Some Stuff You Learn Launching a Service

Okay, so I’ve sent out the alpha invitations and I’m sure glad I’ve only sent it out to people whose support I appreciate because it’s been a very very rocky start. I hope to iron out a bunch of uptime issues and build out some core functionality (I plan on using it myself soon if it doesn’t prove to be so craptastic that I wouldn’t want to risk my reputation with it). You just have to be there to appreciate the futility of spending months on development to be greeted by a 404 or 500 error.

Everything that Can Go Wrong Will Go Wrong at Launch

Whoever said Murphy’s law was dead? That bastard is living large and has a special place in my heart (reserved for people like Hitler). You can develop websites all day and throw yourself imaginary high fives all around until you hit that button to upload your latest and greatest earth-shaking code only to watch your dreams crash in a flaming death.

The web is held together with a mesh of inter-related technologies and web services are no different. The programs we write stand on the shoulders of giants and sometimes they shake us off so we can plummet straight to the ground and splatter our brains on the concrete.

This makes perfect sense since you tend to test your site in very well-defined ways because you’re designing it. You know how it can and cannot be used so naturally you don’t cover everything. Users will quickly help you find faults with your service. It’s been a rocky ride but I’ve got some tricks that I hope will remedy some of the immediate concerns.

After All is Said and Done, It’s a Microsoft World

Even without touching Javascript, CSS is enough to trip you up when designing a site. It’s not so much the fact of learning the code but all the problems you’ll deal with when optimizing for IE. You only have two choices. Let standards be damned and develop for IE to start only to watch things crash and burn when viewed on more sane browsers or developing for Firefox and applying hacks or subtracting stuff as you pull your hair out over IE rendering bugs. You really need to design a website to appreciate the true crappiness of IE. It’s like a software reincarnation of Steve Ballmer.

Oh and watch the colors you use because monitors display colors differently. Yeah, I’m seeing purple when I should be seeing blue on this screen in fact. You need to stick to safe color schemes and also check on a couple different monitors. I don’t know of any way around it.

Feedback or Lack Thereof is Good

Feedback or no feedback is a good thing. It’s easy to picture yourself on a yacht cruising the ocean surrounded by hot babes pouring champagne as your site adds yet another million users when in fact you hear virtual crickets chirping. It’s the hard, cold reality and a sign that you’re not doing something right. Still, it’s so much better than writing line after endless line of code in an eternal feedback loop between you and your disconnected computer. Having something live and at a freely accessible location is fun in and of itself.


The Journey has Just Started (I Hope)

Well, I’m having some fun so far. Just wait until I get a decent number of users and go into the hell that is customer support when they use my alpha service! Stay tuned.

Categories: web

Twitter is…

The beauty of twitter is that it can be anything to anyone but these days it’s starting to feel like a massive link promotion tool. Of course, I’m not saying that my twitter friends are spammers.

I think it basically boils down to several factors. Some people don’t want to bore you with the mundane details of their lives so they offer useful links to you instead. Either that or people are running scripts that post to twitter every time they write something on their blog.

I guess we’ll have to see how it pans out. For one I personally find myself enjoying random twitters via Twittervision because well it’s random and flies all over the globe. It’s like peeking into the matrix that we call the internet.

There’s obviously a lot of golden opportunities for the folks at Twitter but I think all of us agree that none of us know what would be the best way to capitalize on it.

Twitter is…

The beauty of twitter is that it can be anything to anyone but these days it’s starting to feel like a massive link promotion tool. Of course, I’m not saying that my twitter friends are spammers.

I think it basically boils down to several factors. Some people don’t want to bore you with the mundane details of their lives so they offer useful links to you instead. Either that or people are running scripts that post to twitter every time they write something on their blog.

I guess we’ll have to see how it pans out. For one I personally find myself enjoying random twitters via Twittervision because well it’s random and flies all over the globe. It’s like peeking into the matrix that we call the internet.

There’s obviously a lot of golden opportunities for the folks at Twitter but I think all of us agree that none of us know what would be the best way to capitalize on it.

