RexTech

So the fine folks at Blogger.com have released a nifty plug-in that lets you post entries into your blog from Microsoft Word.  I'm using it right now for this blog post.

This is an interesting bit of technology.  I'm sure it can't handle everything Word can spit out but still it's amazing to me that it's even possible to try something like this.  I'll have to try stuff like photos and complicated tables, the stuff Word can do but that are a little difficult to implement in Blogger's built-in edit widget.

This post originally appeared on an external website

Don't agree to anything without a formal specification saying what the customer wants and how they want it to work, then do at least minimal research (including asking more knowledgeable/specialist staff) to see if the spec is even possible in it's current form and to get some idea of timescales i.e. don't wing it, that's fatal. Also be firm about delaying extra features to get current targets met, unless later changes really will require significant redesign work, if so get agreement for the extra time required. Don't expect management or sales to have a clue, especially if they are new to the type of project proposed and any inherent fixed timescales, greenness can lead to gross underestimates to customers before you even know about the project e.g. 4 times on my current project! Lastly forget about extreme development it was and is a dangerous, stupid, wishful thinking fad with no basis in reality, as proven by the _failed_ origin project (funny how that part gets ignored), Internet-time projects still need early planning and design!

This is one of those times I wish I was a better programmer.  Because id software has released the source code to Quake III Arena.  I would love to be able to mod something like that into my own game.  

I'd make the hero look like a middle-aged, slightly overweight, Filipino guy.  And one of the guns would be designed to make all the bad guys similarly fat and slow thus leveling the field of combat.

I'd also maybe give it a different sound track.  The 80's were good, maybe some Culture Club or something.  Now that would be an awesome game.

But I suck as a programmer in general, and I really such at complicated low-level languages like C/C++.  So my dream game will never happen.  Sigh.

This post originally appeared on an external website

IT isn't just for clean, air-conditioned, glass-walled server rooms. It's an important part of the auto racing scene. One of our editors made a trip to a recent race up in Indianapolis and took a bunch of photos from the event. Sorry though, no up-close pix of Danica Patrick.

This post originally appeared on an external website

Every morning for the past week, I wake up and I fire up my PSP to check for the new firmware. Sony finally, a week later than originally promised, has released the 2.0 upgrade for the PSP this morning. Had a chance to play with it for a little while, and I must say the new browser is very good.

Of course it choked on the main site I work on, but it made an effort at least. Managed to display most of the header before it bogged down on something, probably the Flash ad or some such.

Now on sites designed for mobile browsers, like News.com to Go, it was awesome. Text was crisp and highly readable, links worked exactly like you'd want, and the page formatting was very good.

I will quibble that the page scrolling function is weird--I'd have expected the directional buttons or the analog stick to do the scrolling, but instead you have to hold the Square button and then use the analog stick. The analog stick by itself just moves the cursor around within the current page, only holding the square button does it let you scroll up or down the page.

Wish I could try it out more here at work but unfortunately our WiFi here requires you log onto the VPN and then it lets you talk to the Internet. Don't see any VPN function in the PSP, nor would I expect one yet as that's kindof a business thing. But I'm sure somebody will eventually.

I guess the real question for me is, why bother having such capable Web browsing in a little handheld device? I don't know that I'll be doing a lot of Web surfing on my PSP.

This post originally appeared on an external website

Amazon.com has an interesting implementation of tag clouds. It's on books that have the "Search inside" banner, like the Chronicles of Narnia.

Hover over the book image and you'll get a little pop-up. Click the Concordance link and you'll get this tag cloud of the 100 most frequently used words in the book. Click on any of those words and you'll get a search result showing where the word appears in the book.

Very nifty use of the tag cloud. I wonder though if this is a sign that tag clouds have jumped the shark. Or are we still on the early cusp of this whole tag cloud thing?

This post originally appeared on an external website

I was on vacation in Florida and went to Universal Studios. The thing is, a lot of their rides were temporarily out of commission throughout the day, like the Mummy ride that I really had wanted to see.

I wonder if, in the race for ever more sophisticated rides, we've reached the point where the rides cannot run reliably? They have to be babied every day, and can get knocked out by all sorts of bugs.

Mechanical problems are relatively easy to diagnose and fix, computer problems can be a lot more pernicious to track down. Basically I wonder if companies like Universal are over-thinking these rides, making them so complex the parks have to spend a lot of time fixing them due to frequent breakdowns? Which leads to bad customer experience--I know I was pissed at not being able to ride the Mummy. It would have been better to not have the ride at all, instead of disappointing so many visitors.

