It’s happening! Brizzly is doing the first stage of exactly what I was hoping for in my first two blogs here: they are auto-expanding all photo/video links contained in tweets, making a more seamless all-in-one twitter interface.

Now can someone send me an invite? :)

giantrobotlasers:

jonsteinberg:

People on the business side of internet software, constantly bemoan their inability to code.  I’ve been guilty at times of the frequent refrain, “My kingdom for the ability to code.”  However, I’ve found over the past year that the emergence of APIs coupled with eLance (or oDesk or one of the other contractor platforms) have made this expression of exasperation largely hot air.

For $500 and four weeks of late night emails to eLance developers, you can basically spec and build simple, rough apps that knit or build upon open APIs to create things that are interesting and potentially valuable.  To be clear, you can’t build complicated apps or the next Salesforce.com on this kind of shoestring, but you can achieve the kind of learning, vetting, and experimentation that is left undone if you don’t.

I call this process Hackable Business Development.  If you’re interested in a platform or service from an intellectual, career, or partnership prospective, you simply must build on it.  APIs are such a vital part of web business growth and extension, that the API is almost more important than the front-end.  So if you are really interested in a web site, the only way to understand its functionality and potential is to hack on its API.  An API is in many ways the equivalent of a living breathing business plan  - replete with the company’s view of its place in a highly competitive, fragmented world of web services.

I consider this the modern day equivalent of reading up on a business and sketching out your thoughts.

When I was interviewing to work at Majestic Research in 2003, I was asked to write up a few pages of ideas to grow the business.  I did this, but I’m fond of saying “ideas are like water”; they’re everywhere and they flow.  How much more impressive and educational would it have been for me to get a raw feed of API data from Majestic and model it.  I hadn’t yet discovered the Hacking Business Development model, and I was lucky enough to get the job at Majestic, but it would have been so much more valuable for me to have hacked for them.

API documentation is basically written in prose, so that’s always a good place to start.  Instead of taking a novel to bed tonight, take some API documentation.

Interesting ideas. I think even engineers could start using oDesk. I’ve written enough twitter apps to know a framework others should work in. It would make their job easy, and save me time.

I think it was Fake from Flickr who said APIs are “Biz Dev 2.0”. It’s so awesome.

Having _made_ one for Tipjoy, my take aways are basically that you should make an API first - then build your service on top of it. Eat your own dog food. Also, there is a good argument that if you make good coverage in your API, your data representation will be better.

People don’t talk about this specific issue a lot - but they do talk about it indirectly. When a company needs to do a redesign, the representation of the data is what matters most. If you change the data, you need to evolve what is already there, and swap out the new framework quickly. Also, the speed at which you build new tools is almost always related to the quality of your data representation. If you’ve built something good, new tools can build on top of it with little friction from data representation evolution.

Sweet!  In this case, I will have Web 3 beta ready by next Monday.

Or failing that, at least an alpha… or a tweet about it.

whitneymcn:

Interesting thoughts, considering now.

Also: I suppose I’m dating myself by remembering that the first modem I used (yes, to try to hack into bulletin boards in the ’80s) was my best friend’s older brother’s 300/1200 baud. Good times.

gbattle:

Apparently, on August 22, 1991, at 2:23pm EST (exactly 18 years ago) I was commenting on the newest Metallica album on USENET.  Wait, you’ve never heard of USENET?  You never used that NNTP address your dialup ISP gave you in the 90’s (just the SMTP address for email)?  Never played around on DejaNews before Google bought them?  Don’t worry, most people didn’t and haven’t, but there is a cautionary tale to be learned from USENET discussion forums when making predictions regarding Twitter.

It dawned upon me today that Twitter is just a faster, centralized, unstructured version of USENET.  Apparently, many smart people beat me to this same realization eons ago.

