paul reed

An intriguing view of where the world of information and media could go by the year 2014.

• I saw MailboxMap a few months ago; a great mashups that helps you locate USPS mail boxes and post offices, and lists the pickup times. Since I know the details for most of the mail boxes near where I live it was interesting but not immediately useful. However, the other day I in a new part of town and needed a post box with an early pickup. Voila, MailboxMap comes to the rescue!
• BestParking.com finds daily and monthly parking in Boston, New York or Philadelphia.
• Walk Score rates neighborhoods for their walk-ability based on proximity to grocery stores, restaurants, movie theaters, schools, libraries, etc. Seems more useful for towns or suburbs than living here in Boston (which rates a near perfect 98).

More commentary on the state of the web, and some of the dog-eat-dog moves that are going on. On one hand you can’t argue with what Google and the other established players are doing, after all they have a business to support too and people are making money off their services. However, they are shooting themselves in the foot if they create an environment where nobody builds anything on top of their services for fear of being put out of business when the big players feel like wielding their power. Perhaps this is really just another sign that there is no real business models in on the web that have proven successful beyond advertising.

I was getting bored with the monotone, all text design, so I had a bit of a play with the HTML and CSS for the WordPress theme. There are still a few tweaks to make, but it’s mostly finished. If you happen to read this via a feed, you might want to wander over and take a look at it in all it’s glory!

Back in December I circulated a link to Steve Rubel’s post on the death of the page view. At the time it was one of the arsenal of weapons I was using to show that the effects of the technology shift we are within — begrudgingly I’ll acknowledge it as web 2.0 — are far reaching, and fundamentally alter the business model we have grown accustomed to.

I don’t think anyone really expects page views to completely disappear, in the same way that people still talk about hits (albeit never used for anything other that geek bragging rights these days). David Biesel talks about some of the emerging metrics and tools that are trying to provide useful information in the face of page views becoming less representative.

It’s interesting to see this evolve. Given the predominance of ad based business models, this under appreciated area of web technology, must keep step with content and browser technologies.

Try asking Google maps for directions from Boston, MA to London, England. The Google Maps team, once again, proves it has a great sense of humour.

Update: I wonder why they route you via France first?

For some reason my Technorati Profile didn’t have a “claim” on my blog. This’ll fix that!

I dislike the “web 2.0″ label. It is much hyped, and a highly misused term. As a result it is very much misunderstood. Moreover, it has run it’s course having floated around for 2 or 3 years, and still not really found a commonly agreed upon definition. Even as a marketing term it has been diluted to the point of being useless.

Don’t get me wrong, I’m all for grassroots movements and things bubbling up, but this one seems to have gone off the boil. It’s time for something new. Something different. Forget the labels and lets focus on what is really going on.

There has been a radical shift on the web, compared to say 3 years ago. Much of this has been driven by the loosely connected set of technologies with an alphabet-soup of names such as AJAX, REST, SOAP, JSON and their bedfellows. This sparked the development of a set of libraries, or frameworks as some would like to call them, such as Prototype, Dojo, GWT, YUI, and the list goes on. Together this gets lumped under the banner of RIA technology, or the ability to build highly functional, zero-install, applications right there in the browser.

In parallel with this, and maybe as a result of it, there was a social shift with the exponential explosion of content. Everyone wanted their voice to be heard and to contribute to the sea of content. Nobody cared if it was being read, the important thing was to be putting your voice out there. It’s Usenet on steroids, rebuilt for the web with some nifty user interfaces. Blogs and wiki’s are really not new ideas — sorry guys! — but rather this latest incarnation is more palatable for a broader set of users. Hence we have the shift that gets called the read-write web with all it’s user generated content.

But here web 2.0, with it’s fuzzy, ill-agreed to definition, begins to falter. What next? Peter Rip says in his posting Web 2.0 - Over and Out, that really that’s all that web 2.0 is:

Much of the “easy” innovation seems to have been wrung out of the Web 2.0 wave. Web 2.0 was cheap - thanks to open source, simple - thanks to RSS/REST, and distinctive - thanks to AJAX and Flash.

Peter goes on to say:

Now the hard work begins, again. The next wave of innovation isn’t going to be as easy. The hard problems in the [web] are no longer usability or ease of everyday content creation. [...]. Now the hard part is moving from Web-as-Digital-Printing-Press to true Web-as-Platform. To make the Web a platform there has to a level of of content and services interoperability that really doesn’t exist today.

In my mind it is this next version, if that’s how you want to label it, that is much more interesting and significant. It is the entre to the grandly titled Semantic Web, and is founded on the ideas of SaaS or SOA, and better integration of this sea of information. Better yet, it’s already happening, quietly and significantly on the web today. Alex Iskold sums it up better than I could in his post Web 3.0: When Web Sites Become Web Services.

I remember years — and I’m probably talking over 10 years — ago, the idea of one person having only one telephone number. This was when mobile phones were starting to get popular, and faxes were all the rage. Seemingly space on business cards was getting scarce as people began to have more than just a switchboard and extension as their contact number.

Well, good ideas never die, they simply hibernate for a while, and it seems that a company called GrandCentral has woken up to the idea that getting in touch with people can be hard these days, and on the other side people want more control over how they’re contacted.

GrandCentral’s idea is simple: they provide a virtual telephone number that redirects to whichever physical phone you choose to take that call on. They provide multiple ways of taking the call, avoiding the call, or customizing what the caller hears when placing the call or leaving a message. This puts the control of who reaches you back into the callee’s hands, rather than the caller armed with a list of numbers to go through.

If I hadn’t already cut down to one phone number (my work number rarely gets used these days) I’d consider using this service, as much to avoid having to give out multiple phone numbers and explain when they work or should be used. Still surprising how long it’s taken for this sort of thing to pop up …

I recently downloaded a basic Gentoo VMware image. To make sure the system had all the latest patches on it, I did an emerge world, but kept stumbling upon a problem where it would not install the package man-pages-2.23.

Seemingly it didn’t exist on any in any of the portage repositories with 404’s and timeouts trying to fetch it. Curious.

So I searched the forums, and came up with nothing. Nobody else seemed to report that this particular package was in the portage tree but not available from the repositories for download. Weird.

Then I remembered from a previous install on a KuroBox that the first thing you should do installing a stage 3 image, is to do an emerge --sync to let portage do some house keeping and tidying up of it’s internal databases. All fine and well.

So now to do the updates:

# emerge --sync
# emerge -Davu world
...
[blocks B ] sys-apps/pam-login (is blocking sys-apps/shadow-4.0.15-r2)


Aaaarrrgggh! This old problem again, the pam-login vs. shadow conflict. This one stumped me for ages a year or so ago when it first appeared, and there was little help in the forums, but lots of untested ideas. Nowadays the Gentoo forums are littered with threads on this, as it caught everyone running this Linux distro. Unfortunately, the solution is a little lost in all the chatter around how to deal with it, and differences of opinion as to the “best” approach.

I thought that I had blogged about the eventual solution that worked for me, but alas a search comes up with nothing. So, for the record, the command that works for me on both my KuroBox and in VMware is:
# emerge --buildpkgonly --nodeps shadow && emerge -C pam-login && emerge --usepkgonly shadow


There is also a useful guide on the Gentoo Wiki on how to remove PAM.

Your mileage may vary.