The things we have been reading lately

Betrayed by an accelerometer

How to draw a circle

9 Metrics for making good decisions for your startup: From the Kissmetrics blog. This is one of the best blogs on marketing, startups, good web design, and more. It’s worth frequent visits.

A walking man, done in CSS3

Some “leaks” from the forthcoming Steve Jobs biography

How to sort file names: Raymond Chen proves that computer programmers are not normal human beings.

More, more, more: A nice riff on our constant push to do more and be more.

Our latest installment

First there was the Joel test, now there is the Rands test: Go read this now for the sake of your company!

An animated history of the iPhone’s evolution And a lot of other stuff too.

Dennis Ritchie

Paul Rand and Steve Jobs making the logo for NeXT: Look for the link to an interview where Jobs talks about Paul Rand.

A beautiful, interactive exploration of understanding and solving a mathematical problem

Buying lunch in lower Manhattan is not cheap. If you don’t brown-bag, you’re pretty much in for nine bucks, and more if you want to sit down. A meatball sub is $8.50. The Hanover Square Deli does a respectable dduk mandu guk for about $9.50. A meager-looking, but tasty, turkey and guacamole sandwich from British import Pret A Manger will set you back $6.75 (plus tax). Or, if you really want to eke out some savings, you can get a cup of soup and a hunk of bread for about $5.

At Fog Creek, we have our lunch catered every day.  The cost of this to the company is $15.75 per person per day. Not counting drinks and snacks, this amounts to an extra $4,000 per year per employee. So why wouldn’t we just pay people $4,000 a year more? If they want to be frugal, they can buy that $5 soup and pocket the difference.

Well, first off, since lunch is a catered on-site meeting, the cost of that lunch is 100% deductible as a business expense. If you worked here, it’d cost the company $16 to give you $10 to go buy your own mediocre udon noodles.  (Don’t take any taxability or corporate expense stuff here too seriously; I’m not an accountant or lawyer.) Second, Fog Creek’s free lunch is much more because of how it fits with the rest of our workspace and culture.

Free lunch is nothing new. Google and other big tech companies have been giving their employees breakfast and lunch (and sometimes dinner). Free lunch has been around for more than a decade because it works for attracting and retaining top talent. (Or, rather, people think it works.)

Now, the food at Google, and the people who make it, are awesome. The food at Fog Creek is good, but nowhere approaching what Google does. We don’t have Sam, the sushi chef, doing hand-rolls to order. We don’t do miso black cod, lamb shanks, or osso buco. We don’t have a raw vegan station with selections so delicious they attract the most dedicated carnivores.

Here’s what we have at Fog Creek instead: no meetings.

For us, lunch is our only recurring meeting. The only standing interruption in the day/week of a developer here. Everything else is ad hoc or temporary. (One exception is our quarterly meeting to go over financials and to grill the founders with questions.) The default at Fog Creek is no recurring meetings. Once you’ve established that, recurring meetings become the exception, rather than the rule, and tend to wither naturally as their usefulness degrades over time. For example, the FogBugz team is currently doing a stand-up meeting for fifteen minutes every day right before lunch, but this won’t last. The Trello team was doing them before and shortly after launch, but they’ve subsided.

The Kiln team lead assures me that their weekly stand-up is a real recurring meeting, but at some point they’re going to lose interest and go back to coding. It’s what they do. Meetings have a network effect. They need other meetings to legitimize them. If you’re constantly scheduled with meetings, you don’t mind being interrupted one more time. Or, rather, you don’t have the energy to protest. When you proceed from the assumption of no meetings, you have to expend effort to keep a meeting going.

It’s not all a playground, of course. There are downsides. Things, sometimes important things, don’t get communicated to the right people at the right time. More often than not, the worst result is hurt feelings or slight confusion. Sometimes it’s more than that. But we wouldn’t give it up.

For someone who’s used to a standard work environment, it seems silly, cult-like, possibly even daunting, to be “forced” to break bread with your colleagues every day. It seemed odd to me before I came here, but by the end of the first week, nothing seemed more natural. When most of your “socialization” with your colleagues takes the form of mandatory, recurring meetings over a conference table, it’s natural to not want to see them again over the lunch table. People have been making decisions about your time all day; at lunch, you need some time to yourself.

But when most of your time is spent working in a private office, taking breaks according to your natural attention span, having short chats with one or two colleagues, it’s a pleasant prospect to surface for some pleasant conversation about StarCraft or football with nice, intelligent people. And maybe you’ll hash out that new feature, too.

