Member Since: 28th Jul 2002
21st Mar 2019
Fine for cats and dogs.
Hasn't this story been released 10 days early?
13th Mar 2017
I'm talking to the target audience; accountants. My warning is simply that the Cloud is most definitely not the panacea it's often portrayed as being. The reason that large companies such as Microsoft like the cloud is because it locks customers in to a subscription model from which it's hard to extricate themselves.
Buying off-the-shelf, or rather, out of the cloud solutions, particularly from the medium and smaller vendors, needs scrutiny. These people own your data and it may be hard to get it back in the format you need to move to another vendor later.
Most companies need some kind of system to run their business. The most cost-effective is often COTS - commercial off-the-shelf systems. If running in the cloud there are many new caveats to emptor.
And, of course, if choosing developers for a bespoke cloud solution, be careful that they have the experience behind the scenes. There's many a pig in pretty lipstick. What they're selling you now may well be impacted by other customers of theirs as it could well be a shared environment. Or they may well update it in ways that doesn't suit you - just look at the mess that is LinkedIn.
Cloud is good. Cloud really is the future. But see the reality as well as the hype.
28th Feb 2017
I'm a techie that builds this stuff and I have had to go through the "paradigm shift" (yuck) that the Cloud computing architectural change has brought.
We're looking at two ends of the spectrum here; you're looking at general purpose applications, I'm looking at bespoke developments.
My biggest concern is security. You have to take it on trust that the cloud vendors do what they say they do, because you no longer control where your data is and who can get to it. If you have secret data -- e.g. PCI card numbers -- then you must build the appropriate security into the systems assuming that everyone can look at your data.
The architectural changes are also huge. The database tier is provided by the vendors (e.g. Amazon's RDS) and this is resilient out of the box, but your application will need to be written to cater for this - it's no longer a simple connection to a single database running on the same box.
Anyway, I spend a lot of my time integrating with other vendors who have real problems with scaling. There is a large amount of learning how to properly develop to exploit all the cloud has to offer.
I have a deep dislike for marketing gimps who give the alternative truths to the Cloud, painting it with the rosiest of tinted spectacles -- no, I'm not accusing you of this!
Managers and accountants need to appreciate that costs will change. You can't buy the asset and have a fixed cost. Cloud computing is uncharted waters when it comes to costs -- you just don't know what the bills will be as there's so many variables: network bandwidth, virtual server types, CPU performance, storage, memory, etc., etc. Obviously this is offset by not needing "men in white coats" to look after your kit, there's no power bills, maintenance, etc.
But you will get big bills. And you will still need specialist developers, architects, security specialists, network people, etc. Little boutique development consultancies are good places to seek out people with those skills.
The future's bright. For developers:-)
26th Feb 2017
Throughout history computer architectures have drifted between a centralised model and the client-side model. Cloud computing is, in essence, a shift back to centralised computing.
The beauty of this model is the centre's managed by people who are experts in this. Professional administrators, architects and developers.
Now, step away from flogging your latest wares and get into what the end-client needs. End-clients need software solutions that work - everything that you've said.
You, as a vendor, want the client locked in as tightly as possible paying a "subscription" every month on pain of deletion. This is completely at odds with end-clients who do not want is to be permanently tied into a specific vendor, i.e. a monopoly supplier. If they are, then they're tied to whatever that vendor supplies; they've little chance in changing or staying on an old version that "works" for their business.
Similarly what when the aggregated data becomes very interesting to some seriously good hackers: you're stuffed and reliant on that supplier having met their duty of care. Meet collateral damage.
I think Cloud Computing is great. To be able to configure an environment in minutes -- not weeks -- is amazing. To stand up a test stack temporarily for next to no cost is truly amazing. But you must have people who know what they're doing, you know, professionals. These people aren't cheap and are in great demand.
Accountants are famous for knowing the cost of everything. There's a massive value in Cloud Computing; but it's got to be carefully managed and takes a big change in the mindset of the end-client.
24th Feb 2017
One of the main reasons cloudy stuff runs slowly is that it's out of your control and in the control of goodness knows whom. As any cloud-based application will be running at a *much* larger scale, all sorts of scalability issues will be encountered.
A lot of the smaller applications which run on a PC or a small company server simply cannot scale to the application sizes encountered in cloud computing. Thus you need all sorts of specialists to untangle the mess and apply the performance enhancements required. This isn't cheap as it's like having a team of specialist accountants running an investigation and audit. Suddenly you're needing specialist enterprise/solution architects, communications and security specialists, web developers, application developers, testing teams, and a bevy of project managers to put it all together. Loadsamoney.
Then there's the simple issue of many customers sharing the same server stacks: a failure of which will bring down everyone. If you're lucky this will often this can be peak time delays -- HMRC for example. If you're unlucky you could loose the lot.
Somehow all of this gets forgotten in the rush to proclaim the Cloud as The Promised Land.
When done properly, cloud applications are superb. Poorly implemented they're an utter disaster.
What's not mentioned is that Cloud computing is pretty much analogous to the old Mainframe computer model. You're centralising all your data and application development. What goes around, spins around, etc.
23rd Feb 2017
Struth, words like Nirvana.. What could possibly go wrong.
Loose all your data? Privacy? Security? Resillience? Data ownership? Tie ins to specific vendors? Subscription model? Uncontrolled costs? Migration in (and out)? Switching vendors?
One person's Nirvana is another person's purgatory.
You cannot simply switch to "The Cloud" without careful consideration of the implications.
Why is a vendor pushing that particular model. A benevolent vendor? Do you believe in fairies too?
28th Oct 2016
Had to look up the TLA for MTD. Kind of illustrates the problem...
29th Oct 2015
Incorporation is required
The amusing thing is that one can't contract for the gubernment *unless* you work through a limited company. It's required along with all the insurances, etc.
Not that it matters nowadays as since they've implemented their G-cloud bidding system. It's all being outsourced anyway; mostly offshore and frequently outside of Europe to places that pay their staff less (but still bill full rates)
7th Feb 2014
SEO - selling snake oil
This is so wrong on so many levels... SEO is utter snake oil and most of the time is overpriced nonsense. Sure, it has a use, but common sense will almost always give you the same results.
Amazing that their turnover was so high. Oh, of course, this is deluxe snake oil.
What criminal charges are following? If someone stole £140k they'd be facing a stretch of porridge.
Extradition proceedings? Oh, of course not, it's the wrong direction.
22nd Nov 2013
Did they pay tax in Portugal? If they did, surely the double taxation rules apply. If they didn't, fair on the HMRC for getting the tax that's owed.