Sunday, March 29, 2009

Music Sales: Ur Doin It Wrong

30-second samples of songs just won't do for me anymore to make purchasing decision with. They might give you some idea about the bridge or chorus of a standard pop song, but 30 seconds won't tell you if a 10-minute remix is a crashing bore of repetition or a fantastic patchwork of sounds. Especially if by chance the 30 seconds is at the beginning or end of a song.

The 30-second music snippet worked really well in the early days of the web, when Amazon and Barnes&Noble had to, well, at least let you listen to something in their attempt to lure customers into entering a credit card number into this new-fangled form and hope something would arrive. Considering 28k8 was a general speed modem at the time, having to wait for 30 seconds of lousy quality to come in over that was pretty much it.

But it is 2009 now and even a 2-year-old mobile phone can get data in faster than that. So why are Apple and Amazon still thinking 30 seconds will do for a purchasing decision? Let me hear the whole thing, then I will decide. Then again, it is not up to these sites, but the record companies, and we all know record companies seem intent not to actually sell music online.

Right now in the music space every site and store is trying to find the sweet spot of features for selling music, having to navigate between what radio licensing and record labels will allow and what users want. Pandora does discovery of new music based on your own tastes the best, but doesn't let you listen to anything on demand and just grabs something from your likes to play randomly, and is locked to the US only. Last.fm is global, has a lot of information to discover new music with, but then doesn't let you actually listen to it: again, many songs get clipped to 30 seconds or are not available, and noise levels are all over the map. Rhapsody wants $15 a month to let you listen exactly to what you want when you are online, and will let you put all the songs you like on your player, but they those have to be 'Rhapsody-compatible' devices only -- i.e. not iPods, the most popular music player. Nokia Comes With Music gives you unlimited music on your phone and PC for free for a year, but after that you need to subscribe to add new songs, and the music is tied to Windows PCs and Nokia phones. Meanwhile Apple is trying with the Genius for suggestions, has perfect service delivery and integration between downloading, playing, and putting on your portable player, and its music is transitioning to no longer be tied to their own brand of players, but their store has that stupid 30-second rule that has fooled me into buying too many songs whose 8 minute mixes were far worse than the 30 seconds indicated. Lala.com has moved on from its CD-trading roots and now has become a pay-for-songs website with discovery of new music, full streaming on demand of songs for free for a few times, pay to hear more on-demand songs, and buy unprotected tracks easily, but doesn't integrate with portable players and is locked for US only. Spotify is the reverse geography-wise: Europe only, and streams what you want to hear on-demand (and really well too, with a large catalog) for free with an ad every 5 songs, or pay to have no ads. No purchasing, no discovery, which makes most analysts think they are doomed, as ad-supported music seems to not work on the web and them getting enough subscribers to cover licensing costs seems uncertain.

And then there's Imeem, which I find incomprehensible to actually deal with. And then there's MySpace music. And then there's Slacker with its own player that loads music over WiFi but doesn't let you hear exactly what you want because it is a subscription web-radio model, and, and, and many others. And none of them have the whole stack of helping me choose, letting me listen before I buy at a good bitrate, keeping it safe for me on their servers, letting me have and keep it for my players now and in the future, at a great bitrate, without having to worry about which one uses what technology first, with simple clicks. And oh yes, global.

So, Last.fm to discover, Spotify to preview, Apple to actually purchase and get on my iPod, for me here in London. This is supposed to make me buy more music? This is supposed to make me pirate less? This is supposed to not make me hit Usenet with a good one-click download archive newsreader? This is supposed to help me find new stuff instead of playing my old catalog over and over? This matrix of options and restrictions I can't really understand until I sign up and try?

It should be as easy as entering a name in a box to instantly hear what I want in full, suggest new music based on already knowing my tastes from uploading my iTunes or Windows Media Player or Last.fm or Pandora ratings and play-counts, and let me click once to pay for a song and have it on all my players. Globally. And until then, the music industry will just keep dying.

Thursday, March 26, 2009

Has Elon Musk Actually Never Driven A Car?

A touch screen? In a car? As the center-console user interface? Are you shitting me? The one interface that has to be designed to be operatable without looking at it is going to be a smooth piece of glass with different layouts? Will have no tactile feedback for finding the right button just by touch? So you have to look to make sure the AC screen is on to change the temperature of the car? You have to look to select the right radio station instead of gliding your fingers across the edges of the presets?



Center LCD console of the new Tesla Motors electric car


No, seriously, are you kidding me? You have to be kidding me, right?