This post originally appeared on an external website

Now this is a great story--a system designed to alert lifeguards when it detects a swimmer in distress has saved a girl from drowning. It's a system called Poseidon, apparently installed at a number of swimming pools in the UK. When it detects a swimmer motionless for more than a few seconds, it sounds the alarm and notifies lifeguards where in the pool the swimmer is.

This post originally appeared on an external website

It's official--Metrowerks as a legal entity will cease to exist. They were one of the survivors from the heydey of software development tools, focusing at first on the Mac platform but eventually providing versions for many other platforms including Windows and Linux.

They've been struggling for a while now against too many competitors on too many fronts. Microsoft has Visual Studio on the Windows side, Apple provides free development tools for the Macintosh platform, and of course there's numerous Java development tools like Eclipse and NetBeans. Metrowerks had products in all those spaces, that's not sustainable against that kind of competition.

They'd actually already announced the end of their Mac tools last month, and they'd sold off the Windows version some time ago. So the writing was clearly on the wall.

Now it's official--Metrowerks is no more.

This post originally appeared on an external website

The recent Hurriance Katrina tragedy has pointed out once again what a great platform Yahoo has built with Yahoo Store. They had the Red Cross donation site up an running almost immediately after the storm hit. That's just remarkable, that the systems they built are so flexible they can have a customized store up and running within hours.

And apparently the Yahoo Store site is holding up much better to the traffic onslaught than the Red Cross' call center. Much easier and quicker way to get your donation to them than trying to call and donate.

We had a group here at work that used Yahoo Store to build a quick experiment in selling some of their products. They were able to do it without any engineering support, in fact without any tech production support either. Again, a remarkable testament to the platform that Yahoo has built.

Makes me wonder if an online store system might not be a requirement to be competitive in the portal space. Maybe that's the rumoured e-commerce thing that Google's building, because they need it to compete with the likes of Yahoo.

This post originally appeared on an external website

Nice little presentation (in PDF) that one of the ActiveState folks gave recently on Firefox extensions. Also has a download tar file with the sample code discussed in the presentation.

Yet another thing I need to find time to work on, understanding Firefox extensions.

This post originally appeared on an external website

Looks like they're pretty darned close to pulling the trigger and converting Slashdot from table-based to pure CSS-based layout. The code looks incredible--view source on this page:

http://www.slashcode.com/

There's not one blessed HTML table in the entire layout. It's pure divs and CSS. Remarkable, can't wait to see them activate this on slashdot.org.

This post originally appeared on an external website

Ever since id software released the Quake III source code I've been trying to get it to compile. Actually, most of that time was spent trying to get Windows XP installed into a Virtual PC on my laptop.

See, the Windows compiler of choice for Quake III is Visual C++. I haven't had to deal with Visual Studio in many years but the last time I did it was a nightmare. The damned thing installs so much crap, and modifies so many registry keys, that it's essentially impossible to uninstall when you're done. So I didn't want to install VC++ on my laptop. Instead I figured I'd use Virtual PC to create a little Windows XP space where I could safely install VC++ without affecting my day-to-day work.

Turns out apparently my old MSDN key for XP wasn't working, at least not with the version of XP I had. Which ironically was a version from my old MSDN package. Not certain what's going on there but I could never get the installer to accept my XP key.

So I ended up borrowing XP Home from a co-worker, who has a current MSDN subscription. His key worked fine, though it took a very long time to get XP installed in the Virtual PC hard drive. It also took a long time to install Visual C++, for some reason nearly two hours. The damned thing was so slow. I know, VPC's are always going to be a lot slower than the host PC but really.

Once I got my VPC environment squared away the actual compile for Quake III was blazing. Now I just have to figure out how to load an actual game into the Quake III engine. There must be mods or something on the Internet that I can download into the engine.

This post originally appeared on an external website

Crazy--just came across the Delay Tolerant Networking Research Group, who are working on a protocol for deep-space networking. The kind of environment where a communication channel between sender and receiver is not available for the entire duration of the network transmission.

I came across it because I was reading where Google has hired Vint Cerf, the guy who helped design the TCP/IP protocol used on the Internet today. I'm guessing in his new Google role he won't have a lot of spare time for a while to work on side-projects like Delay Tolerant Networking.

This post originally appeared on an external website

Sometimes networking is a big dark mystery to me. Like recently, they made some kind of change to our network that required I update all my configurations in Putty. Instead of being able to use the short name to reach a server (i.e., b2b-dev) I now have to specify the fully qualified name (i.e., b2b-dev.cnet.cnwk). Yeah, we use a completely different domain name internally, which I think is supposed to be good from a security standpoint. Because nobody outside our firewall would be able to resolve that domain.

