Beaucoup d'Idées

iTunes Playcount Threshold

2005-07-21T15:32:00.000-05:00

So I've been kicking this idea around for more than a month in my head but I'm finally getting around to writing it now. It took me a while to appreciate the usefulness of the playcount in iTunes. I'm now a big proponent of making smart playlists and one of the best ways is to combine playcount and at least one other factor. I have a whole bunch by now: four-star songs with a playcount under five (for stuff I like a lot but have kind of forgotten about); all five-star songs with a playcount of ten or greater (the songs I really love); my twenty-five-most-played songs (kind of fun to keep track of); songs with a playcount of zero (music I really need to get around to checking out); you get the idea.

There's just one small problem with the playcount system. It requires that the song actually advance to the following track before it counts. Now, before you try to argue with me that that's what one play is, hear me out. Would you consider playing the whole song less two seconds a play? What about less 5 seconds? What about less 25% of the song? I would like a way to customize what is considered a play. I realize that this could get kind of complicated, so I think by default it should be off and that it should be an extra screen off of the Advanced option tab.

Here's how I envision it working:
First of all, you could make multiple Playcount Threshold rules and could add ones just like you can add fields to smart playlists. However, you would obviously have fewer fields. I think all you would need the following categories to specify, at a minimum: current playcount; rating; last played; and duration. Each of those would need the appropriate limitors: greater than, less than, at least, at most, equal to, and any. Then, once you had set those, you could set the Threshold by percentage of the song played or time remaining in the song.

Now, that may or may not make sense. Let's get some examples on the table: rule 1 - for all songs longer than three minutes, fourty seconds, of any rating, played in the last 6 months, and with a playcount greater than five, once 75% of the song has played, it counts as a play; for all songs of any duration, any rating, and a playcount of zero, once 94% of the song has been played, it counts as a play.

"Why oh why would anyone care about this?" you may find yourself asking. Well, most people probably wouldn't -- that's why it would be set to a default that would be unoffensive (like "Off"). However, I find it to be kind of annoying that if I skip seven seconds of silence at the end of a track, it doesn't count as a play. Furthermore, once I've listening to a song fifteen times, if I only listen to 80% of it, for purposes of iTunes I may want to consider that a play. It's very customizable. Plays matter because they help me make better playlists and because it's interesting, especially in the long run, to reflect on how many times I've listened to a song.

That, plus I'm a little OCD. Whatever.

Redesigned Phone Keypad

2005-06-29T10:53:00.000-05:00

Imagine if we typed on a keyboard that was just in alphabetical order, A-J in the top row, K-S on the second, T-Z on the third. Actually, now that I think about how that breaks down, it probably wouldn't work badly at all. Most of the commonest letters would be on the home row (l, m, r, s) although only one vowel would fall there and there wouldn't be as much back-and-forth (the is, hand to hand) as experts seem to think is most efficient.

Ok, forget all that, that intro didn't really do my point justice. My point is that the telephone alphanumeric keypad is almost absurdly inefficient for typing. The location of S in relation to the position of X is a perfect example. The letter X, rarely if ever used in an SMS message/160 character email, is accessable in one keystroke; the letter S, probably one of the top 3 commonest letters in the English language, requires 4. That just doesn't make sense. There are other many other examples of the innefficiency of the strictly alphabetical keyoad. Ever try typing out "night" in an SMS? 66-444>-4>-44-8 (> means next or right arrow, unless you want to wait the duration for the phone to automatically moves you to the next letter). That's eleven keystrokes, and the "ight" suffix is pretty common, and even moreso the "*ght" suffix. I've thrown together a potential model that I'm calling the AIR keypad (because the first characters of of the 2, 3, and 4 keys are A-I-R). I'm not a lexigrapher or linguist or a typographer or an *ist/*grapher that specializes in keypad layout. I tried to avoid putting letters on keys that are commonly typed in the same word to avoid the "night" problem, and, as much as possible, spaced out common letter combinations among rows to decrease the liklihood of missing keys for their neighbors to the right or left. All the vowels are on 2 and 3 and are divided such that (again, as much as possible) no common vowel combinations are on one key. This may not be the best layout, this is just a rough idea.

As far as punctuation and symbols go, I think keeping the major ones (space, period, comma, question mark, exclamation point, probably in that order if they're not already, I don't remember off the top of my head) assigned to the 1 key makes sense and is convenient. As far as other less common but still useful ones such as quotation marks, dashes, slashes, ampersat, and parenthesis go, they should be assigned to the second of third tap of * and #. "Smart quotes" aren't necessary, but "smart parenthesis, that is ones that can sense if they've been opened in some way, would be convenient and could save a keystroke or two although I suspect not many people use them on SMS messages anyway. As far as accented letters go (é, ç, à, etc), those will be assigned to the taps immediately after the number (tap 4 or 5) on the key to which the letter to be accented is assigned (é, è, are on 1, ç is on 5, etc.).

