Internet Technology

It's interesting how often the question of online versus traditional shopping comes up. A friend asked me this earlier today and I gave him the same answer I've been providing for a decade now.

These days the response seems reasonable, but back in 1998 it was heresy. It used to be guaranteed to make a room full of start-ups and venture capitalists go dead quiet. Of course back then we were in the middle of the dot-com boom, when somehow geeks who don't like daylight managed to convince everyone that their concept of a good shopping experience was somehow universal.

So here it is:

In an absolutely brilliant but evil move, a Trojan fools users into solving CAPTCHA images. Infected users think that they're entering codes to see a model undress, when actually they're helping crackers register for illegal Yahoo accounts.

In principle, the "network is the system" idea has a lot of merit. The benefits of having all your data stored in some reliable, secure, redundant database that's centrally managed and hooking into it with whatever device is at hand — be it a desktop machine or a cell phone — has a lot of appeal. Keeping a system available on the net, up to date with fixes and patches, and secure is no trivial job. It's exactly the sort of thing that should be left to someone who is a professional at it.

[Revised: Less than 24 hours after I posted this, I received a phone call from a salesforce.com representative, apologizing for the misuse of my information. My understanding is that one of their partners is to blame; that the misuse originated outside their organization. I was looking forward to receiving more details in an e-mail, but in preparing to let their mail pass my spam defences, I messed up -- and all mail has been bounced from late Friday through most of the weekend. Hopefully they will re-send it so I can add more factual information to this. At this point, it's clear that at least Salesforce takes this sort of misuse very seriously, and I have accepted their apology.]

I have always wanted to beat "phishers" at their own game. Briefly, a phishing scam creates a page that looks like a legitimate site, requesting user name and password information. The scammers send phony requests via electronic mail under a variety of pretenses, urging customers to follow the enclosed link. Instead of going to your bank or eBay or PayPal, the link goes to their rogue server that looks like a legitimate site and the information is logged there for subsequent criminal activity.

As a rule, if everyone who received a phising attempt (or a mortgage solicitation for that matter) took the time to follow the link, then input bogus data, then the scam / solicitation would instantly be rendered ineffective. The criminals would be faced with sorting through thousands of garbage records in order to locate the actual victims.

Unfortunately as a society, we're don't do all that well at things that benefit the "collective good", so we're stuck with scams in our mailboxes.

But AJAX changes that.

This post is written for developers who use the PHP programming language. It provides a technique for catching a common class of typo by using the __call() __get() and __set() overloading methods.

If You Liked "The Secret" went live yesterday. Talk about a project that expands to exceed available resources, this was it.

The site is for people who liked the movie "The Secret". It offers some of the classic titles in this field of thought, customized with a person's name and downloaded in electronic format. It sounds like a good idea on its own, but it's a surprisingly powerful moment when you open up a book and find it written for you.

As a developer of web sites, I've had experience with quite a few hosting providers over the years. Few of them have been much beyond "satisfactory". After watching how Rochen managed the extraordinary demands of the Joomla community web site (www.joomla.org), and considering that they did this above and beyond the call of duty — as a sponsor Rochen sees no direct revenue from Joomla, I recently decided to open an account there.

The relationship is new, but so far it's looking very, very good.

Those of us who do web development face a daunting task. Even with well defined web standards, every browser implementation has differing levels of compliance, bugs, and differences in interpretation that can make a site that looks great in one browser look awful, or even non-functional in another.

Then there's Microsoft. Despite being a party to the development of many of these standards, they're notorious for going their own way and for ignoring standards when it doesn't suit them. The problem is that Internet Explorer is by far the most installed browser out there, so web site developers are left to accommodate their errors and arrogance.

Okay so it's been a long time since I've made a post here. There's lots of reasons for that, but one of them has been that I just couldn't tolerate Blogger anymore.

After looking at a variety of solutions, I finally stumbled upon Serendipity (www.s9y.org), and it looks like it's going to do the job very well.

There's still a few things to take care of, a few broken links and style tweaks to get into place, but it was easy to install, fairly easy to extract data from Blogger, and it generates decent HTML.

We're back in business!

It's always fun to try to decipher how an algorithm works. It's going to be even more fun to write this post without biasing the results: the requirement is to be abstract without being obscure. It seems now that the main page is responding to content, but only that near the top of the page. All linked content pages are still responding as previously described.

This suggests that the top of a page is what's important, which is an interesting observation, both for those seeking higher placement and for those viewing results. The old adage of "put what you want to say in the introduction" holds true more than ever. I suspect "the top of the page" is the text below the second level header.

If the results no longer trigger off topics related to biochemistry, we'll know this is true. What's missing is a way to discover when an indexing event has occurred.