Anyway, back to the network change. I have no idea what changed, none of the engineers around me knows either. But something got switched last week and now I have to specify the full machine name to log onto the dev boxes. It's bizarre because nothing else except Putty broke. Everything else I use, all the tools and servers I hit to do my daily work, all of them operate the same as before. Why Putty alone would suddenly require the full paths I may never know.

It feels bad, as a former engineer, to be so in the dark about something like this. It is a reminder how much of the Internet world I do not really grok. Which frightens me sometimes--how can I be working in the Internet field and know so little about the Internet?

This post originally appeared on an external website

Some researchers at the University of Pennsylvania have developed a prototype backpack that generates electricity while you walk. About 7 watts apparently, which is more than enough to power small several small handheld devices (a Gameboy, an iPod, a few more). Or of course store the power in a battery pack for later use.

The interesting thing is that it apparently does not require a great deal of extra metabolic effort compared to what it would cost you to carry a normal backpack of equal weight. This could be a great boon to hikers--the weight of your backpack combined with all your walking around could generate the power you need to recharge your GPS and whatever small electronic gadgets you're carrying around.

This post originally appeared on an external website

I kinda hate the JavaScript language, mostly because it's a weird abstraction for me. I mean, you're writing code that gets served thru the Web server but executed on the user's browser. That's just odd, and for my poor little brain it's often a difficult abstraction to get a handle on. It's one of the reasons I write such crappy JavaScript and have to spend a lot of time debugging.

Traditionally of course I would do what everyone else does--embed lots of alerts() to see what the hell is going wrong. That gets a little tedious and in some ways it's bad engineering. Because once you have the code working you remove the alerts and by definition that means you've now changed the code again meaning you really need to run it thru your user tests again. Very very tedious.

That's why I was happy to come across this good article on fvlogger. Finally a coherent way to do JavaScript logging, to help you trace execution and see what might be going wrong with a particular script. I'll have to give it a try on my next JavaScript project and report back on what I think.

This post originally appeared on an external website

We all know that the actual engineers who work at Microsoft are pretty smart. But apparently the recruiters and head-hunters who look for new talent may not be quite as smart. One of them recently tried to hire Eric Raymond.

I mean come on, would the guy who annotated the infamous Halloween document really even remotely want to work at Microsoft? Raymond has spent the last several years as perhaps one of the most vocal and visible critics of all things Redmond. The likelihood of him ever working for Bill Gates is about the same as George W. Bush officiating at a gay marriage in San Francisco.

What lazy kind of recruiter doesn't even Google, er I mean MSN Search the name of a potential recruit? To find out what kind of person they might be dealing with. The fact that they didn't bother is either lazy or stupid or both.

This post originally appeared on an external website

A couple lifetimes ago, I worked as a software engineer in the aerospace world. First at The MITRE Corporation in their Houston office supporting NASA's initial Space Station Freedom project. Then McDonnell-Douglas, also in Houston and also working on Freedom. My last stop before I gave up on the aerospace world was Simmonds Precision, which now operates as Goodrich Aerospace, making fuel systems for fighter jets.

In all those companies, the programming language of choice was Ada. Not because all three companies loved the language, but because our primary customer (the government) had mandated the use of Ada.

Well it turns out there's still folks actively using Ada. Enough so that they actually came out with an Ada 2005 ISO standard, and somebody is actually building Ada 2005 tools. Probably not going to be quite the same size market as, say, Visual Studio.

I must admit I'm surprised the language is still hanging around. I kinda liked it, though the whole package/body thing was a tough concept for a lot of folks to get a handle on for some reason. I'm wondering how much longer it will hang around though. I'm guessing the pool of experienced Ada programmers keeps shrinking every day.

This post originally appeared on an external website

I've been in a lot of teleconference calls here at work lately (one of the many downsides to the new job, where I'm doing a lot more management-type junk instead of real work). There's a noticeable delay in the teleconference calls and that means it's possible for multiple people on the call to start talking simultaneously without realizing it at first. It can be a second or so before they all hear the cacophony, and then there's long periods of silence while everyone stops talking and waits for someone to seize control over the phone line next.

Yes, we've been using a voice-over-IP (VoIP) solution for some time to handle our internal telephony. Makes total sense for a company like us, where we will always have a high-speed Internet connection between offices anyway so why not piggy-back voice as well as data on the same network?

