And yet…

Apple is switching to Intel.

I’ve been looking, but have been unable to find, something that will let me browse my iTunes Library by Album Cover, the way I do with my real CD’s. I think that would be MUCH more enjoyable than looking through lists of album names. Has anyone seen such a program?

Clutter kind of has the idea, but it doesn’t pick up the art from iTunes, hasn’t been updated in almost a year, and, most importantly, you have to choose and arrange onscreen the albums you want to be able to browse through. If you want to look at your whole collection, you’re back to lists of names.

If you’ve seen anything out there, let me know. This might be an interesting project to get me back into programming…

1. Reporting

This is the most important feature of Taskmaster. It stores when you do the things you’re supposed to do and when you don’t. It can then report, in terms of percentages, how good you’ve been at doing these things. This lets you figure out the tasks you might need to reprioritize in the future.

2. Event Repetition

In most such programs, when you set an event to repeat, the options are very strict. You have choices like “every two Mondays” or “the 15th and 17th of every month.” What about options like “2 weeks from the last time you did it?” What about “This should be done everyday, but if you miss one, don’t pile it on top of the next one; just let it go.”

This, I suppose, is my official declaration that I was working on it…but I’m not anymore. I put it on hiatus for a few reasons.

Most importantly, I wasn’t as happy with Cocoa as an API as I was expecting to be. Sure, Objective-C is severely object-oriented, a la Smalltalk, which is great. And sure, Cocoa does a lot of work for you – the example often cited is that you can make a text editor or a web browser with just a few lines of code. And it’s true. You can. The problem, however, is that it does all this work, but the documentation never quite clearly expresses what it’s doing. You’re required, in a sense, to second-guess the myriad methods that come “for free” with the environment on a fairly regular basis. I’m sure once you’re a seasoned Cocoa developer, this all comes quite naturally, but something about it just turned me off.

Secondly, there never seemed to be enough time to work on the project, as I work full-time and spend as much time as I can on my musical endeavors. I decided to focus on smaller projects in PHP, which have been much more rewarding.

Finally, as I’ve mentioned before in these pages, I’m not quite happy with the user interface in OS X these days, so I’d like to see how Panther stacks up in terms of usability before really devoting myself to the platform. But don’t hold your breath.

The source code is released under the GNU General Public License, so if these ideas are interesting to you, here’s the tarball. You can also just browse through the source.

By tying together Amazon Web Services, Kung-Tunes, and a simple mySQL table using PHP, I have managed to speed up album cover lookups on the sidebar there, but more importantly, I’ve got album covers for all recent tracks now! I wrote a previous entry about grouping recent tracks by album – this one is about how to suck the album covers off of Amazon’s servers…

[Update 9/15 2:08 PM: I've fixed a bunch of bugs, including bad character encoding in song titles, database problems with long album titles, and issues with including both frontend pages within the same webpage. Thanks to Etan for catching a bunch of them. ]

The basic algorithm of it goes like this:

Kung-Tunes supplies an artist and album name. searchagent.php does a keyword search through Amazon, which sends back an XML page rife with information. The scripts parse out the URL for the thumbnail image along with the unique ASIN (Amazon Standard Identification Number) for that album. They then save the information into a MySQL table. The next time the page is loaded for that album, the script checks the table for it, saving valuable load time. This info is passed to one of the front end scripts for formatting.

searchagent.php stores the Lookup object and the SearchAgent object. Lookup handles SQL lookups, and, if it doesn’t find anything, calls SearchAgent to look it up on Amazon.

amazon.php is the frontend for the “iTunes is currently playing” section. It basically passes information from Kung-Tunes to the Lookup object, gets back the info, and formats it for viewing.

itunes.php is the frontend for the recent tracks page. It groups the tracks by album, and then passes that info on to the Lookup object. It takes what it gets back and puts in a pretty table.

You can download all three here.

