Fluidinfo and social tagging

Posted on April 3, 2011

For the past few weeks, I’ve been messing around with Fluidinfo as a result of the O’Reilly Fluidinfo Writable API Competition. My entry to the competition can be found tucked away over here. Fluidinfo was started in 2009, so it’s still early in what it plans to do, but I think the competition is one way for it to get the word out and raise its profile.

What is Fluidinfo? It’s sort of hard to describe, because it’s potentially different things to different people. One of the obvious comparisons is to Wikipedia, so I’ll use that as a starting point. Where Wikipedia has articles, Fluidinfo has objects…

Wikipedia has an article on George Orwell’s 1984. The article describes the background of the book, has information on the principal characters, the book’s themes and so on. The article is informative and people who read the article also have a chance to edit it, but that’s probably the extent of their interaction with the article.

Fluidinfo has an object on the same book. On this object, users can add tags they think are relevant, such as whether they read it or whether they thought it was any good. The object in Fluidinfo doesn’t have a synopsis of the book or information on the characters, but there’s nothing to stop someone from tagging the object with this information.

Fluidinfo is more or less built on the idea of shared data. Most companies build their own databases, which are walled gardens. These databases are their intellectual property and they can decide who has access to the data and what sort of limits they want to place on it. But with Fluidinfo, a bookstore might tag the 1984 object with how many copies they have in stock and have another tag that show which shelf it’s on. They could extract this data from Fluidinfo and send it to their website, making it useful for their customers. They could also extract the data to their inventory system, making it useful for their employees. They could also tag the object with the name of their supplier, making it more useful still. It’s not a walled garden because anyone can read the tags placed by other users. Someone browsing Fluidinfo can find out which users thought 1984 was a good book versus those who didn’t think highly of it, or they can find out if their local bookstore has copies of it.

While Fluidinfo’s objects are very different from Wikipedia’s articles, there are some other differences as well. The first difference is the scale or scope. Wikipedia has been around for years and has millions of articles. Fluidinfo has been around a couple of years and has nowhere near the number of objects (thousands, maybe tens of thousands). This is probably also tied to the number of users. Wikipedia has a very large and active user base and thousands of new articles are created every day. If I had to guess, I think the Fluidinfo user base might be a few dozen people at the moment – I don’t have this information, so I’m basing it one the number of usernames (namespaces) I’ve seen when browsing objects. Still, every venture starts small. Even Amazon and Google probably had a few dozen users at some point.

Another big difference between Wikipedia and Fluidinfo right now is how difficult it is to find things in Fluidinfo. Objects are actually identified by an alphanumeric sequence. What makes them easier to find is the about tag. When I was looking for the object for 1984 earlier, I did a query in the Fluidinfo Explorer, which was fluiddb/about matches “orwell”. It would have been easier if I could have just typed orwell or 1984. This is due to Fluidinfo’s alpha status. In their faq’s, they note that text matching isn’t available yet, but should be in the future.

Every object in Fluidinfo has an about tag. This is basically the “name” of the object. Fluidinfo will still identify objects by the alphanumeric sequence, but people (users) will undoubtedly use the about tag to find things. As an example, Fluidinfo identifies the 1984 object as a78d77ce-a055-40e3-97a9-de4223858bd8, while the about tag is much easier to understand: book:nineteen eighty four (george orwell).

As an aside, there’s a very good blog that’s largely about the about tag. It’s written by one of the earlier and most active users – he was responsible for entering the Guardian 1000 books into Fluidinfo. A lot of the posts, at least the ones that most interest me, are the ones that talk about the conventions of adding data to Fluidinfo. Anyone who has spent a fair amount of time at Wikipedia probably knows about disambiguation pages and redirects. They solve a lot of the clutter and organizational issues – someone searching for Mercury might be looking for the element, the planet, a car brand, or the mythological character. The author of the About Tag blog has spent a lot of time on these ideas and his blog serves as sort of a style manual for the about tag.

