28 posts • joined 28 Mar 2007
Mythical Spreadsheet Month
Mythical Man Month is a series of thoughts on two very big projects - the IBM 360 series of computers, and the 0S-360 operating system. These were comparable in size to the big failures we see today. One simple rule is if you have a big-ish project and it is slipping, the more people you add, the worse it gets. However there's much more to it than that, and it should be mandatory reading for all IT Managers. Unfortunately since it was written we have seen the rise of spreadsheet/plan-driven IT. Nowadays as projects slip, the big SIs just add in more spreadsheets, plans reports etc. One reason is you can charge say £400 for someone who produces working software whereas you can charge £1000 for a project manager who doesn't. And you can have as many project managers as you like. You're only limited by desk space and number of toilets. Having worked with EDS, Accenture, C******** and Fujitsu and doubt we will get away from this multi-billion waste. It's not just their fault by the way. The Public Sector ties them up in counter-productive bureaucracy and mind-bogglingly labyrinthine contracts. I don't see a way out until the world embraces Agile, and it won't happen any time soon.
Perhaps Some Journalistic Input?
It would be nice if Reg contributors were limited to those who explain what's going on. We seem to be having to piece together what this threat actually is. It would appear it is actually several code generators that are at fault rather than Flash, but it's not by any means explained. Journalists should seek out the facts and not just repeat hearsay.
I recently bought a £600 HP laptop. What you can't do on the Internet is touch things. I walked up and down PC World, chose the shinyest one and bought it there and then. They gave me Office and Norton 360 for £100 extra. Of course mainly I would buy from Amazon or Dabs.
It will only be a matter of time before Oracle, Microsoft etc will be things of the past. They have to keep changing everything to force you to hand over fresh wads of cash. Surely corporates will see open source as a way off this treadmill. Mind you, the next time an IT director behaves rationally, will be the first.
Already mainly Third Party Software.
"In fact there may be more chance of error as it probably won't be compatible so there'll be room for copy paste errors."
I think the majority of business returns are done directly by eg Sage. HMRC have a Third party Validation System for vendors to test their software for compatibility.
They do make rods for their own backs, though. I just did a Corpororation Tax CT600 on-line which was accepted on-line, then I got a letter rejecting it because I'd sent paper accounts, not using the PDF thingy on-line - I'd not be warned there was going to be a fancy format. It's little things like this that need simple sorting out.
This Doesn't Add Up
Maybe geeks go the do-it-yourself route but corporates want to see a salesman because they need to understand the support model, and that way they also get latest release/roadmap information. So to me it smells of they're not getting enough corporate response to justify even the sales people they've already got. Chucking out other people rather supports this feeling the revenue's not coming in.
What we do know is EDS provided the extracts and they were stored as 100 Zips. As far as I'm concerned that is encrypted despite what all these idiots are saying publicly. The protocol was the discs were sent without passwords. The recipient would ring up/email upon arrival and the sender would then email the password. This is consistent with my experience of HMRC. They were using this dataset because they couldn't get approval to spend money to get only what they needed. This is of course the enormous price you pay for outsourcing. What you could do in a couple of hours previously, now is almost impossible because of all of the layers of bureaucracy that snowball out of control once you get the likes of EDS, Capgemini, Accenture involved. It seems the single mistake here is the HMRC bods didn't imagine the enormous downside of the rather low probability risk they were taking.
Do you get this feeling?
Normally your brain is running an trillions of flops, amazingly decoding mulitple streams in a complex and ever-changing environment, and you don't feel a thing. However every now and again you hit a sentence like this:
"Web 2.0 social software has created conflicting generational user expectations"
There's almost a pain in your head as the logic simply fails to provide any traction and your brain cells spin desperately.
The people who write this nonsense are clever. They know what they're doing. I think we should form a vigilante gang - like the people who blow up speed cameras - and take out PR departments from Oracle etc into the car park...... and give them a good kicking...
Agile - They Don't Get It
I'm a great supporter of agile but it's essentially a reaction to people telling you how to do things with complex, prescriptive, structured methodologies. So why do people keep cynically spoiling the theme by trying to tell us what agile is and how we should go about it in detail??
This is Big News
You don't talk about being taken over if you think your future is rosy. My take on EDS is they persistently get good opportunities but then get into trouble. Some of it is the common problem of under-pricing but I believe their process-driven approach and obsession with getting jobs offshore are leading them down the wrong path. The interesting thing is other big SIs have exactly the same beliefs. Deutsche Telekom can't do any harm to such a significant chunk of our industry.
Bill Did the Decent Thing
I believe Bill went to Seattle Computer and was offerred the $50,000 per customer normal deal. Of course Seattle couldn't imagine any customer could possibly sell thousands let alone millions of pcs, or that they might be called IBM, so Bill trousered all the profit. He did I believe later incorporate Seattle into Microsoft so I guess quite a few millionares were made anyway. I dispute Bill had Xenix at this early stage.