And for one-to-one communication, i.e., a standard phone call, it works like a charm. The problem is on these teleconference calls it starts to act a lot like Ethernet does--lots of stations on the network trying to talk simultaneously, when they all realize there's too many talkers everybody backs off for some semi-random amount of time until one of the stations eventually takes control over the connection. I mean, doesn't that sound basically like a slotted Aloha protocol?

Which is what got me thinking that we really need some kind of token ring type mechanism instead. Only the person with the token can speak, they can only hold the token for a set amount of time before they have to relinquish it to the next speaker.

Theoretically it would be faster. Token ring networks are deterministic and generally use the available bandwidth more efficiently than aloha-type protocols. We could cut the time spent in a teleconference by half at least.

I know, I'm basically asking that a group of humans act like a co-operative network protocol to be more efficient. But it's for a good reason--to make us more efficient in our teleconference calls.

I guess I'm basically saying that people need to be more like computers. Is that so bad?

This post originally appeared on an external website

For some reason it's like open season on Google lawsuits. Probably goes with the territory when you become an 80 billion dollar company almost overnight.

Microsoft just won an injunction in their lawsuit over an executive poached by Google, and now a British company is going to sue over Google's use of the trademark G-Mail. On top of a German company that already won an injunction against the use of the trademark GMail in Germany.

Oh yeah, and of course Geico has a lawsuit that's the real threat to Google's business--trademark infringement when third-party ads show up linked to the search term "Geico". The other lawsuits they can probably fight and let run the natural course, but this Geico one's the biggie I think. Google has to settle this one somehow, I don't think they can afford to risk losing this one in court.

I just love it though when big companies sue each other. Geico vs. Google, Microsoft vs. Google, it's all good clean American entertainment. I should have been a lawyer.

This post originally appeared on an external website

So you've rushed out and bought yourself a nifty new iPod nano. Now what to do with all that flash memory? Well forget boring songs and photos, why not install a Linux image onto the nano?

No, this isn't so that you can run Linux on the nano. What it means is that you can plug into any PC that supports booting from USB, and have your very own Linux OS botting into that PC from the nano.

So why go to all this trouble? That's the part I'm not really clear about. I mean, it would be a lot cheaper to use one of those big flash USB drives rather than an iPod to carry around your Linux image. I guess this is just one of those weird proof-of-concept ideas that geeks get all excited about.

This post originally appeared on an external website

So the rumour surrounding a Microsoft project code-named Sparkle was that they were gunning for Macromedia Flash. But apparently that's not the case. At least not directly.

Sparkle is not a multimedia application, you're not going to use it build online games. It's a way to let creative types not just design the user interface but actually build it. So that the people closest to the ideas can get in and build the UI they envision, play around and see if their ideas really work.

The notion is that you'll eliminate the round-trip between designers and developers. The designer doesn't just draw out the UI they build a running version of the UI. It's a forms designer but with the ability to actual build the interactions behind the form pieces.

Doesn't mean Microsoft isn't gunning for Flash, it just means Sparkle is not their anti-Flash tool. There might still be an anti-Flash project buried somewhere in Redmond.

This post originally appeared on an external website

From their proposal to return men to the moon, it's clearer than ever that NASA has grown too hidebound and bureaucratic to be in charge of manned space flight. I mean look at what Burt Rutan accomplished with his SpaceShip One project. With a few years and a few million dollars he managed to do what hundreds of billions of dollars and decades at NASA have not done--put spaceflight within reach of everyday citizens.

I worked for several NASA contractors in the 90's and that culture is not the kind of entrepreneurial, gung-ho environment you need to tackle the challenges of spaceflight. They might have been once but now it's a bunch of career technocrats who spew techno-babble and Gantt charts but don't really know how to innovate or lead. There's a reason the Space Station ballooned way over budget and decades behind schedule--it was NASA running the program.

I'm not saying we shouldn't spend taxpayer dollars on manned spaceflight. But if we are going to spend the money, don't let NASA have the checkbook. Dismantle the place and start over.

This post originally appeared on an external website

I haven't had to create a PowerPoint deck in many years. One of the nice things about my former life as a coder--you just have to sit thru other people's PowerPoints but you don't have to create your own. Well that's all changed, I'm working on a PowerPoint deck for a presentation to one of our executive committees. They just want an idea of what our development plans are for the sites we run.

So this is the first time in maybe six years that I've had to use PowerPoint, and what amazes me about this current version is how easy the transition stuff is. Like with a couple mouse clicks I've been able to make the bullet points on my slide fade in one at a time, so that I know my audience isn't going to read ahead. They have to stay on the bullet that I'm talking about, and then I can reveal the next bullet.