One can make the argument for free lunch based strictly on developer productivity. A free lunch could give you a hundred hours per year from your best people, time saved in driving, waiting in line, etc. But, consider the depressing ease with which such a gain can be wiped out by a few recurring meetings.

So, for our final post in this series, we do have a number: $15.75 per day. $4,000 per employee per year. It’s a lot of money. But without the rest of the company culture, it would be sort of meaningless. A sense of entitlement grows rapidly around any perk you offer, and lunch is no different.

A lot of this series has been based on getting hard data out there so that developers, our main audience, have an easier time talking to management about some of the things that’ll make them more happy, healthy, and productive. For this post, it’s a bit more difficult. You might get your boss to spend $16 a day, but changing the culture of meetings in any workplace is nearly impossible. (During my tenure, Google tried no-meeting Thursdays and formal meeting-reduction task forces, reminiscent of Brazil’s Ministry of Debureaucratization, to no avail.) It would require rebuilding the company culture from bedrock.

Our bedrock is the idea that, once we’ve hired good people, it’s the effort we make to direct their intention, rather than their attention, that creates value. It’s not just our lunch benefit that springs from that, but nearly every other thing we do.

For Fog Creek, our founding principles, and the pains we take to stick by them, are the price of developer happiness, and that can’t be measured in dollars.

Here is our latest installment.

Why code reviews beat testing: We knew we put code reviews in Kiln for a reason.

Geoffrey Moore talks about core vs. context in your business: This is fantastic talk from the 2009 Business of Software conference. If nothing else, it will make you think hard about your business.

The king of code comments?

The recent Node.js controversy

The always fascinating Richard Feynman: Three videos that use Feynman’s voice to talk about beauty, prizes, and curiosity.

Algorithm is not a four-letter word: An awesome overview of maze generation algorithms.

How to generate music with one line of code

Here’s our latest installment of things we’ve been reading.

Harvey Mudd gets more women: The President of Harvey Mudd college has managed to triple the number of female comp sci majors since 2006.

A proposal to fix callback hell in Node.js

Inbound Hiring: I think most startups do this, but here’s a short analysis on early hires from Gabriel Weinberg.

Why blog?: Also from Gabriel Weinberg. What he’s learned over the last year of committing to his blog.

Try 16 different languages in your browser

Our oldest product

Let me tell you a little bit about http://www.fogcreek.com/:

• It’s old; It’s hosted by IIS6 alongside a horde of other websites.
• It’s the first place people go when deciding whether to use FogBugz or Kiln.  It’s pretty important to us.
• Recently, we started to treat the website as a product. It has received substantial attention, primarily in the form of A/B tests.

We’re building some fancy new web servers for fogcreek.com, but in the meantime we’ve been accepted into the Google Page Speed Service beta. It seemed reasonable that it might provide us with some useful tools, and in the short term it might give us a workaround for some of our IIS6 woes.

There were four things that we hoped Page Speed Service would take care of:

1. GZIP all of fogcreek.com’s static and dynamic content.
2. Set far-future expires headers on our static content.
3. Distribute our static content via a CDN.
4. Fix or remove the broken ETags that IIS 6 Is sending out.

Enter Google Page Speed Service

For this test, we brought Google Page Speed Service to life at w.fogcreek.com. This is a totally unused alias of www.fogcreek.com, so we were able to test PSS on an isolated mirror of our production environment.

After a small amount of setup, page speed service was live!  (Setup basically consisted of adding a Google verification .txt record and a CNAME record mapping w.fogcreek.com to ghs.google.com.)

The comparison tool that Page Speed Service provides was inconsistent, but it looks like there might be a lot to love.

• Its page load times differ from those reported by GTmetrix and local testing
• The measurements change significantly over the course of running it multiple times

But the results look promising.  Here’s a report on Fog Creek’s home page load time.  On the left is the load time without Google Page Speed and on the right with it.  Google PSS gave us a 10% difference in initial page load time, and a 25% difference in repeat page load time. We got that for very little effort.  Nice!

