I’ve been working in the IT business since the mid-nineties and a recurring problem in recent years has been that even magna cum laude university graduates don’t know how to program properly anymore. The reason for that is – they aren’t being taught and the days, when hardcore Nerds like the Hippopotamus Corporosus were foreswearing any sort of social life in favour of all-nighters teaching themselves how to get their puny little computer to do stuff it was never designed to do, are a thing of the past. Kids these days don’t grasp the concept of spending three days of programming to come up with a program or device driver that you can just download from the vendor these days. Today’s generation wants instant gratification.
Back in 1994 I installed Linux on my computer for the first time. It was akin to building a Caterham kit car that came with a few parts missing, like the gearbox for instance, which you then had to build yourself. In my case the system didn’t know what to do with my massive 512KB SVGA card and apparently my brutally fast single-speed Mitsumi CDROM drive spoke only the binary equivalent of Greek. So I spent two weeks designing and programming device drivers for my exotic hardware. Today’s youth would send the DVD back if the system wasn’t installed and ready to go after one hour at the latest.
If you come fresh out of university today, chances are that you have learned to program in Java or C#, which has nothing to do whatsoever with programming. Imagine you were asked to write a program for bolting stuff together. Today’s kids would come up with something like that:
List BunchOBolts = new List ();
List BunchONuts = new List ();
Nuts nut=BunchONuts.begin();
for (Bolt element in BunchOBolts) {
element.ConnectTo(nut);
nut=BunchONuts.next();
}
Voila! Program done in 2 minutes: Instant gratification. Now here’s the problem. They have no control over whether or not the Nuts and Bolts are created properly and with a minimum amount of resources. And more importantly, they rely on the runtime system to give the resources used by the temporarily created Nuts and Bolts back to the system when they are no longer used.
By using ‘convenient’ programming languages, they are isolated from and left in the dark about the underlying processes. Back in the olden days we had to do that on our own and thus had full control, but also full reliability. If our program didn’t work, it was our fault. If today’s script kiddies write a program that doesn’t work, chances are they hit a bug in the runtime system.
The thing is though: Hardware doesn’t provide ‘convenient languages’, especially cutting edge stuff like this year’s engines. You don’t need useless Java monkey’s, you need battle-hardened experienced Assembler jokeys, who know how to shave off a few nanoseconds from an inner loop by eliminating a redundant near jump. You need folks, who know that
MOV BX,AX
SHL AX,2
SHL BX,4
ADD AX,BX
SHL AX,2
is a few processor cycles faster than
MUL AX,80
Renault, however, seem to have hired an unholy bunch of C# script kiddies or some guys, who acquired their hardware programming skills on Wikipedia, because if they were hardcore nerds from days gone by, we wouldn’t have heard the words ‘software problem’ uttered nearly as often as we did during winter-testing. To put it bluntly: A program that after 2 years of development still doesn’t run properly has been written by someone, who chose the wrong job.
Back in the day hardware programming was witchcraft. You either got it right or you incurred quite a repair bill. When I wrote the Linux graphics driver I had to program horizontal and vertical frequencies manually to achieve a certain non-standard screen resolution. I got one number wrong and pushed the horizontal frequency past the limits of the of the monitor’s tube. It started to stink, made a nasty ‘pppphhhh’ sound and caught fire. That meant shelling out 400 Deutschmarks for a new monitor.
But that wasn’t an attempt at ‘programmer’s jackass’. We simply didn’t have any other option. Today, you would write a CRT emulator, that would give you a ‘parameter exceeds limits’ error message instead of a plastic bonfire. But in our days computers had less memory than today’s wrist watches. They simply didn’t have the oomph or the resources to simulate a full-blown cathode ray tube with all its bells and whistles, so we had to play ‘horizontal frequency bingo’. But that also meant I didn’t just run the stuff through the assembler as soon as I thought it could compile. I checked and rechecked things, which is why I burned only one monitor, not five.