This turns out to be very addictive--like I've been playing around with using a different kind of fade animation for the high-level bullets so that they stand out from the sub-bullets more. That's just ridiculous, but PowerPoint makes it so easy and so seductive to try these kinds of things.

Clearly PowerPoint is the devil's work, meant to suck up all my time playing around with fonts and fades and animations and transition effects. In the end I think I've maybe spent 40% of my time on the slide content and the rest on the visual stuff, the fades and animations and such. For a non-artist like me this is as close to artistic expression as I'm going to get.

This post originally appeared on an external website

Very handy collection of links for anybody who develops Web sites. Aimed primarily at us front-end/UI guys, which is fine since there's so many back-end solutions there's no way one page could capture all those links. But it works great for the UI people.

This post originally appeared on an external website

I missed this last week--Slashdot launched their conversion to a fully CSS-based layout. I'm gonna have to get off my ass and convert my personal Web site to pure CSS as well, instead of its current mix of HTML tables and CSS.

This post originally appeared on an external website

So this little story about an Internet-tracking patent awarded awarded to the NSA got surprisingly little hoopla. But it scares me a little.

Now full disclosure--I worked as a co-op student for the NSA. Thought it was a great work experience--learned a ton, got to work on cool technology with cool people.

But anytime I hear about a government agency that can track your activity I worry. On the one hand it's a good thing that they can potentially track the location of somebody trying to organize terrorist activity over the Internet. But what about the temptation for government to use that to track, say, a political opponent?

Probably not that big a deal anyway, since the technique has some severe limitations. Like it cannot really track dial-up users past their ISP's gateways, which could be several states away from the user's real location.

Which makes me wonder why they even bothered to patent it. Unless it's a smoke-screen to get the bad guys thinking the government might be able to track them down even over the Internet.

This post originally appeared on an external website

Nice little demonstration of what you can do with Ruby--this guy shows a little http Web server he wrote in 70 lines of code. The interesting part--he works for Microsoft. I wonder what it means, that a Microsoft programmer is playing around with open source software like Ruby?

Anyway, yet another thing I need to learn, this Ruby language. I'm so far behind it's laughable--I'm a lousy programmer and getting lousier by the minute.

This post originally appeared on an external website

Back in early summer I helped build out a Photo Gallery system for our sites. Basically any url where the filename starts with "2300-" like this one from CTIA.

Naturally since it's a new system I figured we should train folks how to use it. There's a few little gotchas to be aware of, some features that aren't obvious from the content entry tool. So I set up some training classes, lots of people accepted the invites (I used Outlook's scheduling tool). But when the time came for the class only a handful showed up.

Okay, no big deal I figured the ones who had shown up where the folks who really would be using the Photo Gallery tools so I went thru the training and everything was okay.

The News.com folks and the ZDNet folks went off to the races and started using the galleries immediately. But the people on my team, TechRepublic, did not. In fact it's only been the last few weeks that they suddenly realized how useful these galleries can be.

And that's where my annoyance comes in--the very people who failed to show up for my training are now suddenly asking me how to use the galleries. Or they're doing something horribly wrong because they don't understand how it works and so I have to go tell them individually what to do to fix the problem.

If they'd have just come to the training everything would be cool. But no, they decided it wasn't important enough to learn the system until right now, when it suddenly became critical to their daily job.

It's very frustrating, I tried to be proactive by offering the training but the only people who showed up were the folks who probably could have figured it out on their own, the tech production teams from News.com and ZDNet. Isn't that often the way it works--the people who need the training most don't bother showing up?

I'm just going to stop even trying. Clearly the whole just-in-time paradigm has taken over, nobody wants to try and learn the new tools until the exact moment when their boss dumps a huge project on them that requires using the new tools.

This post originally appeared on an external website

Nice little hack--turns out it's relatively easy to spoof the .mac servers or at least the iDisk portion of them. So if you want to archive stuff thru iDisk but not necessarily to the real .mac servers, you might go thru the trouble.

Other than that there's probably not any really useful application for this spoof but still it's cool that Apple used such open protocols to develop their .mac and iDisk apps.

This post originally appeared on an external website

Sony hasn't officially sounded the MiniDisc death knell but I'm betting the bell-ringer is warming up. I loved the format at first, but then Sony made some huge stumbles. No direct support of MP3 comes to mind--you had to go thru that abysmal copy process to re-rip your MP3's into ATRAC. And they just never got enough 3rd party support, in terms of innovative software (WinAmp? iTunes? Anything besides Sony's horrible SonicStage?).