Here are GTmetrix reports on www.fogcreek.com from 9/23/2011.  First, the results without the Page Speed Service:

• Report for fogcreek.com
• Report for fogcreek.com/fogbugz
• Report for fogcreek.com/kiln

And the same tests on w.fogcreek.com, with Page Speed Service enabled:

• Report for w.fogcreek.com
• Report for w.fogcreek.com/fogbugz
• Report for w.fogcreek.com/kiln

Overall scoring improvements:

The Fog Creek landing page:

Google PS +1%, YSlow +10%, page weight reduced from 466kb to 365kb

The FogBugz landing page:

Google PS +1%, YSlow +12%, page weight reduced from 608kb to 379kb

The Kiln landing page:

Google PS +0%, YSlow +10%, page weight reduced from 520kb to 411kb

Interestingly, GTmetrix shows page load times (as measured by the onload event) that are consistently slower for w.fogcreek.com than http://www.fogcreek.com/, usually on the order of about 0.7 seconds.

GTmetrix’s tests don’t seem terribly consistent with webpagetest.org’s (Google’s promoted testing service), which show consistently faster page loads than GTmetrix. But, it’s hard to draw conclusions since incremental rendering of the page is an important factor which isn’t represented by the load time.

For those of you who wish to dig deeper, take a look at the waterfall diagrams of a given page rendered with or without PSS. You’ll notice that the PSS timelines have their content proxied across 4 of Google’s servers, the overall number of requests has decreased, and the waterfall has rearranged significantly.

For example:

www.fogcreek.com/fogbugz, Without Google Page Speed Service:

w.fogcreek.com/fogbugz, With Google Page Speed Service:

So how did PSS handle the goals we wanted to meet?

• PSS has compressed (GZIP) the content that it sends down
• PSS has far-futured the vast majority (but not all) of our static content
• PSS has properly configured our ETags
• PSS is distributing our content via its CDN

Pitfalls

1 - PSS’s lossless image compression is lossy.  Our website uses a tiled, textured background image.  Page Speed Service attempted to optimize it, and in the process altered the background images.  For example, this image:

Became this one:

I’ve put together a test case at w.fogcreek.com/ImageTestCaseForPSS/

As part of the lossy conversion, PSS converted the image from a png into a jpg.  I altered the source image files to be jpgs instead of pngs, and then PSS was able to optimize and bundle them without any difficulty.   I’d also note that the image optimization step (and all of their other optimizations) can be disabled in order to work around problems like this.

2 - PSS supports blacklisting URLS, but not blacklisting HTTPS urls.  I knew from the start that PSS didn’t support HTTPS, and therefore our trial signup forms, e.g., https://www.fogcreek.com/kiln/try/, would have to be exempt from PSS’s optimizations.

PSS supports blacklisting, which allows you to exempt a specific page from being hooked by PSS’s proxy.  Instead it will go directly to your source server.  Unfortunately you can’t configure HTTPS pages in this manner.  If you have a website with HTTPS content, you’ll have to move all of your HTTPS content to a separate subdomain, e.g. “secure.fogcreek.com”, in order to exempt it from PSS.

This was an inconvenient change to make, although not hard. Still, PSS’s support of blacklisting was put to good use.  We use a heavily-modified version of the FairlyCertain A/B testing framework on our fogcreek.com properties.  Server-side A/B test frameworks and CDN’s don’t really play well together, but PSS’s blacklisting allows us to blacklist any pages which contain active A/B tests.

3 - For the duration of the beta, once you bring PSS live on a given subdomain (e.g., w.fogcreek.com), you cannot reconfigure PSS to work with a different subdomain (e.g., www.fogcreek.com) without an additional beta acceptance.  I’ve reapplied to include www.fogcreek.com in the PSS beta, but have yet to be approved.  For this (and only this) reason, we’re blocked on moving forward with using Google PSS in production.  If a Googler is reading this and would like to whitelist www.fogcreek.com for the beta, we’d be much obliged.

 Impressions

Pitfalls and onload-event oddities aside, I’m very impressed with Google Page Speed Service.  I’m happy to see that they’re abstracting away the ever-present need to minify, compress, and bundle your static website content, as well as optimize your images.  In our specific case we don’t really care about those  features – we already optimize our images and bundle our scripts, but it is a great feature for most sites.  Our case is probably more edge than common—we’re using PSS primarily to work around IIS6, and it appears to succeed at that.  PSS’s CDN seems promising, but I don’t yet have any data on how it compares to, say, Amazon’s Cloudfront service.  Setting up PSS on our site was an absolute breeze, however, and  I think it may turn out to be a great tool for web developers and site administrators.

