Posts by John Smith 19 16330 publicly visible posts • joined 10 Jun 2009
As someone observed a long time ago "It's complimentary. It's not free (or cheap)"
Someone is paying for your convenience.
If you're bi**ching about all that spam you have to weed out of your in box....
Where do you think it's coming from ?
You mean the people who probably don't have degrees, just actual skills needed to get stuff done.
This is the real difference between the UK and (IDK) say Germany.
Both offer excellent training for their top graduates.
But Germany makes much better use of those not-academic-but-smart students rather than what appears to be the UK view of "They are basically disposable scum."
You forgot the fear of long term government final salary pensions that cost an arm and a leg to support (courtesy of Gordon Browns desire to acquire some "free money" by scrapping various employer tax reliefs on their contributions).
That's about right.
OT but does anyone see parallels between Carillion and the Royal Bank of Scotland?
1) Go on massive growth spurt using borrowed (but low interest) money
2) Get f**ked when interest rates start to climb for the first time in a decade.
RBS was (to coin a phrase) "The bank for the good times." Will Carrillion be "The outsourcer for the good times"?
Funny you should say that.
Because if you really knew something about this subject you'd know that's exactly what they do.
Most of those companies are hollow. What they really bring to the table is
a) Deep pockets to survive the years long "procurement" marathon mandated because of the it's-a-mult-billion-pund-deal-over-a-decade BS civil service rules.
b) Lots of ex Ministers and Senior civil servants who know who to talk to get write a great sales pitch that ticks all the right boxes.
3) Negotiators to stop their "creativity" being hamstrung by any penalty clauses for late delivery, late delivery (but being quite s**t) or compete non delivery.
4)Negotiators who will ensure that any further changes (including fixes to their failures) are paid by the government.
5)Programme managers with a small address book of good (but expensive) contractors who know what they are doing about those systems (probably ex direct HMG employees) and who will not tell the employer they are not direct employees of the outsourcer.
6)Programme managers with a big address book of code monkeys who know Jack s**t but work cheap and fast (and most of whose code is s**t). That include the number to some offshore "palace" of code monkeys in somewhere like India, Pakistan, Russia, Malaysia or Hong Kong.
7) A PR team that will categorically prove it is not there fault
That's what these Ahole bring to the table.
Yes minister first aired in the UK in 1980, when Margaret Thatcher had just come to power.
And AFAIK outsourcing first started to be looked at as the magic wand that would a)Save money b)Improve services.
It is now 38 years since then.
With honest accounting I'd expect you should be able to answer the question "Does outsourcing save money or not?" with 38 years of data quite easily.
But I bet you can't.
it sounds like a lot of Govt IT procurement is done like a lot of UK Gov Defense procurement.
With a great deal of staff (but not quite the 23 000 the MoD uses) and very little apparent improvement in the end product.
So what do we have here.
1)The Home Office has failed to do its job on immigration control for decades, including the more than 50% of the total from outside the EU, which the UK government had (and still has) total control over.
2) Non existent respect for personal privacy
Once again the hard task of sorting their own s**t out is replaced by cajoling/brow beating someone else into doing it for them.
*Blame Avoidance Strategy #1 is of course "Blame the EU for the requirement/inability to have to do/not do this thing." Delete as appropriate.
and yet nobody is ever, ever sacked on any side."
There are time honoured methods of blame avoidance that the UK senior civil service excel at.
Actually it's worse than that.
This was not a "Green field site." There was an existing system in place, so Project Management Tactic #1 should be in effect even without a useful contract.
PMT#1 is "Develop question list around existing system" IE "What does it do well," "What does it do badly," "What do have to workaround" "What is so bad you cannot work round it and must avoid at all costs" etc.
Isn't the whole point of things like Agile the notion that people don't know exactly what they want in the abstract, but when you put something concrete in front of them they are pretty good at saying what they like and dislike about it?
And they have had something concrete in front of them (the previous system from HP) for years to form an opinion.
IOW The good bits of what they have + improvements to how it does it to remove the rough bits.
That's a baseline, and it seems they couldn't even deliver that.