Wednesday, March 18, 2009

Seismic Shift Shopping!

Remember when SUN was going to buy Apple because Apple didn't seem to have a clue and SUN and SGI were basically serving up the Web so they were huge and rich? And then years later Apple was going to buy SUN, because Apple had a clue again and it proved to be cheaper to serve the web on x86 commodity boxes so SUN had no business model? (Or Apple was going to buy Nintendo. Or the other way around. I love rumors.)

At the very least I can report that when NeXT was purchased by Apple to crawl out of the engineering morass Apple was in, at least one SUN kernel employee told me they were surprised: they were convinced Apple would choose their operating system kernel to put Copland on since Apple's own efforts were going nowhere. That's the closest any of that was ever reality.

None of those rumors ever made sense. IBM buying SUN does.

I can't wait. It's been a while since something huge like computing behemoths colliding happened.

Monday, March 16, 2009

Apple: Morons?

I have to take the Apple-will-become-a-headset-with-remote-standard thing back. They are being morons: in order for a headset with the new super remote control to work with the iPod Shuffle, it has to include a special Apple-supplied hardware chip according to iLounge. That's right, Apple is now designing music players that turn the open headphones market into a Manufacturer Approved Peripheral market, going the same way as phones with their non-standard music jacks. It's just another version of Nokia's PopPort.

The review also says it is simply a bad music player because the remote control makes it complicated to use. The sending-little-beeps-over-the-line sounds remarkably like that CD player with remote I used in 1993, except that had dedicated forward and backwaqrd controls that made those beeps, it didn't make me hit a button 3 times to go skip a track and hit it twice to go back.

I think I need to hoard a stash of 2nd Gen Shuffles.

Edit: Apple has weighed in: there is a proprietary chip in the remote that manufacturers have to license to be part of the Made For iPod program / certification, but there is no encryption on it and can be easily cloned. I still think it sucks because it hinders the adoption of this scheme as a universal plug for headsets that go from phone to iPod to stereo etc.

Saturday, March 14, 2009

Your Texts Could Be Out There!

A Tweet turned me on to Skydeck. Skydeck is a web service that gets between your phone and the network and allows you to manage your voicemail, your texts, and your contacts all from your desktop browser. From reading the help pages, it seems Skydeck does this by
a) becoming the voicemail system for your phone. That is easy enough, the GSM standard explictly allows phones to tell the network which number to call for voicemail.
b) installing a little application on your phone that syncs your text messages on your phone with their servers or, if you do not have the right kind of phone for that, asks you for your name and password to log into your cell-phone operator (aka Mobile Network Operator [MNO]) and sync their servers with your account data and text log
c) wil spoof Caller-ID and SMS ID to display your phone number when you use their web application to make calls and respond to texts instead of your own phone.

This sounded great to me. I am behind a computer at least 8 hours every working day and always actually more, and communicate mostly through texts and VoIP. It would be great to finally integrate the two comms devices I use most, and have a big keyboard for my texts and a big window to manage them in, and read transcripts from my voicemail, and just use my computer to handle calls without switching headsets and everything.

My next thought was that this would totally be the way to go in ten years when mobile coverage is synonymous with high-speed data coverage: no need for software on the phone itself besides a browser and codecs for voice: all of the phone's SMS and voicemail and calling and phone book and calendars would be done by the servers on the network, not the always-memory-starved device. Handsets would become dumb terminals with great screens because you could save on all other components, with user interfaces and software that would simply be upgraded just by updating the web application on the operator side, and all the applications and toys and games and data would be made of little software components of which you can choose whether they lived on your dumb-client handset or the big servers on the network or both and stayed synced, and whichever computer with a browser you walked up to would be also be your full phone. Awesome! I wish I had Skydeck in the UK now.

And then I went through my log of texts I have received -- well, my Inbox -- that my little N73 will gladly keep letting grow for as long as it has memory left. (Which is also why an N73 gets dog slow in opening a new text after a year, unless you clear the Inbox out.) And yes, it was filled with 140 characters or more of jokes, thoughts, business ideas, obscenities, sarcasm, flirts, insults, meetings, fury, navel gazing (hey, I Twitter), and outright hardcore sex that well-rounded adults with an over-reliance on their phones would receive. And I wondered: does this really need to be on a 3d party server?