Things are a little different these days. If the programmer themselves haven’t got a clue how stuff works, what are the chances that their managers have the slightest idea? Most companys think that programming means that a fat guy vomits a few letters and numbers into a file, pushes a button and everything works. Testing these days is considered optional. When was it last time that you bought a video game that didn’t install a first patch within a week of release day? If you buy software on release day these days, you’re a blithering idiot as it is unfinished. The last program I bought that didn’t need patching to repair basic functionality cost me 0,00 bucks, because I had written it myself. I had spent 3 months on programming it and 2 months on testing it and when it it was released, it worked. Today’s industry would have released it after 3 months and then spend 4 months on releasing patches that should have been part of the release version.
This is why Renault doesn’t deserve any pity. They either have completely inept programmers or they haven’t tested their stuff properly. Most likely both conditions apply. As a result of that they now have more egg on their face than a farmer, who faceplanted a day’s load of eggs. Formula One is supposed to be the pinnacle of automotive engineering. Bringing software that is fundamentally broken is completely and utterly unacceptable There is nothing more humiliating than your customer (Red Bull) announcing that they are going to send a task-force to sort out the mess you’ve made of a product you’re asking 20.000.000 items of currency for. Needless to say, money will arrive at Viry in much smaller quantities than originally anticipated. Nobody would pay the full price for a restaurant meal, if they had to go to the kitchen and prepare it themselves, because the cook is hopeless.
Renault puts the powers that be between a rock and a hard place. There are three possible scenarios. The first one is, that Red Bulls hacker brigade is over 30 and played with soldering irons in their youth as opposed to ‘my little pony’. It would be a massive achievement if they would sort out Renault’s software by – let’s say Barcelona, but that would also highlight Renault’s utter uselessness even further.
The second scenario is that Renault keeps making a fool of themselves with catastrophically inferior units, which would mean a catastrophic financial loss as all their customers would withold payment for what would essentially be a broken product. In that case FIA would find themselves in a situation where they would be forced to grant Renault permission to sort out their engines despite the engine freeze, which would lead to much wailing and gnashing of teeth in the Mercedes and Ferrari camps.
The third scenario is the unlikeliest, as it would require that FIA actually grow a spine. Back in the late 1990s Toyota entered Indycars as an engine manufacturer and their engines were ridiculously underpowered and unreliable. After almost 2 years of on-track shambles the organizers set them a deadline and threatened to exclude them from competition if they hadn’t at least achieved an acceptable level of reliability by then. Kicking out a manufacturer for utter uselessness has happened in the past – Andrea Moda anyone?
Renault have done every mistake in the book. They invested too few resources, employed too few engineers and tested on a woefully out-dated and inferior dyno. As a result their engine is best described as utter crap. Like last year’s early season Pirelli tyres it is not fit for the intended purpose and will serve no other purpose than to thrash the manufacturers image. The only team that can make the thing last longer than barely a race distance is Caterham. But to satisfy the units excessive cooling demands they had to ruin the cars aerodynamics and as a result the green monstrosity from Leafield would come second to most of my furniture in the wind tunnel.
If Lewis Hamilton’s and Jenson Button’s statement from the latest test are anything to go by, Adrian Newey has built a stunner, despite the hurried design phase. Even though Vettel was running with a neutered engine on the last day of testing, Motorsport-Total reported that his corner speeds saw the opposition’s jaws hit the deck hard. But all that comes to nothing if your lobotomized PU loses you 30kph on the straights. Newey might have packed the car tighter than David Coulthard’s white trousers again, but it is the excessive need of cooling, shoddy programming and general hopelessness of the Renault engine that renders the RB10 the world’s most expensive paper weight.
For this epic fail they deserve neither pity nor forgiveness. They deserve nothing but ridicule. Unfortunately four teams have to pay the price.
