Author: mike

My first Google Search today:

The problems are many:

1) The results are not relevant

The first result is an article titled, “Present Status and Marketing Prospects of the Emerging Hybrid-Electric and Diesel Technologies…” isn’t an exact match. Sure, there is “hybrid” in the title, but this is absolutely not a product comparison.

The second result isn’t really about hybrid cars at all. Its something about the Department of Public Works Garage Vehicle Acquisition Policy.

The third result, and probably the best of the bunch, talks about Hybrids in general, but again is not a comparison.

2) The results don’t work!

Clicking on the first article lands me at a Google error page stating, “Your search – author:”Burke” intitle:”Present Status and Marketing Prospects of the Emerging Hybrid-Electric and Diesel Technologies t” – did not match any articles. ”

The other two articles are PDF, which is just hard to deal with, doesn’t fit within the screen size appropriately, and is difficult to search.

3) Google’s native results are better

The regular old web search engines include reviews from Motor Trend, Edmunds, etc. Much much better.

Conclusion
This is sort of like Donald Trump’s Apprentice with “Street Smarts vs Book Smarts”. The fact is, that Google results are best served by real web pages rather than this scholarly stuff.

It will be interesting to see how long Scholar stays up top. From my perspective, Google just broke the relevance of their top-placement result by putting this in at all. Why not just let the scholarly articles sort to the top like any other web page? If they are linked to a lot, and have the right keywords, promote them. But why artificially boost junk?

This morning during my usual drive into work I had an idea. Its probably a horrible idea, but I find myself enamored with it enough for the moment that I’ll post it here.

Tags are pretty cool. For those that don’t know, “tags” are an organizational mechanism where multiple users can “tag” items with keywords which are important to them. For example, I may look at a picture and tag it with “yosemite”, while someone else may tag it as “camping”. Both are relevant, and both can be used for others to find the picture in a meaningful way later. Using both tags (and then weighting the most frequent tags) makes for a dynamic organizational tool where each user “votes” with their tags for how to organize. I’m probably fumbling on the exact definition, but thats some of it.

Some site which use tags well include del.icio.us, rojo, and gmail. It should be noted that traditional apps like Outlook have really had tagging for years, although they’ve buried them in the user interface under “categories”, and there is no way for multiple users to contribute.

Idea
So, what if in email, we had a way to associate tags with an email? Each user could “tag” the email, which would send an additional email to the other recipients of the email which would just contain the tag itself. Their mail clients would absorb these tags and automatically add the tag to the user’s database.

This could be acheived using a simple X-header, perhaps
X-Tagging: foo

Why?
This would work fantastically well inside the corporation. A large email thread might be about the “3.0 release”, about a feature called “funk”, and also about the “schedule”. Each of the recipients of the thread would receive it, and tag it as appropriate for themselves. However, all of the other recipients of the thread would also have *their* emails tagged with the other user’s tags. Now, any of the folks could search or pivot-view their email with the distributed tags.

I’m sure this idea is not new.

A few weeks ago, I started forwarding a big chunk of my mail to gmail. I just needed a web-based email system so that I could read it remotely better.

Overall, I like it quite a bit. I think there are a few UI nits, like the reply button being surprisingly well hidden, but its a good product. I even get a lot of spam, and Gmail seems to have the best spam filterer I’ve seen. It blocks all the bad stuff.

But today, I discovered a real problem. I’ve been suspicious for a few days that I wasn’t getting all my email. I just hadn’t heard from people that I knew were writing. At first I thought my forwarding was messed up (how could gmail miss this badly, right?) But today I discovered that its gmail falsely accusing my friends as being spammers.

I took a look at some of the emails that were filtered by gmail, and they look pretty innocuous. And they are from my friends/family directly – not from mailing lists or the like.

Beta.

Just last week I finally got around to writing about Google Earth, and then what should happen but my own company ships MSN Virtual Earth!

Like the Google product, its a fabulous piece of software.

The MSN product, ironically, is a web product. While the Google product is a client that you have to download to your PC. (shouldn’t that be the other way around?) Personally, I prefer the web-based product. It makes it a whole lot more useful, as it just runs everywhere. The Google product was unable to run on my PC at home because my graphics card was too old.

The MSN product is also just a lot more thought out in terms of usability and also for being useful for really finding things. To use it, first, put in the addresses that are relevant to you (home/work/whatever). Then, just type what you are looking for, like “pizza”. Bing- you can see all the pizza joints near your addresses. And its really smart too – it remembers all your locations so that when you come back later, you don’t have to retype your address!

