Notes on Toolbar 6

I tried the quick search in Google Toolbar 6.

Typed "chrome":

#1 result: Search google for chrome

#2 result: [Run] Google Chrome

Typed "photoshop":

#1 result: Search google for photoshop

#2 result: Run Photoshop Lightroom

#3 result: Run Photoshop

So I know Google wants me to search, but it's a bit much, and the sorting isn't very smart. I would type "Lightroom" if I wanted Lightroom, after all.

I expected a more Quicksilver/Spotlight thing, so if I type in a file where I have a local app in my start menu by the exact same name, maybe running it could be done by hitting "enter" rather than "down arrow once or twice or three times, then enter".

Maybe I don't really need another way to search the web. I have a browser for that. This toolbar search is fast, and it could be useful if it really did what I wanted, instead of what it does. As it is, I'll probably uninstall it.

Why you shouldn't look at screensavers

16 hours/day.

70W (video card 30W + monitor 40W).

= $4 /month.

And you probably don't look at it that much.

Use power saving. It's the best screensaver. 

$0.01/month. Energy Star.

Property vs. Contract Rights

To me, the End Days Battle (i.e. where things really go in our economy) seems to be a fight between Contract Law and Property Rights.

Bankruptcy establishes a pretty clear hierarchy: 

Contracts lose to Property. Your creditors get to claim your assets, regardless of the weird contracts you wrote that obligate you to pay all your assets to your nephew.

And I must point out, that in this regard, CDS is just a Contract. It's not Property. And as we've written a lot of bad contracts in the last five years that nobody can pay for, at some point somebody's going to blow up a lot of them. 

I bet that even nationalization can't pay for some of this stuff. Didn't work for Iceland.

Really, some of the contracts just have to be torn up.

Vista vs. Leopard

Microsoft shipped Windows Vista in January, 2007 (or March, depending on how you count.)

Apple shipped Leopard in October, 2007.

So Vista's been out 6 months longer than Leopard.

Stats for f.lux users:

88% of mac f.lux users are using Leopard

27% of windows f.lux users are using Vista. (Most of the remainder are on XP.)

For a paid upgrade to convert at that rate is amazing. Apple is just kicking butt.

[This is just data from Google Analytics. f.lux isn't phoning home with any of the above.]

An open translation service for software?

When you read about the magnitude of trade imbalances in this country, you understand that a lot of what we've been doing in the US is this: we make services and products for people in the USA, and we borrow money from people overseas to pay for it. Everyone's finally noticing that this doesn't make much sense in the long term.

One of the things I learned at Google is that the world is a really big place, and people in other countries like software and services, and we could actually be a very big exporter of software, not just 10 big companies or 8 countries, but even startups...and to 100 countries.

But, translation and localization are expensive things (I read that some big companies will spend nearly $1M to do professional translations in all their languages for one product, with documentation, etc.). There's enough effort required that especially small companies don't do the work in their first year or two on the market.

But--what if the translation part were free? What if there were an online service that would translate all your strings instantly and with decent quality? I think companies like Google might be able to sponsor such a thing, and it might have far-reaching impact for commerce. If we could export (and import) just about everything we make online, it would be a pretty amazing world.

And before solving the general translation problem, I wonder in particular about catalyzing some effort around translations for software, because it's a constrained problem, and there's lots of repitition...once you've translated "Are you sure you want to delete this file?" to 200 languages, you don't have to do it again. Everyone could use it.

Comment here if you have thoughts or a way to make this closer to reality.

Introducing f.lux

We're all in favor of more laptop use by candlelight. Well, Lorna and I are up late a lot using our computers, and we noticed that the monitors looked awful after dark. So a few months ago, we began our mission to fix that.

Your monitor looks like daylight (6500K). This is just fine during the day, but it looks pretty bad at night.

F.lux fixes this. Tell it where you are, and it makes everything better.

F.lux defaults to match halogen lights at night (3400K). Our alpha testers tell us that shutting this off feels like a "punch in the face", so be careful out there.

It works on PC and Mac. You won't notice anything until after dark.

Enjoy!

http://www.stereopsis.com/flux/

Windows 7 on Virtual PC: 21 dead per day

I installed Windows 7 on a Virtual PC on a 3GHz E8400. While this is definitely not "real" hardware (Virtual PC is definitely slow), it's interesting to notice a few things about performance.

1. Opening the Control Panel under Windows 7 takes about 2 seconds. Clicking an applet leaves you with a busy cursor for a second. These times aren't awful (my XP box takes about this long), but it's a notable contrast to the Mac, where the System Preferences is really fun and instantaneous. There seems to be a lot of "web-like" asynchronous UI in Vista and Win7, where you kinda get a frame, but then wait for it to fill in. These 1-2 second pauses make everything seem slow.

2. Windows 7 copied OS X's "keep this program in the dock" (they call it "pinning") very nicely. A migration from quick launch to this new thing would be nice, though. The mouseover screenshots that show you what's going on before you click don't seem to be implemented so well.

3. Being a long-time XP user, I still find the Vista/7 start menu confusing. Suppose you want to do start->run->notepad. You're actually "searching" for notepad, which rather than being instant, takes a second. But the result appears halfway across the screen from where you actually typed your search.

