Author: mike

Over the last few days I’ve been talking to a few folks about the Dual Core chips and how much power they can draw. High end systems are drawing 150-200Watts each. That’s pretty staggering, and the dual core processors are drawing less than the previous generation of Pentium chips.

Today, the NYTimes writes about a Google whitepaper being presented by Google today for improving PC energy efficiency. It may not matter much for a PC or two at home, but for a server farm, which can use 10,000 or more computers, it’s a really big deal. Further, while your home PC is usually idle, the goal of a server machine is to actually have it under fair load 24×7. And more load invariably means more power draw.

To put it into perspective, if you had a datacenter with 10,000 CPUs (and no overhead equipment), and you paid $0.1 per KWh, you’d be facing an electric bill of about $108,000 per month. Again, according to the NYTimes, Google has around 450,000 machines. If they used these expensive chips, they’d be churning through $4.8M per month just for electricity. No wonder they wrote this whitepaper. No wonder they are moving out of California.

Just wait until the Tesla and other electric vehicles appear. We’re worried about oil prices now, but electricity demand is on the brink of massive increases.

See also: eWeek article, cNet “Power could cost more than servers”

MSN has it’s own YouTube clone coming out called Soapbox. I haven’t been able to try it myself, but I thought that B2Day had a great comment about these types of services,

Microsoft will build the technology. But it will be a struggle for technology alone to overcome the social phenomenon of YouTube.

I hear a lot of people complain that Microsoft isn’t innovative. When you try to point out innovations that Microsoft has made, they claim that all the “good innovation” is via acquisition. Even if that were true, does it matter?

The nemesis of any startup is getting good distribution. It’s incredibly expensive and difficult to build great distribution channels. Microsoft, by contrast, has distribution like no other. If Microsoft buys a great product and gives it distribution, why is it that Microsoft shouldn’t get credit for innovation? The fact is that the startup couldn’t get the distribution on their own. Microsoft provided it by buying the company and using it’s distribution. Isn’t that a form of innovation in it’s own right?
Microsoft deserves full credit.

Marc Orchant compares Office 2007 to driving a new rental car. I think it’s got some truth to it. I think the analogy can be extended in that you only rent a car when you don’t already have a car (e.g. Office 2003). 🙂

I actually mostly like the new UI. It looks clean, and it is not hard to use. But a UI doesn’t compel me to upgrade. I don’t really need prettier menus. I don’t even think I need updates to word or excel or even powerpoint. (Perhaps I lack vision on those products!)

What I do need is better email. Why is it that when I add a conference room to a meeting invite in Outlook, I have to manually mark it as a “resource”, or it won’t actually book the conference room? (So many people have experienced this – it could easily be comedy material for the Office Space sequel). Or, if I send someone my phone number, why doesn’t it automatically add my phone to the contact info? Or what about automatically showing me map locations for addresses? Or adding appointments to calendar automatically? Why can’t my wife schedule time on my calendar at work?

Lastly, the .docx filetypes make Word 2007 cumbersome to use for collaboration. Sure, it is an open xml format, which is a great step forward from proprietary formats. So, it is with some reservation that I complain about this. But, if you have Word 2007, and your colleague has Word 2003, I guarantee you’ll accidentally send a .docx file which he won’t be able to read. To solve this, he can install a .docx reader for his older version of Office, but more likely you’ll end up resending the file after manually converting to a .doc file for him. It doesn’t seem right that the default course of action is that you send a document that someone else can’t read. Why doesn’t Outlook figure that out before you send? After all, you probably sent email to this colleague before, and Outlook puts its version number into every mail header, so your Outlook can know what version of Office your colleague has. Outlook could tell you, before you send it, that you are sending a Word-2007 specific file to a user that probably only has Word 2003, and automatically convert it. Now that would be pretty smart! I guess I wish the default were to save .doc, and not .docx. There’s probably an option in there for that somewhere. But the ribbon has so many buttons on it, I’m not sure quite where to look!

On the good news front, Office 2007 does successfully tackle my #1 pet peeve with Outlook: fast email search. That is the best reason to upgrade.

I registered a new domain name recently, and finally brought it online. Not much to see yet, but you can find it here: http://www.sendcarefully.com/

I stumbled across Dapper today, a pretty phenominal UI & utility that can create sophisticated screen scrapes from any website.  I have two reactions:

1) WOW!  These guys did a great job of making it easy to scrape!  Fantastic interface, boiled down to very simple tasks!

2) Jesus Christ!  Who in their right mind would try to build anything off an unknown screen scraper?

The latter is really where I finished thinking about it.  It seems that in our Web-2.0 world, people think that somehow hacks can sustain.  Hacks can’t.  Hacks are hacks.  And fundamentally, screen scraping is a hack.  One small div-hierarchy change and the whole thing breaks.  One slight UI one-off, and the data provided is bogus.  I usually try to avoid being a purist, but screen scraping is just one approach I can’t support.

Conclusion: Dapper represents a lot of very nice work behind what is ultimately a futile effort.