Friday Rant: Why Won’t Google Make It’s Applications Work With Windows Mobile?

Let’s start with a confession.  I’ve bought into the Google ecosystem big-time.  I use Gmail as my primary personal e-mail account.  I use Google Apps for the email on the ochsenhirt.org domain.  I use Google Reader for RSS and Google Maps for directions.  I haven’t yet jumped on board the web apps bandwagon for word processing and spreadsheets, so the added features in Zoho haven’t been enough – on the few occasions I’ve done a spreadsheet or quick document online, it’s been on Google Docs.  Plus Google for search, of course.  With all that said, Google is really starting to bother me.  Specifically, the way that Google’s mobile offerings work with Windows Mobile.  Or rather, the way that they don’t work.  Granted, many of the problems are at least as much the fault of AT&T and/or Microsoft, but Google could make them work better.   So here in no particular order are some glitches in the Google Matrix:

1. Gmail via IMAP results in blank message bodies

When Google introduced IMAP for Gmail, it sounded great.  Finally, I could access Gmail on the Blackjack without using the java application.  You could always use POP, but that resulted in mailboxes that were horribly out of sync.  Plus, you lost the benefit of Gmail’s tagging system.  IMAP promised to be better.  And it is.  Via IMAP, I can grab messages on demand or automatically, using the same messaging application I use for Exchange Direct Push.  Not so fast there, fella.  Most, but not all, HTML messages show up with blank message bodies.  WM5 doesn’t do HTML mail, so I didn’t expect the HTML to come through intact, but I should still get the plain text.  Apparently, Gmail’s IMAP implementation isn’t reporting certain optional fields.  Google apparently didn’t bother to test IMAP on WM.  WM6 doesn’t seem to be any better, so unless a third-party app like Flexmail can fix it, or Google fixes it, IMAP on the Blackjack is fairly useless except as a glorified Gmail Notifier.  Some have reported success using AT&T’s Xpress Mail, but I have no interest in encouraging AT&T to push its own services.

2. The Gmail java application doesn’t work in AT&T’s broken Java

This one is clearly not Google’s fault, but Google could fix it.  AT&T intentionally crippled the Java implementation on the Blackjack, the 8525 and the Tilt (and probably other recent phones like the Moto Q Global and Blackjack II, but I don’t know for sure).  If you try to run an unsigned Java midlet on the Blackjack, it will ask for permission every time it needs to send data to the internet.  This is not part of the J2ME specification, and is not part of the stock midlet manager AT&T uses.  They intentionally crippled it by removing the option to grant permission on a per-session basis.  What does this mean?  It means you have to click OK many, many times before you ever reach the Gmail inbox, and you have to continue to grant permission every time you open or send a message.  That makes the application completely unusable.  The same is true, incidentally, of the Google Maps java application, Opera Mini 4  and anything else that accesses the net.

AT&T did this, they say, for security purposes, but that’s a load of crap.  They did it to avoid cannibalizing the market for their own services.  If you can use the Gmail application, you don’t need Xpress Mail.  If you can use Opera Mini, you won’t be impressed that the Q9 Global includes Opera Mobile.  If you can use Google Maps, maybe you don’t pay $10 a month for AT&T’s GPS service.  There is, however, a workaround.  By installing another midlet manager, such as IBM’s J9 or Esmertec’s Jbed, you can install java midlets that offer per-session permissions.  That shouldn’t be necessary, however, and Google could fix the problem by either (a) offering a signed java application or (b) offering the Gmail application as a native WM application, the way they did with Google Maps.  All is not well in third party midlet manager land, however…

3. The Gmail java midlet crashes Jbed

Among the alternative midlet managers, I like Jbed better, because it renders Opera Mini better than does J9.  Unfortunately, if you try to run Gmail under Jbed, it crashes upon sign-in, and does it every time.  Therefore, my Blackjack now contains three midlet managers, the stock AT&T one that I never use, Jbed for Opera Mini and J9 for Gmail. If Google isn’t going to fix Gmail for all the AT&T customers, they could at least make it work under the workaround.