As far as other, even less common symbols such as dipthongs, Greek characters, and other linguistic symbols go, they can be buried pretty much anywhere. I imagine that these are rarely, if ever used. Assuming the shouldn't be eliminated on English phones completely, their placement is nigh irrelevant. If you're writing an SMS about the 1911 Encyclopædia Britannica, a) your text messages are far to erudite (and pretentious), and b) you can survive with an ae instead of the dipthong if you don't want to just write "encyclopedia."

In case it's not obvious why this matters, SMS is huge and is still exploding in popularity and usefulness. As the inevitability of cellphone service on airplanes looms, anything that makes SMS more functional is to the benefit of all, at least until they get wifi/EVDO on planes - then we can all die happy.

NB: I threw together a mockup of the keypad layout in Paint but it's so atrocious and illegible you're just going to have to wait until I have time to put something better together before I post it. In the interim, the key layout is as follows: 1: same as current; 2: AEU2; 3: IOY3; 4: RKX4; 5: TMC5; 6: LBZ6; 7: SJQV7; 8: GNZ8; 9: HPDW9; and * and #: to be determined.

Update 7/12/05: Added said graphic of my design. I uploaded it as a PNG and I don't know why Blogger automatically converted it to jpg. I could do it with Flickr but I don't like that layout as much. Anyway, that's the gist.

As far as I know, neither this idea nor my layout have been concieved of by anyone else. As such, I declare it and the contents of this post to be covered by the Creative Commons Attribution-NonCommercial-ShareAlike License

Why I love open source software

2005-06-27T20:07:00.000-05:00

The experience I had today with Sprog today pretty much encapsulates why open source software (OSS) is so great. I was going through the Waxy.org links and read an article Andy had linked about Sprog, a program that automates various and sundry tasks. It looks very cool and I'm looking forward to trying it. In fact, it looks so cool that it inspired me to download and burn two bootable open source bootable CDs, Ubuntu (based on Debian, I believe) and Knoppix (Linux). Since Sprog wasn't going to be something I could immediately use, I surfed around it's home page, generally reading up on it. The creator called the name, in part, a "contraction" of sprocket and cog. That made perfect sense to me except it seemed it was more of a portmanteau than a contraction (because contractions tend be combinations of a verb and either a pronoun or a negation). This is a good time to just get out on the table that I'm an English major, major grammar nerd, Wikipedia addict, and obsessive-compulsive editor. So, since I was at work today and wasn't getting any work at all, I decided, "Hell, I'll drop the author an email through SourceForge." So, I did:

I came across Sprog from Waxy this morning, it looks really
great. Enough so that I'm looking into getting a good ISO
bootable Linux to try it out - it looks great. Damn Windows.

This is a quite minor issue but I figured I'd mention it, for your
interest if for no other reason.You say Sprog is "a contraction
of the words sprocket and cog..."; technically, I think it would
be a portmanteau (http://en.wikipedia.org/wiki/Portmanteau)
as contractions generally are the combination of some kind of
word (a pronoun, for example) and a verb. As a combination
of two concrete nouns, Sprog seems to be a portmanteau.
Keep up what looks like great work!

Matt

Not an hour later, Grant wrote me a very friendly email not only recommending a Linux to me, but made a note on the FAQ page mentioning my email and linking to the Wikipedia article for Portmanteau.

What's my point? My point is that OSS authors give a shit. They're friendly even to some random OCD word-checker who can't even run their software natively. Not only did Grant respond nicely, he changed the webspace to reflect my pithy comment. Furthermore, by doing so, he made me feel like I matter, which a) is further motivation to use his product and b) makes me want to try to contribute to his product in the future, because I know he gives a shit.

I know other issues are at play such as how new Sprog is and the fact that it's probably relatively unheard of at this point, but that's not the point. The point is, by showing me that he cares about what I think, he has gained someone who cares about what he does. That's a pretty useful thing in the computing world.

Response to Jason Scott

2005-06-05T13:44:00.000-05:00

This is a response to Jason Scott's essay (an excellent read, he makes great points) about CC Licensing his documentary about BBS.