I think Fluidinfo has a lot of promise, but there’s a lot of work to be done. For my part, I’ll probably be seeding it with some more new objects (books and movies, probably) and maybe writing an app or two to pull stuff out of Fluidinfo, just to show what can be done with it.

Errors

Posted on August 21, 2010

During the past week this website has been down for a database problem. It turns out some of the tables crucial to the running of WordPress required “repair”. Apparently, there is even a “REPAIR TABLE” command in MySQL. I’d never had to use it before, so I didn’t know it existed. Pretty handy thing to have. The tables are repaired and the blog is working again. I’m still not sure why it happened, but if it happens again, I’ll know what to do.

Now I have to clean out the spam queue.

New theme

Posted on August 1, 2010

I’m working on a new theme. It’s sort of work-related, but mostly because I was getting tied of the old theme and couldn’t find one I really liked. Like cigarettes, rolling your own theme seems simple until you try it.

Calling it a beta version is giving it too much credit, but it’s a good experience.

MediaWiki and image gallery columns – mini-tutorial

Posted on March 16, 2009

I’m mainly noting this for posterity, because I couldn’t find a quick answer online and finally figured it out after a day or two.

Last year I setup a MediaWiki site for work to take the place of a site one of the resident’s had setup. The site’s overall look-and-feel has to comply with standards from the School of Medicine, and this year they’ve redesigned everything. I like the school’s new design concept, since the old one was starting to feel dated. Still, it takes work to convert a site from one theme to another.

The previous design took up almost the entire width of the browser window, which means MediaWiki could easily accommodate 6 columns of image thumbnails. The downside is it’s hard coded into the MediaWiki code. It’s a variable, but not one that’s documented. Normally, variables that are meant to be changed in MediaWiki are meant to be included in the LocalSettings.php file or the includes/DefaultSettings.php file. However, in the includes/ImageGallery.php file is $mPerRow = x; near the beginning of the file, where x is the number of columns. It can be set to whatever is appropriate, so I set it to 4 for the new layout.

The moral of the tutorial is that it pays to look in unusual places if you are hunting for some hidden variable.

Dequirked

Posted on June 10, 2008

The site had some quirks to go with the redesign and the software upgrades, but I think they are all taken care of now. The comments are back on, so feel free to try it out.

One of the main reasons for getting MediaWiki going is that it should be a good way for me to make notes while I’m doing research for a master’s degree. I keep running across linguistic terms that I’ve seen before, but don’t recall exactly what they mean (like lemma), so it would be a good idea for me to come up with a glossary of terms and definitions. If my understanding of a term evolves, then I can easily change the definition in the wiki. Writing something down usually helps me understand things better, plus I can put them in easy-to-understand (for me) terms, rather than academic prose.

I also formally applied for the M.A. program yesterday, so that’s out of the way.

Finished the book The Diving Bell and the Butterfly and put the movie next in my Netflix queue. Good book, but short. It was translated from French, but the English was very witty. I assume Bauby’s original text was just as witty.

New design

Posted on June 8, 2008

As mentioned earlier, I’ve been working on a new design, which I was finally able to apply. The redesign of MediaWiki didn’t take as long as thought, but it may have been that I was so into it that time flew by.

There are still some quirks, however. I forgot to add the comment form stuff for WordPress. There’s no convenient way to search on MW or WP. MediaWiki has a ton of different elements I haven’t defined in the stylesheet. I was mainly concentrating on getting the overall design to work, that I didn’t worry about the things that are infrequently used.

It’s a small world after all

Posted on June 2, 2008

Two things: first, I’m reworking the site a bit and trying to develop a look that will work with both MediaWiki and WordPress. I have the WP version started, but not yet finished, which can be seen over here. Given the architectures of MW and WP, the overall look will have to be coded separately for each piece of software. It still has a way to go yet, and there is a glitch in viewing it in Firefox 3 RC1, though it seems to look okay in Firefox 2, Opera, and Safari.