Just another example of Sony having great technology but stumbling badly in the marketing and the support areas surrounding the technology. You have to wonder if they're headed down the same path with this whole Blu-Ray thing.

This post originally appeared on an external website

So I’ve been maintaining two separate blogs over on Blogger.com–one for my personal journal and one for my work journal (which gets re-blogged onto my employer’s TechRepublic.com site). Now the reason for two blogs is that they cover separate categories of content, and Blogger doesn’t have a way to separate categories within a single blog.
Naturally [...]

This post originally appeared on an external website

Gotta love those crazy engineers over in Japan–they have kept plugging away on stuff like hybrid engines and fuel cells.? The former has now definitely hit the big time, with hybrid cars selling like hot cakes and US manufacturers once again scrambling to keep up.
Is the fuel cell ready for its turn in the spotlight?? [...]

This post originally appeared on an external website

Now this is the kind of research I can support — a couple of professors at Penn State believe a hacker could jam every cellphone in Manhattan with just 165 SMS text messages per second. That’s exactly what we need–a way to get people off their damned cellphones.
In the past couple days I’ve seen [...]

This post originally appeared on an external website

You gotta love geeks who know both hardware and software–who else could have dreamed up the lava lamp build monitor? Simple as pie–if your Cruise Control build fails the angry red lava lamp bubbles, but if all is good you get the happy green lava lamp instead.
It’s silly and pointless yet so fun. [...]

This post originally appeared on an external website

Last year none of the contestants in the DARPA-sponsored robotic vehicle navigation finished, in fact they didn’t get very far off the starting line. But this year it appears three of the vehicles crossed the finish line. That’s awesome, autonomous robotic cars managed to navigate across the Mojave desert. We’re still a [...]

This post originally appeared on an external website

Circuit City had a special on the Netgear SC101, a device where you slip in up to two hard drives and it turns them into a network hard drive. Turns out the thing is a piece of crap–I spent an hour and a half trying to get it working without success.
Installing the hard drive [...]

This post originally appeared on an external website

Hey look at that–Sun is showing some commitment to providing Perl support in Solaris 10. That’s cool, I thought for a while they were putting all their arrows behind Java.
Of course I no longer use Solaris–we switched to Linux at work a few years back. Everybody who had a Solaris desktop was moved [...]

This post originally appeared on an external website

Hard to believe it’s been 20 years since Microsoft introduced Windows 1.0. What a piece of junk that thing was–tiled windows, task-switching that wasn’t even as good as third-party tools from Symantec and Borland. Amazing though what perseverance can do–Microsoft just kept chugging away at it and now Windows runs on something like 95% of [...]

This post originally appeared on an external website

Well I finally got my developer invite to Ning.com over the weekend. I was officially on vacation and wasn’t supposed to do a lot on the computer but I went ahead and set up my first Ning app, cloned from their restaurant review sample app.
I’d like to of course extend it a bit–going to [...]

This post originally appeared on an external website

With 1 GB USB memory sticks dropping for under $100 I was wondering when drive manufactures would start pumping out flash-based hard drives. Well Adtron has done it with their 8 GB internal hard drive replacement.
The problem is the cost–8 GB shouldn’t cost nearly $2,000. Not when the aforementioned 1 GB flash memory [...]

This post originally appeared on an external website

So we’re going to experiment with posting some of our content on a Wiki, to encourage users to update/edit the stuff. But I need to figure out which Wiki software to use. I’ve narrowed it down to two–MediaWiki, which powers the Wikipedia, and Twiki, which we use on our Intranet.
I’m leaning towards Twiki [...]

This post originally appeared on an external website

Why is CVS evil? Because it has a very long memory that you cannot ever escape.
I just got an email this morning from an engineer over at Download.com. She?s working to migrate some reporting code (it hits an old Oracle database server that?s being retired in favor of some fancy new high-speed data [...]

This post originally appeared on an external website

I’ve been wanting to get a linux install at home but haven’t been able to decide on which one to use. Then I ran across Ubuntu, and that’s what I’m using now to wrtie this post.
What decided it for me were two things. First is that everything I read indicated Ubuntu was really [...]

This post originally appeared on an external website

I’ve got Ubuntu Linux running on an old Mac G3, but it doesn’t seem to have built-in drivers for the little D-Link Wireless USB adapter. So I can’t get network connectivity to the Linux Mac which is a real bummer.
I dragged the box into the living room, where my cable modem and wireless router [...]

