Blogs

Mobile devices from Apple in the future may have the ability to draw power from the sun for longer battery lives. This revelation came to light today as people at Trade the News (a better article is here) noticed that recent patent applications from Apple included technology to harness sunlight.

While it may be a ways off in the future – photovoltaic cells beneath the touch-screen could add hours to battery lives – making devices like the iPhone much more practical for business users – who have been complaining about the limitations of the small battery for users who are constantly on the go. This is one of the major reasons consumers won’t be leaving Blackberry anytime soon for business communication.

Of course the bigger story here is that with solar energy and using other types of micro-energy devices we could one-day take our mobile devices completely off-grid. It has some appeal, if you can get over the idea that you’ll never again have the excuse ’sorry my iPhone battery was dead.. wasnt receiving any calls’.

All you designers out there.. if you ever were looking for a font that was clear even at very small PT’s, this is for you. Over at WebSiteTips.com, I found a list of pixel fonts – many of them free! Check them out:

http://websitetips.com/fonts/pixel/#pixelfonts

I wrote a fairly basic but handy JavaScript Object Inspector and Logger that works across different browsers. I call it “NitobiBug“.

Read all about it’s features here. I did a video tour also, which you can see here (turn down your volume – its loud!).

Check out the live demo here.

Essentially, what it does is provide a logging utility like Firebug’s console.log that properly inspects objects and shows you it’s members. If you log errors it formats them nicely too. If you inspect DOM elements with it, it attempts to show you where on the page they are and calculate their widths and heights and positions on the page. You can resize and drag NitobiBug around the page, and it tries to remember where you put it.

I use it all the time while I’m working on RobotReplay so I figured maybe other people would too. It’s certainly not the only such tool out there but I think it’s decent. Anyway, your comments are welcome!

There was an interesting post today on the Gmail blog about how they got GMAIL to load faster, especially on slow connections.

A couple of the things they did were:

• Spriting – combining all the small graphics into a single image and using CSS background image placement to split out the individual icons.
• Request pooling – Combining together seperate XHR requests into a single large request and then parsing out the results.
• Cacheing – making more of the resource requests cacheable by the browser (JavaScript, CSS etc). That way when the user reloads the page, it doesn’t need to download the same resources over and over.

To help them with this task they used proxy trace tools like Fiddler, WireShark, and HTTPWatch. I think Firebug would have been a good option here too. Other things they could have done (and may have) are:

• GZip compression – compressing all your static resources with ZIP making them much smaller to transmit. This works on all major browsers these days.
• Conbining JS and CSS resources – By concatenating all your JavaScript and CSS resources into one, you reduce the number of requests needed and really speed up your site.

I felt lucky to get on the Viewzi preview invite list. You can too if it hasn’t launched by the time you read this. Viewzi is a search aggregator that tries to present search results in an innovative and user-friendly way. This is achieves quite well, in my opinion, having studied a few other attempts at this over the years (snap.com, nexplore.com). Here is a summary of my experience on Viewzi today.

Launch Page

Nice and simple. I like the invitation to watch the training video. However, then I went to see it in Internet Explorer 8 (running in IE7 mode) and the whole thing went to hell (see screenshot below)

I was also treated to a JavaScript error. Next, I was curious about the footprint of this launch page. I opened up FireBug and watched the download. This page is 113kb, which in my opinion is too much for a search launch page. I recognize that nobody on a dial-up connection would ever use this site to begin with – fair enough, but under high-load conditions this is going to be an expensive page to serve and probably a slow page to download. Certainly when compared to the 12KB of utilitarian sparseness of Google.com. Anyway, the page did in fact come up very quickly for me so I probably shouldnt complain.

Test-Search “U2″

I tried searching for the band “U2″ and was presented with this results-browsing view. First off – it looks great, and the UI is really smooth and intuitive. However, this was not a search-results page. I think I should have been shown search results right away – as a jumping-off point for browsing these other views. Note: If you DO click on a search results view, further searches are immediately presented in this view.

In general I was impressed with the overall speed of everything. Search aggregators have a rep for being sluggish. I didn’t get that impression here.

Next I started exploriing the different views. Since U2 is a band, I was curious what results the MP3 view would produce:

These were all U2 songs and I could play then directly from the viewzi window. Nice! The other day myself and Mike Han were talking about 90’s rock and we wanted to hear some Nirvana. This would have been great.

Next I clicked on the ‘celebrity photos’ view to see if there were any Bono mugshots.

No mugshots, but these were mostly all relevant. The question is what can I actually do with these results? Normally when I’m searching for images I want them to download. for use in some graphic I’m putting together. This isn’t the view for that, but fortunately there IS another photo view:

This is where Viewzi had search-relevance problems. None of these images were of U2. Oh well.

Back to the other results. Viewzi has traditional text-search results that aggregate Google and Yahoo (is that legal?). Anyway, they were spot on of course – and quite snappy. The other view that really caught my attention however was the Video search:

The video search aggregates a bunch of video services in a really cool browser that actually saves you a lot of time. To me this is one of the key strengths of a service like this.

Overall I’m really impressed with Viewzi. I think it had some search relevance issues with the images but I’m sure they’ll continue to work on that as they move towards release. I think what could really help Viewzi is if they in turn opened up their aggregation capabilities in the form of a set of API’s, and Widgets that other people can use on their sites in the way that Snap.com has done. I don’t think I’ll really switch over from google (not until they get their browser search widget to work) – but I’ll definitely be checking back to see how it evolves. In the meantime, I encourage you to check it out: viewzi.com

Ever wanted to HTML encode strings in Ruby? I did. I found this. It works great!

http://htmlentities.rubyforge.org/

To install just go:

gem install -r htmlentities

Then to use:

require ‘htmlentities’
coder = HTMLEntities.new
string = “<ã ©lan>”
coder.encode(string)              # => “<élan>”
coder.encode(string, :named)      # => “<élan>”
coder.encode(string, :decimal)Â Â Â Â # => “<ã ©lan>”
coder.encode(string, :hexadecimal) # => “<ã ©lan>“Â

Opera has released an alpha version of their answer to Firebug – DragonFly. Its a JavaScript/CSS/DOM/HTML/General debugger for your web apps. It has a lot of the features of Firebug and Drosera, but with some special sauce too.

The really cool thing about this tool is it assists with device-development too. Opera is the browser of choice on a lot of mobile devices, Nintendo Wii, and such. Firefly – er, DragonFly - helps us debug pages running natively ON those devices, which is an outstanding feature. I haven’t had occasion to try it out yet but will soon.