4. The Google Reader mobile site crashes Pocket IE

Google hasn’t introduced a Google Reader Mobile application, but there is a quite usable mobile site for Reader users.  You can view all items, or view individual subscriptions or individual tags (which most people, including me, use like categories).   So far, so good.  Unfortunately, if you try to use the Mark These Items As Read link, which should mark the nine items on-screen as read, all it does is close Pocket Internet Explorer.  That makes it unusable for me, as I don’t want to open each item individually when I’m reading it on the Blackjack.  Opera Mini works just fine, so this is some sort of PIE issue, but it’s still a pain.  Although there’s probably a work around for this, too, under Jbed Opera Mini won’t respond to keypad number shortcuts, so I can’t push # to mark all items as read.  So even the Reader workaround needs a workaround.

5. The new Google Maps Find My Location feature doesn’t work

The new version of Google Maps Mobile includes a neat feature that attempts to use cell tower triangulation to provide a rough location for devices without GPS.  Unfortunately, on my AT&T Blackjack, GMM just says my location is currently unavailable.  Lots of other people are reporting the same problem.  I have no idea whether this is a Windows Mobile problem, an AT&T problem or a Google problem, but it appears they didn’t do adequate testing with WM devices.

With the announcement of Android and the Open Handset Initiative, along with the customized Google apps on the iPhone, it’s doubtful Google is going to work too hard fixing these problems.  That’s unfortunate, as I have zero interest in the Apple product, and Android is a long way away.  There’s no good alternative, but Google deserves criticism for the way their products interact with WM.

How I Got My Hands on a Wii

Wired has a nice explanation of Why You Can’t Get Your Hands on a Wii.  Nintendo says they’re trying to keep up:

Ultimately, Wii production numbers — and the United States’ allocation of consoles — are determined by Nintendo’s home office in Kyoto, Japan. [Nintendo Senior Vice President George] Harrison says the company will continue producing 1.8 million Wiis every month until demand subsides.That should happen next spring. But with many high-profile game releases coming after Christmas, like Super Smash Bros. Brawl and Mario Kart, Wiis could be hard to find well into 2008.

And if you want to put a Wii under the tree, without going to one of those eBay scalpers? “It’s going to take some luck,” says Harrison, who notes that retailers sometimes hold Wiis off store shelves until Sunday mornings, when the advertisements go out in the paper.

It was with much trepidation that I heard just before Thanksgiving that a Wii would be in the letter to Santa this year (although it’s probably still better than the fart machine that was also on the list for a while).   Thanksgiving morning was a bust – Kmart advertised Wiis, but the Lynchburg, VA store (where we were over the holiday weekend) only got 10, and 6 AM wasn’t nearly early enough.  I then tried an online inventory tracker for the Wii, but that was a bust, too. It reported stock at Sears a few times, but that was mostly due to Sears’ busted code. had I been awake at the right time, however, it may have let me get a Wii when Amazon infamously posted some.

We have a Wii now, and we got it the only way that seems to work – we sicked  my Mother-in-law on them.  She was also looking for a Wii for my niece, and by calling around was able to find out when stock was to arrive at Kmart (for Wii #1) and Target (for Wii #2).  So be diligent, and you’ll get your Wii.  It helps to be looking in a smaller market – the “pester the electronics manager at the big box store” strategy didn’t work nearly as well in Richmond (projected 2008 population 1,093,227) as in Lynchburg (230,651).

It’s not necessary to buy the $677 Walmart bundle or drop hundreds on Ebay (which currently has one with a Buy It Now price of $250,000 which appears to be in jest and one at $1299.99 which appears to be serious). I’m out of pocket about $300 ($249 plus an extra Wiimote and Nunchuck).  Now to make my Christmas list (Metroid Prime 3: Corruption, Super Smash Bros. Brawl,  Geometry Wars: Galaxies, Star Wars: The Force Unleashed).

Earth2Tech: Climate Change is Gonna Cost Us, So Give Me Your Wallet

Earth2Tech asks a question: “Would you, for a limited amount of time, give up a small percentage of your income to stop global warming?”