There are many parallels between the two:

• public discussion forum
• constantly streaming information
• short term memory/limited data retention
• ubiquitous access
• multiple clients across all operating systems
• unmoderated discussions
• message size constraint

What’s more interesting are the differences in what USENET offers:

• logical taxonomy of conversational topics organized into newsgroups
• asynchronous decentralized multi-node mesh of servers (like email)
• each mesh-node can deterine its own level of data retention
• threaded discussion groups with attribution
• multiple levels of identity, including complete anonymity
• open protocol, not an API (again, like email)
• optional moderated discussion areas
• client independent filtering methodology
• subscription to newsgroups and threaded discussions over individuals
• larger message size contraint allows for multi-part binary file transfers

But Twitter has some magic all its own:

• “real-time”* delivery
• SMS connectivity
• centralized search
• unstructured format allows developers the freedom to build new, more structured ecosystems upon it
• full unicode implementation (often overlooked and underappreciated)

All of this got me wondering if someone ever considered writing a Twitter to USENET bridge to facilitate distributed storage of the Twitter corpus.  As it stands now, Twitter’s memory via search is woefully limited.  With distributed deep storage, there could be a whole new level of innovation on the deep memory of Twitter.

From the other direction, if somebody added an XMPP** layer on top of USENET, pushing it into a realtime distributed mesh, that could be a powerful open competitor to Twitter with a taxonomy.

All in all, a few early thoughts on USENET and Twitter.  More to come.

* it isn’t real-time

** this is much closer to real-time

(Note: I could have said Twitter is just a big BBS, but that would only date me even more than the 1200 baud modem I used back then to connect to them.)

NOW YOU’RE TALKING.  Yes, please distribute the twitters.  See my first few posts on “Twitter as the backbone of Web 3”

When I first joined Twitter not too long ago, one of my first thoughts was that it reminded me of Usenet.  Definitely as a social culture, as you describe above, but also, in terms of the larger society that’s not interested in this sort of geekery, a similar divide exists between those who like social computing like Facebook, MySpace, etc., but can’t function in an ALL TEXT environment like Twitter.  They seem like different worlds. But technically they’re the same thing - it’s a bunch of status updates with links attached. That’s all social computing needs to be.  Whoever can get social computing down to its barest, most portable essence - status update/blurb, link, security keys, all carried on a “tweet” protocol - can break down the walls between all these closed networks, so that we can enjoy having an “internet” again.

giantrobotlasers:

Tipjoy is shutting down.

You can still sign in and cash out funds. We’ve also made it easy to download all your transaction history. But all other functionality on the site is now turned off.

If you have a positive balance, you will soon receive a formal notification describing the shutdown process, which will include a reasonable deadline before which you will need to claim your funds.

If you have any questions about your account, send us email: help@tipjoy.com

We have decided against continuing to pursue additional funding. After a long and hard look at the market and the situation, we didn’t feel it made sense.

When we evaluate why there’s been so much hype about payments on Twitter, and yet so little traction for us (and even far less for our competitors) it is clear to us that the reason is that a 3rd party payment service doesn’t add enough value. We strongly believe that social payments will work on a social network, provided that they’re done within the platform and not as a 3rd party. “Simple, social payments” is *the* philosophy needed to do digital payments right, but once a service groks that, they need only to implement it on their own. We’ve been the thought leaders in this space, we see the hype and excitement, and yet we know very intimately the difficulties in gaining actual traction. The only way to get around this is for the platforms themselves to control payments - then all people wanting to operate on that platform would have to play along. We believe that a payments system directly and officially integrated into social networks such as Twitter and Facebook will be a huge success.

Thank you to everyone who has supported and helped us along the way.

This sucks.  I maintain still that the interface designers need to get on this.  I disagree that the major players are needed to make this happen.  Client designers who can integrate all these social networks and cool little toys have a land-rush situation in front of them.  Whoever can figure out how to put it all together, with micro-payments a natural part of the “one stream” paradigm, will change everything.   I don’t see Twitter and Facebook being competitors at all.  Tweets should be protocol. Twitter can become the backbone of Web 3, if it figures this out.

blech:

Whenever I have a tag field, I use it, and I try to be consistent. So Tumblr finally having some sort of global support for tagging is nice. Because I was away today, I first noticed it when looking at other people’s posts on my dashboard (although this example is one of my own):

My first thought was “why the fuck are they using a hash in front of the tag? This isn’t bloody Twitter; Tumblr have more sense than to force metadata into the body of a message (well, except for all the post types that don’t have titles, but that’s another matter)”.

