Linux Played a Crucial Role in Discovery of 'Higgs boson' | Hacker News Reader

Linux Played a Crucial Role in Discovery of 'Higgs boson'

ubuntuvibes.com

166 pointsdartttt93 comments

Do we still need to point things like this out? Who here thought Linux didn't play a major role at CERN? Linux is the norm in large data center applications. What percentage of US commodities and securities transactions are touched by Linux? My guess is "most".

There are interesting facets to it, I think. What would they be using if Linux didn't exist? How much longer would it have taken if they'd had to use BSD? Or Windows?

How many of those millions of man-hours contributed to this discovery? How many other major breakthroughs have they enabled? How much amazing potential is locked up in proprietary software, utterly useless?

A major discovery like this is a perfect chance to blow the horn for publicly-funded research, and open source software is a huge part of that.

I'm curious: For the applications at CERN, would BSD really make much difference to Linux?

It would be harder maintaining a large collection of software outside of one of the major Linux distros. Debian/kFreeBSD might have a chance thanks to the GNU userland, but the archive is less well supported than with the Linux kernel.

Why do you say that?

Quite the opposite. This isn't about the userland, it's about the kernel. The linux kernel is where the high-performance cluster computing advances have been. The GNU userland is just a clone of the BSD userland, nothing special there. Moreover, FreeBSD and OpenBSD's packages are diverse, well-maintained, and current. But neither OS's userland nor any of their packaging systems have anything to do with high-performance cluster computing.

I may be wrong but they probably have a support contract with RHEL or similiar or they are using a sponsored distribution that caters to their specific needs such as Scientific Linux. I don't know if similiar projects/teams exist for BSD distros.

I doubt CERN has a formal run of the mill support contract with RH like other large organizations. Scientific Linux is developed (in-house) in partnership with Fermilab.[1]

[1] http://en.wikipedia.org/wiki/Scientific_Linux

They use RHEL too: http://linux.web.cern.ch/linux/rhel/

For offical usage in the context of Oracle production databases and eventually other applications requiring commercial support there is a limited number of Red Hat Enterprise Linux 6 Server, Red Hat Enterprise Linux 5 Server licenses available at CERN. For all other tasks standard CERN Linux distribution: Scientific Linux CERN shall be used.

IIRC, scientific linux is maintained by CERN itself.

BSD kernels generally do not have such good support for hard real time applications. It's been done though, so I don't know if it would have a real impact.

No need for hard real time, first level detector output goes through FPGAs which only pass potentially interesting data to the Linux powered systems which do further data filtering.

How much longer would it have taken if they'd had to use BSD?

No longer, because software was not the bottleneck.

The hard part about finding a Higgs boson is designing the gigantic particle accelerator and the detectors, then getting the funding to build them, then building them, then running them for N hours and analyzing the data. These things took decades. They started in the 1980s, years before Linux.

I'd contend that manpower is the bottleneck, and time spent writing and debugging software is time taken away from designing and performing experiments. How much time, exactly, depends on the precise software.

I'm not saying it's on the scale of years, but that's what I'm opening up for discussion. It's certainly plausible, given the sheer scale of work that's gone into the development of Linux, that there may be features that researchers would have had to recreate at great effort and expense had Linux not existed.

Haven't kept up but 20years ago when I shared a building with these guys the computers were the bottle neck.

The original design for ATLAS had a sampling system where only random blocks of data could be analysed just because the computers (and buses and memory) couldn't keep up - so the data analysis became a sort of monte-carlo process as well. And this was in spite of huge and very impressive arrays of custom FPGAs and Transputers (in the 90s).

computers were the bottleneck

Perhaps, but we were discussing software. Specifically, "Linux" vs "some other operating system".

I agree that advances in computing must surely have a lot to do with the discovery of the Higgs, but such advances are about hardware, not operating systems. Since the 1990s, billions of dollars, both public and private, have been spent to keep Moore's Law rolling forward, and that's most of what makes modern computers more capable.