Simple answer: No.

Slightly more complicated answer: It depends, given that the question is fundamentally unfair and incomplete. Whether I would give up some of my income in order to (maybe) produce some inchoate good (no one thinks we can stop climate change for 1.6% of global GDP, but maybe we can change the rate of change) depends on the relative severity of my sacrifice compared to the benefit produced.  In other words, what are you giving up? I’m willing to give a dollar to a panhandler, and if 999 other people gave him a dollar, he’d have $1,000.  But I wouldn’t give him $1,000 so that those 999 other people could buy a burger off the Dollar Menu.  So don’t just ask me to give up some of my income without telling me what you’re giving up.  And by “you” I mean “the other 6 billion people on Earth.”

The Blackjack II Sucks, Moto Q better?

I know I said I’d never buy a BJ II just to get WM6, but I’ve been seriously reconsidering.  I had to go into the AT&T store (to swap a blue RAZR I got via Premier for a pink one), and the shiny new toys were calling.  Plus, MS just officially released the update to Office Mobile, and I’d love to have HTML mail.  So I started thinking about new devices.  I like the smartphone form factor better – the Tilt is nice, but huge.  The obvious choices are the BJ II and the Moto Q9h, which are roughly equivalent in price.  The BJ is familiar, so I was leaning that way, but I’m dissuaded after reading Omar Shahine explain why the Blackjack II Sucks.  Omar hates the crappy battery life, and that the screen comes on whenever a new email arrives, thus making battery life even worse.  Since Richmond got 3G service a couple of weeks ago, by BJ I battery life has been horrible, so I’d hate to see it get worse.  But this was the kicker:

Internet Connection Sharing is removed from the ROM. Gone, bye bye.

AT&T has always hidden ICS, which allows you to use the phone’s data access via USB without paying for tethering (Omar actually pays for tethering, so it’s even worse for him).  But the file was still there, in the Windows folder.  You just had to make your own Start menu shortcut.  Now, apparently, it’s gone completely.  I use ICS on occasion when Wi-Fi isn’t available.  It’s a very handy feature.  It’s not fast, so I’d never use it over Wi-Fi, even Wi-Fi that I have to pay for, but there are times when there’s no Wi-Fi, and such times are few enough and far enough between to make it impossible to justify adding a cellular data card to my plan.  The Q9h apparently has ICS (hidden, of course).  So if I do upgrade, AT&T made my choice for me.  Hear that Samsung?  Your benighted carrier partner just lost you a sale.  I’m sick of being nickel-and-dimed to death by AT&T just so they can push their overpriced add-on services (data tethering, GPS, etc.).

If anyone has experience with the Moto Q on AT&T, I’d love to hear it.  Reviews have been fairly positive so far.

Microsoft shows off Windows Mobile 6.1, AT&T users would settle for WM6

Engadget Mobile says that “today Microsoft showed off the next version of Windows Mobile at their annual Mobius conference.”  A handshake NDA prohibits those who saw it from saying much more, but it’s apparently an update, not an upgrade.  Call it WM 6.1, not WM7.  That’s all well and good, but I still want to know where the WM6 upgrade for my Blackjack I is.  How ’bout it, AT&T?  Think you can manage the upgrade you promised long ago before MS releases the next version in Q1 2008?  Didn’t think so.

NYT explains why à la carte cable is more expensive

The New York Times’ Joe Nocera explains cable economics:

The reason is that unmoored from the cable bundle, individual networks would have to charge vastly more money per subscriber. Under the current system, in which cable companies like Comcast pay the networks for carriage — and then pass on the cost to their customers — networks get to charge on the basis of everyone who subscribes to cable television, whether they watch the network or not. The system has the effect of generating more money than a network “deserves” based purely on viewership. Networks also get to charge more for advertising than they would if they were not part of the bundle.Take, for instance, ESPN, which charges the highest amount of any cable network: $3 per subscriber per month. (I’m borrowing this example from a recent research note by Craig Moffett, the Sanford C. Bernstein cable analyst.) Suppose in an à la carte world, 25 percent of the nation’s cable subscribers take ESPN. If that were the case, the network would have to charge each subscriber not $3, but $12 a month to keep its revenue the same. (And don’t forget: with its $1.1 billion annual bill to the National Football League alone, ESPN is hardly in a position to tolerate declining revenues.)