The choice of setting up the Page Speed Service on w.fogcreek.com instead of our production subdomain, retrospectively, was a useful mistake.  I appreciate the ease of using tools like GTmetrix to compare the two subdomains directly. I’ve also enjoyed keeping PSS’s configuration changes completely separate from our production environment as well as the ability to share the PSS changes with other people without having to manually configure their browser to proxy traffic through PSS.

Here’s some interesting stuff we’ve been reading lately:

How to keep your mojo when you’re doing SEO day after day: Lisa Barone gives some advice on how to keep your SEO efforts fresh, and how to stay interested when you start to flag. “The result of all of us leading more social media-focused lives is that we’ve tied our personal worth to how we do on these various social media channels.”

Designing the Ubuntu monospace font: ”Making a font takes a really long time…”

Twitter project does for dist. real-time computation, what MapReduce does for batch processing.

Why do Windows functions begin with a pointless instruction? 

The Smashing Magazine Web Design Challenge: A challenge intended to help designers become better at justifying their design decisions.

Can health data predict risk of genocide?: “The results seem to demonstrate that the victim population had been marginalized and denied access to nutrition and health care even before the genocide took place.”

Picking a product name is all agony and no ecstasy. It’s also a giant time-slurping vortex. And in the end, it kind of doesn’t matter.

A product, or a company name, only really needs to achieve one thing: not remind us of some unpleasant bodily function, or the results of a wild and drunken night. But even this seemingly common-sense advice is often and successfully broken. Wii anyone?  No thanks, I went this morning.

Yes the name matters if it’s truly totally weirdly awful, but outside of that it just doesn’t.

Recall that “iPad” was the subject of all kinds of snark and scorn in the first couple of weeks after launch. Who makes those jokes now? No one.

Amazon? I remember lots of people making fun of that name and insisting there would be endless confusion, but it worked out fine, and at least it wasn’t Cadabra, one of the early names Bezos considered, and something sure to upset the Steve Miller Band.

At least a dozen people I know wondered why you would name an e-reader Kindle, something that reminded them of fire. Fire. Books. Fire + Books, bad. Anyone think that anymore? Nope. Everyone just buys them by the millions.

If Google wasn’t such a familiar and ubiquitous name, most people would reject it as a company name on the sensible grounds that it is sillier than a bald cat. But we’re all used to it, and we don’t care. And that’s the point.

And of course there is no shortage of bad-seeming names on really successful products. Dare I mention FogBugz in this context?

We just recently went through the naming agony since we just launched Trello, a tool for project and task and process management, and…well, go check it for yourself. 27,000 users have signed up in the last 24 hours, and they’ve been saying really nice things on the Twitters.

Trello was code named Trellis when it was in development. It got that name because one of Fog Creek’s co-founders, Michael, suggested it as a code name in an early meeting. It was fine. It stuck.

Eventually we decided that we needed An Actual Product Name. Which, of course, we did, but we spent way more time on this than we needed to. Two of the team members took some time to come up with some names. Actually, a couple of hundred names. For one reason or another, these were all rejected or shelved, mainly because all possible domains have been plucked up by an automatic domain-registering spaceship.

Feeling morose over this, they then hired a professional naming person. He came up with about 200 names, some of them very good, and one of them “lasagna”.

Not knowing about the previous naming efforts, I jumped in one day with about 125 name ideas. Then several of us sat down together and reviewed our name ideas, and came up with about 200 more. We tried everything: animal names, plays on various aspects of what Trello does (board, card, list, task), Japanese words, and every combination in between. We threw them all at the wall, from the practical to the nutty. Kardboard, Hippolist, 5 Camels, Listly, Idealist—all were suggestions at one point or another.

Late in the process we thought, maybe we could just use the code name, Trellis. Why did it have to change? But, we couldn’t buy trellis.com. They weren’t selling. We tried to buy trell.is, but it was more than we wanted to spend. We really thought it was important to get a .com domain anyway, so we weren’t thrilled with this option.

The wheels continued to spin and every couple of weeks the product name came up and we would lurch off on another round of fumbling around for a new name.

Work continued on the product code named Trellis and for a couple of months we didn’t have much time to think about the name until the mid-September launch forced the issue: nothing like a deadline to provide some clarity.

So Joel organized a company-wide brain storming session and we got out our markers and some giant pieces of paper, tacked those up on the wall, and came up with about another 150 names. Some really good names came out of that session. It was raucous, and it was fun.

