kontrawize

How much would you pay for a blog in the top 74%??

I mentioned that PubSub have this blog rated in the top 74% of the sites they monitor. Actually, it's dropped to top 84% now. So, you might wonder just what the value of such a highly rated blog is in the free market economy.

Today I stumbled across BlogShares, which has rated the value of kontrawize at $1137.50. To put that in perspective, it looks like they start all blogs out with a value of $1000, so I've increased shareholder value by almost 14% over about 6 months. Assuming the same compounded rate of growth, this blog could be worth about $1300 by August 2006, for a 30% annual gross profit before taxes, expenses, and extraordinary costs.

Apparently, each outgoing link from kontrawize is worth $61.88. On that basis, the Wikipedia owes me an all expenses paid dinner for four at the restaurant of my choosing, including a few bottles of good Australian red wine (although I don't think there will be enough in the kitty for some Grange Hermitage).

Hmmm. So, to test this theory out, I had a quick look on eBay, and searched for "blog". Guess what? Nobody seems to be selling one, which means nobody is buying one. Markets really only work when there is sufficient liquidity in the market, so the market for blogs doesn't look like such a good investment.

Probably better to invest in Grange Hermitage instead.

The Top 74%

I'm interested in weblog statistics, in as much as I'm interested to get some feedback on what topics people are most interested in. Today I came across the PubSub site, which does link ranking of various sites (according to the blurb, over 20 million weblogs, more than 50,000 internet newsgroups and all SEC (EDGAR) filings). According to PubSub, kontrawize is currently in the top 74% of the sites it monitors.

That's certainly a polite way of putting it. They are definitely members of the "glass half full" camp, rather than the "glass half empty" camp. I guess that's the American way, where everything is talked up, you know, everyone is a vice-president of something or other. I would have said it was in the bottom 26%. Let's call a spade a spade.

It's not all bad, though. There have been a few days when kontrawize was in the top 10%. I can't obviously say why, looking at how the statistics do and don't relate to the days I have posted new items. However, I'll be watching these statistics as the year goes on, to see if there is any discernible pattern.

Blackberry exemption for government - is it fair?

I read in Computerworld that the US government, in its varied forms, might be excluded from any shutdown of RIM's Blackberry service in the US. Now, I don't live in the US, and I don't use a Blackberry (though I might get a Nokia E61 when they become available), but when I was reading this story, it made me wonder about the morality of selective application of court orders and/or laws in general.

I can see how it much be attractive for NTP to exclude government users from the action, since the US government is the only authority that might have any means of derailing the legal action against RIM. However, I would have though that "one in, all in" would be a fairer and more morale (albeit tougher) way to apply the court order.

Indeed, I can see selective application of the court order causing more problems for RIM that would a blanket application. A blanket application, from what I've read, will lead to all Blackberry users having to upgrade their firmware. Painful, but everyone would get over it together in one hit. People survived just a few years ago without Blackberries, they can probably survive a week to a month while they are upgraded.

With selective application of the court order, you can bet that government users either wouldn't upgrade their firmware, or would do so slowly. That means RIM having to manage more versions of the firmware running on the system for years to come, and multiple versions of anything always complicate the solution. So the selective application of the court order could actually make the total impact on RIM worse than would a blanket application.

I understand that legal action is allowed to be selectively applied to some people and not others. Is it moral, though? Should it be allowed to happen that way? Should the "one in, all in" rule apply? It makes me wonder.

Subversion 1.3.0, Apache 2.2, SUSE Linux 10

Today I finished setting up a Subversion version-controlled file server. In the past, I have mostly run Subversion on a local machine (i.e. just on my laptop), but I needed to set up a central repository for a small workgroup. In this mode, Subversion runs in Apache, using WebDAV to manage updates and commits of file and directory versions.

I haven't used Apache much. I never liked the number of textual control files that you need to set up. In comparison, in the mid 90s I was managing a Web site that used O'Reilly's Website Pro, which had a fully functional HTML GUI to manage it. 10 years on, Apache still doesn't ship with a GUI. I'm not sure if that is good or bad, but I find it surprising.

Anyway, the good news was that the installed Subversion documentation includes a 'README.SuSE' file that tells you how to set up a Subversion server using Apache on SUSE. It covers pretty much everything. The only thing that is missing from the setup instructions is the following:

• Apache 2.2 has an /etc/apache2/uid.conf file containing the default user and group that are used for running Apache tasks. That particular user and group aren't set up in SUSE by default. So, you should create the user and group that you want to use for running Apache tasks, and then replace the ones in uid.conf with the ones you have chosen.

Where is the nearest hotel to where I'm going?

