breakfast first

Flickr, Yahoo, and Testing

Frank points to an interesting interview of Eric Costello on the development and evolution of Flickr. What caught my eye was this statement from Eric:

But we also have a very agile development process. We deploy code to the site maybe 10 times a day on a busy day. And we’re constantly adding new features, small and large, even though lately it’s been relatively small features, sadly.

I assume that most of the ten changes on a busy day are cosmetic, but still, that's very rapid, and it got me wondering about their development process. It turns out that Cal Hendersen of Flickr is giving seminars (in London) on their development process, which Tom Coates writes about here.

I'm most curious about Flickr's approach to testing. In my experience, changing a production site ten times a day can only work if you have very effective automated testing. This is even more crucial when you're dealing with a distributed app, where developers in one area can make a perfectly reasonable change which passes their unit tests and local functional tests, but has unexpected consequences in wider system integration. Without thorough, automated functional tests, it's only a matter of time before you'll get blindsided. If anyone has details on Flickr's strategy for testing, please share.

How Flickr's development process will mesh with Yahoo's, now that they've been acquired, will also be interesting to watch. From reading the above articles, my guess is that they'll probably blend well. My understanding is that Yahoo does not formally QA much of the software deployed on their site, but rather relies on the developers of each property (as the various components are known) to test the software themselves and deal with operational issues that arise. That's an interesting approach, but it can lead to wide variance in quality between properties. But since this already seems to be Flickr's modus operandi, and since the Flickr team seems to really care about quality of user experience, I'm sure it will work out well.

Baby digs the K&R

For the littest nerd in your life.

(via DaddyTypes)

For the Ava fans in the audience

Friends and family have been asking for more, and more regular, updates on Ava, so I set up a blog for her. It's password protected to keep undesirable elements out. To get in, enter the name of our oldest cat as the user, and the name of the city where Miss J and I first lived after we got married. If you have any trouble, send me an email.

PDF and PS

Every time a PostScript file lands on my desktop, I'm amazed that Adobe hasn't chosen to support the format in Acrobat Reader. Certainly, early on in the life of PDF, Adobe had incentive not to support PostScript in Acrobat -- many free apps could write the PS format, and Adobe's strategy was to make money on writing the files, not reading them, so supporting PostScript viewing in Acrobat would have cut into PDF's base.

But now, 12 years after PDF burst on the scene, the battle has been won. PDF has gone far beyond its page layout roots and has spawned a whole family of products for Adobe, while PostScript has generally been relegated to a few dusty corners of academia where TeX still reigns. (I'm talking mainly about text documents here.) I was curious to see how severely PDF trumped PS in the popular vote, so I checked with Google on a few random terms:

word	Number of PDF docs	Number of PS docs
dog	1,930,000	18,100
topology	1,080,000	160,000
election	4,330,000	15,300
electron	2,580,000	95,100

As you probably guessed, it's a landslide.

So what's a poor user to do when he wants to read a PS file? Assuming he doesn't have a copy of Illustrator handy, he's going to need to find and install GhostScript and GhostView. GhostScript is a very powerful and flexible application (it has no problem handling PS, EPS, and PDF files, among others), but the viewer lacks the polish of a commercial offering. From my perscpective, forcing consumers to use a third-party app such as GhostScript tarnishes Adobe's reputation -- it makes it appear as though Adobe has turned its back on its firstborn.

Obviously it would require effort for Adobe to add PS support in Acrobat Reader, even though all of the hard work is already done. There might even be a bit of customer support cost to go along with it. But in the long run, I imagine those costs would be a drop in the bucket compared to Acrobat's revenue and it would generate goodwill among the academic community.

So come on Adobe, dance with the one who brung ya.

A Tale of Two Startups

Stealth(ish)-mode startup Zillow is getting a lot of buzz in Seattle these days. This real estate startup, founded by Richard Barton, the founder of Expedia, promises a "little revolution" and many seem to think he can deliver. (Not everyone is convinced, though; David Chase makes an interesting comparision between cracking the real estate establishment and cracking healthcare -- and remember the promise that Healtheon once held?)