[edit: it's finished now and looks fine in Firefox 3 RC1]

Second, coming in this morning I saw bicycle commuter coming in a door I was passing by. However, what caught my eye was the bike jersey he was wearing. I saw the logo of Bristol Brewing’s Laughing Lab Scottish Ale and did a double take. I’ve never seen Bristol beers here in Indy, so it was strange to see the jersey. I stopped and talked with the guy for a few minutes, and it turns out his brother works for Bristol Brewing as a QA person (what a job that must be :-D)

Apple tart

Posted on February 26, 2008

I’ve had the MacBook Pro for over a month now and since I have more experience with it, it’s probably ime for a progress report or something. As an aside, I keep wanting to call it a Powerbook because it reminds me of the Apple G4-era Powerbooks from a few years ago.

Typing still sucks. I’m not sure why, but I must hit some keys lighter then others when typing and the letters don’t show up. On other computers I get a lot of typos from pressing the adjacent keys, but with the MacBook the problem is random letters are missing because I didn’t press a given key hard enough. Also, Apple advertised it as having a full-sized keyboard, but they probably meant a full-sized laptop keyboard. I’ll talk about that more in a bit.

Horsepower. For most of the past month I’ve used the computer mainly for listening to music and watching DVDs. Hardly taxing for a capable machine, so I decided to put the CPU to use and installed the BOINC client for SETI. A few years ago SETI had the SETI@home project which harnessed the collective CPU power of the machines with clients installed. It was a well-known example of distributed computing. However, some people felt the idea of looking for aliens was idiotic and came up with their own distributed computing clients that worked at protein folding or gene mapping, with the hope of finding a cure for some disease. Eventually BOINC came about as a standardized client for these projects, so a person only needs to install the BOINC client, then they can choose between dozens of different projects for their computers to work on. Since I work with a lot of computers that often it idle, I installed clients on them too, but the MacBook is second fastest of them, following a quad-core Xeon server.

Sound and video are great. For a laptop, the speakers are surprisingly good, while the screen has higher resolution than my Linux PC at home. The screens at work aren’t worth mentioning. The screen is capable of HD, but the DVD will only play regular DVDs, not HD-DVD or Blu-ray. I don’t know if they can be bought as external peripherals.

The battery life doesn’t seem that good. I have Dell D630 laptop at work that was bought in September. The Dell has a smaller screen, but it also has a Centrino Duo CPU as opposed to the Core2 Duo in the Apple. Overall, they are about the same though, with only minor differences, IMHO. With both of the fully charged, I unplugged both at the same time. The Dell lasted 56 minutes before the warning LED came on and an error message popped up. The MacBook lasted 71 minutes before an error message popped up. To be fair though, I should have waited until each shut off from lack of power. I had a previous laptop that would keep goin 30 minutes after it showed a similar error message.

I said in a previous post that the Mac seems pretty worthless for web development and I stand by that for now. For the way I do web development work (i.e. with a text editor) Linux is far better. Plus, the tools in Linux are free.

I recently bought Poser after not using it for years – they don’t make a Linux version. I’m trying to rebuild a rusty skill set there. I used Poser 3 or 4 years ago and now they are on Poser 7, which came out last year. It also seems like it’s changed hands a lot over the past several years. It almost seems as if every time they come out with a new version, the company gets sold. Some company called e-frontier developed Poser 7, then the company got bought by Smith Micro for something like 6 million dollars last autumn. Sounds like a drop in the bucket for Silicon Valley prices. Poser and the Mac seem to go well together, though it helps to have a two-button mouse attached rather than using the trackpad with one button. Also, I’ve only had Poser for a few days, so the impression may change later.