Of course, the real math is even worse for the consumer.  According to the recent Warren Communications study, there are about 67 million households that subscribe to cable service with at least 36 channels.  For the week of November 12-18, the top-rated prime-time cable channel was the Disney Channel, with an average of about 2.9 million viewers. For that week, less than 5% of cable households tuned into the #1 cable channel (ESPN’s at number 3, or 4.1%).  If you go down to number 20 (the History Channel), it’s only 1.5%.  Would any of these nets really get 25% of the total cable households to buy them a la carte?  It’s really pretty simple math: the revenue required to run the channel stays the same, but the customer base over which such costs are spread goes down dramatically, so the price goes up (also dramatically).  That’s not to mention the sharp cut in advertising revenue with far fewer potential eyeballs, and any increased overhead costs for the cable companies, who would have to implement a far more complicated billing system (and they can’t even get the simple one right much of the time).

A la carte programming sounds great in theory, but it would be bad in practice, and wouldn’t save anybody any money.  that’s not because the cable companies are evil, greedy bastards (at least not this time), but simple economics.

Update: As pointed out in the comments, the numbers above are average viewers (which I noted when I quoted them). Unique viewer data isn’t available. I stand by my original point, however. You’re not going to get from 2M average viewers to 12M or 15M subscribers for most channels. Even the #1 cable show (almost always Monday Night Football on ESPN) gets only about 10M viewers. Is the total ESPN subscriber base significantly broader than the pool of people who watch MNF? You’re certainly not going to get to the numbers needed to make the fee per subscriber cheap.

Universal CEO: We’re Not Malicious, We’re Just Stupid

In an interview with Seth Mnookin in this month’s Wired, Universal CEO Doug Morris takes a new tack in justifying the music industry’s heavy-handedness with its own customers.  You see, they weren’t malicious, they were just stupid.

Morris was as myopic as anyone. Today, when he complains about how digital music created a completely new way of doing business, he actually sounds angry. “This business had been the same for 25 years,” he says. “The hardest thing was to get something that somebody wanted to buy — to make a product that anybody liked.”

And that’s what Morris, and everyone else, continued to focus on. “The record labels had an opportunity to create a digital ecosystem and infrastructure to sell music online, but they kept looking at the small picture instead of the big one,” Cohen says. “They wouldn’t let go of CDs.” It was a serious blunder, considering that MP3s clearly had the potential to break the major labels’ lock on distribution channels. Instead of figuring out a way to exploit the new medium, they alternated between ignoring it and launching lawsuits against the free file-sharing networks that cropped up to fill the void.

Morris insists there wasn’t a thing he or anyone else could have done differently. “There’s no one in the record company that’s a technologist,” Morris explains. “That’s a misconception writers make all the time, that the record industry missed this. They didn’t. They just didn’t know what to do. It’s like if you were suddenly asked to operate on your dog to remove his kidney. What would you do?”

Personally, I would hire a vet. But to Morris, even that wasn’t an option. “We didn’t know who to hire,” he says, becoming more agitated. “I wouldn’t be able to recognize a good technology person — anyone with a good bullshit story would have gotten past me.” Morris’ almost willful cluelessness is telling. “He wasn’t prepared for a business that was going to be so totally disrupted by technology,” says a longtime industry insider who has worked with Morris. “He just doesn’t have that kind of mind.”

