No more (FEED)readers?

Pluck RSS reader is going to be discontinued and the commenters agree that it is a sign of the consumer feed readers becoming (or rather already being) commodities, with this space headed to be dominated by the big players.
It is sure that this market is crowded and it’s also clear that acquiring users has been very hard, both for small startups and for big enterprises (see the warm acceptance of the google reader).

I agree that the best opportunities will be in providing white-label reader technology for smaller customer facing sites. I find also extremely interesting the opportunities for aggregation, clustering and filtering solutions for businesses which will enable better use of the knowledge spread through the blogsphere and will allow building smaller specialized bizspheres connecting enterprises with customers and partners (and even competitors).

But on the other side I see this as a market that is still in its infancy, I think that most of the internet users (even those reading blogs daily) barely know what a feed reader is, and don’t use one regularly.

Maybe they won’t need to know, and feed reading will just be another feature of email clients, but the danger is that there will be almost no research or experiments on how to build better feed readers, so the users may be forced into sub-optimal experiences, which in the end will limit the diffusion of blog content.

I expect the solution to come from the bottom, from open source products on the desktop and smaller players on the web which will have the creativity and the ability to challenge the status-quo.

P.S. see here for my views on the RiverOfNews style of readers

Removing modality

I read today the humanized blog and I find it’s (very) true what Aza Raskin wrote in Monolog Boxes and Transparent Messages: the best interace should aim to minimize the interruptions in what the user is doing.

When you are “in the zone”, especially when writing or designinig or programming, any interruption in your flow of tought is dangerous and makes you to lose that special status of sharp, focused attention that fosters creativity. Modality in an application interface causes these breaks the main activity and should be reduced when/where possible.

I know that there are many points that can be made better in TheTyper to make the writing esperience even more seamless, and I’ll look into ways to reduce the modal interactions that can interrupt the writer. The first task is to embed the spell check into the main editor, enabling some kind of check-as-you-type mode of operation.

I don’t see many more sources of modal interactions that can be removed from the interface, since all the rest of the popups are for operations which aren’t usually done while writing (e.g. changing the fonts or exporting the document) but nevertheless I’ll need to think a bit more on these issues.

Ideas for RailsConf 2007

RailsConf 2007 will be in Portland, Oregon on May 17-20, 2007 and the call for proposals has just opened (the deadline for submissions is November 27, 2006).

In the meantime the conference wiki is open for suggestions on what you’d like to see/hear. I have added two ideas to the wiki:

• “flash” presentations: it would be really nice if there was a session (or two) of 5 to 10 minutes talks each one presenting a tip or recipe.

• “instant demo” area: why not put a projector somewhere (in the lobby or hallway or whatever) where anyone could show what it is doing with rails.

what do you think?

The rise of application-dedicated servers

I always hoped that someone would put into the market a hosting solution that would have made possibile to sell dedicated web applications were the end user had its own application space with its own database: it seems that that day is coming closer, at least looking at the offering from Media Temple, Engine Yard and Amazon.

Coupling some kind of virtual private server or the like with a low enough price and flexible configuration and deployment tools would enable a small software shop to provision dedicated solutions to clients which would not accept having thier data on a shared machine, and would not require a full dedicated server (specially for the management burden associated with it).
Of the three i mention above Amazon EC2/S3 is the one which seems to have the best automation tools (you could easily write a service which automatically deployed the server on behalf of your customers) but it seems to operate at a lower level than the other two.
Engine Yard has nice management tools (albeit without any API it seems) but it seems to offer (too) higher profiles than would be needed (and the prce tag is also higher).
Media Temple Grid Server has the right price and its basic plan features seem ideal for hosting a single application but there is no mention of API support or reseller tools.