It’s so-so for Blender. Using Blender it best to use one hand for the mouse and one hand for the keyboard, because Blender has a lot of hot keys or key sequences that control actions. One of the biggest reasons for using the keyboard is to use the number pad to go between alternate views, but the MacBook, with it’s full-sized laptop keyboard, the number pad is an alternate function of the main keypad. This is common with laptops, but it makes using Blender very difficult. I’ll probably try plugging in a full-sized USB keyboard and see how that works.

So far so good, but after a month I’m only giving it 3.5 of 5 stars (I’m a tough grader), because it seems mostly average.

Geeky ponderances

Posted on November 1, 2007

I’ve been reading a bit about rich internet applications, which are net oriented applications that can be run on the desktop, as opposed to a browser. The information I have been looking at is mainly dealing with Adobe AIR (Adobe Integrated Runtime) and Mozilla’s XULrunner. Both of these are runtime environments that allow code to be written once then run on multiple platforms (Windows, Mac, Linux, etc). This is actually what Java was originally intended to do.

Adobe AIR was released as a beta earlier this year and there are already some apps made with it. I downloaded one called Bee, which is supposed to be a place where people can update their Flikr or WordPress blogs. I wasn’t able to get it to work with my site however. I have XML-RPC turned off (it’s a security thing) so that may be why Bee isn’t working, but I am not sure.

XULRunner should be differentiated from Webrunner, which is also a Mozilla project. Webrunner was recently renamed Prism and from what I understand, it is simply a very lightweight browser that is set to only open one page, which might be handy for a bank or an email setup. XULRunner, however, is a runtime environment so people can create actual programs to run on the desktop. The best example I’ve found so far is Songbird. Songbird is a music player that can play local files, but it can also play songs over the net and display web pages. It’s also skinnable, though they call them feathers instead of skins.

I’m not sure how different they are from Java (well, JVM), but my experience with Java applications in the past was that they were slow and didn’t look that great. however, that might have been more of a problem with the application developers rather than Java itself.

Even though I’ve let Giraf/Lingo/whatever languish, I still hope to do something with it, though I am not sure what yet. At first it was going to be a website I developed to be accessible from a mobile, then it was going to be based on MediaWiki. Now, I wonder about AIR or XULRunner. Of course, it would help if I knew what I wanted Giraf to do.

I would like software that has a multilingual dictionary. If the dictionaries are in a standard format, I would like to be able to download the dictionaries so the software can read them. However, I would also like the software to be able to read dictionaries that are posted online. In this case, it would probably be best if the dictionary format was in a text file. I also like the idea of flashcards. I like the idea of people being able to use the software to create their own dictionaries and their own flashcards. I think writing new words down reinforces them in the vocabulary. I like the idea of using the software to display conjugations. It’s hard to decide which aspect is most important. I’ve seen lots of multilingual dictionaries, and even some flashcard type software, but I don’t think I’ve seen anything that displays conjugations, so that’s something else to consider.

PHP 5 and WP 2.2.2

Posted on September 2, 2007

As I’m prone to do, I decided to give myself a project this weekend. It involves the Lingo/Giraf thing I was working on months ago. My coding skills aren’t quite that good, so I figured I would see if MediaWiki 1.10.1 could be adapted for use in that project. It looks like it works okay on my PC, but with the site on the PC, I can’t check it with the Palm T|X.

Long story short, I upgraded the whole website to PHP 5. As suspected, it broke some things, namely WordPress. I upgraded WordPress to 2.2.2 to see if that helped, and it didn’t really. What I had to do was remove the SpamKarma2 plugin along with the wp-contact-form. The contact form isn’t a big loss. SpamKarma2 was a mixed bag. Ever since installing it, I would get all kinds of fatal errors relating to memory (e.g. tried to allocate …). Plus, I found myself logging in less often because I knew SK2 would take care of the spam. I got lazy.

Anyway, WordPress seems to be back to normal more-or-less, minus a couple of plugins. I’ll probably have to sort out problems as they occur.

Next Page »

© Copyright gottahavacuppamocha • Powered by Wordpress • Coffee Candy is based on the Eye Candy theme.