Maybe they were just that ignorant, but it’s hard to believe.  Morris wasn’t prepared for a business that was disrupted by technology? What business isn’t disrupted by technology?  Besides, the music industry had already been disrupted by technology several times over. Cassette tapes replaced LPs, and were in turn replaced by CDs.  By the time Napster hit the web in a big way, the stand-alone record store was already on its way out, replaced by Best Buy and Walmart and Amazon, which were able to leverage supply chain management and diverse merchandise catalogs to undercut the Towers and mall record stores of the world.  The difference is, these disruptive technologies increased industry profits, and the industry could manipulate the technology – there’s a reason CDs aren’t scratch-resistant anymore, and a reason that CDs became more expensive while the price of other technology was falling.

The MP3 changed all that, as it unbundled music production, manufacturing and distribution.  With no physical product, the industry lost its control, and with it its ability to turn the new technology into new profit.  But they should have seen it coming.  Consumers were ripping CDs before Napster, playing them on their PCs (Winamp just celebrated its 10th anniversary) and their MP3 players.  Napster and its progeny didn’t create file sharing; they just made it easier.  So the industry damn well should have known change was coming.  Even if they were in fact that clueless, the success of Napster should have clued them in – customers wanted their music digitally, they wanted to be able to play it on any device at any time.  Lots of them also didn’t want to pay for it, but many of them would have willingly paid for their downloads if they had had the opportunity (I said exactly this on the old blog in 1999 – give me a way to buy downloads and I’ll give up Napster in a heartbeat).

But you can’t compete with free, right?  Of course you can.  What you can’t do is wish the problem away, or offer a half-baked “alternative” that is clearly designed to increase industry profit at the expense of satisfying customers.  Why did iTunes succeed where others failed? Partly because it’s part of the larger iPod ecosystem, but partly because it’s so easy to use.  Millions of tracks at your fingertips, easily searched, with guaranteed quality and speedy delivery.  P2P file-sharing is free only if the value of your time is $0.

So was the industry stupid or malicious?  Probably a little of both.  There’s just no way they didn’t see this coming.  maybe they were late on the draw, but they knew eventually.  They chose litigation over competition and denial over innovation not because they were clueless about the technological seism underway, but because they didn’t see an easy way to use it to create additional profit.  They couldn’t stop the oncoming train, of course, so now they’ve given away millions in revenue they could have made over the last decade and are being dragged kicking and screaming into the 21st century

Scoble hates on the Kindle: doesn’t do enough irrelevant stuff

Scoble doesn’t like the Kindle much.  If by “not much” we mean “everyone associated with the project should be fired.” He’s got an advantage on me – he’s used one for a week, which is exactly a week longer than I’ve used one.  But I still say his criticisms are (largely) bogus.

Scoble: No ability to buy paper goods from Amazon through Kindle.

Me: Robert gets off to an OK start here; it’s odd that you can’t use an Amazon product to buy other Amazon products.  But is the Kindle an all-purpose internet device or an e-book reader with internet functionality designed to complement e-book consumption?  It seems that Kindle’s mobile broadband and web access is really intended to get content from the interwebs onto the device so you can read it. Everything else is secondary (but it’s still an odd choice on Bezos’ part).

Scoble: Usability sucks. They didn’t think about how people would hold this device.

Me:  Fair enough if true.  Ergonomics are important, and others have complained about button placement. It does seem ergonomically superior to the Sony reader and much smaller repurposed devices (PDA, iPhone, etc.).  An e-book needs to be roughly the proportions of an actual book, in my opinion.  There’s a reason books are shaped like they are.

Scoble: UI sucks. Menus? Did they hire some out-of-work Microsoft employees?

Me: Robert is starting to go off the tracks here.  This is a variation on his later beef about touch screens.  Not everything is an iPhone. Not everything should be.  Sure, a UI based on big icons and a touch screen would be great, but for 90% of the time you’re using the device, you don’t need to touch the screen (it’s an e-book reader, remember?).  The primary requirements for the display are (a) that it is easy to read, which means black text on a white background and (b) it doesn’t drain the battery in two hours.

Scoble: No ability to send electronic goods to anyone else. I know Mike Arrington has one. I wanted to send him a gift through this of Alan Greenspan’s new book. I couldn’t. That’s lame.  No social network. Why don’t I have a list of all my friends who also have Kindles and let them see what I’m reading?

