Emil's Wicked Cool Blog

Jeff told me he was going to blog about this, but it doesn't look like he's done it yet, so I will. Jeff has been looking for a notebook recently, and since I work for Intel I'm pretty close to that particular market. I'm also working on the Verified By Intel® program, which deals directly with the notebook market, so I'm pretty clued in.

So Jeff emails me on July 3rd to ask me what the T2050 processor is. He found a notebook that includes the processor, and can't find any information on Intel.com...I verify his results, find processor information on wikipedia, and then the fun begins. I can't find information anywhere internally, which is a huge problem for me because if I can't find information, that means that Intel can't process returns on the product and there is no support documentation on the product.

Jeff emails customer support, and after being routed around (support is looking at the same things I am, so they're running into the same problem), basically gets back a non-response around July 11. I have a few internal contacts that Jeff doesn't have access to, so I email them, but Intel is in the middle of a major launch at this time, so I don't get back any response. Finally I hear back on July 25th, and the information I get back is somewhat disturbing. I find that:

• The processor is "off roadmap", meaning that it wasn't planned at all, but rather put onto the market at the last minute, without any supporting documentation.
• It's OEM-only (I support "boxed products" - these are products that you can buy in a box online or through a retailer). Being OEM only, you can only get the processor if it's already in a computer. This is why it's not in my tool for returns, but it still should be in for technical support...
• The person responsible for worldwide support of the processor didn't know anything about the processor until he bumped into one at Circuit City.
• This apparently happens all the time with desktop processors.

What a way for a company to support its products. Hopefully the drive to efficiency will help...

I generally dislike really long blog posts. I think my last two were too long. So this one won't be.

My last post also reminded me that I've been meaning to post on another subject - that of software updates. These are (sometimes) necessary and usually desirable functionality, as long as they don't get in the way of the user trying to do their job. As mentioned previously, I think Adobe does a great job showing how not to do this (modal dialog box, even when viewing in a browser, usually asks to install a bunch of unneeded crap, etc.).

One thing I see consistently done wrong in these update programs is the fact that they tend to ask you to update when you start the program. If a user started a program, do you think they want to update the program, or use the program? (that's rhetorical) Unless the update is critical, the update process should ask to update the program and/or its components when the user requests shutdown of the program. Here's the complete guidelines that I think should be followed:

• If the update isn't critical, wait until shutdown to request the update. "Critical" updates include:
• Security-related issues that can expose the user to remote attack
• Stability issues that the user has experienced. You would know this only if a crash reporting facility exists and the user has opted into the use of the facility.
• If critical, state so and offer the update on application start
• On shutdown, ask the user to update, but close the application if no reponse has been received after some time (how long?)
• Only include updates for existing components - don't ask the user to install Google Toolbar (unless of course the application that they're updating is Google Toolbar)
• Allow the user to choose which components to update. It's ok to use an opt-out model for this. Include a "don't ask me again" facility along with a way for the user to go back and install the "don't ask me again" updates (e.g. through a specific update application menu item)
• Try to avoid forcing an application restart. If you're shutting down the application it may not be a big deal, but if the user selects that menu item or the update is critical, this becomes important. It's really important if the application is a service or an operating system. Yes, Microsoft does a horrible job at this...
• If you have multiple applications in your company, use a consistent service that can update all applications. I do give credit to Microsoft on this one as they are working on getting a single update service. Ideally, there would be a single service/paradigm used across companies (e.g. MSI/Installshield for installations). Market opportunity?

Of course, I could be wrong, but I don't think I am. Let me know.

As I am waiting to respond to a friend of mine today because my browser won't respond, I'm reminded of a few really stupid things that hopefully will become less common in software in the future.

The specific incident today was that I wanted to click on a link (Outback Steakhouse Menu), but the browser stopped responding. Was my computer slow? No. Was it doing something else to the point where it couldn't handle my command? No. What happened was that I had been downloading a 50MB file from a web site. The other browser window just finished downloading the file...so the last step in the process is to copy the file from the temporary directory to the final download directory. During that step, apparently all other browser windows stop operating.

These issues make computing frustrating, and to the casual user, makes the computer "slow".