Excellent op-ed piece. I hope the FIA actually doesn’t do anything and let nature take its course. If Renault doesn’t fix its problem, then they will slowly begin to lose their F1 customers, and will fade away from the world stage. I’m sure it wont affect Clio sales either way, but their reputation as one of the best ever F1 engine makers will be relinquished to the history books, as Ford, Climax or Alfa. Time for someone else to take the throne.
ha! epic thrashing! lol
Yeah well imagine it’s your favourite team and driver that was affected by Renault’s shit power unit. You’d be the first to scream for allowances for Renault to improve their systems throughout the year.
Oh wait… Um. Nevermind. 😉
My sentiments exactly. They deserve to look rather silly for a at least the 1st 4 races before any “special permission” is granted to them for improvements. I read the Renault had only half the number of technical staff on their engine project as Mercede did. Surely everyone knows, even newbies to the sport, that the 1st rule of Grand Prix racing is ‘go hard, or go home’. I honestly feel that if Renault are given a special dispensation to do any work at all on their engine, then the other manufactures should be allowed to work on the same areas as Renault, as we all know that they will not only improve reliability but performance too. Let’s face it, they have form!
Brilliant piece – I stood up and applauded after reading – awesome piece. This is Formula 1 guys – the pinnacle of the sport – shape up or ship out!
“second to most of my furniture in the wind tunnel”. Simply the best!!
Mr Hippo,
You forgot this 😛
http://i.imgur.com/Owxi9.gif
This is absurd.
The software problems will be fixed.
But if there are significant h/w problems, what is the advantage of allowing more than a third of the grid to show up race after race in under-powered, unreliable cars?
F1 would be a 14 car grid essentially… That would kill F1.
I’m not in favor of killing F1.
I might be.
Most people here (?) agree that F1 needs fixing. And with that we mean real fixing. Not double points for the last race, or DRS or sprinklers.
This is a good time for a proper collapse. A good time to rise from the ashes!
However… It might just mean that it slowly, slowly dies. Probably that would be bad.
What is the advantage of reducing the field to 14 regular F1 cars plus 8 slow, unreliable cars that have a significant h/w issue for the next 19 or more races?
The advantage that F1 has this year is related to Skinner’s operant conditioning, in the sense that variable reinforcement has been found to have the greatest effect on behavior. Basically, it’s the psychology behind slot machines.
The fact that in any given race we *really* don’t know who will finish should be a big positive for the audience, and the fact that the Renault is behind and mostly guaranteeing someone other than Vettel winning the first few races is also a big plus, Of course, as you point out, if the results become too predictable you lose audience, so ideally you want the situation to be fluid from race to race and for the viewer to never be sure if they will get what they want, i.e. “their” driver or team atop the podium.
The narrative of RB getting “punished” by the rules, while not strictly accurate, will also be a big draw IMHO. Look for all the broadcasts to run with that as a story line all season till they manage to win a race.
I sincerely relish the thought of seeing the Red Bull freeze on the starting line when lights go out in Melbourne, the Toro Rosso opting in shame for rolling their car from the pits instead of the starting grid, the Lotus catching fire 3 laps on and the Caterham being lapped by a Marussia. I’m not sure why.
Perhaps because the RB is a net liability to F1, with Horner cajoling with Mr E and Todt for unacceptably favourable conditions (read: payouts). Maybe because Renault has been sneakily improving their power plant during the last freeze to the point where they’ve become utterly dominant the last four years. Perhaps because the insolvable Lotus shipwreck needs to burn for F1 to finally realize that the incredibly inequitable distribution of revenues is unacceptable: I see no reason for there to be more than a $10m delta between the payout to Caterham vs the payout to Red Bull. (It would be incredibly sad though if Enstone burns down along with Lotus.)
The point being, evolutionary pressures manifest themselves in adversity. Survive those who are most adaptable (and resilient to change). Perhaps F1 needs to let go of a couple of teams or manufacturers to allow for new blood to trickle in (Honda anyone?), perhaps in more equitable conditions than what Marussia, Caterham and HRT have experienced. (Equality of chances is superior to doling out cash for “historic” reasons.) Perhaps Red Bull needs to go the way of Benetton, and seek marketing opportunities elsewhere without ruining a sport.