Funny you should say that.
At lease one artist programmed a system to be a creative artist, using a plotter as their canvas. Interestingly it was in his style.
Likewise one of the theses (looks wrong but I can't be ar**d to check if it should be "Thesi" or even "thesii") on the idea of (basically) "English as a computer language" in the mid 70's retired (at UC London) and retrained as an art historian.
The trouble with this idea is that on the surface it sounds quite impressive.
But as others have noted while the answer is directly embedded in the text it really is a case of simple pattern matching.
Humans have some quite amazing abilities where language (both spoken and written) are concerned.
1)Learn a language without being taught a formal grammar, in the computer language sense of the term (or often the rules of grammar, which is a different thing). This is fortunate, given various attempts to write full grammars for human languages are f**king huge.
2) Parse without backup. This used to be thought impossible. Mitch Marcus, during the MIT "Personal Assistant" project (still waiting for one of these) demonstrated otherwise.
3) Parse with minimal lookahead. Marcus reckoned you needed 3 "phrases" but later work at Edinburgh found that 2 were sufficient (but a phrase is not just a single word).
4) Do it all without a complete dictionary of the language.
Which is fortunate, given it would be obsolete the day it was declared "complete."
Humans are really good at recognizing when something is a noun, verb or adverb, even if they've never seen that word before "I just drumped a can of lager off Jack" "Pass me the fribble, said Sarah" are legal sentences with (AFAIK) meaningless words in them, but you probably worked out "Drumped" is some kind of verb for some behavior (not quite sure what) and fribble is a noun for something (not quite sure what). So you know enough to reason about something you don't actually know anything about.
5) Relating new stuff to what they already know or have experienced. With this person relates new stuff to what they know, building out a web of what they need,and relating the new stuff to the old.
Note that property "enough." If you know someone only knows 1 person called John, you (probably) know who they are talking about. If you know they know N people called John (or Jon, or Johanne) you know you're going to have disambiguate and apply some context.
Likewise if you're a molecular biologist and someone gave a paper on plasmids you'd know what they're talking about, as would a plasma physicist. Except that they would be thinking about a different kind of plasmid, and vice versa. This is one of the reasons why any approach starting "We build a complete lexicon of the language" is likely to fail, as is something that does no spell checking/correcting (look at the number of short words where a single letter change changes the word, or the context of the sentence, completely).
Humans brains are multi layer neural nets, but they have the ability to restructure that net based on new input, not just to add another word, but another whole concept, that can provide leverage for all future sentences. Instead of (To be. Verb, Personal,Impersonal, Plural, Historical, Current, Future) it spaws a "Subnet, verb detecting" to find one, and a "Subnet, verb handling" to identify what it is and how to deal with it in terms of the persons view of the world (which I guess is what a verb "means")
I don't think any software efforts have tackled this, and until they do they are always going to be limited.
Note, despite all those 100s of 1000s of word dictionaries different systems have created or look through the fact is about 50% of all English text is made up of 300-350 words, most of them < 8 letters long. The "Semantic Primes" theory has identified 63 words (or concepts) that are universal across all known languages, and which every other word or concept can be expressed in (of course it's taken since 1972 to build that list, which is a bit long for the average AI research grant application).
Do you think, IDK, maybe figuring out how to leverage that fact, and those 300 odd words, could be quite a good idea?
$Deity, wading through all that ancient AI s**t is depressing as f**k. The endless philosophizing and psychobabble. Computational lingusticians are worse than economists for forming tribal groups.
I really do need to get a life.
Given El Reg's nanny with buttplugs at Whymo engineers house story.
Here's an idea, El Reg.
Feed the story through a text to speech engine and listen to it.
What your writers eye has skipped over your Editors ear will (probably) hear as a very dud note indeed.
Hear the phrase "the locked were changed, " and you should immediately be thinking "WTF?" or "Warning, noun, verb disagreement. Parse failure," if you've just waded through a load of English parsing AI papers and wondering why the answer to "How do you parse 'police police police' " is not "WTF are you talking about? Explain yourself better."