Categories: web

Keeping it Minty Fresh

Despite the lack of updates I’ve been (ironically) thinking about keeping things fresh, for web services that is. This is something that affects services that cater to users and rely on the community to support it.

It’s hard to imagine a world where MySpace, Facebook, Flickr, or whatever doesn’t dominate but back in the day Yahoo and GeoCities was king. Once you have a trillion registered users it really doesn’t matter what you do because you’re still going to generate traffic and still going to generate money even if you don’t add any functions or refresh the design.

Still, there’s a definite point where things go “stale”. It has nothing to do with how usable the site is or if the design is kicking. It’s more of a vibe you get. We see so many services launch with flashy web 2.0-inspired designs that don’t pack any thunder and vanish as quick as they come. We also see things like Twitter that wouldn’t actually look out of place in the 90s but making waves in the present.

I think it happens when the people who should be the most passionate about your product start going through the motions. The point where they don’t care what you say or do. Unfortunately, once you reach that point it’s like a marriage on the rocks. You can separate or go to counseling but you’re basically biding your time before the inevitable.

It can be a mix of things but for me I start noticing the disconnect when I no longer care about new features or using the service just wears me down. The usual progression is from anger to frustration to apathy. Like they say about love/hate relationships, sometimes love and hate are two sides of the same coin. Apathy always kills.

This sort of ties in to my previous post about features. When I find a new service or piece of software I usually get excited when I see an active developer continually improving the product. If they’re open to suggestions from the community it makes it all the better. Still, as it reaches maturity both sides start getting cautious about “fixing what isn’t broken.” Too much of a change can unleash a massive flamefest (especially if it means removing something). Yet the funny thing is after a good while, people simply stop caring that much. It gets to be “one of those apps” that are out there. They don’t drop by too often or have simply stopped coming. Nothing makes me lose interest in something more than having to wade through a million abandoned account names just to set my login name. Again, there’s a vibe.

To come to think about it I can’t think of one “social” website that’s held my interest for more than a couple years. There’s always so much talk about “makeovers” and “reinventing yourself” in real life but what’s the equivalent for the web? Something like Netscape where you re-appropriate a domain and give it a whole new purpose to exist? Or maybe follow the GYM way of buying little startups so at least your portfolio stays fresh.

PS Yes, I’ll get back to your comments sometime soon hopefully and yes I appreciate them

Categories: web

Coda: The All-in-one Web Development Application for the Mac

Web development forces you to be a jack of all trades by its very nature. You have to at least have a rudimentary knowledge of html & CSS, how to use a database like MySQL, how to use a unix system and maybe even some design. This is all on top of whatever language it is you develop the main application with whether it is PHP, Java, ruby or perl. The typical workflow is just as chaotic.

At any given moment you’re switching between your code editor (which makes or tries to make life easier on your fingers with code completion and syntax highlighting etc.), a terminal application for connecting to your server and even manipulating local files (if you’re on a unix system), transferring files via FTP, some kind of a html/css design app to get the looks right, a browser to see the changes, looking up lots of documentation for the coding stuff, and maybe even a database admin app.

Naturally, many a web developer has wished for “an integrated solution” where they can just use one application to do it all rather than swimming through a sea of windows. This is clearly the territory of IDEs and a holy grail of sorts.

coda.png

Coda has set out to do just that for the mac. I must say that I was intrigued when I first heard about the product. I already use a mix of tools including iTerm, TextMate, CSSEdit, and Parallels for Windows IE testing (to make sure uncle Bill doesn’t throw a monkey wrench in my plans) along with Firefox/Safari (this sounds like a post waiting to happen). Sure it gets hectic sometimes but it’s hardly a chore.

Here’s where Coda’s falls apart: You’re never doing everything at once. Whenever I’m doing development I do it in a variety of modes. Sometimes I’m just straight coding and other times I’m tweaking the design (or lamenting the lack thereof) with lots of reading in between (you tend to look things up lots when you develop).

The truth is humans are only good at multi-tasking when it comes to goofing off. Surfing the net, chatting and listening to music at the same time is a walk in the park because none of them require focused concentration. So if you want to be productive as a developer you want to focus on the task at hand and avoid intentionally switching contexts.