Me:  These are actually two different complaints, but they come from the same place.  Just like everything need not be an iPhone, not every service should be Facebook.  This mad rush to create social networks around everything diminishes both the social network and the service.  There’s plenty of other avenues to share your reading list with your friends, and Amazon rightly chose to concentrate on the core functionality of the Kindle instead of the latest Web 2.0 faddishness.  It’s a far more significant accomplishment that you can download e-books into your library wirelessly than it would be to create a Kindle social network that most people wouldn’t use anyway.

Scoble: No touch screen. The iPhone has taught everyone that I’ve shown this to that screens are meant to be touched. Yet we’re stuck with a silly navigation system because the screen isn’t touchable.

Me: Wrong, wrong, wrong. Touchscreens are a good idea for some devices, but a bad idea for others.  In some cases, adding a touchscreen would make the device harder to use (imagine replacing the TV remote with a touchscreen). In this case, adding a touchscreen would require compromises in other areas. It would make the device significantly more expensive.  It’s already too expensive at $399. It would decrease battery life, as the reason the battery lasts so long is that e-ink only draws power when the display changes (ask Steve Jobs about design compromises driven by lengthening time between recharges). Finally, the screen on the Kindle is designed primarily for displaying text.  Everything else is secondary, and for now at least, there’s no such thing as an e-ink touchscreen.  And for an e-book, e-ink and menus trumps touch.

The Kindle is clearly an imperfect device, that is incrementally better than what came before it. The technology still isn’t there to make it good enough to replace paper for me, and the economics of the Kindle Marketplace are still off.  So there’s a lot to criticize.  That it isn’t an iPhone/Facebook/browser that happens to also display text isn’t where to start, however.

Another cheap laptop on the way

I’m apparently becoming king of Cheap Laptopia.  You’ll recall that I bought the Acer 5315 at Wal-Mart for $348, concluding that it was actually pretty good for everyday use (web, e-mail, office documents).  You wouldn’t want to do any serious image or video editing on it, and I had to spend another $150 to get Vista Home Premium and 2 GB of RAM installed.  But still, decent performance for under $500.  I was going to add a laptop on Black Friday, maybe the $399 Sony that Best Buy supposedly will offer.  That one is a 1.7 GHz Athlon dual-core with 1 GB RAM and a 120 GB hard drive.  But that involves spending the night at Best Buy, so instead I just ordered a Dell Vostro 1000 from Round Rock for $449.  You can actually get it for $399, but moving to $449 gets you Vista Home Premium, which is cheaper than a Windows Anytime Upgrade ($79).

The specs are not half bad:

• AMD Athlon 64 X2 Dual-Core processor TK-53 (1.7GHz)
• 15.4 inch Wide Screen XGA LCD Anti-Glare Display
• 1 GB DDR2 RAM (Dell skimps and uses two 512 MB sticks – boo!)
• ATI Radeon Xpress 1150 256MB HyperMemory integrated graphics
• 120 GB 5400 rpm hard drive
• 8X DVD+/-RW w/Double-layer DVD+R Write Capability
• Dell Wireless 1390 802.11g Wi-Fi Mini Card

The nice thing about the Vostros is that they come free of trialware and crapware.  No need to uninstall Norton because its not on there to begin with.  I’ll bump the RAM to 2 GB, and use 667 MHz sticks instead of the stock 533.  That will add about $50.  Grand total should still be under $500.  Preliminary ship date is 11/28, so I should be able to compare the Dell and the Acer next week.

In Cheap Laptopia, you can’t do much, but you have more money for beer.

Update (11/26): Changed my mind about the Vostro.  Dell’s already shipped it, but I’m going to return it.  While I still think a basic laptop is good enough for my needs, I’m going to hold out for one that is a bit lighter and has a better processor.  I’d also like to find one with the Intel X3100 chipset instead of the ATI set in the Dell.  I was using a Toshiba with the Radeon Xpress chip, and the Acer with X3100 seemed better on that front.  Maybe I’ll even get a dedicated graphics card.  I struck out on Black Friday, but there’s always Cyber Monday.  Right now I’m tempted by the Gateway MT6916 ($759 at Circuit City) or M-6823 (currently $849 at Best Buy).  Those both come with a Core 2 Duo and X3100 graphics.  The 6916 has a smaller hard drive but more RAM.   The HP Pavilion dv6500t has also gotten good reviews (about $750 with a Core 2 Duo, 120 GB hard drive and 2 GB RAM).

