Ionrock Dot Org

Traditionally, I'm pretty slow to adopt technologies in a meaningful way. Honestly, RSS totally skipped my radar, in that I just wasn't very impressed. Atom (for whatever bizarre reason) managed to make a blip. Podcasting is another technology I'm only currently becoming interested in. It is not something that is especially ground breaking on a technical level, but I finally have started to see some of the social implications that actually applies to me.

One idea that I would love to see more of is a mixtape podcast. The problem is always going to be copyright issues, but conceptually, I'd love to see people podcasting playlists of bands they are interested in. Searching has never been my strong point, so finding new music has been relegated to friends sending me mp3s or my wife pointing out a band. As feeds have replaced a daily routine of news/planet sites, I would hope that podcasts could offer similar benefit.

What really frustrates me about this idea is that the looming question of copyright. While in some ways the concern is valid, I think it is rather unfounded when it comes to podcasts, simply because a podcast is build on the idea of a feed with audio. In that way, the big benefit of a podcast is that the original resource is still the primary source. Feedburner is a good example of a service that provides something of a content delivery network and, more importantly, accurate statistics on how many subscribers. When it comes to music, this is a perfect web based model to emulate the radio listener system or Sound Scan records in record stores. If labels or artists provide "podcastable" links others can use for a mix tape, it presents an opportunity to spread the music, and more importantly, show actual interest in some artist. MySpace has most likely been doing this since its inception, which can be concluded by the emergence of albums that are released by MySpace. My guess is MySpace keeps reports on the number of plays for certain artists that allow them to make truly educated guesses based on market interest in order to put out records, that for all intensive purposes, should be relatively profitable.

The obvious problem with this model is MySpace holds all the cards. That makes it a problem for the artists because there is no way to link interests geographically or compare specific markets. For example, if one band has sold 1000 CDs without touring and living in a small town, there is a good chance that band is a diamond in the rough and could do well nationally. Likewise, some band living in Chicago could be selling huge amounts of CDs (as seen in their Sound Scan score) and touring like crazy, when in fact they are simply gaming the Sound Scan system. The MySpace lock in is also a problem for MySpace in that its own statistics are only as reliable as its community provides. MySpace is not the same as LinkedIn in terms of the social trends for honesty.

The positive side of this whole issue is that it is only a matter of time. From a technical standpoint, the pieces are in place. There needs to be a social change that presents the benefits of podcasting playlists and sharing music in a way that allows keeping tabs on who is listening (from a statistical perspective). At this point I'm just waiting for the flavor of the month to hit the right mix of form and function the public at large will grab on to.

Permalink | Comments

Permalink | Comments

Over the past week or so I've been making some changes to my website. I've more or less completely rewritten everything with a single focus on moving my data to external services. For blogging, I'm using Posterous. It is pretty slick because you just email them and the blog gets posted. This is not really a new idea (using email as an interface), but it is really clean and simple. This simplicity is very much a requirement because I needed to be able to effectively cache the results and I didn't want to jump through hoops getting little XML details fixed. This fit the bill nicely. For commenting, I'm using Disqus. This seems to be a pretty popular option among some folks and you can export the comments, which seems helpful. I've never been a big fan of comments, but I've been thinking more about it and having a little discourse seems positive.

The way the system works is I have a script that runs and downloads feeds from various sources. It then adds the entries to a SQLite DB. The rest of the site only reads from this DB, so I shouldn't have to worry about concurrency issues. The script that downloads feeds allows filters to be applied to the feeds. The filter can do whatever it wants, so in the case of Posterous, I strip out a few links that get added to the end of my posts. I'm using feedparser which is a first. I've used it before for experimenting, but most times I'd just skip it in favor of using Amara. The nice thing about feedparser is I can use a parsed version of the content. This makes it easy to keep all my entries in xhtml (as in type="xhtml" in Atom speak), which really is just how Atom fits my brain. If I ever get into things like SVG, this seems helpful, but I don't see myself venturing into those waters anytime soon. RDF and Semantic technologies, yes, but beyond that, I'm keeping things pretty basic.

One reason for this change is that I've been making an effort to expand my horizons a bit. I've been focused on XML technologies pretty heavily for a few years now and it feels like it is time for a change. In some ways, my design changes are a sign of my focus on pragmatism. The recent iteration of my site was very easy to put together and it feels rather full featured compared to my previous site. Also, in using Bright Content on and off, it was always very difficult to pick up. I made some bad decisions along the way that added quite a bit of cruft and confusion that made it unmaintainable. This redesign is something of a reset on that to focus on better ideas that use more of the web and its services. In doing so, I've scrapped some flexibility, but that was part of the plan.

This "plan" I mentioning is really just me thinking about the work I've done in the past. XML has always been my hammer, which is not what I wanted. I've come to the conclusion as well that XML as a tool is different from the ideas behind many XML technologies. In some cases, I think the non-XML equivalents might be a little better at times. JSON is a good example of this in that it makes for much better object serialization format in the sense it is modeled after a real programming language (Javascript) while still allowing the benefits of a real document. Another good example is XPointer. HTML links work very well and in the vast majority of cases, the functionality is effectively the same. Microformats are another area where XML community ideas have been picked up by non-XML (HTML) markup.

The biggest problem I've had with XML is that it forces me to use XML as a model. I'm not saying that this is a bad thing, but it does mean that the development environment is no longer the king. I'll be the first to admit this impression is largely because I still have a great deal to learn about my language of choice and XML. The problem is learning both at the same seems to mean neither become my forte. My pragmatic decision then is to try and stay in the Python (and Javascript) world as much as possible. So, far this decision has been very fruitful in that I'm not only become more proficient in Python, but it offers new ideas for how I work with XML. A win-win in my opinion.

My recent reflective trend is because I recently had a birthday. I'm officially 30 years old. I can't say I'm old, but I don't feel young anymore. This is troubling, but I'm sure it will wear off quickly. In all honesty I'm frustrated I haven't done more. I'm not complaining or anything, because life is great, but it feels like a wake up call to get a little more serious about getting things done. The changes in the site is simply a reflection of that desire, so hopefully there will be more changes to come. Wish me luck!

Permalink | Comments

Permalink | Comments