Like the multi-tasking human equivalent, when you try to do everything at once you run the danger of not doing anything well. This is where I see the greatest danger for the future of Coda.

Take any two tasks in web development, like coding the app and tweaking the CSS or manipulating files in the terminal. These are very different activities. Each activity has its own context. This is why having several different specialized apps open a good thing because each of them are tailored to specific needs. Cramming all this functionality into a single interface is a very tricky enterprise. Making things easy to find and useful while fighting clutter is a difficult task for any application with a decent feature set. Coda, by its very ambition, seems destined to fight an uphill battle against bloat and clutter.

Another great shortcoming I see with Coda is lack of any integration with source code management like subversion. Any developer that wants to hang on to their sanity is going to use version control to ensure that they can rollback releases or collaborate with other team members better. Sure, you could just do it from the terminal bundled with Coda but it pales in comparison to TextMate where you can edit your commit message within the app, not to mention all the other common tasks.

I will admit that Coda is an appealing proposition coming from one of the best indie mac developers (making it all the more enticing) and that it’s a phenomenal product for a first release. Still the $79 price tag just seems wrong. Sure, buying TextMate (around $40), CSSEdit (around $30), and Transmit FTP ($30) will come out to a cool $100 but each of these specialized apps do one thing and do it very very well. However, if you do any semi-serious web development, the chances are you own several of these apps and even their competitors.

Is Coda enticing enough to relegate these superb apps to the back seat? Will it revolutionize your workflow so much that it seems like a bargain? I’m not sure it will and many people wont bother finding out. Every developer cherishes their tool of choice and hard to switch someone once they’ve “found their app”. I think that for a start, if Coda was in the $20-40 for an introductory period, some developers might say “why not” and add it to their toolbox while the offer lasts but otherwise happily code in their current environment blissfully unaware of Coda.

I do think that Coda holds a lot of promise for the future. In the immediate future I’m excited by this release because it’s sure to spice up the competition.

Announcing Coda

Did dodgeball Even Have a Chance?

The founding developers of dodgeball.com have left google for good. I guess free food and a nice banner over your cubicle is not enticing enough to keep the entrepreneurial types satisfied. Twitter and dodgeball look like a tale of two cities on the surface but lets be realistic. Dodgeball was on the scene a long, long time and had plenty of opportunity to gain traction. Was dodgeball the twitter that could have been? I don’t think so.

Basically, dodgeball is a location-based social networking site for mobiles. You register, register your friends, and then “hook up” with others in your extended network. I’ll admit that the idea is enticing but it’s also why it didn’t succeed.

Let’s recap:

  1. It’s geography-based
  2. You’re not going to hook up with a friend of a friend in LA if you’re based in Brooklyn.
  3. It’s tied to a platform
  4. You basically hook up with someone in your area via a text message. Although there’s obviously a web interface, it’s main function is as a mobile application.
  5. It’s only for hooking up
  6. The main focus is for using it to “hook up” with friends. Now how often is that going to happen? What’s stopping you from flat out spamming 10 friends directly via SMS if you’re about town and want to grab drinks?

So basically, it looks great on paper but they’ve narrowed their audience so much that it’s hard for them branch out and grow. Why’s everyone on twitter? Now, it’s because everyone’s on twitter but when the service launched nobody knew what the hell it was for. Most people ignored it. But then it started catching fire.

The beauty of twitter is that you can use it for anything from anything. You can twitter from your phone, the web, IM, and many more. You can use it as an IM, mini-blog, social network, or newsletter among other things. Yet twitter doesn’t even have a decent splash page telling you what it’s about. People figured it out for themselves and each person uses it just a little differently. It’s simple and yet diverse.

Having said that, would dodgeball have been better off not getting acquired? I think so. These guys suffered a lot under a corporate bureaucracy. Shocking that it was Google but since when did multi-billion companies act otherwise? Too bad the founders didn’t see the light earlier and just take the money and run. I’m sure this bitter experience will fuel their desire to succeed and this time they wont make the same mistake.

Categories: web