4. Libraries. Microsoft's brought back some of the Vista beta features and shows composite views of your Pictures, Music, Videos. Seems nicely done, but again asynchronous. For my collections it can take 5-10 seconds of waiting before I see content. During this time, the content pane fills up slowly, but mostly doesn't tell me it's working to find more stuff. I guess I don't want to replace synchronous folder views with 10 second async search views. Frustrating.

I'm impressed, as I've said, with Apple's Tiger to Leopard upgrade. Mail searches and Spotlight are now instant rather than taking a few seconds as before. Microsoft seems to be making some fixes speed-wise, but this core idea that it's okay to make me wait 1-5 seconds when I click, is really wrong. Seriously, this is why the web is winning, because Google cares about wringing out every 10ms.

I think it's a Steve Jobs quote: "How many people did you kill today?"

If Windows 7 and Vista waste 100 seconds a day, and there are a half-billion Windows users, that means 21 lives wasted per day.

Too many, I think.

JPEG XR nearing standardization

Per the JPEG press release, JPEG XR is nearing the final stages.

What is JPEG XR? It's HD Photo, from Microsoft, now on its third renaming. Likely this standardization process will put all patent worries to rest.

I don't see much mention recently of T.851, the simple improvement to JPEG, with better lossless compression than Huffman, and 16-bit support.

Things I like about HDP/JPEG-XR: high bit-depth support, streaming all the way to lossless, fast decompression. Things I don't: quality.

Sachin Garg has done a reasonably good study of this, and at typical bitrates (say, 0.5-1.5bpp), JPEG actually beats HDP/JPEG-XR: http://www.imagecompression.info/lossy/ and it would do even better with T.851.

Revisit: holding my data hostage

I wrote an article 8 years ago after a frustrating time with some backup software: Holding my Data Hostage.

The idea at the time was that subscriptions ('software as a service') left user data at the whim of service providers, that they could impose arbitrary fees for access to your data. I had in mind a subscription version of Microsoft Office, and annual payments for access. In contrast, I liked the idea that I could "buy" a perpetual license to access my data, which is the model desktop software followed at the time. With some work, you could more or less move your software from PC to PC, and your data with it.

But another thing has happened in the interim. PCs are cheaper. The Mac, iPhone, and web services are much more popular. And the problem is related but different. The industry's challenge has been clear for 5 years or more: make it easy to move user data across devices, laptops, desktops, smartphones, etc. And I believe the desktop software industry, by not adapting at all, puts itself at considerable risk. (Think "record industry" and "selling CDs".)

Since 2001, software "activation" has become prevalent on the desktop, with Microsoft especially (and Adobe to a lesser extent) making it punishingly hard to move your data and licenses between PCs. (Think "DRM" and "MP3s".) I neglected to get an OEM Office with my new laptop, and have since found that it can cost up to $400 to get a copy. It's a pain to move Photoshop to a new desktop, because I have to somehow boot the old one to unauthorize it, even though the hard drive doesn't work anymore.

In the era of $299 netbooks, $400 is quite a tax for simple access to your data. But my laptop is missing a $400 copy of Office, because it doesn't make sense to buy it for a computer I use 10% of the time.

Currently Google Docs (and Zoho, etc.) provide enough functionality that the basic uses are available online (for free, rather than for $400), and for more serious work, I can use my "main" desktop. But what a sorry state of things!

It's clear that software delivery today is broken. I can't get the experience I want, everywhere, because I want to spend about the same amount of money I used to on Microsoft's products, not 3x as much. I think they should be driving their revenue growth by making my life more convenient, not less.

And I think it's actually fair to say that Microsoft's monopoly in Windows and Office is making a smooth transition to the "right model" more difficult than it should be.

Microsoft won't upgrade your XP system to Windows 7. Settings and third-party apps must be re-installed. (It usually takes me a month to do this.) 

Apple will do a great job upgrading your Mac, in an hour. I just upgraded Tiger to Leopard, and everything works. But Microsoft makes a huge amount of money from new computer purchases, so it's incredibly hard for me to understand why they make upgrading so hard. If I could replace my PC every 18 months instead of every three years, I probably would.

My wife has a 3-year-old PC, and she can't even upgrade the parts, because her XP license considers a new motherboard an invalidation of the license. She won't lose her apps and settings if she keeps the old desktop, so she does.

Here are my solutions (we'll call this the "How to put the Cloud on the Desktop"):

1. Extend OEM licenses (purchased with a PC) to "guest" licenses on other PCs. Automatically. Make a 3-year time limit (after which you can only use the original PC without paying more).
2. Replace activation with authentication. iTunes just works better than Windows activation tied to a specific 3-year-old motherboard. 
3. Make licenses transferrable and renewable for a small fee when you sell a PC. ($10 gets you a wiped PC, with secure delete of your old data.)
4. Deliver software through the cloud, automatically, when you login.
5. Make upgrades from PC to PC (with legacy settings preserved) easy.
6. Make data sync to the cloud and other PCs.
   
7. Make settings sync to the cloud and other PCs.

The demands of data movement and a focus on user data as the most valuable commodity will actually drive this change, and it seems like it's going to happen fast. But the analogy with the record industry is clear. 

The cloud is going to be disruptive to the desktop software industry, and mostly for the wrong reasons. The software is still valuable, but the delivery must change. Failure to adapt to it could displace some major players, if they do not make accommodations to making peoples' lives easier, instead of burying their heads in the status quo.

 I do hope it's not another 8 years.