140Char

There’s something about Twitter’s recent decisions that just doesn’t make any sense to me at the moment.

It’s experiencing huge growth and celebrity adoption, it’s appearing in mainstream media (Even The Sun is carrying stories about Twitter),  and it has some money in the bank.

But one of the big advantages of the Twitter model is that it can use the huge advantage of external developers and applications undertaking the risky part of establishing themselves before acquiring them or introducing something similar.

After all, Twitter acquired Summize to become Twitter Search.

But since then, things have been a little different. First Twitter acquired Values of n, the company behind Stikkit and I Want Sandy, with Rael Dornfest joining Twitter’s user experience group – and the former services being shut, which caused some outcry, and a group of developers to come together to create an open source version of Sandy.

Recently Twitter introduced a ‘suggested users‘ page for new registrations and alongside the options to find other members. Which seemed like a reasonable idea to encourage new users to start following and interacting with other members.

Sadly it’s rubbish. For two reasons.

• There’s no relevancy. It’s offered to new users before they have filled in profile details or sent any messages, meaning thatthe suggestions have no context, and are pretty much useless. And although many people have said this could be the start of monetisation, how much money do irrelevant friend suggestions make?
• There are far better options out there already. For new users with no context, why not simply let people take a look at relevant categories on Twitter user directory Twellow? And if you’re looking for relevant suggestions for people to follow, there’s already the totally awesome MrTweet. (Interview here)

The only time I've mentioned a whale, it had 'fail' in front of it!

Much better suggestions and information from MrTweet

And now there’s concern over the decision to limit API calls from applications. I’m not aware of how many calls are average, so take a look at Jesse Stay’s thoughts on the SocialToo blog, Mark Evans at Twitterati, or SocialToo advisor Louis Gray (Who very kindly recommended and linked to 140char today!)

The idea from Twitter’s perspective is to ensure reliability – which is certainly understandable giving the problems that sometimes occur – and the only services that aren’t viable any longer are those which notify you of people unfollowing, such as Qwitter (no longer with us) . But as Marshall Kirkpatrick writes at ReadWriteWeb, Twitter applications are developing incredibly fast and this could hurt anyone trying to offer something radically new. And as Rafe Needleman points out at Cnet, it seems very odd that Twitter hasn’t used this chance to partner with external developers.

On the bright side, the API limit should force more efficient use of the Twitter API, which will benefit everyone, and the Twitter Firehose and OAuth support are due within around a month.

Hopefully that will mean Twitter can stop worrying about the scalability and learn to love external applications again. Especially as they’re building financial models (e.g. Twittad and Magpie), and monetisable services (e.g. Stocktwit) which show where the cash is for Twitter without the internal team having to experiment at all.

The trust that many people have in Twitter has been shaken recently by three major events – but there’s one idea that could solve some of the problems.

The events have been:

• A major Twitter phishing attack, which resulted in spam Direct Messages from comprimised accounts.
• High profile Twitter accounts being hacked – apparently by a fairly simple brute force dictionary attack on someone with access to Twitter support tools.
• Strange Display Errors – which turn out to be due to the combined forces of mass tweets from Macworld and CES.

Stopping anyone with admin access from using a password like ‘happiness’ should cure point number 2, and deadling with mass traffic is something that only Twitter itself can solve.

However, the loss of trust in applications is something that effects the whole Twitter ecosystem, as Mark Evans writes on Twitterati. And even implenting the much-requested OAuth as a technical solution doesn’t guarantee a rogue app can’t affect people. (via the MrTweet Blog)

So what’s the solution then?

It’s a simple idea – there are a lot of sites currently listing Twitter applications as soon as they become available to be the first to carry the news, and also to be a useful resource.

But what about an agreement between some of the Twitter bloggers and established app developers to implement a testing and approval procedure – a relatively simple process which could then list approved and tested applications, and allow them to display an badge of approval.

What gives bloggers the right?

The reason for pulling together reasonably prominent bloggers to implement approval is that we have something to lose if we’re not utterly honest – anyone can update the Twitter wiki with a link to a malicious application, but if 5 prominent Twitter bloggers did it, we’d all lose trust and social reputation, so it keeps us honest.

So what are the benefits?

• A list of Twitter applications which are being used and monitored to ensure they work as stated
• An independent approval system by people with a vested interest in keeping things honest
• More authoritative testing, and a larger quantity of apps being tested than each of us stating individually which apps we use – and a safeguard in case we’re tempted to recommend something without taking a proper look because we’re busy or going on holiday that week.
• And it means developers can display something to give them a trusted status without the need for a paid store (like the iPhone store), or worrying about being tarred with the same brush as malicious scammers?

So I’m throwing it open – good idea or bad? And are my fellow Twitter bloggers interested?

Want to spread the word? Copy, paste and tweet:

A quick and simple solution to sort the trusted and honest Twitter apps? http://bit.ly/vL48

Following my previous post on the implications for Twitter of the first large scale phishing attack, I’ve seen a few interesting responses:

First up, @benbarden responded to my concerns over short urls by suggesting that people could host their own, e.g. 140char.com/link1 etc.

A pretty cool idea, and one that Ben is apparently running on a site already (I might have to beg him for a guide!). The only flaw is that a lot of people run hosted blogs, and will therefore still be at the mercy of shortening services. But for those of us paying hosting costs it’s worth considering.

Then the always friendly @mingyeow from MrTweet asked my opinions on a blog post ‘Addressing Privacy Concerns‘. Suffice to say it’s a very eloquent explanation of how and why the developers of one application are aiming to keep your accounts safe:

One of the points raised is that MrTweet will support OAuth as soon as it becomes available, although it won’t answer every security question, because, as they quite rightly say, securityand convenience are always a trade-off.

There’s some really interesting debate around the use of OAuth from both Jesse Stay on LouisGray.com, and Dave Winer.