Discussion on:

History repeating: Why cloud computing could revisit the mistakes of the 1980s PC boom

the leaders of sub-groups within the organisation making decision that affect the whole organisation based only on their sub-group's limited needs. This often results in a decision that is very harmful to the overall welfare of the whole organisation; and that's what happened in the 1980s. One sub-group got a set of PCs that best fitted their sub-group, while another did the same, and then they found out the two answers weren't compatible while another answer would have given them both a 95% fit and been better for the overall organisation. Today we're seeing a lot more of this "stuff the organisational needs, I want" attitude with the proliferation of mobile devices and BYOD - each person gets what they want and don't care how it fits with the organisation's needs or legal requirements. Cloud computing is very much the same way.

Any decision about IT MUST be made by someone who knows and understands the FULL technical, legal, and operational ramifications of what the changes will mean for the overall organisation. This very rarely happens when the decision is amde by a single unit head in isolation.

Sharing data between different applications and business areas.
At least with an in-house I.T. department coordinating interface development and maintenance, the Finance department can send information to Accounting, and warehousing, and manufacturing, etc. and they can all speak to each other intelligibly. Few, if any, cloud services provide that capability across the entire corporate spectrum; and if they do, they certainly aren't going to be cheap, or as responsive.

I was working on mainframes at the time and remember this all too well. Want-2-be programmers coding spread sheets, amoung other things with no knowledge of test procedures etc. It caused a lot of problems with us trying to prove our data was correct.
And now the soapbox: Don't put your corporate data in the cloud! The data is a corporate asset. Can you afford to lose access to it or have it stollen?

Not that I have the answers, but more to clarify the questions.

The cloud is a sexy name for outsourcing, and there are significant challenges with managing both the delivery aspects of ANY third-party relationship as well as the technical, data management, and security of the company data, while making sure systems are available and meeting the business needs.

The Billion dollar question, is when push comes to shove, can IT say the simple word 'NO'.

If IT cannot say no, then it's game over, period.

The flip side, of course, is how IT can re-brand and reposition itself to call what it does something sexy like 'private cloud' so that all the suits can tell their golf buddies that their business is all cloud-based.

IT needs to take a page from Microsoft's playbook...Adopt, Embrace, and Smother. I say that with sarcasm, but I don't think it's an understatement to say that we are at a turning point...we see both this lemming-like rush to the clouds as well as the rise of 'cop in a can' GRC platforms. He who falls for anything stands for nothing.

The whole cloud kool-aid is a major shot across the bow to corporate IT.

The rise of the desktop PC is a VERY good analogy....and also consider what happened AFTER the initial love-fest with the IBM PC/AT running Lotus-123.

The good parts of PCs were adopted by IT, and the tide went back to IT. The server room filled with mainframes and mainframe controllers became filled with racks loaded with Compaq ProLiants running Netware, then Windows NT server, and now VMware, Windows, Linux.

Not to digress, but how different is a virtual server hosting a virtual desktop so different from an IBM mainframe connected to a bunch of color terminals? The answer is it's not. History repeats itself, and those who do not learn the lessons taught by history get to repeat the same mistakes again and again.

I think cloud-based services have their place, for example, web-hosting has been around for ages now and that is where this all started. Have lots of virtual web servers sat on a box connected via the internet.

I'm not sure the example provided is a good one, you've got to be pretty sloppy with your web control and so on to let the sales manager get to the point of doing that.

But it is definitely overhyped in a major way, I mean you can buy a 1.5TB external HDD for $100. Worried about losing it? Buy two. When you factor in the cost of having an internet connection and renting that space in the cloud, you begin to see what a ripoff it is.

There are so many downsides to the cloud though, legal liability for example, if so and so user uses your cloud-based service for kiddie porn and they turn it off, what then? What if the server is located in a foreign country? What about having to access data over a shaky internet connection instead of having it with you? What if you're in Cuba or another embargoed country? What if the law places restrictions on what you can do with the data (e.g. healthcare or govt. information)?

I read in the EU there is the Acquired Rights Directive which basically makes outsourcing very difficult.

The list goes on and on.

This is already happening on the BYOD front. There's a lack of integration and control of these devices as soon as they hit your business. Microsoft is trying to change this with Windows 8 PCs/Tablets and Phone but now IT has to convince the CEO that his Apple product doesn't belong! IT leaders need to take the power back and cure the disease before it spreads.

From someone who used to sell time sharing (TSL, ADP Network Services - remember them?) with dial-up to SaaS (databases, project management, specialist software - and raw computer power). The wheels came off when customers looked at their monthly bills and thought they could buy their own PDP-8 and put it in a cupboard for that amount (what infrastructure?!).
PS - who remembers the mini-PDP-8 in a (large!) briefcase? What was it called?

Everyone who warned about putting your data in the cloud has valid concerns which I share -- security being primary. Let's not forget, however, why the "cloud" has become such a popular idea. There are merits, especially for SMBs.

In real estate its all about location, location, location. Likewise, regarding corporate data, itss all about access, access, access. We in IT need to acknowledge the benefits that the suites see and offer them those same benefits. For some a public cloud infrastructure makes sense, but for most a private cloud makes more. We simply have to show the board that we are interested in empowering them to meet the companies objectives, and we want to help them do that in a way that doesn't end disastrously.

There will always be chaos when a business is outsourcing non core functions whether they be IT or anything that needs expertise. The business is then dependent on the supplier for advice and this, of course, will be that the supplier's product is the best there is. Getting competitive tenders without the in-house knowledge to assess them means that the cheapest is chosen no matter how bad it is.

attended in the early 1990s. The Australian Defence Force was very seriously looking at outsourcing what they could, all under the idea it would save money. Part of that was what they called the Commercial Support Program, the concept was to identify certain tasks done by uniform personnel that could be done by contractors or civilian employees. The bottom line was to identify what we could push outside the uniformed military structure so we could save big bucks, with much of it going to private enterprise if we could.

Part of the week long event was for us to split into work groups and do a presentation to the conference either for or against outsourcing. My group included a number of one and two star military people and a couple of civilian equivalents. I put forward my thoughts and almost needed an ambulance for some people before I could finish speaking. In the end, my thoughts were the basis of our presentation, and I had to deliver it. It summarised as:

1. Huge savings could be made by outsourcing ALL military operations by employing mercenaries as needed or by coming to an agreement with another country that had a huge military organisation, such as the USA.

2. We felt the above option was not viable due to the lack of control and responsiveness.

3. We needed to clearly define what were the operational and security requirements of our military were and what was needed to meet those.

4. We needed to reduce uniformed military activities to those that were needed to meet the requirements above, their in combat field support, and their training to meet those.

5. Anything else needed to be reviewed to see how the loss of control and responsiveness would have an effect on items 3 and 4, and all the rest could be outsourced in some manner.
.............

I think it's wise to say that all these points apply to ALL outsourcing decisions and need to be fully addressed before anything is outsourced, the only thing that needs to be added is to include legislative requirements in point 3.

The way I see it the problem really comes down to this, HR, accounting, sales, etc. people within on organization think just because they can put a PC in their house or on their desk that they know as much about IT as the people that work in IT. It would be like someone in sales that keeps their checkbook at home deciding they knew enough to select a new general ledger system without talking to accounting.

You also need to remember that IT is there to support the business. If HR has a need that is not being met by the IT department then they are going to be looking else where. Open communication and trust need to be built between IT and the departments it supports or money will be wasted on unnecessary or underutilized technologies.

Bill