Well, first of all, would I even be allowed to use their service? Let me translate Skydeck's Terms Of Service into English:
You agree not to do any of the following while using or accessing the Site, Content, Services, Your Data or Skydeck Reports:

Post, publish, or transmit any text, graphics, or material that:
You can not send texts using Skydeck that
(i) is false or misleading;
i) contain a lame excuse (a.k.a. a lie) about why you are or were or will be late or will not show up;
(ii) is defamatory;
ii) describes someone as a ho;
(iii) invades another’s privacy;
iii) describes exactly why that someone is a ho and how many were involved;
(iv) is obscene, pornographic, or offensive;
iv) describes exactly what that ho did;
(v) promotes bigotry, racism, hatred, or harm against any individual or group;
v) includes a reference to a blonde joke, or the whole joke, relevant to that ho;
(vi) infringes another’s rights, including any intellectual property rights; or
vi) describes how stupid the ho's latest idea for a new kind of mobile sex chat web service was;
(vii) violates, or encourages any conduct that would violate, any applicable law or regulation or would give rise to civil liability;
vii) well, no, that would be awful, I am glad none of my texts ever do that.

In other words, Skydeck doesn't want any normal humans to use their service for what they actually text. Especially anyone between the ages of 7 to 50, but I bet this generation won't grow out of calling people ho's after we turn 50.

But still, suppose we break these Terms Of Service, which I guarantee you 99% of Skydeck's subs using their text facilities already have done. Facebook has already brought home for the most people that Internetting your non-work life -- not even your private life, but just what you do outside work -- can kill your reputation. Oops you were drunk on camera once and a friend uploaded it, and now the Internet will a) never forget b) let anyone find out by entering your name. Blogs suddenly also blurred the boundaries between having an opinion and getting fired for it because your employer thought it would make them look bad. There are real repercussions to uploading your life to the cloud. (Meanwhile, the cloud is also capable of forgetting exactly what you wanted to keep. It's never fair.)

What I have put up on the Internet in my younger more naive days is bad enough, if my text logs for any reason would become public I would be simply mortified. Skydeck has a privacy policy and a security policy and writes about how they are so aware that they need you to trust them. Most of their security policy is bog-standard baseline stuff that any website that needs a password to access it should do, but there is one place where they go the extra mile: Skydeck claims all your data on their servers is encrypted. I hope that means the data of the account holder is encrypted with the password of the account holder so that all accounts are encrypted independent of each other, and not that all account data is encrypted with the same key, because that would still create a single point of vulnerability. But their privacy policy is clear: Skydeck will give your email address and your other data to 3d party service providers (calling out mailing list administrators as an example) and warns that your data could be hosted on any server in the world. They're therefore really open about the fact that an awful lot of companies that are not Skydeck -- and in my opinion 'an awful lot' is more than 0 -- will be entrusted with some or all parts of the data. Skydeck is not specific here, and every extra company is an extra vulnerability. They do say that the account holder gets to say whether another company gets access to the data or not, but that means that not only do we need to trust Skydeck -- about which we know very little -- we now have to make trust decisions about 3d party start-ups as well, with very little information, except their privacy policies. Policies which are non-binding, by the way.

Skydeck even makes the unfortunate claim that
we treat your cell phone records with the same respect that your bank treats your financial records.
Banks in the US and UK have recently been found to have made public the personal of how many account holders? By losing tapes and not shredding papers? Banks are not your golden standard for privacy, people!

Neither the security nor privacy policy has any teeth. Skydeck is not willing to put anything on the line here for you, not making any guarantees, just setting their own limits of liability (and they seem to be capped at $20). Seriously, go through your text logs and wonder how many of those you want the world to be able to read. Most of you will be ok, but many many of you would be toast. Red-faced, I-need-to-move-and-change-my-name toast. I can't think of a jucier target for hardcore lulz than those servers once Skydeck and their competitors takes off.

When I mentioned this in my first Tweet response, I was told, well, Skydeck isn't doing anything more your MNOs would or are doing. Well, actually, no, that's not true. MNOs are under stringent FCC and FTC regulations. And from the regulatory point of view in the USA, a text is a private as a phone call. AT&T is not allowed to just record your call, and T-Mobile is not allowed to start mining your voice-mails for data beyond what they need to keep their network running. If any of those companies wants to listen in or share any part of the content of these communications, they need legal papers. When AT&T got caught recording customer calls to hand over to the a legitimate government security organization, court cases and legal debates ensued and a law had to be specifically passed to clear the network operator of guilt (which up to that law they were swimming in). When I worked at an MVNO I came up with a service that would keep our subscribers safer by having our computers monitor their texts. When I presented it Legal said basically NFW, it would mean we were eavesdropping on our own subscriber's conversations and we could not legally do that. Being an operator comes with being in a framework that has real teeth to create real protections for consumers. Skydeck can go get and store your stuff because you give them permission to act as your agent, yet since they are not an operator, they are not in the regulatory framework of phone privacy set and enforced by the FCC (but still, are they allowed to store texts other people sent to you? That's different from you giving them permission to store the texts you write). It's like your email provider. Google says they are keeping my mail safe too, but geezus, man, what if they don't? And I know they are mining it, they are quite open about it. Meanwhile the postal mail has its own policing environment to enforce nobody but you reads your mail or tries to defraud you. Should we really not think a little more about these things before we click yes?

