Malaysia government to draft insane Act for Computing Professionals
plus.google.com
plus.google.com
So what qualifications do you need to have to become a member?
Just cash and a valid diploma. Hell its probably easier to testify as a software expert in court for a member who pays his fees but works as a rug dealer than as a 10+ year network security analyst who isn't a member.
Any academic who supports the current implementation should have their credentials questioned, it defies logic that is attained at a high school level.
its legal basis is stated here
http://www.emo.org.tr/genel/sss.php?grubu=%DCYEL%DDK%20%DD%D...
basically if you engage in anything that is considered to be an engineering task (inc. computer/software/biomedical) you have to be registered. They are basically taxing you every year for graduating from your discipline.
edit: They claim to have 40,000 members I think, at 65 a pop that would be 2.5 million dollars a year. Not bad. Of course, as many Turks try to avoid taxes a lot of them skip these payments as well.
I know very little about Malaysia, so I can't say for sure whether licensing engineers/architects/lawyers/doctors has been a net positive there, but I strongly suspect that it has been, just like everywhere else in the world that has tried it. It is certainly not "insane".
Remember, lots of people in IT don't even have degrees in their respective fields -- thats the beauty of IT.
Similarly, the problem with IT (from back street "computer repair" guys, through to "systems integration engineers" from large consultancies) is that there are too many dodgy chancers who have no real idea what they're doing, and who end up bodging everything together with spit and bailer twine.
Suggesting that this is a problem that the industry has had decades to solve on its own and has failed, and that it might now be time for the government to get involved is certainly not "insane" either.
IT is an immature profession, change is rapid, best practice are constantly changing, client expectations are insane and unbounded, and there are many, many people in the industry with a university degree that can't program their way out of a cardboard box, while there are people with no formal education and a few years of experience that are better than most veterans.
The skill of a programmer depends too much on creativity, so a license can only guarantee half of the skillset, not the entire skillset, and therefore it's worthless.
For example, in the UK a three-year accounting degree only shaves ~6 months off the time needed to become professionally qualified, and many accountants studied something else at university anyway.
At my first job after graduating we had a self-taught architect/god/guru who never bothered with primary keys. One of our systems started slowing to a crawl. I added primary keys, and some indexes and suddenly I was a whizz kid. Another time I went for a job interview with an architect who was an actual architect (i.e. he studied designing buildings, not software systems). He didn't understand the concept of database normalization. Needless to say, I didn't take the offer.
Some of the best, and worst developers I've worked with had no degrees. On the other hand, CS majors were generally a "safe pair of hands", especially after they got a few years experience behind them..
If credentialling guaranteed a basic level of competence, with, or without a degree, I think it's a great idea.
CS programmes everywhere are getting blasted by the industry for not "producing" graduates that meet their requirements, therefore it's reasonable to assume that a certification process would suffer the same fate.
How could such a process meet the demands of a rapidly changing industry where different players have completely different requirements, where the language-du-jour is constantly changing, where best practices are outdated in a few years, and where the only way to survive is to constantly educate yourself?
I don't see how an IT license could lead to anything but a sort of tax on corporate IT companies that wish to do business with the government.
I believe in private accreditation (in computers this is already common for some jobs, through accreditors like Microsoft, Cisco, etc.) and letting end clients make their own judgements as to whether private accreditation is necessary for their choice of professional.
I do not believe in public accreditation, which is often a single-system for diverse fields (eg. architecture) and actually limits people from calling themselves by a commonly known job title (eg. you cant call yourself an 'architect' unless you pass this single [often poor] accreditation procedure).
But licenses encourage bribery, nepotism and other means of gaming the system. The effects are quite contrary to what people want.
Instead, grant these licenses by default without fees or limits, then let the employer decide if a candidate is good enough. Let the market decide if the employer is good enough. And if people's lives are threatened because of somebody's incompetence, then invalidate his license for a couple of months or permanently. If people's lives are lost because of incompetence, then throw that somebody in jail.