No.
As she tells it the kids went through their parents stuff.
She was just tidying it up afterward. Kids, eh? What you going to do.
She just happened to keep contemporaneous notes of everything she heard and did (which re therefor her personal property and not on behalf of her employer) while doing so.
This is because she is a)Got a bit of APD b)Didn't trust her employer c)Planned all along to collect dirt either for herself or on the behalf of others for a spot of "Biographical leverage."
I imagine a court will have to decide which of those reasons is/are the right one/s.
And an inability to prioritize relevant information over irrelevant trivia as well.
My first (stream-of-consciousness, naturally) reaction was "She sounds like she's somewhere on the APD spectrum."
Is "Lack of empathy" a character trait you want in someone looking after children? Lack of sympathy perhaps, lack of empathy, not so much.
Her hiring (at a time of maximum confusion) sounds quite opportune, almost ideal if you were planning to hit someone with a large lawsuit later. How did she pick this lawyer for the job?
I smell a rat, and I'm not just referring to any hints of her former employers personal hygiene either.
Who decided how important the program is?
Why that would the NSA of course.
No doubt when the relevant Committee Chair asked "How important is this to US safety" and NSA head replied "Very. It's saved a huuuge number of lives. We can't say how many because, y'know, security, but trust us, it's huuuuuge."
No. I don't believer there has ever been an independent cost benefit analysis of how much it costs, how many plots (in detail) have been found and stopped, and how many American citizens have had their details recorded (forever) in this database to do so.
FTFY.
As a non-American let me see if we can agree a few things WRT to broad band access to the internet (although TBH it could apply to other areas as well).
Competition lowers prices to the consumer.
Lower consumer prices are a good thing.
Monopoly suppliers of anything are a bad idea.
Many parts of America have no competition.
Many places have a monopoly supplier, whose installation costs were sunk decades ago.
Some way (or ways) needs to be found to encourage competition (or even delivery) in these areas with no supplier or a monopoly supplier
One option, with little or no cost to the Federal govt, would be to allow (not encourage, simply allow) local communities to set up their own ISP's* to be run to break even for the benefit of those communities. *
While existing suppliers may wish to deny further competition that is a wish, not a right of those suppliers in a free market
Yes I do believe that the US Congress could get agreement on the above, given (IMHO) the fairly mild criteria and general support for free markets (which I thought they are all in favor of) ?
But IRL do I f**k. Monopolies never give up a monopoly (obvious or effective) without either a massive battle, or being given the chance to form an even bigger one.
IOW "A local broadband, for local people." If they can't break even they go out of business. If their prices are too high people go back to the incumbent, like people do when there is real competition. If the incumbent drops its prices to an uneconomic level either its a)Predatory pricing (FTC violation) or b) It's been gouging the customers at the existing price level, and those (lower) prices should be permanent.
If people want a better service they should always remember the motto of New Hampshire
But it sure did hit their benchmark targets, didn't it?
I'm curious. How many systems don't start with the assumption (and it is an assumption) that the OS code is absolutely correct, and should therefor have full access to the machine?
Because IRL, over and over again, we've seen that is complete bo***ks.
We know every substantial piece of code will have bugs in it, and they will escape detection whatever level of testing it was given before release (which appears to be f**k all in some cases).
So what OS developers plan for bugs in the code?
Sense of the term.
IIRC it's pretty much the sense used by the USAF in Vietnam for their drones.
Fly pre-programmed course of altitude, speed and bearing way points, then either execute landing "subroutine" or eject sensor package (the most valuable part) and self destruct.
In 2018, perhaps describing that as "autonomous" is a bit generous.
This is the # 1 way to CYA. But make sure you keep an off site, off line copy, to avoid "The email archive got deleted" defense.
One hopes eventually the smarter PHB's will cotton on that what they say has consequences and the more accurate they can describe what they think is wanted the better chance they have of getting it.
At least that's the more positive view of doing this.
you need mechanisms in place to handle that.
Except most British firms I'm aware of are a bit s**t about capturing requirements, let alone tracking them, seeing how the system delivered conforms to them, what's outstanding.