I’m no expert, of course, but this should NOT kill F1: if F1 were to die because of this, then it’s not worth calling itself the pinnacle of anything. In any case, subsidizing industries that went broke is a big no-no of hardcore free market economics. If Red Bull and Renault, among the most favoured of actors, screw big time within the regulations, they should be made to pay the price (i.e. public humiliation). (Think of McLaren who took one with their MP4-28.) So lending them a helping hand in the circumstances may not be the most useful use of the sport’s resources.
Great comment on a great article.
At the end of the article, Mr. Hippo makes the point that Renault deserves nothing but ridicule.
If they’ve only s/w problems, they’ll be solved in a few weeks.
if they have h/w problems also, the implication is that they should not be granted a modification by the FIA. That is absurd, as we all know it will happen anyway. It will happen for performance equity, and to entice other manufacturers.
The institution of F1 wants new champions…. Don’t expect any favours for RB and Renault…
Many of our new readers may not realise that the Hippo is die hard Vettel and RB fan….
Loved it oh wallowing one (applause applause)
Yes the Hippo is a diehard fan of Vettel but the real essence of the Hippo is the rant…. we always look forward to that
Well thank you oh gavel wielding one. Although it got quite a slating on reddit though. I nearly laughed my derriere off when some guy made my point by showing he’s exactly the sort of clueless kid coming out of university these days.
I avoid forums per se…. The abuse I got early doors when I tried to publicise TJ13 on forums made me seriously wince and want to curl up into a foetal ball….
It’s a whole F1 underworld I just don’t get…..
Then it is just as well you keep avoiding the forums 😉
You might be interested in this, particularly the posts from ‘enotsne’….http://forums.autosport.com/topic/191766-renault-problems-are-catastrophic/
I was going to argue with you throwing all the blame on Renault, but then again, if they can’t build an engine to satisfy the tight design of their factory team, something went wrong.
In 1983, Barnard told Porsche I don’t want a flat engine, I want a compact one to fit in my design. McLaren dominated for the next 2-3 years.
Either Renault are truly incompetent, or the fact that engine and team factories are separated by water played a big role.
Great piece of work though! Well done!
Just a comment if you allow me for the future. I thought the paragraphs talking about IT were a bit too long. I liked the analogy, but half the length of the text on IT would probably work as well.
I for one liked the IT bits. They rang a lot of bells from (somewhat tangential) personal experience. 🙂
While I cherish the thought of RBR struggling this year, I fear them becoming involved in the programming task, because it is likely to produce results.
I believe most university qualified programmers (like my son) do actually cover off assembler type stuff. I don’t think that’s the issue. I think the issue is French. Most modern programming languages are not written in French needless to say, and I think it’s a significant problem to write in C sharp or Java for that matter without a better grounding. That said, I am assuming that the software WAS written in France. Most likely Mercedes software was written either in Oxfordshire or Germany. German of course as a language is incredibly precise, so it doesn’t pose the same contextual problems.
It’s most likely why Ron Dennis choose to replace Martin Whitmarsh by a frenchman, just to ensure that the whole team cannot understand the Racing director…
A great article Constable Hippo 😀
Particularly enjoyed metaphors such as “…..the green monstrosity from Leafield would come second to most of my furniture in the wind tunnel.”
LOL
As a software engineer of 14+ years i agree with some of your points. However, to lay blame soley at the engineers/testers feet is not only unfair but most likely inaccurate. No the issues plaguing Renault is probably what’s plaguing most companies today and that’s lack of leadership, management and policatical infighting. For instance if we were to take your software analogy further and look at a company like Microsoft, you’d see that they have no shortage of talented people. But how is it they can take several years to develop something like Windows Vista only to watch it fail miserably upon release? Or to completely miss the boat on mobile devices only to watch your competitor take off and run with it?