Very inspirational and well-written. I'm a big supporter of open-sourcing/CC-licensing projects and products, but I do think it presents a problem. Open-sourcing/CC-licensing works really spectacularly right now because it is a practice known of and embraced by a comparatively tiny minority. However, there's a whole world of infrastructure out there on which peoples' lives, jobs, and families depend. A comparison that strikes me as useful is to the practice of blocking internet ads (or PVRing and having commercials automatically deleted from TV shows, for that matter). I use ad-block and a handful of other utilities in Firefox to block basically every kind of ad, from pop-unders to banners to Google text ads. For me, this makes browsing the internet a fantastically more enjoyable experience. However, much of the internet (and the ability of individuals to have their own sites) is dependent on ads. Practically speaking, it's unlikely that my practice will be widely adopted, hence the infrastructure will remain sound and people who want to start messing around with the internet can sign up for a free (read: with ads) Angelfire account and start messing around with HTML exactly like I did in 7th grade. If, however, everyone started blocking ads and it stopped being possible to generate revenue through their use, that would force a lot of great websites to go in one of two directions: charging a use/membership fee, or shutting down; it would also keep a lot of interested casual computer users from ever having a good opportunity to learn hands-on about this technology. These days I write in XHTML and CSS, but I doubt I ever would have been able to get my foot in the door if I had to start out paying for hosting and bandwidth. One of the best things about the internet is its cross-culture, cross-class accessibility, and it would be a real shame if that fell by the wayside.
This issue is much like the issue of the CC license. On a small scale, the release of media and software with CC/GNU licensing works very well, makes a good statement, gives everyone access to the product, and results in whole myriad of other positive outcomes. Jason suggests that more media in particular be released in such a manner. This is a good idea in principle, but economically speaking, it would be the equivalent of suggesting deficit spending as a business plan, and long term, in a capitalist country, that won't work.
I don't know what the solution to this problem it. The outcome may be that those "in the know" will continue to know about ways to avoid ads or obtain free content and only the poor schlubs will get stuck with the price while we take it for free.

I'm not presenting this completely fairly because there is definitely room in the world for both kinds of licensing. However, the CC licensers and open source zealots (I'm one myself) talk in terms of principles ideals - if that's the direction we want things to go, that's great, but we must think about the implications of a large-scale implementation of this system.

NB: I don't mean to be a naysayer or a critic - I applaud his move away from materialism and his stand against highly punitive copyright laws. I just think it's something to think about.

Song Meanings Follow-up

2005-04-18T01:48:00.000-05:00

Well, SongMeanings took a step towards what I suggested in my previous post and it's definitely a good one. I'd like to see it happen in another way, but I think lyric spam is a legitimate concern for a site that hosts something as popular as lyrics. On the other hand, SM could, unlike Wikipedia, just not allow unregistered users to start add new bands or edit lyrics and then use one of those "are you really a human" confirmation codes and then block spammers if they become a problem. I could see this being problematic because people are stupid and two pages could be started for the same artist if some idiot goes on and tries to find "Beetles" lyrics, doesn't find them, and then starts a page for the Beetles and then submits a bunch of lyrics. Wikipedia doesn't have this problem that I know of, but without sounding too elitist, Wikipedia probably gets smarter users on the average. Nonetheless, I think the system I proposed in my previous post wouldn't be overly problematic and would be overwhelmingly good.

This Would Rock

2005-04-05T21:40:00.000-05:00

I want to combine two things I love: Wikipedia and SongMeanings. I don't want a user-driven music encyclopedia or anything like that. I want SM to adopt the wiki system of generating content. The current system is slow and doesn't take advantage of user initiative.

If you want to submit lyrics for a band that's not yet on SM right now, here's what you have to do: post the band's name and a website proving its existence in a particular thread in the message boards (that frankly, isn't always that easy to find). A couple of days later, if you remember, you check back and probably someone from the SM staff has approved your request, and then you can submit lyrics. This system is bad for at least three reasons: one, it's a pain in the ass for someone at SM to deal with; two, it doesn't take advantage of the impulse of the listener who just got into a band to post lyrics from their songs because he has to wait two plus days to do it; and three, once the lyrics are submitted, even if they're wrong (more common that you'd imagine), they're relatively set in stone. I say relatively because while I imagine if something were grossly incorrect it could be changed, I see wrong lyrics for songs that have been around for ages and never corrected despite corrections in the comments area below the song.

Here's the system I propose: to start a lyrics page for a band that's not on SM yet, you are sent to a page where you enter the same info you would have had to post in the current system, that is, band name and website proving its existence. As soon as you submit, the band's lyrics page is created immediately, at which point you can begin to submit lyrics. If someone doubts the existence of a band (I feel like this would never happen, but who knows), it could be flagged for review and someone at SM could check it out. If a band is flagged and checked and it exists, it would be considered "confirmed," so it couldn't be flagged in the future. The lyrics would then be generated the same way as new articles on Wikipedia and hence, could be edited very easily when bright chaps like ruben submit completely incorrect lyrics. According to his SM journal, ruben has apparently been offered the spot as a lyrics mod at another lyrics site. Maybe the other lyrics he's submitted are good or something, but I digress.