But I've heard very little about another Seattle real-estate startup, Redfin, other than from satisfied users. Their site ties together MLS listings, FSOBs, satellite imagery, previous home sales, tax records databases smart searching and a slick interface to create a one-stop shop for buyers. I found it invaluable in my own search, as have many of my friends. So why is this company getting no press despite having street cred?

One indicator might be this: Redfin feels more like a service than a full product. Their technology is wonderful, but I don't know if they can parlay it into a viable company. I can easily see established brokerages (first and foremost, Windermere) benefitting from Redfin's technology -- perhaps their future lies in licensing, rather than end users. Will they play to their strengths?

Just curious

In thinking more about the Costco discussion earlier this week, I wonder if anyone can tell me the average length of time non-institutional investors hold individual stock? This data must be known. I'd be curious to see how behavior has changed over the years.

Tech Exodus

Good article in John Cook's VC Notebook column today on the migration of high-tech workers from Boston and the Bay Area to Seattle. He focuses heavily on the executive angle in the article, but I think the more serious and telling story is what's going on in the lower ranks.

Right now there is a crisis rapidly emerging in the Bay Area. Largely driven by the most expensive housing market in the nation, many established technology rank-and-filers are finding California simply too difficult to bear, despite its many virtues. Insane housing prices, high taxes, overall high cost of living, and terrible public schools are pushing many in our industry to look for good alternatives, and Seattle is at the top of their list.

If innovation and start-ups are the essence of Silicon Valley, then these workers are quite arguably the Valley's most precious resource. Start-ups in particular thrive on the energy and experience of people who have been around block enough times to know where the pitfalls lie and how to avoid them. These workers' wisdom allows companies, at a very real and day-to-day level, to rapidly get ideas to market by tempering just enough the youthful idealism and exuberance at the core of every start-up with the prudence of the possible. As much as VCs and executives get the press, seasoned workers are the lifeblood of the tech economy.

I am among those who have, heartbrokenly, left the Valley upon realizing that they cannot afford a regular life there. Reactions to my departure were mixed: some -- the wealthy, who had made their fortunes from start-ups -- seemed perplexed; others -- the believers -- suggested I suck it up, make it work, hold on for the payout. But many more understood where I was coming from, and where I was going to.

I found I was hardly alone in my decision. While I was finalizing my decision, I learned that three other colleagues at Danger had independently made the decision to move to Seattle as well, all within months of each other. (Three of us work remotely; the fourth moved on.) My wife and I were simply following friends who had moved a here a year before us. Now, in a small circle of friends, two more couples will be moving up within months.

California's loss will be Seattle's gain as much of the muscle of innovation migrates north. Capital, which is fluid, will follow the creative workers, who are not. No doubt the recruited executives will follow, too, as Seattle's start-up community grows in stature and as more money flows in (Washington already ranks third in the nation in terms of the number of deals; fifth in terms of dollars). And, ready to help out, a fresh crop of prime raw talent is produced yearly from the CS department of UW. The breeding ground of innovation is more fertile than ever here.

If trends continue, where will California be a decade? It's hard to say. The money will always be there, as will the young and eager. But by then, the majority of those with critical knowledge may have moved away. Perhaps the continued rise of remoting, coupled with the relatively short distance between Seattle and the Bay Area, will make this less of a concern. Regardless I anticipate more green in the Emerald City's future.

Miss V, please take note

With a blog name like "Breakfast First", I feel obligated to report on innovation as it relates to the most important meal of the day. So behold this wonderful contraption: a toaster, capable of handling bagels, that can also cook an egg and breakfast meat for you. Not a bad design, either, and it looks like cleaning -- normally the big "gotcha!" with food prep products -- won't be too difficult. The site says it's available in September; expect a full review here shortly thereafter.

(via linkmonger)