This post originally appeared on an external website

Metadata is, literally, data about data.  So you'd think that it would stick close to the original source data.  That isn't always the case, as I learned to my chagrin this past weekend.

So here's the story--a few weeks ago the cheap old eMachines I use as an iTunes+iPod host went ga-ga.  Not totally ga-ga, but something was horrible wrong with the networking.  iTunes was unable to connect to the iTunes Music Store, nor could it look up tracks on new CD's thru Gracenote (which I still think of as CDDB).

Now that makes iTunes pretty useless, so I tried everything I could think of to fix it.  Disabled the WIndows Firewall, turned off Norton Anti-Virus, blah blah blah.  Nothing worked, so I finally decided I would need to do a restore on the PC.

First of course I backed up my entire music library.  Took 4 DVD+R's and several hours but I got it done.  Or so I thought--I assumed that backing up the physical files had also backed up the metadata associated with the files.  That seems like a logical assumption--everything else about a track (artist, genre, album, etc.) is stored in the MP3 or AAC file.  It's logical to assume my rating for each song is also stored in the track.  I mean, all that metadata looks the same in the iTunes UI, why would I think differently?

Turns out of course that's horribly wrong.  After I restored my PC and re-installed iTunes I discovered that my ratings were not copied over to the fresh install.  So I've got over 4,000 un-rated songs I have to go thru now.  I have to re-rate them all so I can get them to sort properly.

This is my new crusade--software developers should keep metadata physically attached to the source data files.  So poor users like me don't get stuck.

I noticed this morning that Yahoo thinks I will like Memoirs of a Geisha. It claims this is based on the 60 other movies I've rated on the Yahoo Movies site.  But the thing is, Geisha hasn't opened yet so there are no other user reviews to compare against my profile.  So how on earth can they project that I'm going to like the movie?

They must have some kind of psychic technology, to be able to guess I might like a movie that nobody has actually seen yet.  Or, the possibility that is more disturbing to me, perhaps that little "You'll probably like this movie" box is sponsored.  Like maybe the movie studio has paid Yahoo to promote the movie at certain members such as myself.

In the past that little recommendation box has been pretty accurate--when it says I'll probably like a movie I generally have, and when it says I probably won't like a movie I generally haven't.  But now I'm going to have to take its advice with a lot larger grain of salt.  Clearly it does not operate the way I had thought (comparing my prior review profile against the profile of other people who've reviewed the movie).

I used to work as an ad support engineer.  That means I was the guy who would get paged if the ad engine at CNET.com went down.  It's been years since I did that but I still love to see what's happening in the online ad world.  Which is why I'm excited by this new page-turn ad unit that CNET is launching today.  Here's a sample, the Circuit City ad:

http://www.cnetnetworks.com/advertise/adgallery_2005/standard/mpu/circuit_city/index.html

It starts out the same size as an MPU (the large square-shaped ad that CNET started and is now an industry standard), so you'll probably start seeing it on TechRepublic pages that support the MPU. 

What I like about it is that it's interactive without being in-your-face like interstitial ads or the ads that take over part of the page.  You can browse the pages if you like, or just ignore the ad.  And for a retailer I think it's perfect--the Circuit CIty example is a great showcase for what you can do in this new ad format.

2005 was a tumultuous year in the IT world. We saw the buzz growing around this whole Web 2.0/AJAX thing. We saw further delays in the next version of Windows, though it finally got a real name instead of a codename: Vista. And we saw spyware enter the mainstream when Sony put out millions of intentionally infected music CD's.

That was 2005. It's the end of the year, time to start worrying about 2006. So what will be the big headlines of 2006?I have no idea--what do I look like, Gandalf? But here's five headlines I know you won't see next year.

5. Google surpasses Microsoft in stock market

Google stock would have to climb from around $400 to around $900 per share to beat Microsoft's lofty stock valuation. Or Microsoft would have to lose over 50% of its stock value. Neither of those things is going to happen. Yes Google's had a tremendous run in the 15 months since it went public but sanity has to click in at some point and burst this price bubble. They're an important part of the IT landscape but they simply do not have the kind of real immediate economic impact that Microsoft has on our industry. At least not yet, not in 2006. Maybe 2007.

4. Bill Gates retires from Microsoft

Now while I don't think Google is quite ready yet to surpass Microsoft, they are clearly the biggest challenge the Remond behemouth has ever faced. More serious than Netscape, or IBM, or Borland, or even all three put together. The recent moves at Microsoft, such as elevating Ray Ozzie, show Gates clearly wants to win this battle. He is not going to leave this fight unfinished, he's going to stick around a few more years.