Other annoyances:

• Modal dialog boxes. These are barely tolerable under normal circumstances, but are horrible in applications with multiple windows when the modal dialog box requires input before any operation can commence in any window. Adobe Acrobat does a fantastic job with this one...they have a window come up for their product updates that sometimes pops under the browser when you view a PDF through the browser. It can hang Firefox and can appear to hang Internet Explorer.
• Waiting to have it "perfect" before displaying results. Windows explorer is a good example - some directories will take 30 seconds or more because I have large files and it wants to do something (to prepare thumbnails maybe?). I don't even need the icons right away - just show me the file names and fill the rest in on your own time!
• Lack of user feedback. This just happened to me also. Click on a link (Quicktime movie) and nothing appears to happen. It's actually downloading but I don't have any feedback, so I click it again, and the download restarts, but I am completely unaware of this. I see others struggle with this one. My favorite is when someone launches a really big application but there's no feedback, so they launch it again, and repeat until eventually the computer brings up the application, but then brings up like 10 more windows and is completely choking on memory.

Other examples aren't coming to me, but I'll try to post when I run across one. Feel free to leave comments with other good annoyances that appear to slow/stall the computer, but really aren't.

Maybe this is just a recent phenominon, but I've been watching and/or participating in IPO investing (sometimes trading) for the last year or so, and have a 100% success rate in making decent money. We've been selective, but made investments in Under Armour (UARM), Master Card (MA), and thought really hard about Sealy (ZZ) and Burger King (BKC). Here's the formula I've (we've) been using:

1. Absolutely must be a well-known company. If the company name isn't recognized, then I don't usually give more than a cursory glance at the rest.
2. Business model needs to be solid. We passed on Vonage (VG) because they have little to offer customers to stay with them, and the big incumbants can (and are starting to) roll out similar services as comparable price points. There's no reason for customers to stay.
3. The wife needs to agree. Not to be flippant here...a second or third person brings a different perspective and will often kill some reasonable concepts (in our case Sealy and Burger King).

UARM and MA were both fairly short term investments, and you can certainly argue with me there. UARM was held for about a quarter until I got sufficiently concerned about Nike coming in to destroy them.

MA is a decent long-term investment, but I had to go on margin to buy the stock, so I sold after a month and having made some significant money. Even without the margin issue, our #1 rule has been "don't get greedy".

Now that I have a fantastic reason to use the new camcorder, I have been making some recordings and otherwise playing with the features. I recently posted two videos to Youtube.com, and the front-to-back process was really straight-forward...technology should continue to put consumers at the center of their world like this:

1. Plug camcorder into computer, turn it on
2. Load Windows Movie Maker
3. Tell WMM to import video from the camcorder
4. WMM detects the clips on the camcorder
5. Take each clip you want to save, save the file as a seperate WMV
6. Upload WMV file to YouTube

While the above process is really cool, here's how I could see it optimized even further:

1. Plug camcorder into computer, turn it on
2. Load Windows Movie Maker
3. Tell WMM to import video from the camcorder
4. WMM detects clips, but also stops when hitting the end of the tape that's been recorded on
5. Take each clip you want to post, post the file (menu item/context menu) directly to YouTube - if login required, WMM can request, or integrate with the upcoming Infocard

So I ran into a really weird situation with the camcorder I recently purchased. The camcorder has two data ports, one is Firewire, and one is USB. The camera itself can do three data-related things:

1. Transfer recorded video to the computer
2. Transfer pictures to the computer
3. Use camcorder as a web cam

In order to do #1, you need to use the Firewire port. This seems logical but a bit antiquated. Originally, Firewire could handle speeds of 400Mbps where USB could only handle 12Mbps. Apparently Sony decided it was easier to force folks to use the Firewire port than to have to explain the differences between USB 1.x and USB 2.0, which can handle speeds of 480Mbps.

All fine so far. I bought the firewire card and cable and installed it into my computer. Now here's where things get weird. Firewire does not allow you to do #2 and #3 above. So now I need to actually use both the firewire and the USB cables in the computer if I want to do all three operations. How weird. Maybe Sony is trying to transition off Firewire in favor of USB?

Whatever the reason, it's a very awkward customer experience, and I personally think they should stick with USB only and suck up having to explain USB 1.x vs. USB 2.0. I would think they'd enhance the customer experience while reducing both manufacturing and support costs.