If its possible for a multi-billion dollar company like Microsoft to fail, what makes Renault any different? So while i agree that not every Renault engineer is a “rock star”, i also think its unfair to level blame on just one group within a large organization. I mean, who planned this project? Who designed the systems? Who created the test plan? And finally who signed off on the final iteration of the engine?
Oh and with all due respect Mr. Hippo, coding drivers for your dvd drive is a far cry from writing complex engine management systems. And i’m pretty sure they don’t code in C#/Java or any managed language for that matter on these systems. So i’m pretty sure your theory about “script kiddies” working on this stuff is way off.
The C#/Java analogy is of course not an accurate description. What I’m getting at is, that proper hardware programmers – and that’s what you need for such a task – are hard to come by as today’s university education does not help in that.
I worked in a hardware programming job in 2012 and if I told you what they paid me, you’d empty your bowels in a rather rapid fashion – it was an obscene amount of money, simply because they had failed to find a hardware coder for over 6 months.
I think I adressed Renault’s lack of leadership by mentioning that they did every mistake possible – too few engineers, too small a budget and a pre-historic dyno. Even if they had the best coders in the world, it wouldn’t have mattered. I chose to concentrate on the software side, because that’s the one part where I can make an educated observation and because it was such a prominent problem during the tests.
A textbook example of argumentum ab auctoritate.
Ouch…
😉
Perhaps The Judge and Hippo have more in common than they think?
Come on, Fat Hippo, don’t hold back; say what you REALLY think!
The only thing that can save Renault now is hiring a couple of real experts from out there in cyberspace, people who studied IT twenty or so years ago and know just what needs to be done.
Hippo.
Rindfleischetikettierungsüberwachungsaufgabenübertragungsgesetz. I think you have been eating impure Bull !!! Red meat is bad for you ha ha!
Seriously, I think you might just consider, that over the last few years, Renault were way ahead with programming their engine control system. If you look at page 87(?) diagram of the current F1 Tech regulations, you can use your programming experience to see that there are going to be a substantial number of control loops. Not an easy programming task. That diagram is far too simplistic. If you design an engine(or anything) that operates on the limits of its performance envelope, then the control systems have to be perfect, or you will let out the magic smoke. You might have missed it, but Remy Taffin mentioned that they were having problems in getting their software to talk to RB’s. Programmers huh! Reaches for his copy of Control Systems Engineering.
ps Nicely written. Who said Germans don’t have a sense of humour……
Wait…. That was supposed to be funny?
Iain:R8 I have experience with the sort of programming used in control units. While I haven’t been working with F1 engine units, I did so on industrial installations. Renaults software broke down the second they activated it in the Caterham at Jerez.
If only half of what Marko said is true, Renault have seriously effed up. They didn’t test the engines even remotely adequately and the same apparently happened with the software, which is absolutely and utterly unfathomable as the software is developed and tested in emulators before it is installed in the engine. In an emulator you can subject the software to conditions it will never meet in practice.
Two years ago I programmed a controller unit for an industrial gas production line. Imagine it would have been tested for real on the real application. A bug would have meant a huge BOOM and a couple of dozen corpses. When we installed it for the first real life test it worked and all subsequent modifications were improvements, not bugfixes. That hadn’t had to do anything with me being brilliant. It had to do that the software had undergone all theoretically possible scenarios in the emulator. You have no idea how often I got the “critical condition” error message before it was deemed stable. Each of these messages would have meant a catastrophic explosion had we tested it on the real thing.
There is simply no way to come to any other conclusion than that Renault have done a biblically rubbish job.
A German Hippo with humour….. My god!!!! The pink flying elephants now appear normative…..
“Rindfleischetikettierungsüberwachungsaufgabenübertragungsgesetz”
Btw, that’s a real German word. It means “law regarding the delegation of reponsibilities for inspection of the labeling of beef”. It’s a law passed through the Bundestag a couple years back during the BSE crisis.
I say Renault and the teams that chose them deserve to suffer. I also hope they don’t get any special update permissions, after all F1 is a competition, and all the socialist fairnes garbage makes me want to vomit.
