Why it always rains in The Cloud

c: | f: /

We’re told that cloud computing is the end of the road for bought software and piracy. That it’s the way of the future. So why does it feel like the 1990s all over again?

Cloud computing sucks for office. It sucked in the 90s and it sucks now. The only differences between 1994 and 2012 are processor clock cycles, broadband, and marketing budgets.

The two leading companies trying to reinvent the Brave Old World of thin client office productivity — essentially dumb terminals that run apps and store data on servers instead of on the machine’s own drive — have both approached it wrongly.

Micro$oft: I need a dollar

As usual, the software giant have taken a problem they created — piracy — and have tried to eliminate it by stopping people owning software. Instead of buying Office for $60 you can now rent time and disk space on an online version. If you only need to read documents that’s currently $8 a month but if you want your users to be able to author documents and contribute to articles — let’s face it, the majority of use cases — it’s a whopping $20 per month per user. For a small business with 50 employees, that’s a grand a month.

In comparison, if you wish to licence the bought product of standard Office for those same employees, Micro$oft will cut you a licence deal for a ridiculous fifteen thousand dollars. Allowing for loss of interest on the up-front payment, that’s equivalent to about 18 months of online renting before you break even.

The third option, which is the one I’d use, is to give each employee $60 to buy their own personal copy of Office and ask them to install it either on their home machine and work from there, or their work PC. Payback in around three months with no ongoing costs. For even greater frugality, Office 2003 is only a couple of dollars per copy and works fantastically well.

It doesn’t matter whether the online versions work or not and allow collaboration that free accounts on box.com and asana.com can easily replicate, the fact the service is so damned expensive is enough to turn people off the idea.

Google: Somebody that I used to know

As usual, the search giant think they can fix a problem by making it look cool, employing overpaid, half-assed interface designers to screw up even the most basic user interaction tasks, and then telling everyone how great the company is.

The advantage over Micro$oft is that GoogleDocs is free for Google account holders. But having to endure the tools on a routine, often daily, basis I can report that they really do suck the fun out of writing. In exactly the same way Micro$oft did when they “improved” Word from the solid, reliable, predictable and functional Word 2000 (and to some degree 2003, aside from image support) to the flaky, nebulous, usability nightmares of Word 2007 and 2010.

GoogleDocs is slow. It’s unwieldy to organise documents. Formatting of text and spreadsheet cells is circa 1993. There are outages. It stops. It stutters. It crashes. It’s illogical. The only useful thing is the interaction of data collection forms with spreadsheets, but getting it going quickly is like wrestling an oily goat on a skidpan.

The time I spend fighting the tool to create a half-decent looking, functional document to share with colleagues, I might as well have chiselled my words into a stone tablet and ferried my workmates to my house to discuss it over pizza and a whiteboard.

Cloudy with a chance of Illudium Q-36 Explosive Space Modulator

While more and more marketing and development money is being thrown at coercing people off their PCs and putting more information in the public domain — where it’s owned and managed by people covered with flimsy data protection / usage legalese — the tangible benefits of businesses doing so are not keeping pace.

If I was more cynical I’d say it’s like the companies are trying to solve a problem that doesn’t actually exist, then trying to entice people to use it.

What is the purported reason behind sharing everything? Collaboration is the main driver: work on a single copy of a document and get notified of changes, while chatting in real-time over virtual tea and biscuits. It sounds great; but so did Google Wave (actually, that was great). It sounds efficient. It sounds cost-saving. It sounds like the answer to everybody’s sharing itch.

But the bottom line is that it just isn’t.

It probably comes down to the amount of wasted configuration time and wasted time waiting for the service to respond, let alone wasted time punching the table or shouting at the screen when organizing / sharing documents using the array of badly thought out icons and buttons, or troubleshooting why people can’t access documents they clearly have access to.

It’s like hotmail. The entire experience and nebulous response times; having to click things two or three times before the click registers; the confusing interface and constant, tiny interruptions or text caret jerkiness that cumulatively irk me. It’s like being repeatedly poked in the face with a cocktail stick by an invisible assailant, and after a few hundred such pinpricks all adds up to leave a nagging feeling that the software isn’t robust enough for mainstream usage and that at any minute your document is going to burst into a shower of 1s and 0s.

It’s always been the way, and was the main reason thin client failed before. Put a shared pipe between two endpoints and, no matter how fat the pipe, the application is always at the mercy of the connection, not to mention the server load, authentication overheads, available slots, retries / SYN-ACK notifications, and variable routes for packets to travel between endpoints.

The deciding factor is confidence. Fix that — which is a tall order by any stretch — and people will flock to the tools. Until then, give me Word 2000, email, and a couple of free file sharing / collaboration tools any day of the week and I’ll outperform the cloud. With room to spare.

Your two cents

(required)

(required, never made visible)

(optional, linked with rel="nofollow")

(required)