A heated but calm debate - PHP/MYSQL v's custom program

christillis

I asked a question on yahoo answers a while back

To be brief, I asked why NHS couldn't have used a web browser based platform...so they're not stuck on XP for eternity :queasy: Someone came back and said there's too much data and it would be a security issue. If you're interested, linkie:

http://uk.answers.yahoo.com/question/index;ylt=Al87hU1rgE7BzqmTbvxEswwhBgx.;ylv=3?qid=20090127010553AAF7noO

Just out of curiousity, what do you think?

Cheers!

Chris

rulian

this may sound dumb but what is NHS and what are you trying to do with it?

NogDog

NHS = my high school. 🙂

Weedpacket

I already knew, but between http://www.acronymfinder.com/NHS.html and the mention of "patient database" and "x-rays" - I think it at least rules out the National Highway System.

laserlight

Ah, so Dr NogDog from the High Energy Magic Dept. deals with patient databases and X-rays.

christillis, mind providing a summary of the points? Frankly, I find the discussion on Yahoo Answers rather hard to follow, and it looks like the topic assumes that everyone knows the background details.

christillis

NHS, unless you've already checked, is National Health Service. Instead of the american system (this is a slight guess), instead of insurance we pay directly to central health care. Strangely it's proportionate to income - mine's around £200 but someone working on the till's in a supermarket may pay £50-100. To clarify, it's one of the biggest embarrassments of the UK and many go for private insurance...so they don't die in the reception of an NHS hospital.

Anyway,

I'll have a read of what was said and summarise for you laserlight - it's the least I can do for someone as helpful as yourself! 😃

christillis

Right, here's the round up:

I complained that the NHS has spent 12 Billion pounds on a database for medical records, I complained and though it could be done using php/sql, etc.

A reader claimed there was no way it could hold the quantity of information needed, I said, "well eBay seems to work OK". He then said the NHS records for the United Kingdom are greater than all of the items listed on ebay.

I then accurately guessed that eBay has 20-30 million items (the uk ebay), took into consideration encrypting those and yes, it seemed fair that perhaps whole 'system' is needed. The system I refer to uses VPN's and programs at either end, as opposed to a central system with bank level's of security (my idea).

So the question is, is there ANY situation that web browser script and any form of a DB cannot handle? Why should big companies depend on massive programs? I'm not attacking the program makers, merely questioning the purpose of them when it's only a DB that's being handled.

Perhaps I'm overthinking 😃

PS Laserlight, you're right about yahoo answers. Although there are some areas where grammar skills are quite high and so it's easy to follow. However I'm sure the "relationships" area is full of, "diz boy held my hand dos he fanci me?" :quiet:

Weedpacket

christillis wrote:
only a DB

Only a DB? You're talking about something that in this case would be a massively complex beast. At the very least I'd expect solid quality of service: an assurance that if the interface application claims something has been done then it really has been done and nothing short of a multiple-target nuclear strike would erase the fact.

laserlight

christillis wrote:
I complained that the NHS has spent 12 Billion pounds on a database for medical records, I complained and though it could be done using php/sql, etc.
(...)
So the question is, is there ANY situation that web browser script and any form of a DB cannot handle? Why should big companies depend on massive programs?

Does your 12 billion pound figure come from the National Audit Office's report on The National Programme for IT in the NHS? That report does "complain" that the projected cost was rather excessive compared to the expected financial benefit, but it also gives a picture of something inherently more massive and complex than suggested by "a database for medical records". (On a personal note, a friend of mine really did develop, on his own, "a database for medical records" for his father's clinic using PHP with a MySQL backend.)

I think that it is good to clarify what exactly you mean by "it could be done using php/sql, etc". Are you proposing that every front end to the system should be implemented in PHP, if possible? Are you certain that none of the contractors used and customised existing relational database technology? What is the "etc", e.g., would the use of Enterprise JavaBeans qualify?

christillis

Weedpacket;10904474 wrote:
Only a DB? You're talking about something that in this case would be a massively complex beast. At the very least I'd expect solid quality of service: an assurance that if the interface application claims something has been done then it really has been done and nothing short of a multiple-target nuclear strike would erase the fact.

yeah sorry wp, it's not 'just a DB', I'll immediately apply for 'understatement of the year' award! :quiet:

laserlight;10904480 wrote:
Does your 12 billion pound figure come from the National Audit Office's report on The National Programme for IT in the NHS? That report does "complain" that the projected cost was rather excessive compared to the expected financial benefit, but it also gives a picture of something inherently more massive and complex than suggested by "a database for medical records". (On a personal note, a friend of mine really did develop, on his own, "a database for medical records" for his father's clinic using PHP with a MySQL backend.)

I think that it is good to clarify what exactly you mean by "it could be done using php/sql, etc". Are you proposing that every front end to the system should be implemented in PHP, if possible? Are you certain that none of the contractors used and customised existing relational database technology? What is the "etc", e.g., would the use of Enterprise JavaBeans qualify?

Sorry laserlight, when I say 'php/sql' as an alternative, you have to appreciate my relative lack of knowledge of the subject. I have just dipped into this area of IT but I can already see the potential for it. I'm not aware exactly how the NHS system is constructed, I just assumed they have hired a large company to create an RMS standalone program, rather than a RMS that can be accessed via a typical web browser. Hope I'm not tying myself in knots here! :rolleyes: I could just imagine an entire program being written for it and thought that was a bit silly compared with the alternative, but what do I know! ("very little", I'm sure will be some readers internal diologue!).

