So Comcast finally went public with a bandwidth cap: 250 gigabytes / month. That's about 50 DVDs or 100 hours of high def XviD per month. It's no accident that Comcast, a cable company with government-granted monopolies on video distribution, is taking actions that limit the use of the Internet for distributing video.

Every network port on your computer has a MAC address, a unique 48 bit identifier. It's a bit like an IP address but lower level; your wireless or ethernet delivers Internet packets to you by knowing your MAC address.

It's very important that every computer on a network has a unique MAC address. So important, that all network hardware has a unique ID burned into its firmware (one of only two sources of standard unique bits on a typical PC). However, it's common for routers to support "MAC address cloning", where your router impersonates your computer when talking to your ISP. That feature was placed there to work around inflexible ISP networks and network policies, and it's mostly useful.

But MAC address cloning can be quite harmful, as I learned today. See, I cloned my laptop's MAC address to my router. Then two years ltaer I cloned my laptop's MAC address to my new router, too, in my new house three miles away. Miraculously this worked fine for a year, until this morning. When my network connection would go down at random intervals. I'm guessing the layer 2 stuff wasn't broken by the duplicate MAC address but rather it confused some DHCP housekeeping in my ISP's network management back office.

Three cheers to my wonderful ISP, sonic.net, for helping me figure out this bizarre problem. I love that when I call them I get a tech who will happily discuss DHCP leases, MAC addresses, and non-standard router firmware wtih me. They were as mystified as I was at first, but talking it over at the phone we figured out something was going bad with address assignment. No way either of us could have figured this out without working together.

It's time to retire RSS. Atom is superior, better defined, mature, and a proper standard. RSS is hurting blog readers' user experience.

The problem is that the major blog platforms that support Atom are publishing multiple feeds in the autodiscovery section of the blog. So when a user goes to subscribe to the blog the browser presents a completely confusing choice of multiple feeds to subscribe to. That choice is meaningless to blog readers and just causes unnecessary anxiety. Publishing platforms should simply advertise the Atom feed and be done with it. Yeah, sure, keep the RSS link working for backward compatibility if you need to. Just stop advertising the link to the RSS.

Here's the advertisements for major publishing platforms

• Blogger: 3 choices. Atom, RSS, Atom comments.
• Typepad: 5 choices. Atom, RSS 1.0, RSS 2.0, Atom comments, RSS 2.0 comments.
• Wordpress: 1 choice. RSS 2.0. Atom is supported but not advertised by default.
• Tumblr: 1 choice. RSS 2.0.

Ironically, many of the RSS links are published through Feedburner which has some automatic feed rewriting that ends up serving Atom anyway.

Congrats to Wordpress and Tumblr for not presenting a confusing choice, but you're offering the wrong feed type. Blogger and Typepad, your user experience is awful. I sympathize with the blog vs. comment problem, but at least get it down to two choices.

PS: no criticism intended to the history of RSS. The original Netscape work was revolutionary and kudos to Winer and other early blog pioneers who saw the value of a standard syndication format and promoted it. Atom represents the industry maturity of the syndication vision of RSS; it's time for the progenitor to retire.

I feel a bit bad about picking on Google Doctype's HTML on launch day. Having launched products at Google I know there's nothing more obnoxious than some know-it-all harping on some tiny problem with the product you've worked for a year to launch. But the irony of the problem was too much to let pass. It's fixed now, a plain ol' HTML link. Yay! Since I picked on Google's newest product let me pick next on their oldest product now; web search. And the ugly URLs it produces.

The first problem is the search page URLs are too big. If you go to the plain ol' google.com home page and search for "RFC 1738" in Firefox 3 or MSIE 7 you end up at here:

More inside ...

Google was very lucky to hire Mark Pilgrim, web standards expert and cranky genius. Congratulations to him on for the launch of Google Doctype, a website that documents fancy HTML, CSS, and DOM tricks that we all use. The current practice of rich webapps is way beyond the official standards and it's good to see Google take some leadership in documenting how things should work.

So it's a bit mind-boggling to me to find this bit of HTML on the front page (URLs elded) that breaks the most basic element of the web; clicking on a link.

What's that? Well, it looks like a link to the actual important content; "Browse Google Doctype". But because someone put an onlick handler on it, it's not actually treated like a URL when you click on it. Instead, when you click on it the Javascript is executed to navigate you to the URL. Unfortunately, when you shift-click on it the Javascript also navigates you. Rather than doing what the user would expect, open the link in a new window.