This is a system based on meritocracy and punishment in case of big failures. It works a lot better than a government imposed barrier to entry.
It's such a huge problem that the advancement of technology surely has been completely stunted the world over. /sarcasm
The problem is what you are licensing exactly. I don't know how many "certified" developers I've worked with who were complete freaking idiots. A license isn't some sort of quality signifier, it's just more bureaucracy and government involvement in a field that moves too rapidly for certification to keep up.
There aren't that many "dodgy chancers". They don't survive in the industry, otherwise the industry falls over on itself.
It's only those of us who see spit and bailer twine as all we need that can manage to turn a hacky property of silicon into NAND gates and from NAND gates an iPhone.
If you still doubt it's spit and bailer twine, consider this picture: http://www.don-lindsay-archive.org/talk/transistor.gif
The real professionals are those that can pair their twine with a slick marketing document.
Are you seriously suggesting that the same train of thought that "rubs wool on LCDs because it 'makes them work'" was responsible for turning silicon into an iPhone?
I'm suggesting that knowing that something does work is far more important than knowing how it works, or why it works. Whilst I'm sure LCD manufactures have their scientists on understanding that part of the manufacturing process in the meantime they'll do their voodoo wool rub because it works. Maybe the wool rub with the plastic in the LCD disrupts the higgs field, who knows.
Rubbing wool on an LCD because it works is an example of spit and bailer twine 'engineering'. Similarly, the first transistor was a massive hack, but a huge step forward for technology.
I'm not advocating ignorance, I'm just saying that those who are ignorant can make huge steps forward, eg. the first guy who rubbed wool on an LCD. I'm sure the original facebook code had SQL statement embedded next to HTML; everyone knows it's a bad idea, but despite that you can put the foundations in place to make a huge business. The important thing is solving problems, not the elegance of the solution. Be a wright brother, make an airplane despite the fact that no one knows how.
Solve a problem first, make it elegant second.
My own objection would be: does it really work? If we can determine that yes, in fact, it does, then I want to know why, but I'd continue doing it until we know why. When you know why, you can make it better. Until then, you're still just shotgun testing. Not that shotgun testing can't work, but if you can do better, and you can afford the resources to learn how to do better, you probably should.
I would agree the best is the enemy of the good enough, but also of the better--if all you know are so-called "best practices" and never deviate you'll never improve. You have to answer the question for your particular case though of whether deviations are likely to result in an improvement any time soon.
I don't really understand why, in this particular case, some people would stop the method without a replacement, just on principle of not understanding it. This isn't the first case I've heard of such things though--I've heard of some developers on a team deleting large swathes of code they didn't understand, replacing it with something they wrote that doesn't even work nearly as well. (One of my roommates says for a robot project one of his teammates took apart a working breadboarded controller but still hasn't replaced it with anything.)
> Solve a problem first, make it elegant second.
"If you nail simple and fast, elegant is usually forced to come along for the ride.." ( http://twitter.com/#!/1/statuses/122368755962621952 )
I don't think elegance is at odds with solving a problem, but in general I agree that getting something working that's good enough now beats something that would have been better later. But not always, and possibly not even most of the time, as the time spent in the programming world rewriting and reinventing things shows.
My uncle was hired in 90' as a consultant engineer in a factory (they produced parts for Deawoo car factory in Poland).
The plastic parts of the inside part of the roof was heated up, and pressed into proper shape, and the process was somehow undetermined - sometimes the product was ok, often it was not.
Workers at the line developed some routines that helped (voodoo, like touching the machine in just the right moment, after a few seconds of heating, or making short pauses after a few roofs produced). They argued whose "magic touch" is better :) That helped (according to them), but still big percentage of roofs were broken.
My uncle found the source - one temperature detector was not connected to the machine, the cable instead touched the case of the machine. But that machine was used for a few years already, and the "voodo" was helpful in geting the job done in the meantime.
