Songza

Songza puts every other online music service to shame. It’s of questionable legal nature, but poses no risk to end users. Songza indexes the web for unprotected MP3s and lets the user play them in an extremely elegant flash interface. It’s not possible to download the music, but it’s great when you want to play a particular song, or demo a song in it’s entirety before buying from iTunes or the like. Perfect for finding that one song to play during the party also.

TripIt

TripIt lets you forward any confirmation email from dozens of travel sites and automatically extracts all the important elements and organizes it neatly. For someone who travels a lot, it’s invaluable — lets you forward airline, rental car, and hotel confirmations and have it organized by date elegantly.

Earlier today, Firefox was not rendering a page correctly, so I switched over to Internet Explorer hoping it would. As I did so, something dawned upon me: I haven’t seen an online advertisement in over three years, let alone clicked one.

For example, one website that I frequent, GigaOm.com has a front page that is riddled with advertisements when opened in Internet Explorer, but none in Firefox. The reason? AdBlock Plus. AdBlock is an extension for Firefox that dynamically blocks advertisements using wildcard filters, etc. For example, three filters that were automatically applied to block all ad’s on GigaOm.com were “/magic-ads/*”, “.googleadservices.”, “.fmpub.net”.

GigaOm — Internet Explorer:

GigaOm — Firefox with AdBlock Plus:

Similarly on TechCrunch.com, a site I frequent more than 3-4 times a day, the filters that dynimcally blocked advertisements were, “.snap.com/$script”, “/ads/*”, “/openads/*?”, “.fmpub.net/”, and “http://ads.”

TechCrunch — Internet Explorer:

TechCrunch — Firefox with AdBlock Plus:

Similarly, NYTimes.com:

NYTimes — Internet Explorer:

NYTimes — Firefox with AdBlock Plus:

As AdBlock becomes more and more popular, the return that advertisers are receiving for their money is falling. Even though publishers will tout subscription numbers and daily uniques, the actual number of users that see the advertising is impossible to gage because AdBlock Plus blocks advertisements after the page has rendered, thus still loading the content, but just preventing it from displaying.

So back to the title? How can Microsoft Kill Google, and also entirely shake up how the internet works? Build AdBlock functionality into Internet Explorer 8. Sure it’d lose advertising revenue from Live.com, Hotmail.com, but is that Microsoft’s core business?

Microsoft has one advantage over Google that people always forget about. They make real products that people pay real money for. They aren’t resellers or organizers of information. But rather, they produce and sell products.

Kill online advertising and power returns back to the people who produce products that people demand. Game over. Checkmate.

I’m not sure how I feel about it yet. I like the lack of clutter, but for people with relatively tiny profiles and no applications, it feels like there is too much white space.

Screen shots are below:

Remarkable new search demo website called Powerset released today. It’s still in beta, and only indexes one website, Wikipedia, today. As they get more funding they will add more servers and computing power to index more websites.

Instead of doing simple keyword matches like Google, Microsoft, and Yahoo do today, the engine tries understands what you are trying to ask and does contextual searches.

Ex) Do a search for “when did earthquakes hit Tokyo?” and compare the results to Google’s results for the similar query. It takes synonyms for the word “hit” and looks up words like “struck” and “strike”.

Ex) Do a search for “politicians who died in office” and compare the results.

It’s really remarkable, even though Powerset currently only indexes one massive website.

This technology will certainly disrupt the status quo and has the capability to do what Google did to the internet nearly ten years ago. This company in the next year will get bought out for $1 billion+ by Google/Microsoft/Yahoo or remain independent and seriously garner more and more attention.

It’s the great thing about technology, competition, and entrepreneurship. A company with 15 employees and less than $20 million in funding released technology that will shake Google with its $180 billion dollar market cap

iTunes and iPods/iPhones have the ability to remember where in the book you stopped should you pause or disconnect your iPod. This is tremendously useful because I drive different cars and scrolling through a 54 hour long audio book to find my spot is a pain. The catch is this functionality only works on audio books that are in a particular format that only Apple or Audible.com support.

Well I get my music from a variety of sources, much of it in MP3 format from rented audio books. iPods don’t remember where you stopped on an MP3 file, only on the M4B format. I did some Googling and as recommended by Michael Alderete I downloaded and used MP3 to iPod Audio Book Converter. It converts MP3 audiobooks to the M4B format, can combine multiple MP3 files into a single file, and is free. Best of all, it works wonderfully.

Technical note: While my new photo gallery still looks similar, over 60% of it is completely rewritten. More efficient coding, pagination, gzip compression, URL rewriting, object oriented coding, brand new libraries, inline tagging, single page rebuilds, and a shiny new admin interface. flickrStash, my application that powers this site will be available for purchase October 16th. For more info contact me.

There are two potential scenarios that will result from this update.

One possibility is that this new update may drastically reduce the number of page views. Since Facebook’s main source of revenue is advertising, they rely on traffic which is measured in page-hits and unique visitors. While the number of unique visitors is not likely to drop, the number of page-hits may drop since people may sign on the homepage, read the updates, and sign off. They will have no need to page hop and view all their friends’ profiles to check for updates.