Because that would usually (to PHB's) translate as "Spend money on non-core software that does not directly help the business."
And so the fail continues.
Or allowing Airbus to play the "No, that's our final offer. Accept it or we walk away from any further production" move.
The question of course is wheather or not Emirates believe they really would do that.
I kind of think they might.
Wonder if he's the guy who did the video column in Dr Dobbs Journal back in the day.
Which would suggest he's been around this block a few times.
I also wonder if people how much not like a TV camera the human eye really is?
Starting with the fact the rods and cones are not laid out in a square pattern, but an exponential spiral, starting at the centre of the FoV.
And then it gets complicated...
Which probably explains why it's called "The HP backdoor."
Obvious caveats.
Who's still running 2004 code releases in their switches in 2018?
Are the standard defaults set to allow this to be accessible?
If not how often are switches configured IRL that make these vulns accessible?
Was this found as part of an ongoing actual code audit or did someone just stumble across it?
I'm highly suspicious Lenovo is organized enough to be doing a methodical code audit.
FTFY
Funny how with data "protection " legislation the clauses that allow government civil "servants" and their alleged "masters," to do WTF they want when they want are very carefully drafted to do so, while the actual protective ones are ambiguous to the Nth degree with many, and varied exclusions, reasons-not-to-report certain behaviours etc.
They are not.
There are 2 parts to this.
How good a price are they getting and how much support do they get to make their products.
Part of that process is the fact that (at present) the UK is inside the "tariff wall" around Europe. So if UK prices are competitive with the rest of the EU they will sell into the EU.
Put them on the outside of the tariff wall and then cut their support and hey presto they're uneconomic. Likewise "less favored" and lowland beef and sheep are partly supported because the UK likes the countryside a certain way. And of course UK livestock care standards are very high. That is not EU legislation, that's UK legislation, so Brexit won't mean "Stuff em in the veal crates. The good 'ol days are back." :-)
The Agriculture & Horticulture Development Board actually did an impact assessment of Brexit on British agri & horti culture (because, y'know unlike David Davis they though having some idea WTF was going to happen might be quite useful) and estimated that worst case the most efficient 25% of all British farms would continue to be profitable.
Of course the other 75% of UK farms would be f**ked. Interestingly under all scenarios considered the "Barley Barons" do less well than most other kinds of crops and worst case "Hard Brexit" farm income goes negative. Pig farmers and potato processors incomes rises under all scenarios. But the across the board farm income drops unless the most favorable, most generous EU terms scenario happens.
Looks like Bacon and ready cut chips are going to the keys to the future British diet.
Brexit is a "major transformational project"
Wheather it will "transform" the UK into a free market utopia or crony capitalist s**thole is unknown. But
<gollum>
We wants it
We needs it
We must have hard Brexit.
</gollum>
Without which implementing most large instruction sets (all mainframes and x86 architectures) would be an even bigger PITA.
There is at leas one memoir of one of the staff at TRE during WWII which I stumbled across in an old copy of New Scientist called something like "Biologist at War." The UK was so short of physicists and electronics engineers they were re-training anyone with some kind of technical education to help with new design.
"The US Federal Communications Commission (FCC) does not disseminate pricing data or track broadband availability by address. Additionally, service offerings follow no standard speed tiers or definitions (such as the specifics of video or phone service bundles)."
So that'll be "Sweet" Pai earning his money then.
US Big Cable remind of the UK Banks when they were asked about their views on allowing Credit Unions to relax membership conditions. Keep in mind both have ignored the areas where these services have grown for decades IOW the can't be a**sed to support the area themselves, but they don't want the area supporting itself.
And very handy
The payoffs for this tech are
1) That in principal you don't have to point the spacecraft to do nav tasks
2) It can work outside the solar system. It's the nav system for going 10s of light years.
That said in space with a big enough star catalog (and modern computers have the memory for quite a big one) visible stars should be able to do the same.
I think there's also some benefit in principle from making the equipment quite small, but obviously that's for V2.0. This was getting the outline of the software developed.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