Sorry to go off on a tangent, but your assembly language example piqued my interest. As more of a mathematician than a programmer, I wondered is there a reason that
SHL AX,4
MOV BX,AX
SHL BX,2
ADD AX,BX
wouldn’t be faster still?
Good spot 🙂 Indeed it would be faster by a few tact cycles and is in fact the exact snippet that one would find quite often in early 90s DOS programs as part of subroutines to write text mode output directly to the video memory.
It’s exactly your mindset that has gone missing a bit over the year – looking at a piece of code with the instinctive first thought being: Hey, can’t that be made a bit faster still?
I used the slightly longer one as it is emphasizes the fact more how several operations can still be quicker than a single instruction. Great to see that some still spot it
This is the biggest load of nonsense on the subject of programming I’ve read in a long time. Where to begin? 1) They are not programming in C# or Java. 2) There are these things called “compilers”. 3) Even when you were programming in assembler there were layers of abstraction between you and the 0s and 1s. 4) “If today’s script kiddies write a program that doesn’t work, chances are they hit a bug in the runtime system.” – Highly doubtful. 5) This implied notion that something written in assembler either works perfectly or not at all is complete fantasy.
Thanks Dan, Let me answer some of your points:
1) I never said they’re programming in Java and C#. My main gripe is that those are the languages universities use to teach programming. It used to be C, which is the better way as C is close to the architecture and requires good skills in memory management. Once you know C, learning Java or C# is easy, but if you started the other way round, you’ll have a jolly hard time getting to grips with C.
2) I challenge you to research the version history of the gcc compiler. Even the best compilers produce code that is inherently slower and less optimized than assembler code and compilers can occasionally generate faulty binary code from perfectly sound source code, hence the research of the gcc history. Also, things like ECU’s are not quad-core processors, but small little micro-controllers where you might be pushed to avoid any unnecessary instruction.
3) Nope, no abstraction layer in assembler. Each assembler instruction transforms into a corresponding machine code instruction.
4) I suggest you work as a IBM Websphere or Oracle Weblogic admin for a while. You’ll notice how many Java applications leak memory like a colander, mainly due to a buggy runtime.
5) You missed the important part. The one about testing 😉
Sooooo I think what you are saying is Neo became the One by allowing Smith to integrate him thus leaving Smith without a purpose which therefore placed the systemic anomaly back in harmony and thus negating the purpose for the One’s reemergence?
Couldn’t have explained it better m’lord 🙂
So if you don’t think they’re programming the car’s systems in high level languages, what are you complaining about? You think they’ve hired programmers straight out of uni with no experience in C?
Yes, Renault operated on a tight budget. One of the first things companys do to keep costs down is hiring younger, less experienced and thus cheaper staff in the hope they’ll ‘grow with the challenge’. But these young people are sent out into the professional world with much less comprehensive education than, say, 20 years ago. That imho contributed to Renaults software debacle.
It’s only 22 years since I finished A level. We had to pass exams in Math, German, English, Russian, Biology, Chemistry, Physics, Geography, Astronomy, Art and Music. Can you make just one compelling argument for why the current generation gets an A level by just passing exams in German, Math, English and one subject of their choice? You can get an A level without having had to attend any lessons in natural science. How can they be prepared for live the same as we were? If anything I pity the young ones, because they are sent into an ultra-competitive working environment with one hand tied behind their back
I’m sorry dude but the idea that they’re hiring OO programmers to program systems on F1 cars is utterly ludicrous. Programming has branched out massively in the last two decades, and whilst it’s true to say most are high level programmers (because that’s where the demand is), people with the skills to program in lower level languages are also far, far more numerous than they were 20 years ago. It’s as if you think they don’t teach this stuff at uni any more. I’m sorry but that’s just totally, utterly and completely incorrect.
