Cloud Computing, Microsoft and the "Ray Ozzie Has a Plan" article... [long post]

[sarlacii]

I have just finished reading a very interesting article, entitled "Ray Ozzie Has a Plan" in Wired Magazine 16.12. A link to the article on the wired.com archive is here Ray Ozzie Wants to Push Microsoft Back Into Startup Mode. It makes for some interesting reading... Microsoft's plan for the future, which millions of users around the world will be adopting (or maybe not...)

To kick this discussion off, let's theorise that there is a company in Africa (say, South Africa) which comes with about 70 PC users. Okay? The situation is quite real:

1. Currently using Office 2002.
2. Currently installing, say, an ERP package that requires an upgrade to at least Office 2003 to get full functionality... export to Excel etc.
3. Who still retails Office 2003 in reality? No one... .
3. So, pricing for 40 licenses of Office 2007 comes in at about R198,000 (approx. $20,000) - you go for 40 instead of 70 to try and limit the damage. This amount of money is enough to employ a graduate electronics engineer for a year, to design a product that your company could sell for a decent profit for years to come, i.e. it is a lot of money to spend on buying something that you in effect already have.
4. This is South Africa, so internet access is relatively good compared to the rest of Africa, but it's still comparatively expensive (R250 for a 1Gb per month 384kB ADSL connection).
5. You learn that Microsoft, in their arrogant way, have decided that cloud computing is the way to go. You will soon be faced with a new choice of Office, all others obsolete and no longer supported, that runs in a web browser and stores all your important info on a server somewhere outside of your country.
6. You are still trying to run a business on XP (not supported) and perhaps now Vista too, since those new laptops you got came with it pre-installed and there are no XP drivers. You tell DELL that you are happy to spend an extra $30 on the "XP downgrade" option on your PC's and laptops, if available.
7. You start to realise that you may not want to embrace "cloud computing"... it will cost you a lot of money... again. And "software assurance", like protection money, balloons the cost of those 40 licenses to R700,000 ($75,000) over three years, so that isn't really an option either.

So facts about Africa (and probably South America too?):

a) cloud computing of the degree envisioned by Amazon.com, Google and now Microsoft is not big here.
b) You still need an OS to run your PC, and that OS needs to be stable and acceptably fast... cloud integrated be damned.
c) Bandwidth is expensive and often unreliable.

So the hypothetical company is basically being told that they do not feature in Microsoft's plan for the future (hell, just getting your windows update takes forever on a daily basis, never mind trying to access your "online" document).

What choice are we being given here? In the article, Ray Ozzie, Microsoft Chief Software Architect himself (!), tells us that Vista is "deeply troubled". So we are being told, to our faces, that the OS we paid money for is crap in many many ways, so badly so that MS is abandoning it basically. And we do not get a rebate. Here is "Red Dog"/"Azure"/"Whatever", use this instead, it costs more than the last one.

Linux, Apple, and any other OS... the door to the desktop is finally open... the pain threshold of company executives has been reached... a move away from MS dependence is the only viable solution. Lesson 1 in "how to screw up a software monopoly" is - thankfully I'm sure most will agree - complete.

Opinions? Flames? Let's hear it.

[GNU-Fan]

I use Free Software for a number of reasons. One of the more important is that I want to be in full control of my data and computing. I am not going to hand over any data which is of no concern to strangers until there is a compelling technical reason for this (e.g. the way email works).

Thank you, but my desktop can take care of my letters and spreadsheets just fine.
I am not to decide this for the African company though.

[sarlacii]

Well, as an addition, I went off to Google today after work and signed up for Google App Engine. Nice idea if you're using a shared internet term at a cafe (if you trust them not to have a key-logger) but not sure what the application will be for most.

The spreadsheet runs okay, small delays when entering numbers in cells, slightly longer when running a formula. Anything bigger, like a long recursive formula, takes a few seconds. Compared to running a spreadsheet locally it sucks.

Still, interesting.

[Bemk]

I don't like the idea of cloud computing either. I don't feel like placing my documents on a remote server, to which I don't even have physical access. I've got a server next to me and it works as a web-host. If I want to get my files into the world I'll use that.

I am working on the web-Linux project though, but that's because it's not written as a cloud computing system. We write this as a system which people can use to learn about the GNU/Linux way of doing things. Also it helps me understanding the Linux system better.

The web-Linux project is not a cloud computing system but it comes close, however I hope it will encourage many people to try this on their Windows box and start trying Linux for real after a while.

You can see the web-Linux project at WebLinux - Linux live on the web!. I don't think it will be nice for people with slow internet connections, because it will take ages to load. It's all JavaScript!