Anyway, it occurs to me that this discussion is a classic bikeshed. Rather than debate tricky and obscure issues like the quantum mechanics behind the Higgs phenomenon, the engineering of supercolliders, the politics of getting supercolliders funded, the ins and outs of data processing algorithms at CERN, or the techniques of manufacturing transistors with a 20nm gate size, we have fallen back to debating whether or not the license used for the operating system was vitally important.

If CERN had had to implement a Unix entirely from scratch in order to do their job, they would have done so. It would have been a minor side issue. Indeed, from a certain point of view, that may be exactly what they did. Why discuss how Linux was vital for CERN's scientists, but not the other way around? Perhaps the reason why Linux was so good for high-energy physics is that high-energy physicists built it that way?

Personally (my PhD is in experimental physics) the interesting bit is all the vacuum pumps, magnet control, beam dumps, control systems as well as the detectors.

Whether some semi-imaginary particle fits the parameters of some semi-imaginary theory is pretty uninteresting ;-)

> Whether some semi-imaginary particle fits the parameters of some semi-imaginary theory is pretty uninteresting ;-)

I think they prefer it if we call such things "complex" rather than "semi-imaginary" ;-)

> complex

I found that once you get past the technicalities and when you get back to a bird's eye view, interactions among elementary particles follow almost naturally and looks deceptively simple [0], and the postulate of the Higgs boson existence seems like one† of the obvious solutions explaining mass.

Now when you get back down in the trenches and have to properly define it theoretically I'm positively convinced it is another matter entirely in terms of complexity.

† Then again, we previously tried to explain how light could be propagating through a postulated medium called luminiferous aether. Hopefully it looks like the LHC experiments are not going the way aether experiments did.

[0] http://en.wikipedia.org/wiki/Standard_Model#Gauge_bosons

Read the Wikipedia page on the Compact Muon Solenoid. There were some significant computational challenges involved, including how to (pre)process and store terabytes of data per second. These are challenges that push the envelope at every layer of computing, from hardware to the OS and application levels. It's not an informed proposition to say that the OS was some replaceable commodity with an unimportant role -- that's like saying Linux is unimportant at Google or Amazon. The ability to tune the kernel wad definitely crucial.

I would also question whether 20+ years of devlopment on a private UNIX clone would have produced a better result than Linux. There are plenty of private UNIXes but sarcely a convincing argument that a single one is better than Linux in some meaningful way.

You're opening it up for discussion with the implication that somehow Linux provides superior tools. However, despite the strong implication, you're unsure of what those could actually be.

This is a leading line of discourse -- you're strongly implying that Linux permitted more rapid development, but you don't appear to have any supportable justification for why that would be the case.

You wound me, sir.

It has been noted by others (in the article, for example) that Linux is the undisputed king of high-performance computing, in the public sector at least. My only assumption is that that is not random, that there are reasons for it.

I'm not trying to lead the discourse anywhere -- I'm certainly not a Linux fanatic, if that's what you're implying. I just think it's utterly plausible that Linux, and the open source community in general, saved these people a whole bunch of time.

> My only assumption is that that is not random, that there are reasons for it.

It's complicated, more so than can be adequately addressed here, and far beyond a simple question of Linux specifically "saving [people] a whole bunch of time".

I would imagine having an Open Source kernel is an advantage for very niche applications like this because you can add/change features that would be useless or detrimental to 99.9999% of users but would be useful in this instance.

I have no idea whether that was part of the reasoning for this or not though.

> How much longer would it have taken if they'd had to use BSD?

Macs (BSD-based) are popular at CERN as well:

Smashing Research at CERN http://www.apple.com/science/profiles/briancox/index2.html

> What would they be using if Linux didn't exist?

Probably Solaris.

Doubtful. Seeing as they maintain their own Linux distribution, they would probably be maintaining their own BSD flavor instead.

What would Linux be good for if Tim Berners-Lee didn't invent the Web to make CERN more effective?

