Sunday, October 30, 2005

Openness outside of technology

Maurice, my friend who has been working on the Openomy Ruby bindings, started blogging the other day and today wrote about openness in business:

"Think of all the knowledge stored in the minds of your employees-- unlocking that and sharing it would make everyone more efficient. Just look at the success of the open source development model or of the many community-based online businesses-- no single person's knowledge (manager or programmer) can match the amount of knowledge and expertise stored in a large community."

Read his blog or subscribe.

Monday, October 24, 2005

First Official Openomy API Libraries (Ruby)!

I'm really excited to tell everyone that today Openomy released its first official API bindings.

A close friend of mine, Maurice Codik, wrote the bindings which make it extremely easy to call the Openomy APIs natively from Ruby. Now your next Ruby on Rails web app can easily integrate with Openomy.

Maurice is releasing this open-source under the LGPL license and will be supporting it for bugs and feature upgrades. He can be contacted at maurice.codik@gmail.com. If you want to help him out, email him there.

He's also keeping great documentation on the Openomy Documentation wiki under Openomy_Ruby.

The file itself is located at http://www.openomy.com/api/libraries/ruby/openomy.rb.

Update: Also, as a side note, if you'd like to contribute to Openomy by developing and providing APIs in other languages, simply send an email to iseff@iseff.com (put the word "openomy" in the subject somewhere) and I'd be happy to provide you with everything you need. Thanks.

Why are there so many POSTs?

I've had a couple people ask why there are so many POSTs (rather than GETs -- i.e., links) -- that is, why there are so many submit buttons in the Openomy beta. Here's the answer.


Google has reintroduced their Google Web Accelerator with a vengeance. It was evil enough the first time around, but this time it’s downright scary.

In version 1.0, web masters at least had a fighting chance as the GWA identified its requests with a “X-moz: prefetch” header (as prescribed by Mozilla). Sure, everyone in the world had to change their web applications to fit Google’s vision of a perfect world, but at least they could.

Not so for version 2.0 of this virus. It ships with a brand new mutation: The header is gone! There’s now no way to identify a pre-fetch from a regular request, which means that it’s no longer possible to block the GWA.

If you're going to program AJAXy (even in the slightest), you need to realize at least some of the security implications. I have a feeling we'll be discovering a lot of XmlHttpRequest holes in the next few months and years (sort of like this one), so at least give it a little thought. That starts with realizing there are very poorly written apps out there and taking the simple steps to protect yourself.

Wednesday, October 12, 2005

Is Flash back?

To me, Flash completely died after the era of the Flash intro. But I just watched this video which demos MeasureMap and the new FlexBuilder from Macromedia and I must say, it looks real slick. The integration with HTML and ease of use is pretty compelling. Watch the video and decide for yourself if Flash is back.

Monday, October 10, 2005

eHub Interview is up

Just over a week ago, I participated in the first wave of eHub Interviews. That was really exciting, as it was my first interview. Plus, they put me in the same group as Writely, Kiko, BlinkList, Plazes, Netvibes, and other great sites. That's quite an honor.

For those who don't know, eHub is a comprehensive list of all the new web 2.0-ish applications. It's a really great place to check out new and exciting things. It has also become a really popular site in the last couple weeks.

So, go read the interview and come back to tell me what you think.

Documentation, more beta invites

I've begun working on the documentation for the Openomy APIs. They're now available at http://documentation.openomy.com. I've only started with the methods so a lot more will be added over the next few days.

Also, I'll obviously be changing the design of that site sometime soon, but most likely not before I get through with beta testing. I'd rather have a working service than pretty documentation.

Lastly, I'll be sending out some more invites today. The first few people found some good bugs, and I expect to find others. But the site is appearing more and more stable.

Friday, October 07, 2005

First beta invites pending

I'm finishing up some loose ends and then I'll be sending out the first batch of invites. This batch will probably be small (somewhere in the two- to three- dozen range) and will come with some major disclaimers. But it's really exciting.

If you happen to get an invite in your inbox, I hope you enjoy the service. It's really beta and basically presentation-less. You'll probably see a lot of enhancements over the next couple weeks. Please give me feedback by emailing iseff@iseff.com with what you'd like to see happen first.

Monday, October 03, 2005

Beta is NOT the new prod

Pink is the new black. Alexandra Kosteniuk is the new Anna Kournikova of chess. * is the new *. But beta is NOT the new prod.

I hinted at this in my last post when I said, "Openomy will be released from beta when I feel it's stable, not perfect."

I want to write this post in agreement with Sergi, to whom I responded last time. It's clear that others feel the same way.

Too many sites are coming out in beta and staying in beta for far too long. I really dislike this trend. Beta now means "polished and almost perfect, maybe missing a few features" rather than "unpolished and with bugs."

When Openomy launches shortly, it will be a version which I'm sure will crash every now and then. It will also be totally unpolished, basically presentationless. All the functionality will be there, but it won't be pretty and it will break.

That's okay by me. I want to learn how and why it broke by my users. I don't have the time, money, or space to hire people to come to my studio apartment and perform usability testing. Instead, I'll open the site to the world and let live users do my testing. As 37Signals would say, I'm working with my constraints.

As users, you should expect this behavior. Don't be upset if a page doesn't load or an API call doesn't return what it's supposed to. Instead, please just contact me and let me know what you did and what happened. I'll get right on it and have it fixed shortly.

So, please, if you're site is beta, at least let your users know why it's beta. That's all.