My second thought was to try clicking on the links, and it turns out that despite the post (which I eventually got to after wading through some Facebook nonsense and promotional rubbish) claiming that tags are normalised, they’re not. Or maybe they are, but not all tags. In any case, the ‘victoria line” tag page is coming up empty.

Still, one hurrah for highlighting tags, a half hurrah for a global tag search (that sort of works), and a big fat raspberry for pointlessly aping Twitter’s use of # to indicate a tag when there’s no bloody need to.

I was pretty excited to see Tag Channels today.  I’d was just NEEDING them yesterday, and wasted quite a lot of time researching to find out that you CAN’T do tag channels — well until today.  So this is overall great news to me.

As for the HashTag approach (which originated with IRC), I suspect the reasoning is to emphasize that on Tumblr, these are not “ordinary” tags just for filtering your own posts, as on every other blog.  It’s a different thing happening, that’s pretty cool, and imagine they hope will completely change the way people interact on Tumblr.

That said, is there still a way to filter my own content with tags?  Do I make a tag without the hash?  Maybe that’s the difference?

IMO, Twitter doesn’t need to fuss with “friend-management”.  I could even see Twitter abandoning the friend-management business once interface designers figure out that that’s where the business is… ie. filtering/sorting and distributed privacy/security, as well as providing a nice selection of interfaces for navigating your “one stream” of friend activity.

In my opinion, Twitter would be better off tweaking the API to become standard internet protocol for the social web, so that everything rides on “tweets”.  (See the previous posts.)

blech:

Roo Reynolds posted a pair of photos (1, 2) to highlight a feature he says Twitter needs: searching the updates only of people you follow.*

I don’t disagree, but Twitter might find implementing followed-only search difficult. This is because Summize (the acquired company that provides their search functionality) is somewhat disconnected from Twitter proper: it still has own domain (which stayed up on Thursday), for example. Incidentally, this disjoint nature means that people who run private (hi!) aren’t indexed at all.

It’s also why deleted posts aren’t deleted on search, but then another failing actually saves you: the search index is only goes back a few weeks.

Probably the only practical solution in the short term, given I doubt Twitter have the time or inclination to fix these issues, is to fix it yourself. Simon Willison has started running his own local DB, largely for search purposes. It’s tempting to do the same.

* I commented over there, but thought it was worth pulling out for a larger audience.

Thanks for doing the research! Interface design is where it’s at.

blech:

I went through pages of history so you don’t have to. Sounds like it’d be pretty easy to copy.

Finally broke down and wrote my own stupid twitter client on the train… it does groups! (July 3rd)
@rboulton as a matter of fact I did :) mongodb + about 100 lines of Python (July 3rd)
Having your own Twitter client that saves tweets to a persistent store is AWESOME - group support, search over just people you follow… (July 4th)
my dumb home grown Twitter client is now powered by djng! (July 4th)
@TrevorGerzen it’s really simple - just a script that dumps tweets in to mongodb every 60 seconds, and a one-page Django app that shows them (July 4th)
@rboulton @cackhanded haven’t decided if I’ll release my client yet (July 6th)
@JimPurbrick my custom Twitter client now archives everything from people I follow (in MongoDB), it’s a really useful feature (July 8th)

mikemorrow:

Unless this already exists, in which case you’ll tell me, right?

I want a “trending topics” list for just the people I follow on Twitter.

I want to be able to login somewhere and see a word cloud of what my friends are talking about this morning, what memes are going around, whose birthday it is, etc. And variable by time period, so I can see what my friends are tweeting about this week, this month, or just overnight.

I’m giving this idea to the universe for free…who can make it happen?

This is an interface question that Twitter interface designers should want to tackle, along with URL-elimination, and in general, filtering/sorting/tracking of flexible friends-list activity.  This stuff shouldn’t be particular to Twitter either.  I should be able to have one “Dashboard” or whatever — it’s up to the interface designers to be creative on that front — to see what’s happening on Tumblr or Twitter or Facebook or Digg, all mixed together however I like, and instantly sift and resort my friends activity by groups, tags, post-type, etc.

See my previous post for the WTF.  (Has that become my tag line?)