If TBL is representative, CERN would be running on NeXTStep if it still existed, or FreeBSD as the next best thing.

> utterly useless

Proprietary software can be pretty useful to people who find it worth paying for.

BSD and Windows don't even belong in the same sentence. Where did you get the idea BSD is like Windows?

The network stack? Well, I guess that's been removed, but it took a while...

I also think we need less stories that linux played a major/significant role in X, but for different reasons. I think we need fewer "tool y played a major role in my success" stories. The tool fetish is distracting.

The scientist in the linked story comes out and says that he is not sure about the differences between BSD and linux and that windows could have filled the gap in principle.

Tools are tools. Whether its a hammer or a Scientific Linux Cluster the significant factor is the user, the carpenter or the physicist. The search for the theory of everything is not a logistics/supply-chain-management problem it is a question of physics.

Not so sure I agree with that in general. What about the LHC? That is just a tool, too. Is that too not worth mentioning?

I did not say that tools are never noteworthy. They are important but they are also a means to an end. The LHC is hardly an average run of the mill tool; it is rather exceptionary. So I'm not sure why you think it applies "in the general case."

Well, because you claimed pretty strongly this:

"Tools are tools. Whether its a hammer or a Scientific Linux Cluster the significant factor is the user, the carpenter or the physicist"

I agree with you that tools are a means to an end and that obviously the LHC is a lot more important that the choice of OS. I just think you were a bit too dismissive of tools in your post, though perhaps that wasn't your intent. If someone is going to rah-rah a tool, in this case Linux, they should be pointing out what role it played and what about its use was essential, not just that it was used.

The ability to use Windows in principle was likened to being able to also use a raw Turing machine; i.e. that it could have worked but would have been completely inappropriate.

I think that Unix-alikes are going to dominate any sort of high-performance computing; and I think it's more than just the network effects of the current user base. The scriptability and composability that permeates Unix-alikes is essential for automating tasks across thousands of nodes. That's not to say that it's impossible in Windows, but it's not as much in Windows' culture. Being able to ssh to any node, occasionally draw graphics remotely, all these things are hugely important. The main advantage of Windows, point-and-click administration and lots of existing software, are not advantages when you can't point-and-click thousands of machines and you have to write all your software anyway.

Yep right if these were the 90's,the heading would have been worth its pounds and it was made a cliche in the 90's itself.

Sadly when I did my trainship there in the ATLAS group (2003-2004), many researchers were moving to Windows as their desktop environment.

I still remember the discussions between still writing papers in LaTEX or move to FrameMaker on Windows.

So one tells the truth about what the reality is at CERN and gets down voted, great!

Even some MSFT divisions openly admit to using Linux in-house. So, I'm pretty sure s/most/all/.

Hah, care to provide some references for this statement?

At least Microsoft Research in Cambridge is using Linux, and have admitted though. They develop ghc, the most prominent Haskell compiler, there.

That's kinda lame reference since MRC is pretty isolated from the corporation to do science.

Completely agree with this. If you look at some of the CERN infrastructure pages, you see they use both open source and commercial apps wherever they make sense. For example when I checked out the DB infrastructure a while ago, they seemed to use a lot of Oracle[1] (with some MySQL in there as well), Scientific Linux for some stuff and RHEL where needed ( presumably for support reasons).

[1] https://openlab-mu-internal.web.cern.ch/openlab-mu-internal/...

Just to point out open-source does not equate to non-commercial. Even if all you do is to report bugs, barter is a valid form of commerce and you are exchanging time for a better product. Also, Oracle Linux and RHEL, despite being paid for, are open-source. Scientific Linux is compiled from Red Hat's sources (as is Oracle's, mostly)

I see this more as a "team builder"; a reminder why we are involved in our software communities. It's great to take a moment every once in a while (especially something historic like this) and give ourselves or others some recognition.

Savvy managers and organizations know when to give praise-- they key thing is not to trivialize it. This seems appropriate to me.