On the SQL side, you need to create a table by the name of amazon with the proper fields. This SQL query should probably do it:

That Date field isn’t currently used in the code, but it’s there to deal with Amazon’s caching policies, which I haven’t quite figured out yet.

On the Kung-Tunes end, the scripts assume “^t@@ #p^p#p @@ #a^a#a @@ #y^y#y @@” for currently-playing track and for recent tracks list. Encode HTML entities is off, as they seem to muck with the Amazon searching.

To use my scripts, you need an Amazon Developer Token obtainable at Amazon’s Web Services site. (Since I’m making the source available, I’ve taken mine out of the code.) You also need Kung-Tunes, a great piece of GPL software that runs on OS X, uploading iTunes track data to the web server of your choice. And of course you need OS X – not on the server, just on your own machine. And finally, the willingess to futz around with my code to make nice with your server configuration.

This code is covered under the Creative Commons Attribution-NonCommercial-ShareAlike License. I wrote a previous entry about why I chose this license and how I think it works in terms of PHP scripts. If you don’t agree with something in the license or my interpretation, please don’t use the scripts. Make a comment here or there, I’ll look into it, and we’ll work something out….

Enjoy!

[Update 9/28/03: I noticed a couple of downloads of the code in my server logs recently, and my curiosity is just killing me. If you do end up downloading and using it, please make a comment or send a trackback. You're certainly not required to, but it would make me smile. Thanks!]

Now the term “derivative works” is a bit of a sticky wicket – I will repeat here the license’s definition of the term:

“Derivative Work” means a work based upon the Work or upon the Work and other pre-existing works, such as a translation, musical arrangement, dramatization, fictionalization, motion picture version, sound recording, art reproduction, abridgment, condensation, or any other form in which the Work may be recast, transformed, or adapted, except that a work that constitutes a Collective Work will not be considered a Derivative Work for the purpose of this License.

The problem is that this is not at all clear with respect to the intricacies of code inclusion, library linking, server-side scripting, etc. The terms “based upon,” “transformed,” and “adapted” seem to be made intentionally vague, possibly to reflect the different kinds of “Derivative Works” one might build from a PHP script vs. a blog entry vs. a sound clip.

I contend then that a Derivative Work, in the case of the code I will be releasing under this license today, refers to any system of which this code, or some modified version thereof, is an irremovable component, e.g. a system that gets Amazon reviews for the albums you’re listening to, or a form that asks you for an artist and album name and spits out the price. What it does not include is unrelated content that may, due to the nature of the web, have to be included within the same file as the code. This I would view as a Collective Work.

Since Derivative Works have to be released under the same license, you can make changes to the code to fit your environment, use it along with the content on your website, and you are obligated to release only the code that relies on my code and the modifications you’ve made – your own rants and unrelated scripts are safe under whatever license you’ve been using them with.

Now I’m no lawyer, but I believe that the things I’m saying are in accordance with the actual terms of the license I’m talking about. If something seems awry to you, please make a comment here, and I’ll look into it. The code will be up a little later today…

Then there’s the GNU LGPL, or GNU Lesser General Public License (previously known as the GNU Library General Public License). This is closer to what I want, in that one can link to my libraries without having to release one’s own code, but any modifications directly to my code (bug fixes, etc.) do have to be released under the LGPL. The only problem is that anyone can link to it for any purposes, including commercial ones. Under the LGPL, some fool can slap on a frontend and start making tons of cash off my code! I don’t want that!

So moving on to the Creative Commons Public License schemes. These guys offer some flexibility in terms of attribution, commercial use, and shareability. The problem is that the shareability distinction is not subtle enough. Under the Attribution-NonCommercial license, your code can be modified and publicly used by anyone, and those modifications are lost to you. So if someone else fixes some bugs or markedly improves my code, the license permits him or her to keep the changes to the source code private, while still using his or her new code on a public site. On the other hand, under the Attribution-NonCommercial-ShareAlike 1.0 license, anyone who uses my code to create “Derivative Works” is obliged, as with the GPL, to open up those Derivative Works through the same licensing scheme.