22 years ago was 1992, only a few years before I did my exams. I’m pretty sure you would NOT have had to take 11 fkin A-levels when we only had to take 3. Are you getting confused with GCSEs? I think this just further illustrates how out of touch you are with the current state of education.
I finished my Abitur (German A Level) in 1992. We were the last ones before the A level was ‘reformed’ in 1993. The deal was simple: We had the following subjects: Math, German, English, Russian, Chemistry, Biology, Geography, Astronomy, History, Physics, Art, Music, Sports. Latin was a voluntary subject. Each except Sports had a final exam. Major subjects, like Math, German and English were 4 hour written exams, most of the others 1 hour oral exams. For each subject you got a final mark between 1 (very good) and 5 (failed). If you failed a single subject, you failed the A level. If you got at least a 4 (sufficient) in all of them the median of all determined your graduation mark. Some university subjects, like Law, Medicine and CS were subject to ‘numerus clausus’, meaning they required a certain graduation mark, 1.8 for computer science for instance.
In 1993 the system was ‘reformed’. Out of all the subjects people could select 4 major and 2 minor subjects, with German, Math and one foreign language (English, Russian or French mostly) being mandatory, so people could select one major and two minor subject of choice. What that meant was, that you could pass A level with German, English, Math, Sports, Music and History as your subjects. Tell me that’s not a massive step back 😮
Ah sorry, I assumed when you said A-Level you were referring to the British system. Ignore the previous last paragraph. The rest still stands though.
Assembler programming is no longer taught at my old university. (It was already a voluntary subject in my time). C is now voluntary and the mandatory programming language is Java. Some smaller colleges like Köthen still offer System Programming as a specialisation, but very few take that up. Graphics&Design and Database programming is all the hype these days.
I was hired for a hardware programming project in 2012. At age 37 I totally wrecked the average age in the team. They were all 50+. They were so desperate to find hardware programmers, they offered me obscene amount of money. I earned more on a single day than a minimum wage worker in a whole month 😮
The company that hired me back then established their own education program. New recruits are taught system programming for a year. That’s how bad it is these days.
There may well be a shortage of systems programmers, but that does not mean they’re less numerous now than they once were. Far from it. There’s far, far more demand for them now than there ever has been. F1 teams represent the cutting edge of technology. I assert that they do NOT hire programmers that don’t know what they’re doing whilst the qualified systems programmers go off to make dishwashers. If you think otherwise I’d like to see some evidence because it strikes me as a pretty outlandish claim.
Well, all you need to do is look at the result. Renault didn’t even test the bloody engine properly and the software broke the minute it was activated. Tell me how that can happen if you hire people, who know what they’re doing?
My claim that people come from university these days with knowledge that is so far removed from what is needed in the industry is an observation I make almost daily. I started my own company in 2004. One of the first big projects was me being hired to build up the development department of a company, who had hired 10 uni graduates and then realized, none of them had the faintest clue how to write System libraries in C. And it has been a common occurrence since then.
I have since then developed a ‘standardized’ test to gauge the quality and knowledge of a new recruit:
1) write a program in C that takes a number X (1<=x<=20) as input and calculates the first X prime numbers and stores them in a linked list. Time: 2 hours, System: Linux
2) What is a CISC processor?
3) Write a DLL that enables a program to send an email via the MAPI API. Time: 2 hours
4) Name a more performant substitute for the statement:
MOV AX,0
Most never make it past 2) and I've gotten the correct answer to 4) exactly twice in 8 year even though it is the easiest one of the lot.
So your logic is; Renault messed up on the software, therefore they must have hired people with little or no experience in low level languages? As a programmer, does that logic strike you as sound?
My logic is rather simple. Merc and Ferrari got it right, Renault got it wrong. The only real difference between them is man-power and budget and I have yet to meet a company that doesn’t go for hiring greenhorns as the first step to save money.
After the V8 freeze Renault downsized massively, so they must have hired and unless France has become the country where hardware hacks grow on trees, their team is very likely to be having quite a few people, who are only mildly qualified for the job.