Cheers!

christillis

Additional:

Part of my doubting the file size of the database was this x-ray sent to me by my doctor (I asked for it for insurance purposes)

The original was less than 200KB's despite the 'printed size'.

laserlight

christillis wrote:
I just assumed they have hired a large company to create an RMS standalone program, rather than a RMS that can be accessed via a typical web browser.

If I remember correctly from my skim through that report, there may be at least one web based subsystem.

christillis

laserlight;10904521 wrote:
If I remember correctly from my skim through that report, there may be at least one web based subsystem.

Ah...in that case a moderator should delete this thread and I should apologise to all for replying. Sorry. 🙁

NogDog

Having had experience in working on a number of systems for the US DoD, my guess is that, yes, 12 billion dollars/pounds/euros is very excessive for the software, if that is all that the money went to. I suspect it included many other things in addition to software, though it still may be grossly over-budget.

When such things do go grossly over-budget, it more often than not is due to poor requirements specification on the government's side, equally poor requirements analysis on the contractors' sides, political in-fighting and changes in personnel on the government's side, changes in direction from the applicable legislative bodies, etc., etc., and so forth that lead to terminal "moving requirements syndrome". Pair that up with all the other problems in government contracts (bribery, kick-backs, nepotism and cronyism, the "old boy network", etc.), and you have a formula for financial disaster. And all the "red tape" involved in the process means that everything takes so long that by the time the system is fielded, it's several years behind the current technology being used in the private sector.

Weedpacket

I still recall the New Zealand Police force's big INCIS project; I wasn't involved, but it was a nice idea - an integrated intelligence coordination system for the entire force (considering that at that time the high point of police IT was the criminal records database stored on a twenty-odd-year-old computer). The project started in 1993 and was originally going to be OS/2-based, with the principal consultants and developers being IBM, natch. Later on (partly to do with Microsoft pulling out of its OS/2 partnership with IBM) it was ported mid-stride to Windows NT desktops and Solaris servers and Oracle databases.

Seven years and sagans of dollars later, the Police finally had email. $DEITY it made me want to smack them.

rulian

I can easily explain the budget if its anything like the US, people who have nothing to do with the actually making or planning of the project pocket about 60% before the money ever sees the light of day. 30% of it goes to paying off all the people who are in control of the money or regulation of the projects creation. 5% simply "vanishes" and about 8 million goes into buying the 10 million on hardware and paying off the 100 or so people in charge of designing and programming the project. Which is of course at thsi point over budget at which point the heads in charge will go back and ask for more money, rinse and repeat.

rulian

ohh and I forgot, somehow 30 million will go to alaska to build another 10 million dollar bridge to a manmade island with nobody on it.

cretaceous

can I add a postscript that actually the NHS is a remarkably good system that is the envy of many other countries - quality of service is generally excellent

(though it depends extensively on the Hospital and general part of the country)

its the biggest employer in the uk
it costs the UK horrendous amounts of money

we do not pay directly towards it - taxes are apportioned as government thinks fit (I think chrisellis is confusing "NI" payments - they are nominally for health and pension but really just another tax)

it means anyone can get treatment for any illness at no cost -
be a poor person in many other parts of the world and then make sure you never get ill

but the supamegadatabase, he is bonkers
just like all the government run IT schemes

Roger_Ramjet

Yes, the whole programme is about much more than just a db. It may have started out that way but as someone pointed out is has become an exemplar of moving requirements syndrome. It does not matter what was intended, today it is the whole of the NHS that is being computerised in every function. :eek:

Now, the people who were originally brought in as consultants were all excellent professionals with backgrounds in eg large scale banking transactions. They told the powers that be that the scope and scale proposed even at that time were unfeasable. Of course the politicians, Bug-eyed Blair, refused to listen. Instead they threw so much money at those people that they could not say no. :evilgrin:

The reason that it was winders based are historic: this has been going on for nearly a decade and when main contracts were first awarded web-technologies were in their infancy. The companies who won the massive contracts to deliver it were winders based consultancies anyway so they chose what they knew. Now, many web-based systems exists in specific situations, eg the junior doctor application system that hit the news because it fell over on the first day and no-one had a back-out strategy in place so hundreds of junior doctors could not be hired. :rolleyes:

Now the British Computer Society through its specialist groups have been vociferous in our criticism of both the program and the process. The best academics and industry experts are almost unanimous in their view that the project is just not possible, for all the reasons that many of you can guess. But of course there is just too much money in it (contractor on £2000 per day !!!) and too much political will for our opinions to be heared.

For me the best thing to have come out of it is the hundreds of millions that some of the main contractors have lost. Serves the greedy buggers right.:p

Weedpacket

Random association: Windows-based applications tend to be vast monolithic structures that do everything themselves, expanding until they can do email. This is because it's very hard to do interprocess communication in a Windows environment.

Unix-based applications have a much easier time of being smaller, more focused on doing a single task and doing it well, and hence more robust. This is because IPC is much easier in the Unix environment and there are open protocols in place that everyone gets to use.

This NHS system is Windows-centric in more ways than one .... 🙂