Yes - they should investigate the root cause earlier, but the owner wasn't interested in investing any money in the tooling after initial investition, and just blamed "uneducated" workers for big percentage of broken products.
Did you ever read the name of this site? It is called "Hacker" News.
Learning Java at college doesn't change anything for the computer repair guy. I know what government means then, you have to hire a 4000$ guy to repair your computer who doesn't know how to do it.
The best you can find is probably someone who does this out of joy and knows most of the current hardware.
That's a terrible idea (and doesn't appear to be the solution that is being proposed in Malaysia.)
I remember ten years ago when people had certifications from either Sun or Microsoft. They were completely worthless, and the people that had them fell in one of three categories:
1) The ones who got them as a joke, but were actually competent
2) The ones who got them seriously, to try to hide their incompetence
3) IT consultants who did jobs for big, boring, government agencies, or big, boring corporate IT departments and got them out of necessity to score those contracts.
I see no reason to believe that a government backed IT license wouldn't suffer the same fate, and since those licences were useless, the government one will be, and therefore the suggestion is insane.
However, there are industry certifications that aren't worthless, I place high value on the top Cisco networking certifications, but it's also very important to note that that part of IT is the most like conventional engineering and client expectations are sane and bounded. I see a pattern. Do you?
The scary thing was that you could pretty much take those exams cold if you worked with the technology at all, yet you'd meet these extremely nervous looking people in the waiting room at the testing center talking about how they hoped their 6-week prep class would pay off and they'd be able to pass this time.
For example, should I need government licensing before I'm allowed to sell a $0.99 iPhone game? Is there any real public-interest reason for such a licensing regime?
I tend to think certifying the safety/quality of products or systems is a better approach than certifying people. In the U.S., for example, it is not common (contrary to some countries) for engineers making consumer products (as opposed to civil infrastructure) to be licensed professional engineers. Instead, the products themselves are tested; for example, a toaster has to pass various safety tests before it can be sold.
University places are allocated based on racial quotas. Therefore, the people who are most concerned by this legislation are ethnic Chinese, who are more likely find themselves disqualified from doing certain work due to their inability to access higher education places.
Alternatively, businesses may be forced to hire degree'd people to sign off work as theirs even if these people didn't do any of the programming.
If this bill is passed, it will not only hurt people of a particular ethnicity, but everyone and the whole IT industry in the country.
Given the track record of the incumbent political party, I strongly believe that most of the 'registered' people will be the people who are very close to them. The 'registered' people will make use of their status to get themselves into rent-seeking activities to take advantages of the companies/people who genuinely want to start up an IT company or involve in an IT project. This will in turn hurt the IT industry in Malaysia, which is already very weak due to current rent-seeking activities. What will happen eventually? More brain drain to the country, and not to forget a stagnant, if not backward-moving, IT industry in Malaysia.
You only see the racial thing here because you are not looking at the bigger picture. It's true enough that the 'registered' people will most probably be ethnic Malays, but those are ethnic Malays who are close to the incumbent party. The majority of the ethnic Malay will get as little as the ethnic Chinese get.
EDIT:
> University places are allocated based on racial quotas. Therefore, the people who are most concerned by this legislation are ethnic Chinese, who are more likely find themselves disqualified from doing certain work due to their inability to access higher education places.
It's true that qualified ethnic Chinese are having less chance to enter the public colleges, but most of them who can't get into public colleges go for private colleges or study abroad instead. In average, the literacy and qualification of ethnic Chinese is higher than other ethnics in Malaysia. So it's untrue to say that they are concerned due to their inability to access higher education. While there probably is an issue of education equality but the bill does not relate to the issue.
ps. no higher ed means no qualifications which means no work.