Higgs Boson Played a Crucial Role in Discovery of 'Higgs Boson'.

Scientists working at CERN, Geneva have announced the discovery of Higgs Boson which is considered to be one of the most important scientific feats in understanding the creation of Universe. It is called 'God particle' because it plays central role in discovering itself.

Scientists from CERN just posted on Reddit their thanks to Higgs Boson:

I don't see any CERN related things here, so I want to mention how Higgs Boson (specifically, lower energy one - the only one currently discoverd) had a vital role in the discovery of itself at CERN. We rely on it every day in our analyses to enable particles in atoms of our brains and computers to gather mass. Without it this discovery would truly be impossible. Thank you, Higgs Boson!

Yeah, the news kinda comes down to: "CERN data centers run Linux", which is cool, considering the work that is done there, but not really that surprising at all ;)

I think we should take a step back and also find a moment for another unsung hero, water.

Water also played a bigger role in the discovery of Higgs boson than Linux. Without water the researchers would surely have died of thirst weeks or days after starting work at CERN.

In "water" we trust.

On that note - what about paper and pens. Would the discovery have been possible without paper and pens? We'll never know for sure.

I see a lot of comments here saying that the fact that they used Linux is more or less irrelevant. This is probably true to a large degree since Linux is basically a generic white label operating system that you can build on top of, so if your doing something this niche there is little reason not to use it since most of your clever stuff is probably custom anyway.

However there is definitely a large PR angle to this too. Let's imagine that they had used Microsoft Azure to crunch all the data or iPads to view/analyze it do you not think that MS or Apple would make a big deal out of it?

Whilst most HN readers probably understand the reach and dominance that Linux has in a huge number of areas, most people do not. I've heard it said many times even from people who work in tech "Linux is just a hippy OS for geeks and will never be taken seriously". LHC is now one extra thing you can add to the rebuttal.

But does that add or detract to the image of it being a "hippy OS for geeks"? I would wager most people know even less about particle physics and what the average practitioner does than they do about software engineering. It very much preserves the idea of it being an OS for geeks, by geeks.

If what you meant by "[most people do not] understand the reach and dominance that Linux has in a huge number of areas" was more in the sense of "Linux is a viable platform for building our business on" than "It's almost the year of Linux on the Desktop!" I apologise :)

These kinds of articles were much more important/interesting in the 90s when it wasn't clear that Linux was ever going to be adopted by business or become mainstream. For example in the late 90s when the Burlington Coat Factory started to roll out Linux clients to all of its warehouses, it was a big deal.

Now Linux is everywhere from smartphones to supercomputers, and has pretty much outlasted or killed off all its former proprietary UNIX competitors. It'd be news if AIX, Tru64, SCO, or Solaris had anything to do with the discovery. We can stop doing this now, we've won.

I was writing scientific X-ray detector software in the early 90s, and it all ran on ... OS-9.

(And for the newbies around, that's got nothing to do with Macs)

I know my TRS-80 ran OS-9.

OS-9 was pretty cool considering it was a fully fledged multitasking OS that ran in a very tiny amount of RAM (I think 128K was the minimum?).

Of course we ran it in 8MB on 68020s which was simply luxury.

"In terms of data analysis, Windows could be used in principle. We could also use some type of device that manipulates symbols on a strip of tape according to a simple table of rules. [...]"

The age old dilemma: Windows or Turing machihe?

When CERN presented their findings they used the Comic Sans font, so I guess you could say that Microsoft also played a role in discovery of the 'Higgs boson'. Ha!

Maybe the EU geocities was rife with teenage girls using Verdana, and Europeans secretly snicker at slide decks made by American physicists.

Can anyone comment on what sort of software they use for collecting and analyzing this kind of data? Python+numpy? Matlab? Gnuplot? In a past life I was a neuroscientist, and although you could theoretically do experimental neuroscience entirely using open source tools, almost no one does in practice.

They make their own data analysis software called ROOT. It is C++ based. I got to use some of this when I was an undergraduate in the late 90's. Looks like it is still going strong.