This box is to answer Mike’s question.

whitneymcn:

marco:

… all tr.im links will continue to redirect, and will do so until at least December 31, 2009. Your tweets with tr.im URLs in them will not be affected.

We regret that it came to this, but all of our efforts to avoid it failed. No business we approached wanted to purchase tr.im for even a minor amount. There is no way for us to monetize URL shortening — users won’t pay for it — and we just can’t justify further devleopment since Twitter has all but annointed bit.ly the market winner. There is simply no point for us to continue operating tr.im, and pay for its upkeep.

There are two stories here: how sad it is that this type of tool, despite being extraordinarily cheap and easy to operate, isn’t worth running for one of its most successful players, and the inherent danger that this shows in relying on shortened URLs for anything other than temporary, disposable use.

Third story: “we just can’t justify further development since Twitter has all but annointed bit.ly the market winner.”

1. I think that bit.ly’s decision to extend what it meant to be a “URL shortener” rather than just offer another cheap and easy to operate entry in the marketplace probably accounts for much of their success, but that’s another story entirely.

2. It’s interesting to consider the ecosystem that has developed around Twitter. This service that many people consider worthless, operated by a company that has yet to make any real money from the beast, is legitimately a kingmaker of sorts for a variety of other services. Worth some thought.

In “my model”… URL-shortening services evolve into URL-eliminating services in conjunction with Twitter interface designers.  There’s no reason to see a URL.  It’s already being condensed into near-nonsense.  Eliminate it.  Just give us a little icon to know there’s a link.  This is in the interest of interface designers.  There’s no market for more internet gibberish — but there is a market for its elimination.  Down with URLs, hurray for link tracking.  Of course there’s a market.  The next phase of industry is with the interface/browser/”web operating system” designers.

As for Twitter… establishing itself as a human-driven search engine is a good first step.  I believe the bigger step is to altogether stop being a “micro-blogging platform” and work the API so that “tweets” are standard internet protocol — the atom of networked social interaction.  (see my previous posts for the WTF)

.

My original article “Why Twitter is the Backbone of Web 3.0” was too long. Lemme simplify.

Round 1: Make a Twitter client that duplicates the functionality of Facebook, but in a massively-distributed way. ie. Make the Twitpics and YouTubes and Audio Streams all show up in the interface right in the feed. This is easy to do, and with competition should provide lots of reasons for Facers to move on over. Make it work with ONE LOGIN.

Once you solve that….

Round 2: Change the Way the Web Works.

• I wanna log into my computer once a day. I don’t ever wanna log into yours.
• This means no joining “networks”. This isn’t 1984 — We have an “Internet” now.
• When I send a message, I want it to be instantly available to all my friends in its original form.
• I don’t wanna type in a URL or paste an embed code ever again. Grab and send, receive and forward, NO CODE. Icons OK.
• I don’t wanna visit your static web page. TMI. Waste of bandwidth.
• I’d rather depend on my web client and my network of friends to filter your tweets.
• I wanna be able to quickly clip and re-assemble a media presentation from streamed web content right in the client/browser and “tweet” it.
• My client should automatically unpack my friends’ tweeted multimedia presentations, filtered how I desire.
• A “tweet” is the atom of Web 3.0. It consists essentially of 1) a link to a presentation file, 2) a text annotation, and 3) links to security keys at both the source and the forwarding “tweet provider”.
• A “presentation” is analogous to a “web page” of the next generation. It’s a loose container of web clippings, all pointing to the original source files, with instructions for clipping and (optionally) reassembling them.
• On the receiving side, I want the option initially just to see tweeted content elements in my browser, sorted how I like. (ex. only photos and annotations arranged in a Facebook-Feed-style layout.)
• But if I’m intrigued I could gesture to see more of the presentation. It may resemble an adhoc “web page” presentation, but it could also show up in my client simply as media elements separated how I like too see them.
• There are no “play/pause/stop buttons” to fuss with in the presentation. Redundant. Just use the keyboard or operating system controls.
• In general, interacting with web pages is old news. Rather than managing back-and-forth and URLs, browsers will become more about tagging/filtering, presenting/tweeting, and just socializing.

Comments? Questions?