This is where the problem comes in – the term “Derivative Works.” These Creative Commons licenses define a “Derivative Work” thus:

“a work based upon the Work or upon the Work and other pre-existing works, such as a translation, musical arrangement, dramatization, fictionalization, motion picture version, sound recording, art reproduction, abridgment, condensation, or any other form in which the Work may be recast, transformed, or adapted.”

How does this work in terms of code? If File B has a line in it that says include ('File A');, is B then a derivative work of A? Or does the content of A actually have to be placed directly into B?

Common wisdom about the GPL seems to imply that if File B links to Library A and Library A is GPL’d, File B must also be GPL’d, even if these are just executables linking to one another. This is basically why the LGPL was created. It stands to reason then that when code is being literally inserted from one file into another, whether by the PHP interpreter or by human hands, that this would still apply – that File B would be thought of as derivative. Therefore, neither license is acceptable – either I lose out on changes to the source code, or other lose out on using it.

So what is a boy to do? Where do I find a license that will allow me to prevent commercial use and force published modifications to my actual code to be released, while allowing people to link to it without having to release the rest of their code? I know it’s not very much code right now, but these criteria seem pretty important to me, and I’ll probably want to use them again and again.

And PLEASE don’t say I have to write my own license. Cause I just can’t deal with all that legal crap….

[Update 1:07 AM: I've been doing some more reading about Java and the LGPL , and I'm beginning to think that, at least in terms of PHP, the license I'm looking for does not exist. It seems to be that if the code is included, as it is in PHP, rather than linked as it might be with compiled code, clause 5 of the LGPL, which is what lends it its special power, does not apply, and clause 6, which makes it basically like the GPL, takes precedence. My guess would be that all existing licenses will take this stance, as anyone using my code is essentially incorporating it into their application, rather than linking to it externally. I will probably have to compromise, but in which direction?]

They’re at the end of every category page. So I can now keep tabs on all the new furniture in Manhattan. Or internet engineering jobs in Queens. Or casual encounters in all of New York. Cause those shits are funny.

I have tried to keep myself sharp in programming by teaching myself new languages: last summer I was working on C++, and recently I’ve been doing Objective-C. However, it would seem that anyone looking for a C++ programmer expects experience with teams developing Windows applications, or something of that ilk. Objective-C, on the other hand, is not going to get me anywhere, and, besides, I’m mad at OS X right now, so…I’m gonna delve deep into PHP. Gonna write a lot of it. Gonna know it cold.

Here’s where you come in. Code is only useful when it does something. So, the question for you all is what it should do? What kind of features would you like to see on this website? What kind of data would you like to see manipulated? Ratings? More magical Itunes trickery? Any and all suggestions are appreciated. I also encourage people to tell me I’m barking up the wrong tree and that there’s a much better way to get a new job, complete with explanation and the time I should show up at the interview.

Thanks.

But in another sense, this is an obsession I’ve always had. The information that’s available all around us could tell us MUCH MUCH more than we already know. First of all, there’s our own data – our e-mails, documents, etc. Zöe is a program I installed recently which is hard to describe. Its author describes it thus:

“So what is Zoë? Think about it as a sort of librarian, tirelessly,
continuously, processing, slicing, indexing, organizing, your messages.
The end result is this intertwingled web of information. Messages put in
context.”

It looks like this:

Zöe only scratches the surface of what can be done with a small portion of our own data – our stored e-mails. We have a tendency to think in terms of limited paradigms: search, search by content, search titles, subjects, etc. We forget that there are patterns we don’t see: ways in which we can pieces of data to each other.

WikiWebs are another good example. People often refer to these as webpages which are freely editable by all, but they forget the coolest part – links everywhere that form automatically just by formatting your words in the right way! The data stored in the Wikipedia is not just a set of articles – it’s a whole intertwingular interweb! It’s what the WWW should really be all about! Not static sites with flashy designs! Content! With lots of relevant links to related content! And lots of exclamation points!