3. Cisco doesn't buy any companies in 2006

Please, this is a no-brainer. Cisco has been built on acquisitions, they're gonna keep buying in 2006. I have no idea what, but there's always interesting technologies out there percolating away in some small start-up. Lots of companies just itching to become part of the networking behemoth. Maybe they'll buy Sysco to put an end to the confusion.

2. Apple gains market share on the desktop

No a thousand times no. Much as I loved the old iBook I used to have, much as I still love my 3-year-old iPod, Apple is charting a course towards desktop oblivion. Their core audience is long gone--companies like Adobe make all their key multimedia apps available simultaneously for both the Apple and Windows PC. Why would anybody spend more on the Apple hardware when they can run the same software on much cheaper PC's?

So there's no more advantage to using the Apple platform for multimedia, and for everything else the software choices strongly favor the PC. Yes the iPod is a winner for Apple and should continue to be a winner, but in the desktop world Apple's best days are a distant memory.

1. IT job worker satisfaction grows

Sorry, but if you thought your job was tough in 2005 it's not getting any better in 2006. There's more disruptions than ever happening simultaneously--Web services, Linux, open source, out-sourcing, wireless, the list is seemingly endless. Changes that we have to navigate successfully to help our business partners. Yet the demands from those business partners keep coming--tighter cost controls, better return on investment, doing more with less.

These forces have been squeezing, and will continue to squeeze, the IT worker. It's a fast-paced industry we live in, and it shows no signs of slowing down in 2006. Hang on to your hats, it's going to be a bumpy ride.

Steve Case, the architect of the AOL/Time Warner merger, is now advocating that they split the company apart. Not just separate AOL from the rest, but also break up the remaining Time Warner itself into three businesses: a cable unit, a content unit, and a Time, Inc. unit.  He makes some good arguments about the pieces being able to compete effectively as separate entities instead of divisions within a larger organization.

But you have to wonder--this is the guy who drove the merger in the first place.  His judgement at the time was that the merger was a good thing.  Based on the results, I'd have to say his judgement was in error.  So why should we think he's got the right idea this time around?

I do happen to think AOL could do a much better job on its own.  Especially their efforts to make AOL.com a destination--it would be a lot easier to set up content and technology deals with other companies if those potential partners didn't have to worry about the big Time Warner parent muscling in on the deals.

But I"m not so sure about splitting up the rest of Time Warner.  Could Time, Inc. really survive as a separate entity from the rest of the media conglomerate?  Could the content companies survive as an entity without the safe haven provided by their cable division?  There seems to be real synergy on the content/cable side, splitting them apart could be costly.

We're thrilled to see other business models based on the Wisdom of Crowds concept.  Sure, voting on stupid stuff doesn't really help anyone, but how about turning traditional businesses like the IT analyst business upside-down by making actual IT professionals (in aggregate) "the experts" rather than so-called gurus who have never used any of the products they're advising folks about.

This is just what some execs from Forrester, META Group and AC Nielsen did in forming Evalubase Research. There are no hands-off analysts, just the actual hands-on IT professionals "voting" on products and vendors they've used. Anyone that participates gets a free scorecard and limited research data access. Using this Wisdom of Crowds method, we understand that Evalubase already has more products peer-rated than on all the Gartner Magic Quadrants combined.  

We're thrilled to see other business models based on the Wisdom of
Crowds concept.  Sure, voting on stupid stuff doesn't really help
anyone, but how about turning traditional businesses like the IT
analyst business upside-down by making actual IT professionals (in
aggregate) "the experts" rather than so-called gurus who have never
used any of the products they're advising folks about.

This is just what some execs from Forrester, META Group and AC Nielsen did in forming Evalubase Research.
There are no hands-off analysts, just the actual hands-on IT
professionals "voting" on products and vendors they've used. Anyone
that participates gets a free scorecard and limited research data
access. Using this Wisdom of Crowds method, we understand that
Evalubase already has more products peer-rated than on all the Gartner
Magic Quadrants combined.  

Well I just got the official notice that TR is discontinuing this blog service.  Haven't used it in a while so for me it's not a huge loss.  I do feel badly for folks who invested a lot of content into the blogs--doesn't look like TR's providing a mechanism for saving off the old blog posts.  At least it wasn't mentioned in the email notification I got.

I am looking forward though to the new site they're launching.  Maybe finally, the search system I was working on when I left six months ago will see the light of day and finally get out of the quiet beta test it's been in for a while.