We argued, we lobbied, we pointed out when a possible name rhymed with wee, and then we voted. Of our choices, about none had available .com domains. The whole thing was getting depressing. And we were out of time.

Trellis was still one of the contenders, so, after all of these sessions, and all of this brain-storming, and the many hundreds of suggestions, Michael just decided to start playing with variations on Trellis. He came up with a number of them, and then looked for domains he could buy. Trello.com happened to be one of the domains he could get for a reasonable amount of money and we decided it sounded good. We had made a grand wandering journey all the way from Trellis to Trello. That’s like leaving New Jersey and getting to New York by way of Kinshasa. But here we are.

After all of that it’s clear that the name just isn’t that important.

Gold Bond Medicated Powder. What’s a Gold Bond, actually? And Medicated Powder isn’t a name, it’s a thing. Might as well call your product Lined Paper With Three Holes. But Gold Bond Medicated Powder works, doesn’t it? No one ever thinks about it, they just get some if they have itchy feet.

We ended up with a name very similar to the one we started with, one that all of us like just fine, and one which is just going to be what people call our new, supremely free, project management app. I think everyone would have gotten used to Hippolist too, but it doesn’t matter, because it’s called Trello now.

In the end we should have just had Michael do the whole name picking in thirty minutes and been done with it.

Last time, we took a look at the exact cost of the contents of our private dev offices. For this post, we promised to take a look at the price of the offices themselves.

Since the beginning, Fog Creek’s promise has always been that every developer gets a private office with a door that closes. Don’t want a private office? You get one anyway. If you want camaraderie, you can walk down the hall, put your witticisms on company chat, or store them all up and let fly at lunch.

In the previous post, the underlying message was, “Give your people the workstations they need to be comfortable and healthy.” The message in this post is not, “Give your people private offices.” We’re satisfied enough with the benefits that we’re going to continue it, but we’re not here to proselytize that policy.

The point this time is this: your office space and how you design it is an expression of your priorities as a company. It speaks to everyone who comes through the space, every day. Do not be shy about spending money on it.

Let’s get the number out of the way. We spend a bit over 6% of revenue on office space. Compared to other companies we’ve surveyed confidentially, that’s on the high side, but not by much. And we’re growing fast these days because of the stunning success of Kiln, so we won’t be above 6% for long. We’re based in Manhattan and, as you might imagine, office space is very expensive. But that’s what the company was founded on: a good job for awesome coders in New York City. Without making light of the achievements of our neighbors, we didn’t want to be the best place for devs to work in Hackensack. So it’s time to pony up.

Of course, a lot of companies who are competing for the same talent are actually venture-funded startups. They have different needs, revenue/reward models, and external expectations than we do. I’m going to go out on a limb and say Fred Wilson might take issue if you buy the $10,000 espresso machine before showing a profit. We’ve been profitable since inception and so can work a little differently.

Also, the price tag on our office space is not only because we have private offices. We need a lunch room and kitchen to accommodate everyone having lunch at the same time, away from their desks. Our summer head count can grow by ten because of our summer internship program. Okay, so we don’t actually need the saltwater tropical fish tank or the marble shower or the library.  But we’re also expressing our company culture by how we structure our office, and if that can keep us happy and motivated, plus attract more smart people who share our values, that’s very much worth the extra money.

Here’s a few good snapshots of our office, if you want to see more.

Next, the cost of lunch, and its primary benefit: no standing meetings.

Hey, we’re going to try an experiment for a while. We read tech and software stuff here all day. In fact, I sometimes think we don’t read anything else unless it’s the lolcats blog. We’ve decided each Friday to put a short list of links up to some of the interesting stuff that we’ve been reading, and that we think you, the legions of Fog Creek blog readers, might like to read as well.

So without any additional ceremony, here we go:

How Hubspot plans for innovation: A short primer on how Hubspot tries to make an environment that leads to innovative new products. We’ve done something similar at Fog Creek which has led to Kiln, and to another product coming soon (in just a couple of weeks, actually).

37 signals on A/B testing: We’ve become big fans of A/B testing here. BIG. A/B testing will surprise you. You should be doing it a lot. Read about what 37signals learned.

Peldi on startups: This is from last year’s Business of Software conference, but a video and transcript just went up today. Peldi sounds like the name of a soccer superstar, but he’s, in fact, a very entertaining startup founder.

PyPy is faster than C: The title speaks for itself!

Tagged pointers: How Lion implements tagged pointers for better memory usage.