Firefox 3 Beta 1 released

 

Monday night, the Mozilla crew officially released the first beta of Firefox 3. This is hardly new news, but I was traveling yesterday so this was my first chance to check it out.  I had tried some of the early alphas of 3.0 (Gran Paradiso), but haven’t really used it much lately.  Many of the changes from 2.0 are under the hood – check the release notes for details on security and performance changes.

As has already been much discussed, the new version makes significant changes to the way bookmarks and related information are organized.  At the end of the URL address bar is a star:

Click the star once to add a bookmark to the base bookmark folder using the site’s reported name as the bookmark name.  Firefox will report that the site has been added with a different star:

Click the star again, and you can rename the bookmark, add it to a folder or add tags for metadata:

That is a handy organizational tool – you may never need the Organize Bookmarks menu again.  Those tags come in very handy in the new Places feature.  Firefox 3 adds a new folder to the Links bar, conveniently called Places.  It’s full of information about, um, places:

Here you’ll find quick access to your most recently bookmarked pages (Firefox 3 calls them starred pages, which makes sense given the new way to add bookmarks), recently visited bookmarks, most visited bookmarks, recently/most frequently used tags and overall most visited pages.  This could be useful for navigation, but it will also just be interesting in the way Google Reader trends is interesting – what sites do you really visit the most?

The other new feature that I liked immediately is the elimination of the extension whitelist.  You’ll recall the extension two-step from Firefox 2 if you added an extension from the developer’s site or somewhere other than the Mozilla Addons site  – click allow, add to whitelist, install again.  Now you just click Allow in the warning bar that appears at the top of the screen and the install proceeds automatically.

Like all Firefox betas, this one’s not quite ready for prime time.  If you rely on extensions, most of them will not work – only Adblock Plus worked upon install for me.  Firefox will offer to check for updates, but this will most certainly fail.  Some developers have posted beta versions on their own sites, so check the Addons page for links.  I successfully updated Download Statusbar this way.   If the extension was updated for the Firefox 3 alphas, you can make it work on the beta yourself (this may also work to go from 2.0 to 3.0b1, but it may not).  Here’s how to do it using a stock XP install:

1. Download, but do not install, the extension file (it will have a *.xpi filename).  You can right-click on the install button and choose “Save as…”
2. Rename the file, changing *.xpi to *.zip.  The xpi files are just compressed files, but most decompression utilities don’t know this.
3. Unzip the renamed file, and open the install.rdf file in a text editor.  Notepad will work just fine, but third-party editors are easier to work with due to the way Notepad deals with line breaks.
4. Search for something that looks like this: “<em:maxVersion>3.0a8</em:maxVersion>”  This entry tells Firefox not to allow installation on anything more recent than 3.0 alpha 8.  Change it to 3.0b1.  You can also change it to 3.0 to make it work on anything up to the 3.0 final release, or an even higher number if you want, but I change it to this because you never know whether it will actually work with later versions.  Plus, hopefully the developer will release an official update.  Save the install.rdf file.
5. Select all the files and folders in the folder you unzipped in step 3, right-click and choose Send To > Compressed (zipped)Folder to create a new *.zip file.  Windows will by default give it a name based on one of the files in the compressed archive.
6. Rename it back to *.xpi. I usually just rename it to what we started with in step 1.
7. Drag this new *.xpi file and drop it on an open Firefox window to start the install.  After a restart, you should be good to go.  This process worked for several of my extensions, including Forecastfox, Linkification and Stylish.

The onle thing I had hoped the developers would address that they didn’t was memory usage.  Firefox is still a memory hog (120 MB right now, with  five tabs open and seven extensions installed).