This hit home for me when the blogging network I was on which I made many sexy gossipy angry lulzy entries, stringently protected to be for my best friends' eyes only, suddenly got bought by a company outside the US in a politically dicey environment. This new company has no legal obligation to enforce the privacy controls I signed up for. Of course they didn't switch them off, they are not stupid, but I realized then that the trust relationship I entered 6 years ago with a scrappy start-up is not the same I am continuing with another company far away run by who knows who. Skydeck promises that if they get bought or go bankrupt they will give you time to delete your account, which again puts them ahead of many Terms of Service documents out there, but that is all: they do not say how much time is enough time, and the terms really have no teeth, especially if they go bankrupt. Your data is on their servers, and if they think 2 hours is enough time, two hours between their email going out and making all your texts public is all the time they need to take. Yes, this is an unrealistic and absurd scenario. And I doubt Skydeck will be the company where this chain of event happens. But the absurd happens, so it will, to some server where you have entrusted your private data, sooner or later as we migrate more and more to having lives on and off the cloud. It's gonna be messy.


Thursday, March 12, 2009

The Era Of Stupid Phone Dongles May Be Ending

Nokia N-GageNokia N-Gage
Image via Wikipedia

When I tested the original Nokia N-Gage -- no, stay with me, I swear this will get relevant to the real world -- one of the accessories included was this insane headset with two standard headphone plugs attached to it. For real. I am not entirely sure what the thinking was there, maybe the idea was to finally have a phone you could easily put your own earphones on instead of being confined to expensive accessories that had the manufacturer's proprietary plug. But since standard headphones did not have a microphone, I guess the N-gage needed a second jack to plug the mic into, and therefore the headset accessory in the box was this integrated earbuds / microphone monstrosity that needed two plugs, color-coded so you would know which one to put where. As an accessory it was totally in-line with the N-Gage itself: seemingly based on a sane list of requirements of which the implementation was then completely overthought and overdesigned on a tangent of what was happening in the real world, to then never be validated by the target audience before release. "Innovative shape optimized for gaming, keeps costs down using standard S60 parts, uses cheap standard media, appealing to gamers" turned into side-talking, needing to take a cover off to switch games, load times to get the game off the SD card, and go-faster stripes styling not seen elswhere besides late 80s BMX dirtbikes.

No phone is taken seriously as a music media device unless, through some dongle or by direct design, headphones or earbuds with a standard 3.5mm stereo jack can be used. But that standard plug does not include a way to use a remote control, or a microphone, two things that music phones could always really use. I remember using portable CD player in the early 90s whose remote on the wire worked by emitting little beeps that the player listened to and interpreted, while other devices with wired remotes always had strange plugs. The designers had to compromise.

Well, Apple released an iPod Shuffle yesterday with no buttons on the main device, and migrated the controls to the wire on the headset. Not the first software audio player to do so, but the first one to really need it: the main device is just too small for buttons. Sure you can buy a dongle (this is all sounding awfully familiar) to use your own headphones, and the dongle will cost extra and include the remote on the wire. But the actual jack used is a simple variation of the standard headphone jack, so simple any manufacturer of headphones can copy it (5 rings instead of one: left, right, control, mic, power). And manufacturers now have. Because the Shuffle will sell so many units that making add-on products is not a niche. This jack is now a consumer standard. With microphones as well that will work on other iPods. A new standard that other portable media manufacturers could have set, and simply never did.

Yeah, we may have to buy a stupid dongle from Apple now for our preferred earbuds and headphones, but in a few years we won't because standard earbuds and headphones will have the controls and mics built-in. And almost all other players and phones -- starting with cheap units designed by rock-bottom Chinese outfits who will always eschew proprietary lock-in for cheap availability -- will be made so that those headsets and remotes will work the same too.