[Freston]

In my opinion the whole cloud computing thing is more of a hype than anything else.

[Bemk]

I think the total noobs are going to like it, but for most GNU/Linux users, nah, I don't think so. I think most GNU/Linux users will want to have their documents on their own drives. Cloud computing will become popular but no real geek will use it.

[Cool01]

For me, "cloud computing" is just a fancy name for the natural extension of where I have been going with my computer use for over 10 years now. The evolution runs like this:

- You find that you want a digital device to take notes on the go. On the one end, if you are doing heavy work, you might go for a laptop, and on the other end, you might chose a Palm device, or something like that.

- As time goes on, you find that you have more than 1 such device, and that synchronizing starts to become a chore.

- At the very least, you end up with a WiFi router at home so you do not have to wire everything up every time you need to move a file.

- Once you have the WiFi router and you know where you are sticking your files on your main "whatever-it-is". This could be a single PC that you use for most of your work, or, if you are a bit further along, possibly a small local server. Today, you might get into a NAS. Note that in the beginning, you might not make it accessible through the Internet.

- Eventually, you might leave your "whatever it is" connected to the Internet.

- If you are a business, you might get tired of maintaining it, so you might contract an external server.

Somewhere in the last 3 stages you are really doing the basics of "cloud" computing. All the stuff about apps is really secondary. It is mainly a question of data location and access, vs system complexity and maintenance. That is to say, convenience vs. cost and complexity.

As for the term "cloud" computing. It's a buzz word. Could be worse . . . .

[peonuser]

We have our own cloud computer servers. Along as we can secure the pipe for ourselves there.is no issue. Since no one uses the land lines anymore it may be the most secure way to telecommute.

[questio verum]

Sarlacii,
Thanks for posting the link. I enjoyed the read. I don't agree or approve of many of MS'es tactics, but I don't want to stick my head in the sand either. As for cloud computing; I think it's inevitable. Users will want the freedom to compute on the go, and companies will want the increased productivity it could offer. And both of them will want the increased interconnectivity. Obviously, this comes with a potential (some would say 'certain') downside.

The three biggest impediments, as I see them, are security, access to connectivity, and price points. I hope I'm wrong, but I'm guessing MS will wrangle the chip makers into some kind of cooperative arrangement. Whether Google, Amazon, and the rest buy in, is too uncertain for me to speculate on. I think Google needs to assume the lead on this a.s.a.p. though, or they will become irrelevant.

A subset of the issue of security is the issue of trust in the SaaS provider themselves. This is shadowy. Am I paranoid for not trusting MS to stay out of my data? I think not. But many will blindly offer up their data without a second thought. Yup. There's one born every minute. I suppose MS will offer up some kind of confidence scheme in order to euthanize discretion and common sense. The enterprise clients with highly classified data will probably keep that in-house. Same for .mil's, .gov's, research orgs and big-name content houses.

So if the security question is addressed, the next question will probably be price points. MS will probably offer some kind of price incentive for bundling Live Mesh with Red Dog. Not sure exactly how all that will fall. I agree with Google's CEO though... MS will push market share until they get the numbers to co-opt standards. They will probably offer the product only in the largest markets first, then spread out.

My third concern, not so much for me, but for emerging markets; is the question of connectivity. The whole idea is dependant on connectivity. Does MS plan to subsidize the internet? That would be a sales job indeed to sell a product to customers who don't have the vehicle to take advantage of it. Hmmm... Let me think about that one.

This is all built on the presumption that MS succeeds with this scheme. They may not. It's a very different world from the one that saw MS'es last truly successful OS release. I personally hope it doesn't happen, at least not with MS at the helm. Open-source is starting to get some play in the press. Under the right circumstance, with the right players, I could see taking advantage of the concept. I would never offer up my vital data though. And never anything of even passing importance, without a back-up.


As for Ray Ozzie, I respect the man. He's earned his stripes, even if he does spend too much time on the dark side. He's not a 100% dyed-in-the-wool microsoftie, which is a good thing. I just don't want to see this particular initiative succeed in the way he wants it to.

qv

[Bemk]

I just hope I'm not going to be the one who's running a dail-up connection in South-Africa then, when computers start coming with on-line UIs.

Would be a bit expensive don't you think? It will also be a bit slow, because of the low internet speed. If cloud computing is going to catch on it won't be the CPU dictating how fast your work gets done, but it will be the internet connection and the amount of users on-line. If every one starts doing cloud computing that will mean a whole lot of users which means the servers will be busy. I just hope the servers ISP can keep up!

I'll do my damned best to keep my OS on my own system. I think with GNU/Linux that won't be really much of a problem.