You have to keep in mind: They didn’t hit a minor snag. Their software is a complete and utter failure to the point that RB now says – ‘get out of the way, we’re doing it on our own’. Thats the most spectacular failure imaginable. That could not have possibly happened, had they people on their payroll, who really know a byte from a mouthful of food.
“That could not have possibly happened, had they people on their payroll, who really know a byte from a mouthful of food.”
Could not *possibly* have happened for ANY other reason? I would argue that yes it could. If you’ve only got a limited amount of time, you can have the best programmers in the world and still cock it up. Software development can fail for literally thousands of reasons. It seems to me you’ve made one massive assumption that they’re hiring hacks and then used that assumption as the basis of a massive rant about how things were better in your day.
We are talking about close to two years of development, Dan. No way a software can fail like that after such a time if there isn’t a massive deficiency.
4) xor ax,ax
It’s 1 byte less.
Am I the third?
You are indeed, although the reason is wrong. It’s not the byte saved, but the processor cycles that are saved 😉 Now do the other three and you’re hired on the spot 🙂
As an under-30 year-old still reeling from his first day spent burning each finger and inhaling solder fumes, I understand what this hippo is getting at.
Also, I need a loooong mud bath. I fixed that bloody charger though!!!
“We are talking about close to two years of development, Dan. No way a software can fail like that after such a time if there isn’t a massive deficiency.”
You don’t actually know how long they’ve been working on it, nor do you know how many people they’ve had working on it, nor do you actually know the details of what the problems even are. You’re entire rant is based on one initial assumption.
Phew… but all I can say is: So far it looks like a pretty sound “Initial assumption”… 😉
I could not disagree more.
Well sir I disagree with you, and agree with him. I’m a developer, and analogy wise I don’t care how old a developer is or what language they code in, I’m a real jerk when it comes to code quality. This code is obviously crap, because a truly good developer writes modular code that can handle unforeseen edge cases.
I’m really sorry to be disrespectful but this is just utter bollocks. You’re implying that code quality is the only thing that comes into it. They’re having software problems, therefore the code must be poorly written? Errr…… no. Unless you’ve actually seen the code I would assert that you are not qualified to make even the slightest comment as to the quality of it.
I must say I’m astounded that there seem to be several people here who work in the software industry but seem to think that poor programming is the only reason a software project can fail, when in reality there are SO many other reasons. How can you be so presumptuous? I don’t get it.
Quiche eaters.
http://www.pbm.com/~lindahl/real.programmers.html
Forza Fortran… 😉
Wow this thread is still going. Time to fuel the fire a bit more. Robust discussion is never offensive. Hope it stays that way!
Hippo said: …..”software is a complete and utter failure to the point that RB now says – ‘get out of the way, we’re doing it on our own’…..
You seem to have forgotten, that Remy Taffin is on record, stating that the Red Bull and Renault software is not talking to each other properly. Engines and chassis are different specialities. Who is teaching who?
Hippo said: ….”No way a software can fail like that after such a time if there isn’t a massive deficiency.”
To repeat my earlier statement, and reading the tea leaves; Renault has gone for a very advanced design. This will take development time. A few days track time was never going to be enough. Newey acknowledges the potential engine advantages. Look how long it took them to get the blown diffuser and engine mapping working properly. But more importantly, look at the end result. It’s worth considering the aircraft Industry at this time. Airbus planes have had a history of ‘alleged computer problems’ yet they have more programmers than the entire F1 field. In addition they have a similar sized parallel group, creating test software to test the other software. Better not forget the outside audit team as well. As for battery ERS problems, think Sony and HP laptops, and Boeing 787 Dreamliner. None of these issues are new to high tech engineers. We just get our heads down and fix them!
Of course I could be 0 or 1.
“Always program as if the person who will be maintaining your program is a violent psychopath that knows where you live.”
Martin Golding
You’re soo ridiculous
you’re the only one who fails epicly here and deserve neither pity nor forgiveness. You deserve nothing but ridicule you morron.