The MSN product does lack the whizzy “fly” feature, which is no doubt why the Google product requires such a heavy duty graphics card. So, I miss flying. But when it comes down to it, flying is neat, but again doesn’t help you do anything really that useful. The MSN map is essentially a “flat” earth, while Google is rounded. I’d expect MSN to improve on this in the future to give you more and better options.

Anyway, both of these products are really really cool. These are going to change the way we do address lookups. Both still lack decent driving directions, as far as I can tell, so Yahoo! is still in business for a while.

If you haven’t tried Google Earth, you should do it right now. I know I’m late to blog about this – as the product has been available for quite a while. But I simply can’t remember the last time I saw a new application that was so cool. Seriously, if you haven’t tried it – go try it right now!

You can fly around the globe – downloading real satelite images of everywhere you go. Lookup your childhood home, your college, or whatever you want to see. It is really impressive.

MSN has its own MSN Virtual Earth coming soon (as announced widely). But its not available yet. So, until then, Google Earth is pretty darned neat.

About a year ago, I bought a small domain for personal use. I registered it through a small ISP, who did the domain registration for an additional $5 or something cheap.

About 2 weeks ago, I received a letter in the mail from Domain Registry of America, saying that it was time to renew my domain. I want to keep it, so I filled in their web form for an additional 3 years. I knew it was more expensive than others, but I didn’t want to go through the headache of changing it.

About an hour later, I received a phone call from my credit card company stating that I had suspcious activity on my card – it was domain registry of america. I told them I authorized it and didn’t know quite what to make of the fact that this small transaction had been noticed by them.

But the next day, I received email from Domain Registry of America stating that I had to do more work to transfer my domain to them! ACK! Its at this point I realized I had been snookered. DROA was not my registrar at all! They were just some company that looked up my domain, noticed it would expire soon, and decided to try to steal my business. They flagrantly made their mail look like it was from my current provider rather than being up front about switching to them. Boy, did I feel dumb!

Searching on the net now, I realize there are many resources which also show them to be a fraudulent scam. They send out mail to everyone pretending to be your domain registry service, and try to just steal as much business as they can get:
From Domain Avenue

Pixagogo has pictures of the mail the send

And another one.

To their credit, I called up DROA and told them I felt duped and that I wanted my money back. Today, they refunded my credit card, and I’m back on my merry way. (I plan to renew through GoDaddy, who I’ve always liked quite a lot)

With Lookout we spent a ton of time just figuring out what is the “right” way to extend Outlook with a .NET based addin. This was a time consuming process because there was a lot of misinformation out there, and documentation was sparse. So, it was a lot of trial-and-error and head-banging which finally got the product out.

This week, however, the MSDN/Visual Studio teams published a new product called the Microsoft Visual Studio 2005 Tools for the Microsoft Office System – Outlook (beta). Wow! That is quite a name!

Overall, this tool looks great! It makes a lot of things really easy, which Lookout had to stumble through. Here are some of them:

Creates the Shim For you
This is a great feature because your addin will no longer need to be loaded by .NET’s mscoree.dll. If you load via the “old way” (with mscoree), then your plugin inherently cannot be run in high-security systems, and mscoree is a generic loader and is not signed. By having a shim, you can sign the shim, and then be securely loaded into Outlook. A great whitepaper was published a while back on how to do this, but the new VSTO tools now do it for you for free.

Creates a Separate AppDomain for your Addin
This is a huge new feature. By creating your addin in a separate AppDomain, it is much less likely to have conflicts with other plugins loaded into Outlook. I can’t tell you how many times different plugins that didn’t properly implement the ReleaseComObject logic hosed Lookout and we had to take support calls. With AppDomains in place, these should be a thing of the past.

Handling of Outlook Shutdown Cases
Getting Outlook to shutdown when plugins are loaded can be tricky. Making it work in 3 versions of Outlook (2000, XP, and 2003) is a process of walking through a minefield of random bugs. Fortunately, Outlook 2003 works reasonably well, but there are still a few well-documented gotchas. The new VSTO IStartup interface completely unloads the AppDomain in the shutdown cases, which should make addins no longer need to hand-craft these solutions.

Overall, kudos and thank you to the VSTO team. They didn’t have to build these helpers and tools, but by doing so they will enable a fleet of new applications that can be much more robust and interoperable. This should be a great thing if you are interested in .NET-based Outlook development.