It's bad enough that this article starts by misleading you into thinking the French have come out top in an IT efficiency audit but unfortunately it subsequently fails to correct that clearly so you have to finish up reading it twice. Even then I don't really understand what it's saying. In fact it's more a bit of french-style philosophizing. What actually was the ranking about and how does that affect us?
I was on a multi-million EDS contract that was clearly bound to fail. The salesmen had sold the impossible and then went off for a month in the Carribean. A disaster followed. It's about time this was stopped. Hopefully this will teach everyone a lesson.
Plus ca change
This could easily have been Cobol to SAP or Oracle Forms to Oracle Financials. Truly gargantuan amounts of money have been wasted migrating 'legacy' systems. In almost all cases you would see the same pattern - refactoring your existing code/logic is going to be a better choice. Use something new if the problem is entirely new otherwise don't.
Before you get too carried away, just remember Flash cannot directly replace RAM or disc. It can only be written to a finite number of times. OK that might be a million writes, but that can happen quite quickly if you're close to the operating system. Maybe OK for storing emails but other things might change more rapidly. Adding and moving items might only occur on a daily basis but what happens if you have an index? How many writes are there to that bit of the 'disc' per year?.
For future reference it's Capgemini not Cap Gemini.
For large companies you can find that data centre issues (power and heat dissipation) can mean tens of millions of expenditure. Also in outsourcing deals (common for the moment - surely will die out soon?) astronomic charges can be made per box so reducing the number of apparently insignificant servers can result in millions upon millions of savings. It only makes sense in that world but that world is very big for Intel etc.
Asus Drowns Foleo at Birth
A few hours after they announced this, Asus revealed the EEE at £100 or thereabouts. That was the moment the curtains came down.
Incidentally, is this the Worlds Slowest Site?:
I think you'll find transponders use their own altimeters set to 1013Mb. ATC see a corrected figure.
Everyone wants to be Larry Ellison
You do get the feeling MySQL is moving relentlessly towards a more traditional business model. Good news for Oracle. They charge more but give you more features. Not so good news for MySQL - EntrepriseDB, or somebody, might suddenly seize their crown.
XML is Questionable
You are quite right to point out that this is a programming task and nobody has justified why XML has been chosen. The idea of having a BPEL layer is of course persuasive. If it appears as a visual interface, you can argue it doesn't matter what intermediate representation exists underneath but usually it helps to see the underlying code to check the system is doing what you thought you asked. Also you can often cut and paste things better in text form than with a graphical editor. My concern is XML is an OK idea for certain types of data message and happens to work well for web pages but people seem to think it is somehow magic. I just don't see what you're getting by using XML for BPEL.
What's an 'addressable market'?
A slight exaggeration
Whoever said the Capgemini deal had gone to £8.5bn was distorting the truth. The contract was around £400m per year and crept up to double that at one point but it is entirely wrong to extraoplate that over the remaining 80% of the contract. Spending is much lower this year.
A revolution , but when?
The image doesn't do service to these displays which are actually high resolution/high contrast. The other key pluses are you only need power when the image changes and you don't need backlighting - so potentially very low power consumption. The downside is it still takes over half a second to change a pixel. I'd love to see a general purpose computer with this display. Presumably other people have judged it is not acceptable to users but the promise of something that runs off solar cells is very seductive.
Obviously not quite there yet...
"MySQL does not manage relations"
Nice one. We have an article on relational database becoming part of the furniture and we find someone who thinks a chair is what you put the washing in.
I love lightweight computers and I love e-Ink. Trouble is all these things are five times as expensive as they should be, so no-one buys them. This will be the same.
Enormous missed opportunity
The idea that if something is not 100% reliable then it must be worthless is laughable except when it comes to lie detectors. I therefore suspect people just don't want to be caught out lying.
If we routinely lie-detected all possible suspects and then concentrated on those that had 'problemmatic' readings we would get a lot further with crime detection than we do now. Admittedly some people can lie their way to clean readings and some innocents get caught but no-one has developed those ideas to explain why it's just rejected out-of-hand.
It's the same with DNA databanks, RFID chips etc. People are not protecting freedom. They are worried they are going to get caught doing things they know are illegal.
Missing the point again
You keep on writing articles about agile development. You like the subject area but cannot seem to embrace the essence of it. Your article doesn't say anything. It just seems to be filling space.
An agile project clearly needs to be delivered on time. Without a reasonably clear idea of time and deliverables then you are not talking agile but something called Soft Systems Methodology. An agile project approaches the end point in a different way to a structured methodology. The end result is however essentially the same except much quicker, cheaper and more enjoyable for the people involved.