A big part of SM is the user comments, and the current system, while it's not ideal (I'd like to see the comments threaded à la LiveJournal), works well enough.

If this system were adopted, I suspect the quantity of lyrical content on SM would explode at a rather rapid rate, especially at first, which would rock, since that's a big part of what the website is for. Secondly, it would result in more accurate lyrics and save the SM people, who seem to be endlessly busy, some time and headaches. Since wiki is free and open-source, this system wouldn't cost much (besides the time spent customizing it for SM) to implement. SongMeanings is a great webpage but one in need of a serious overhaul to bring it up to speed with current Internet technology.

Irony on April 1st

2005-04-05T00:26:00.000-05:00

Google Gulp. Ha ha et cetera, that's funny. What I find to be funny is this part of the disclaimer:

No personally identifiable information of any kind related to your consumption of Google Gulp or any other current or future Google Foods product will ever be given, sold, bartered, auctioned off, tossed into a late-night poker pot, or otherwise transferred in any way to any untrustworthy third party, ever, we swear.

God, the irony kills me. If only Google would have a privacy policy this honest and forthright (and reassuring) for their real services. Don't get me wrong, I use and love Google, gmail, et cetera, but do you really think a cookie that doesn't expire until 2037 is necessary Sergei and Larry? I'm not conspiracy theorist, but Google Watch has some points that can't be refuted.

Welcome & Active Playlists

2005-03-25T20:49:00.000-05:00

So, welcome to Beaucoup d'Idées. That means "many ideas" in French, for those of you who don't know the tongue. I like French a fair amount, although I don't really plan on pursuing it anymore in terms of studies. The French will have to forgive me for not putting the d' in my URL - beaucoupdidees looks kind of awkward and sounds kind of like Diddy, and I don't want to give that man any free press. Anyway, in case it isn't obvious, I named my blog that a) because I have a lot of ideas floating around in my head and b) because everything else I wanted was taken and this handle doesn't suck. Thanks a lot elucidative.blogspot.com for taking my first choice; I love that word. At least I have that handle over at LJ; even though I never use it, I'm never giving it up!

Anyway, enough with the introduction nonsense. I am addicted to iTunes and my iPod, so much so that I'm getting this shirt. As such, don't be surprised if a lot of the ideas I post about have to do with one of these two things. This is especially going to be the case early on because I already have a bunch of ideas about these products floating around in my head. The newest idea I had while I was driving today was a new kind of playlist. If you're familiar with iTunes at all, you know about smart playlists & standard drag-and-drop playlists. Then, when you're using you're iPod, there are on-the-go playlists. What I'm proposing is a combination of the standard drag-and-drop and on-the-go playlists: Active Playlists. Before you (I guess there isn't a you yet, since this blog is new and all) comment to tell me that you can add and remove songs from an o-t-g playlist...yes, I know - that's the provenance of the namesake. I want the way to mark one (or more) playlists as "Active" and then be able to add to that. It would be a small but cool tweak. Basically, here's how I imagine it working: you could mark a playlist as Active in iTunes and then when you held down the select button on your iPod, it would switch to a new screen where you could pick whether to add this song to your active playlist or an o-t-g playlist. If you didn't have any playlists marked as active, it would automatically go to an o-t-g playlist. Then, and this is crucial, when you sync your iPod with your computer, it will update the computer version of the active playlist and vice-versa. Usually, if you change a playlist on your iPod but not your iTunes, it goes with what's in iTunes, thereby erasing what's on the iPod. This playlist would stay active until its status was changed.

This would be useful for me, and I imagine other people, because every month to six weeks, I make a new mix. For example, in my iTunes right now I have End of Freshman Year Mix, One Month Left Mix (that's one month of summer, for clarity), Happy New Year Mix, et cetera. These were all complied over the course of the aforementioned time period. The content of these playlists is mostly newish music that I've gotten into during that time period and the ability to work on the playlists from my iPod or through iTunes would be awesome.

This could be useful for people who don't use the method of mix creation I do, as well. For example, I (and apparently many people) listen to their music on Shuffle. I do this both on my iPod and with iTunes. You could make an active playlist of songs that you hear on shuffle that you don't know but stood out as ones you liked to come back to and listen to later; add to it in iTunes or with your iPod.

I don't have any sort of connection to Apple or anything, so the chances of this happening are kind of slim, but I sure would enjoy it. Wouldn't you?