"Notariusz" is regulated job, there is constant amount of them (you have to pass very hard exam to become one, and mostly being a notariusz goes from father to son ,because others won't pass exams somehow). People are required to hire a notariusz to check important contracts, and notariusz gets percentage of contract worth for a work, that any intelligent people will do anyway (like checking ground register before buying a house, etc).
I do think they are sometimes usefull, but why regulate them? Why are they mandatory? People that want to risk should be able to risk, people that want to be safe can pay more to be safe. In IT various schools and certificates exists, so people can prove, that they are good enough to do some job. There are bad schools and better schools and certificates, why do you think state will be better at assesing people qualifications than private company (which sole buisness model will depend on being good at that). Why is state always trying to do others work?
I'd much prefer free market in such matters. The same goes for engineers, architects, etc. Just make sure people are responsible for their errors, and there would be cheap but risky architects, and expansive but safe ones. And insurance companies would require you to hire these with respectable certificates.
No need for state to mess with this.
So banks that lend more money than they have should be controlled by someone. And banks are only a few in a market, big, and can easily ignore new companies wanting to certificate them, making these certificates worthless. So there is need to regulate banks. With workers it's not a problem, so there is no need for state to get involved.
Regarding insurance companies - I don't think there is inherent need for regulation. A few broken insurance companies won't destroy economy. For their clients it will be a problem, but that's also true for other activities we currently don't regulate.
At first, insurance companies wasn't regulated, and they somehow managed to work. Now people expects state to regulate them, so we trust unknown insurance companies too much, so if we were to stop with regulation, many people will lose. But if we did not started regulating them before - this would not be a problem.
If there were some sort of magic certification regime that would guarantee, a programmer knew what he was doing, you can be certain such a certificate would already exist.
You also seem to have just ignored all the problems that exist with mandatory doctors and lawyers guilds. For example the AMA hasn't allowed any new Medical schools to open in the US since the early 80s.
There is a reason doctors and lawyers are so expensive, it's because their guilds have an incentive to keep them artificially scarce.
Right now IT is one of the few professional fields where you can essential apprentice your way to a job. I can guarantee that were we to establish a mandatory governing body, a degree would be required.
Hell, in Florida they require Interior Decorators to have an approved degree (very narrow definition of approved degree too, I might add).
In Georgia (where I live), the board who overseas General Contractors, ruled they have $1,000,000 in assets in addition to the insurance they were already required to have to cover damages. The only reason it was passed was to protect the jobs of established contractors, who were being forced to compete with lower wage Mexican labor.
That kind of thing will happen to the IT industry if you regulate it similarly.
Now, I'm thinking twice.
After reading the draft Act, the people responsible for this are still leaps and bounds behind the curve, by the sheer fact that something like this can even reach a "draft" stage.
Basically, they're setting up a list of accredited computing professionals and companies, and the kinds of services that they can perform. This is a public register, and anyone can look at it.
Imagine you've just started a project, and you're qualified for Java and some database. In the project, you end up using a bit of Python here and there to help with some scripting because it is pragmatic and it makes sense. If someone notifies the "board" that you did this, you could get into trouble. And the "trouble" may put you behind bars for up to 3 years.
Take another situation: I end up hiring a bunch of Ruby developers, all of whom are on the register. Along comes a client who needs some work done in PHP, which naturally, being good Ruby developers, they can pick it up quite easily. But to conform to this new "Act", I will have to turn down the business as I don't want to risk going to prison.
Where's the sense in all of this ?
I want to check the doctor tomorrow. How can I be sure that he is a real doctor and not just a "will heal for food" man.
That's a serious problem. Not every boss/CEO/manager can assess an IT consultant proficiency. They need a proof that they can do the job.
The problem is that academy is broken. The certificate you get (degree) from University is not a real mirror of your capabilities. What government is trying to do is to solve a problem with a broken solution.
Since then I'm not so sure. The definition of "developer" can get a bit murky. For instance I now work in an R&D team; the researchers need to be able to code up prototypes, but the important thing for them is to be good at research rather than coding.