I travel a number of times each year to conferences and technical meetings. I don't want to travel far each day (I really just want to walk), so I always want a hotel that is near the venue. However, it has never been so easy to find out where the nearest hotels are. Travel agents really haven't picked up on this one. They can get you a hotel in the same city, but they usually haven't much of a clue about where in the city the hotel is, and certainly not how far it is from your venue. Some of the travel sites like Expedia will show you a map of the hotels in an area, but they don't help you find where your venue is. It can still take a long time to work out where your venue is in a hotel map of an unfamiliar city.

So I was impressed when I had a look at Google Maps today and noticed that you can use it to find hotels. Here is what you do:

1. In the search box, type in the street address of the venue near which you want a hotel, and press Search;


2. If necessary, click on the location (red balloon with a black dot in the middle) to get an Address pop-up bubble;


3. In the bubble, click on Set as start location;


4. Type "hotels near here" in the search box and press Search.
   

Now the map will show you where the nearest hotels are. Cool. It works better for some parts of the world than for others, so it depends where your venue is. I guess there are no guarantees that it shows the best or most affordable hotels. Still, it is far and away the fastest way I know to find a nearby hotel for a venue.

Please switch to FeedBurner feeds for kontrawize

Today I set up some FeedBurner feeds for this blog. The feeds are

Atom 0.3: http://feeds.feedburner.com/kontrawize-atom
RSS 2.0: http://feeds.feedburner.com/kontrawize-rss-2-0
RSS 1.0: http://feeds.feedburner.com/kontrawize-rss-1-0

If possible, I would be grateful if you could use these rather than the direct 'kontrawize.blogs.com' feeds, since this will give me better (anonymous) visitor statistics. I leave that to your discretion, however. Thanks in advance!

XML 2005 Reloaded (& XML 2006)

The XML 2005 conference proceedings and presentations are now available. I encourage you to have a browse if you couldn't attend; it was a great conference.

For me, the 3 most memorable and noteworthy presentations, of those I had a chance to attend, were probably

• Henry S. Thompson: Names, Namespaces, XML Languages and XML Definition Languages (Proceedings | Presentation)


• Erik Meijer: XLinq: XML Programming Refactored (The Return Of The Monoids) (Proceedings | Paper)


• Henry S. Thompson: Functional XML: A preliminary sketch (Proceedings | Paper)

and congratulations to Henry Thompson for scoring 2 out of 3! The Microsoft XLinq stuff is also very cool. There are still open questions about what the approach should be for XML, but I give them a lot of credit for trying to think outside of the current boxes (no pun intended; Don Box has championed this work at Microsoft).

David Megginson will be chairing the XML 2006 conference which is in Boston in December (not Seattle as originally announced). Pity, I was hoping to get a chance to see Eve Maler and her band at one of their gigs in and around Seattle. Still, I haven't been to Boston either, so I'm looking forwards to seeing somewhere new in December. See you there, I hope!

[ Technorati tags: xmlconf2005, xmlconf2006 ]

Checking your e-mail inbox with Java

I'm having a bit of trouble with an e-mail account today, unable to download messages. The problem with these kinds of issues is that it is hard to be sure if the problem is with the mail server, or with your e-mail client. To check that it wasn't my e-mail client, I used some code from a jGuru JavaMail API tutorial.

40 lines of code later (copied directly from the tutorial; I just added a couple of import statements) and I was able to confirm that it was the mail server causing me the problem, not my e-mail client. Very useful.

[ Technorati tags: Java ]

Do you market to your buyers, or your users?

Tim Bray wrote a piece questioning why Intel feels it needs to market its wares at a simple level that avoids going into details of bits and bytes. His argument was that Intel's customers are Apple, Dell, HP, IBM, etc., and that these folks do care about the bits and bytes.

However, I think this misses the difference between buyers and users. Intel's immediate customers care about bits and bytes, for sure. However, what worries a lot of companies these days is the risk of disintermediation, a loss of a direct connection with the end user. If you lose that link, then your product can be replaced easily by another (perhaps cheaper) product, without the end user knowing or caring. Nobody wants to be selling a commodity product if they can avoid it, nobody wants to be selling just on price. They want a real brand, something that end users specifically ask for, insist on, and won't part with their money unless they get.

Hence all of the "Intel Inside" advertising, and the new slogan "Leap ahead" (not that it grabs me personally). Intel wants people shopping for a computer to insist on Intel. Perhaps more importantly, they want CEOs and CIOs and CTOs to put "Intel" into their corporate RFPs that Apple, Dell, HP, IBM, etc. will potentially bid for. It's a very big deal, if you can get it right. Whether Intel have it right is something we'll have to judge over the next few years; AMD gave a very strong performance in 2005, so the future will be interesting.

Note: I don't give investment advice. This is just a personal comment, and should not be interpreted as investment advice.

Pigs can't fly, but ...

... they can glow in the dark. No really. Follow the link, have a look for yourselves.

I wonder how they get to sleep?