, though; just because Google doesn't need to win smartphone dominance doesn't mean they wouldn't like to. And in the wake of their FroYo demos at I/O, that seems increasingly likely. Dan Lyons (Fake Steve Jobs) makes a lot of good points in his Newsweek blog:

  • Google's technology is way ahead of anything Apple is offering, or likely to offer. Streaming music from your desktop is only one example. Google, not Apple, is offering what customers want.

  • Apple's response to Google's claim that they are shipping more phones was "so what, we have more market share." Lyons says he's heard that before, it's the song of a company that's losing and in denial. I've heard it too. Lyons is right.

  • It's easy to think that Apple fell apart in the late 80s and early 90s because a clueless Coca Cola exec booted Jobs and took over. But the real story, if you're old enough to remember, is that Jobs mismanaged the company after a series of stellar technical triumphs. History appears to be repeating itself.

I am genuinely sad about this; Apple is a great innovative company. There's no reason they can't do everything Google is doing. Analyzing each players' strengths, Apple really understands user experience and design. They have a lock on that. Google really understands cloud computing and connectivity. However, it will probably be easier for Apple to get up to speed on the connectivity issues than for Google to get Apple's design sensibility. Nothing Google is adding to Android is fundamentally that difficult, and Apple has no shortage of engineering talent.

But--and this is important--Apple will not be able to take Google on in the areas of connectivity and cloud computing as long as they insist on a closed platform. Not because Google's FroYo features can't be implemented on a closed platform, but because it just wouldn't occur to you to do so. Furthermore, you can only go so far telling customers that you know what's best for them. I hate Flash almost as much as Steve Jobs, but you know what? If I were building a platform, supporting Flash would be a requirement. Flash is everywhere. Getting tied up in a pointless fight with Adobe is silly. Vic's daughter is right: she wants the toy that can run her favorite online games. That's going to be an Android phone, not an iPad or an iPhone. Apple is insisting on playing the game in a way that they can only lose.

Having said that, why is Apple so interested in HTML5? Why are they supporting it with almost as much energy as Google? I think Steve Jobs really understands that HTML5 is the "right thing" for the future of the web. Apple is not going to drop native applications. But Jobs has always had an uncanny sense of when things are done right.

Although Google doesn't need to "win" the battle with Apple, Apple's hysteria, along with its insistence on fighting the wrong battles, means that Google has a decent chance of winning. HTML5 may be Apple's last chance to change their ways, and make decisions that aren't dictated by their desire to control the platform. If they don't, they will lose, and that would be tragic, both for Apple and for users.

Older posts are this way If this message doesn't go away, click anywhere on the page to continue loading posts.
Could not load more posts
Maybe Soup is currently being updated? I'll try again automatically in a few seconds...
Just a second, loading more posts...
You've reached the end.

Don't be the product, buy the product!