I’ve long been a fan of Bloglines, the web based news aggregator service. I usually point my Mac friends to NetNewsWire Lite and my frustrated PC brethren to Bloglines for subscribing and reading blog content.
With Bloglines latest announcement of their support of a web service interface to their content, they have made their service more useful while mitigating the lock-in risk. This is great. Google, you listening?
I love Google’s GMail service. A great next-generation email interface that is both simple and powerful. I logged into my Yahoo! email today for the first time in awhile; it seems archaic now, painful. But will i switch to GMail for my everyday client? Not likely. Google does not have an interface that will allow me to back-up or download all my email.
If Google decided to change their terms of service, discontinue the service, or change the service offering in some obnoxious way i would have no way to grab my current and archived email. [I know that tools exist to do this and i also know that those violate the TOS and that Google could change their code and break all of them at will.]
With Bloglines i can now easily use their great service to manage my reading list, grab the latest items for changed blogs and easily grab my subscription list (akin to my email above) to keep a current back-up.
I’m going to modify my home-grown aggregator code to use the Bloglines Web Services API in liu of my own database.
Aside: David Winer in his “coffee notes” dated September 28th, 2004, could not understand why aggregator writers would want to use or support this service. Why? Why not? What are the real risks? What if Bloglines decided to ban Scripting News from Bloglines users? Does that prevent me from subscribing to Scripting News? Of course not. That’s FUD Dave. Bloglines service does not somehow automatically make RSS obselete or unusable. (Insert THINK flag here) Brent Simmons said NetNewsWire would support Bloglines users, not require his users to use Bloglines.
Aside 2: BitTorrent is not a panacea. It makes the assumption that whomever is using the system will continue serving that data for sometime after they have downloaded it. Many torrent hosting sites set quotas to make sure folks do this. Applying this model to RSS feeds doesn’t seem to make since to me since the data size is relatively small. It will take longer to start-up the torrent-fetching process than to actually grab the data.
It took two minutes to execute ‘/bin/ls’ on a Globus based grid. Seems pretty silly to use a grid to run ‘ls’ now doesn’t it?
0 comments:
Post a Comment