Saturday, March 07, 2009

Crossing The Streams of Online Culture

When Twitter started getting noticed, I got an account. No idea what it was really for and what to do with it, I just didn't get it, but it was hot and it was there and now I had signed up. I decided I would be a modern day Jenny Holzer instead, except I could never think of anything as deep to say (and that is saying something) and besides, she has her own Twitter feed already. Twitter's IM gateway still worked so it would nudge me to update, and since I would never be at my desk my IM would respond with my IM away message, which therefore got tweeted a number of times early on until I shut that off. Yeah, I didn't exactly ramp up the followers with that.

I also had a personal blog, and quickly decided that making my readers have to follow me in two places was just not what I wanted. I auto-post my flickr feed onto my blog, so why not my Twitter feed too? After all, it was all my stuff, right? I didn't like any of the tools to crosspost Twitter onto blogs, so I rolled my own at the end of 2007, AutoPostBot, just for me, (and jokingly defined Web 3.0 in its manifesto). I wasn't sure if I was hurting my blog community (which in this case is very tight-knit, almost insular) with these very short updates that were so different from the long-form updates blog communities are used to -- other bloggers posted very short updates to their blogs too after all -- but bloggers in that community were already complaining about Twitter gateways like LoudTweet.

The problem I had with Twitter was its lack of dialog. It seemed so one-way, a tweet, with no way to comment or respond to them, and thus very unsatisfying. My re-posted Tweets actually got comments on my blog, just like every other entry, so clearly people wanted to talk about these updates. Humans like to talk to each other, not just shout into the larger voids (unless they actually are Jenny Holzer). So of course the Twitter community created its own dialog style with the @ tweets and topcis with # keywords to create structure, and thus the tweets themselves became a dialog with threads. The website supported it, and suddenly Twitter clicked for me: it was one big 140-character chaotic chat party, with the odd celebrity thrown in to converge around, that one could use to get some attention and give some props.

Getting deeper into Twitter culture meant that my Tweets were becoming more tailored towards Twitter culture in a way I cannot define, but saw happen. Suddenly having them appear in my personal journal made those entries seem disembodied and off. Even though AutoPostBot did not repost @ Tweets -- those are tweets directed to someone else -- and only reposted my general tweets, I started using # keywords and abbreviations that made them look really off in my standard journal that did not have 140 character constraints. Kinda stupid, actually.

I learned from this that to manage your own personal feed reader to be a mix of different streams and dialog groups and cultures for yourself to follow is do-able, but creating your public feed to be such a mix makes it difficult for the readers of that feed to keep track of who will say and see what and where, who people are, and why you are writing like you do. The narrative of the feed becomes very disjointed with regards to the primal human need to communicate and socialize. We tailor how we listen and talk to the medium we are using and who expect to be participating as well, after all. Putting the odd picture on my blog from Flickr is not a big issue, but grabbing an outtake from a whole other set of conversations and only showing my side during another public conversation, well, it turned out for my readers to be veering into the territory of having to sit next to someone talking into a mobile phone on a bus. ("Uh huh. Uh huh. Like, so, like, she used the big knife? Uh huh. She clean up the blood, right? Uh huh. Uh huh.") You sign up for that experience when you follow people on Twitter, and Twitter makes it easy to follow the @ comments to see who the tweet response is too. But readers of my personal blog did not sign up for that kind of conversation.

I shut down AutoPostBot, and shelved all plans to develop it further. I did keep Facebook set to repost my Tweets as Facebook status updates, and that works just fine: the format is also very short for status updates just like Tweets, the gateway does not repost @tweets in Facebook so only stand-alone Tweets get shown, they are exactly what people expect from a status update, they still get to comment, and the stream of updates go so fast that a tweet quickly moves from view. Some dialogs cross well.

PS: @blinker : you are just too kind.

Wednesday, March 04, 2009

What Age Do We Live In?

Whenever he heard the phrase "Space Age Technology" to tout the marvel of some product, my partner at the time liked to add "That means it's 30 years old."

"Mobile-phone Age Technology" just won't sell mops, though. Doesn't roll off the tongue either. "iPod Age Technology" works better but is totally trademark encumbered. The "Green Age Technology" is not happening and too close to the 60's "Green Revolution".

Technologies are really running together too fast to be in any kind of age. The "Broadband Age" is already being eclipsed by the "Wireless Broadband Age", although only in selected areas, of course, and neither will do much as a slogan to sell potato-peelers on TV.

I'll take suggestions, but I don't think we are in any kind of technology age right now. Just more tech, and faster.