http://blogs.zdnet.com/projectfailures/?p=560

Because in the end, Twitter is a communication medium.

You compare it initially to MS Word – which granted, is another tool for communication – but I’d say you missed the mark there.
If Word crashes and a Windows box blue screens, unless you are Bill Gates standing up in front of an audience (where have we seen that again?) there is no immediate impact.
But Twitter is a real-time app – and there are presently bugs that could have serious consequences in a large organization relying upon it as a communication tool.

Let me back-track for a moment…
The issue that I mentioned to Michael that ended up with him directing me here was as follows:
Presently, Robert Scoble (@scobleizer) follows 6,956 other twitterers and is followed by 6,880.
I follow Scoble, and he in turn, has me followed.
But the majority of the time when I’m viewing Twitter through the web timeline, @scobleizer’s tweets do *not* show up in my timeline. Neither the public nor the @others messages.

This occurs not just to me, but to several other folks I’ve discussed it with @aruni, @maxweb, @mackcollier to name a couple. All of us have @scoble followed and are followed in return. All of us have experienced the fact that most times, his tweets don’t show up on our timelines – and then from time-to-time, suddenly do.

Since Robert is a rather prolific tweeter – this can mean suddenly having 2-3 pages show up in your timeline that weren’t there the last time you checked.

Clicking thru to http://twitter.com/Scobleizer will show the volume of posts we are missing… but that’s usually only something you become aware of when you see a one-sided conversation.
In the case of tonight, I saw @mkrigsman tweeting @scobleizer, but not the responses – that’s how I knew to go to his page to see what I was missing.

From an Enterprise perspective – this could prove devastating. Because the assumption is that I’m seeing, say, Manager X’s tweets like everyone else is – only I’m not – until they show up en masse. Suddenly, I’m a time-loss – because once I discover that I’m not reliably seeing tweets, rather than checking the feed, I’ll be checking individual pages regularly. And Manager X? He’ll be ‘double checking to see if everyone got that message.’
This defeats the efficiency of a real-time communication tool.

Now, I’d suspect the issue has to do with how Twitter load-balances their servers – and it looks like it’s not getting replicated across from one to another reliably… but I don’t have a single schematic to look at to verify this. I’ve just dealt with some massively large distributed load realtime databases in my experience and know what it looks like when that’s fouled up.

I would suspect that it’s not just the case with Scoble, but others on my feed as well. But at the moment, we’re less likely to notice it in those who don’t tweet as frequently… it’s only notable in a case where someone is as vocal as Scoble and yet not present in the timeline.

So would I find it “good enough” at an Enterprise level? Absolutely not. The measure of a good tool is whether it increases productivity or decreases it. If you can’t rely on the message to get through, then you will start to add in redundancies – be it an IM program, emails to check up on reliability and message heard, or the good old ’sneaker net’ method of walking down the hall and saying “did you see my Twitter?” – and thus the tool proves to be a detriment rather than an asset.

Good enough for social media? Yes. Good enough for real-time Enterprise communication? No.

Thanks for your kind mention of the LAFD in your blog. For a brief yet detailed overview of how we’re using Twitter at the Los Angeles Fire Department, please read this article in the most recent issue of Governing.

We’ve got a lot of great web 2.0 projects planned for the year ahead, and welcome you to join us as we venture into the world of connective technologies.

Respectfully Yours in Safety and Service,

Brian Humphrey
Firefighter/Specialist
Public Service Officer
Los Angeles Fire Department

Sure, Twitter would be more useful in an enterprise / corporate environment with groups and finer grained permissioning (implementing groups is a long-standing request).

But doing so would change twitter from what it was, and from what many many people use if for now – a light touch, relatively disposable, ambient intimacy mechanism designed to bring and remind us of social and human interaction. I don’t use twitter to communicate vital info, I use it to make social gestures that bind me to friends and colleagues – to share jokes, occassionally highlight information, take the piss out of a colleague sitting 3 desks away, and being gratified when I hear a real laugh, and a virtual one. Adding groups and permissions would mean my twitterverse would be split and suddenly I have to make decisions about which group(s) I’m posting to, rather than a binary post/don’t post decision which is the current situation.

So yeah – for me Twitter is easily good enough, because I’m using it for what it’s best at – not as a mission critical channel for sharing vital company secrets. And while enterprises may not need mechanisms such as this to encourage social interaction, the human beings that actually make up enterprises certainly do.

While we continue to debate this issue in the abstract, I think it would also be useful to discuss what’s ACTUALLY goin on. To wit: who ARE the 700,000+ current Twitter users, and with whom (or to whom) are they Twittering?

Some of them, presumably, are independent consultants like me; but a lot of the tweets I read strongly imply that the Twitterers have “real” jobs in “real” enterprises. And they’re clearly twittering to people OUTSIDE the firewall — e.g., to you and me. What does that say about security? We all agree that Twitter basically doesn’t have any security, so I think we’re getting a strong message that 700,000 enterprise-dwellers are not that worried about the security risks that Twitter may or may not pose.

Meanwhile, what we don’t know is how much intra-enterprise Twittering is going on. If employees A and B both work for enterprise X, do they Twitter to each other while they’re behind the firewall? Is it possible to license, or buy a “customized” version of Twitter that ONLY operates behind the firewall? If so, that might eliminate some of the security concerns that Michael has.

Bulletproof IT represents a reliability goal and is not a statement of perfection.

you say ‘However, for software to gain adoption by the enterprise, those boring qualities become show-stoppers unless they are complete and well-implemented.’

but as we all know (and is mentioned in the post) – most enterprise software is layered on unreliable, unstable software – typically manufactured by Microsoft.

I’m pretty sure that for most organisations time lost to desktop crashes, Exchange outages, network and internet outages is of a similar order (or worse) to ‘just good enough’ services like Twitter. And by most orgs, I mean the average..not the top 1% with clued up IT people who actually know their stuff.

This notion of bulletproof enterprise IT is, for the most part, a complete myth, and is normally used as a self-serving excuse / barrier by enterprise IT to prevent and hinder.

– As Ed correctly asserts, “good enough” involves a delicate balance of trade-offs between function, goal, and cost. So the real issue becomes how and where to draw those tradeoffs. It’s a matter of judgement made by the developer, with respect to a particular audience at a particular time.

– I never said (or even thought!) that “that reliability/security/availability/stability is the most important characteristic of a software system or product, if not the only one worth discussing.” To the contrary, those infrastructure qualities are pretty darn boring and uninteresting. However, for software to gain adoption by the enterprise, those boring qualities become show-stoppers unless they are complete and well-implemented.

– The real question is this: “Is Twitter ready for the enterprise?” When I use Twitter at home, yeah, it’s good enough. But since I like Twitter, and want to see it serve as a component of broader changes taking place in the enterprise, I am forced to say that Ed’s perception of “good enough”, in this case, is just not good enough, since it ignores what I cover in the next bullet.

– Finally, and this is the most important point, to properly address these issues, we must look more broadly at computing and IT changes taking place in the enterprise. The relationship between business and IT in the enterprise is shifting dramatically, and this is where we must look to properly answer the “good enough” question posed by Ed. These shifts are a key part of Naked IT, as I have described it on my blog.