I know, it's a little thing. But it's a horrible little thing, the kind of thing that so many "smart websites" do wrong and break the web standard UI. My understanding of Google Doctype is that it's a whole project about helping developers avoid this kind of mistake.

I've had 4 email addresses for my personal life: @reed.edu (1989-1994), @santafe.edu (1994-1996), @media.mit.edu (1996-1999), and @monkey.org (2000-present). I take some pride that all but the first address still works. But all the old addresses get is spam; 2600 in the last month. That's 2600 spam messages that got through my gauntlet of spam filters. Most are bounce messages for spam that was forged from my name.

I think it's time to stop maintaining those old addresses.

I had the most annoying problem in Firefox 3; Google Reader stopped working. I'd click on a blog and no items would show up, even if I started Firefox in safe mode.

Turns out I'm not the only one with this problem, but it has an easy solution. Press Ctrl-0 while on the Google Reader page and all is fixed.

What happened is you accidentally changed the zoom level of the page (via Ctrl-Scrollwheel or Ctrl-Minus or the like) and some bug in Reader's HTML and/or Firefox's rendering causes all the content to disappear. There's a Firefox bug filed, but they're pointing the finger at Google.

PS: dear Google Groups team, it is unacceptable for new messages posted to a group to not show up when I post them. I don't care if the backend datastore takes a minute to commit the data, figure out some way to make it visible immediately.

RSS 2.0 is a bad format. I just helped Andy debug a problem with his linkblog's feed. Google Reader was sending folks to his own domain rather than directly to the link destination. Why? Because RSS 2.0 is stupid.

The problem is the guid element in the feed was being used instead of the link element you'd expect. Why? Well, read the spec:

There are no rules for the syntax of a guid. Aggregators must view them as a string.
If the guid element has an attribute named "isPermaLink" with a value of true, the reader may assume that it is a permalink to the item, that is, a url that can be opened in a Web browser.
isPermaLink is optional, its default value is true

Follow all that? guid is defined to be any ol' string. Only later we learn that by default, it's assumed to be a URL that feed readers may use to override the other URL in the entry. In other words, the default behaviour of guid is broken and every RSS 2.0 feed should probably be setting isPermaLink to false on every single entry.

Most people have probably never seen this bug because on a typical feed the link and guid both point to the same URL.

My ongoing rants about iTunes may be getting old, but I continue them because Apple has a reputation for well designed software. The truth is a lot of their software does not work particularly well, particularly on Windows.

Today's fiasco was when I plugged my iPhone in to charge the battery. Only it didn't charge. Why? I was in a hurry, so I didn't wait for 30 seconds for iTunes to start up and didn't notice that it wanted an upgrade. Which left the phone mid-sync and some stupid feature in the iPhone means it didn't actually charge the battery while waiting for me to click a button on the computer at the other end of the charging cable. Nice engineering, guys.

I approved the update. 65 megs of download without a functioning progress bar and the upload process stole keyboard focus from my other apps three times. My desktop reloaded several times too, and along the way iTunes once again stole my .mp3 association and littered my desktop and launch bar with QuickTime advertisements. Then it demanded a reboot, which I refused. The iPhone upgrade failed with an "error -50", whatever that might mean. It worked the second time, although the phone was deactivated for about a minute with no useful warning.

The iPhone has a terrible user experience on the Windows desktop. It's embarassing. And because the iPhone is a closed environment, I don't have any alternative. Apple's linkage between products is the sort of thing that got Microsoft in trouble.

I recently switched my IM client from Trillian Pro to Pidgin, née GAIM. Trillian hasn't had an update in years, the Jabber support is bad, and it was acting wonky. Time to switch to an open source alternative.

Be careful what you wish for. Pidgin is amazing in a lot of ways; it supports a lot of protocols, has simple plugins, a solid communications library, etc. But it also has the breathtaking hostility to usability of so many open source apps. A lot of the GUI details aren't quite right and the docs are downright arrogant.

To be fair, Pidgin mostly works fine out of the box. Except the font is too small. No big deal, every app lets you configure the font, right? Ha! Just try the documented solution. Apparently I'm supposed to learn about GTK, and themes, and figure out where GTK places config files in Windows, then edit them with a text editor and insert 500 bytes of configuration directives for "imhtml-fix".

There's a proposal to add font controls in Pidgin, but the answer from the developer is "I think more people should know how to modify their system configuration settings". Because you know, chatting with your friends without squinting should require you spend hours researching some GUI toolkit no other app on your system uses.