‘AOL Journals’ To Bring Blogs To Millions (TechNews.com)
AOL will give members three ways to update their blogs — through an online template with blank boxes for text input, through AOL’s instant-messaging system or by telephone. The phone option will be available only to subscribers to the extra-cost “AOL by Phone” service, who will be able to leave voice messages that will be posted as MP3 sound files.
Dang! I should have patented the invention of using a VoiceXML application to post to a blog by telephone when I had the chance. Actually, I would be pretty disappointed if the U.S. Patent Office granted a patent for phoneblogging. Nonetheless, given the kinds of ideas Amazon and others have recently been able to patent, I wouldn’t have been shocked if they would have given me a patent for the invention. Based on estimates I have read that an independent inventor can expect to spend about $10,000 just for the US patenting process, there was no way I would have gone for the patent unless I planned to license it or turn it into a real business.
The AOL by Phone voice portal was spawned from AOL’s purchase of Quack.com back in August 2000. I can’t remember who TellMe and BeVocal’s other competitors were back then. I didn’t start doing VoiceXML development until sometime late in 2001.
I started working on PhoneBlogger in late October 2002 and released it in January 2003. Between vacation, my real job, and not having a real plan for what to do with PhoneBlogger, I took far too long to finish it up. If I had been able to work on it full time, I’m sure I could have completed it start to finish in less than two weeks, maybe even less than one. That’s far more of a tribute to the richness of the code libraries and tools (VoiceXML, Python, xmlrpclib, TellMe VXML hosting service, Lame, SoX, etc.) I was able to use than my coding skills.
The biggest nightmare by far in developing PhoneBlogger was dealing with XML documents in JavaScript. I estimate that I spent about 1/4 of the total time writing code and unit tests and then debugging what should have been some really simple code for reading and parsing XML. JavaScript desperately needs better APIs than the DOM. In hindsight, I would have figured out another way to deal with the config info, but I kept feeling I was just a couple lines of code away from getting it to work. One of the hardest decisions for a developer to make is when to abandon an approach. This time I let stubborness get the better of me.