The other possibility, which Facebook sees more probable, is that users visit each individual profile to view what updates were made. They are banking on user interest in knowing what friend A wrote on friend B’s wall. This may be true, may not be.

My prediction: Mini-feeds will stay but within the next month we will see options that allow users to unclutter their homepage. We will see options to hide certain types of information from the front page (such as hide ‘Group Information’) or even features that allow users to customize what information they are interested in on a per-user basis. Users may be interested knowing who wrote on their significant other’s wall or who recently befriended them, but may not be interested in updates from friends back in high school.

However, don’t ever expect the Facebook to stop collecting this data. This relational data is what makes Facebook worth the billions it is today. Their revenue stream doesn’t rationalize their market value, only the information they currently possess.

Contrary to popular belief the world will go on, Facebook will still grow, and the majority of users will start to like the mini-feed. I know I do.

But what’s really cool is not the design or features, but the backend. Everyone who follows technology knows that flickr is perhaps the best photo sharing site out there. Not only are their prices a steal, but they have a phenomenal API which allows outside access of images and information. Additionally, there is no limit image size or storage space which is great for people like me with over 16 gigabytes of images.

While the API is great for few images or extracting a minimal amount of data, when you try to show the amount of data my gallery does you need 5/6 API calls really slowing down the browsing process. Additionally, your images are dependent on flickr being online, so if flickr goes down, so do your pictures which can be frustrating at times.

Well my solution is caching. Caching everything. I created an interface which builds my galleries, caches my images, my tags, into an MySQL database. Additionally, the system is two way, so if I add a tag locally to an image it automatically gets added to flickr’s database.

What are the advantages? First the most obvious is speed and reliability. Browsing is definitely faster then a fully API solution and my site isn’t dependent on flickr. The second advantage is my viewers have access to full size images on flickr should they want them. I don’t have the means or storage space to store 16 gigabytes of images on my own server but using flickr certainly allows me that flexibility. Finally, having a locally cached version allows me to add functionality which flickr doesn’t yet support (tagging people in the photography).

Is this a violation of the flickr TOS? No, certainly not because in normal browsing there is no load on flickr’s servers. No API calls nor image source includes. Everything is locally cached and loaded from my own server.

So enjoy the pictures. Check to see how many pictures of you are tagged! Digging through many albums to find pictures of yourself or that hot girl you forced to take a picture with you during Halloween is a lot of work. This makes it simpler.

If there is sufficient interest from developers for my code, I may fine tune it and consider releasing it to the open source community.

Using flickr’s API it’s really simple to achieve what I’m trying to do. I simply set the taken date as the uploaded date and all is good. Sure, it might appear to the astute viewer that you simultaneously took and added the picture to flickr, but who really cares?

My class of choice to work the API is phpflickr, a class written Dan Coulter. Obviously replace the keys with your values. I’m not going to explain how to use phpFlickr, but if you need help setting it up, read the official documentation.

Code is avaliable here.

Previously, I used to run sub domains for my various projects (such as poker.askthefool.com) but a few weeks ago I combined all of them and made them subsections of this personal site. Now again, I’m partially regressing and separating my personal website from these applications. I almost felt claustrophobic merging the applications and experiments with weblog content; so now, codeispoetry.org has been started.

Currently it only has several applications. My recent State of the Union database, Sudoku solver, and various poker applications. Expect more in the coming months.

This stuff isn’t terribly useful, but some interesting things can be seen. For example, the Flesch-Kincaid grade level of the address has been going down. There are multiple explanations for this. The first being presidents are getting dumber or the second being that as more people are becoming involved in politics, politicians are trying to lower the difficulty of the material. In 1790 a few hundred people watched the State of the Union and yesterday there were 62 million viewers.

Play around with it when you have some free time. The site is located here.

There was Sudoku puzzle in the back of Delta’s Fly Magazine on the return trip home yesterday and I decided to program a solver to help me solve it rather then solve it myself. The exercise turned out to be much simpler then I expected. A web version of my solver is available here and Sudoku puzzles are available here.

Check it out, and if you go to USC, consider buying one or four.

Also, you may notice the diffrent approach to the home page. By default I’m listing excepts from the last 2 entries. Underneath there are links to the last 10 posts. The new addition is the “Recently Noted” section. This is just a compliation of links that I add while surfing the internet. If something catches my eye as cool, I’ll post it there.

I’m going to keep this thread active for all redesigns. That way I don’t clutter up real content. Feel free to leave comments.

What users tend to ignore is that the people that create these problems are also rational people. Why design viruses and spy ware for a small percentage of the market, when you can design it to attack a large percentage of the market?

While it’s easy to criticize Microsoft’s Internet Explorer and start using Mozilla Firefox, you have to remember that Internet Explorer is over 5 years old and Firefox is less then a few months old and constantly changing. Firefox is a lot newer and controls a relatively insignificant portion of the market which is why we haven’t started seeing Firefox related security issues.

This is not to say that every user should embrace everything that is Microsoft. Users, just as software developers and hackers, are selfish rational people and should embrace whatever product does the most for them. Whether today it is Firefox or tomorrow Internet Explorer 7, realize there are outside influences affecting the success of every product.

It’s common to be the punching bag for an industry when you control the industry.