If you are interested, it is all open source and available here: http://root.cern.ch

They have not declared it to be "Higgs boson" yet.

Correct. We've found something that could potentially be the Higgs, but we do not know with certainty that it is the Higgs.

Code_primate explained it succinctly: http://www.reddit.com/r/science/comments/w0tty/higgs_boson_c...

They have indeed discovered a Higgs Boson they're just not sure if its "the Standard Model" Higgs Boson yet.

The have discovered a boson, but it's unknown what it is. There's just three things we know about the new particle. Its mass, its charge (0) and that it has integer spin (0 or 2).

I just wish Google had named their "mobile operating system" as Linux/Android instead of just Android, and have a "Powered by Linux" on their homepage, this way many more people would realize that they use Linux more often than they think. The problem is that then we would have purists that would call for a GNU/Linux/Android and others would ask for Powered by libc, engineers, your utility company, lunch from the Burger Joint Inc, pigeons, etc ;)

Yes, Linux/Android! That sounds so sexy! I wonder why they didn't do it. :)

That's GNU/Linux/Android to you.

Except that Android doesn't (sadly) have GNU userland.

That's because every technical decision in Android has been done by lawyers. Or at least it seems like that to an Android driver developer.

If I had a buck every time Android does something stupid (their libc, bionic, is the primary culprit) that works just fine in GNU land, I wouldn't be developing drivers for Android.

Actually, that's because the designers of Android took the gamble of replacing the Linux userland with a touch-oriented layer built on a managed language runtime.

They did this instead of trying to evolve it toward a vague goal in that general direction, or trying to drag along legacy apps.

In hindsight, it's obviously the right thing to have done. But it was a gamble at the time.

> Actually, that's because the designers of Android took the gamble of replacing the Linux userland with a touch-oriented layer built on a managed language runtime

No, I wasn't talking about the UI and the Dalvik virtual machine. I was talking about what's underneath.

That darn libc, for example. Every operating system needs some kind of user space standard library that deals with system calls and such. Android could have used glibc or uclibc like all the other Linux-based platforms do. But no, they wrote their own. To avoid GPL licence.

Which libc you use does not make a difference to your customers, so you won't be adding any value by rolling your own. Instead, you get decades worth of maintenance to take care of. And now Android has this crappy libc that they don't do a very good job in maintaining and lots of little things are broken. If I had to guesstimate, they have used at least a million dollars (and counting) worth of developer time on that libc with absolutely zero value created.

Android is full of these silly examples of things gone wrong. As a customer, you won't see them. As an application developer you might get a glimpse but if you stay within the Dalvik sandbox, you're relatively safe. If you go NDK, try to port an existing software to Android or work with the internals, may lord have mercy on your soul.

It's saturday. I don't want to talk about Android any longer.

I was under impression Android is a Java VM (Dalvik) running on top of Linux kernel, meaning it could theoretically work with any base OS capable of running JVM.

While it may use some linux-only features, is it really hardwired to it so much it cannot be ported elsewhere?

The evolution of social and scientific computing has made nix a part of everyone's life. I believe this is due largely to the fact that nix works in a way that agrees nicely with the aesthetic determination we as hackers bring to our code.

find / -name "higgs boson"

Wasn't it RHEL/Centos based Scientific Linux, and had nothing to do with Debian/Ubuntu/Failbuntu?

Great news everyone computers were used to do that thing that everyone's talking about! Aren't we so cool that something we talk about everyday is used in other things? I just get so excited when I think that the little machine that I've got on my lap now could be used for amazing scientific discoveries. Nobel Prize here I come!

Agreed. It is just like saying that pens and paper had an important role in the discovery... Great news.

Not just computers though, a computer with an open source operating system that most hackers on this site would like to see more adoption of. This is a triumph for the open source community. (Not to impugn their previous achievements of course, this just goes with the current trend being the Higgs potential discovery)

My inner RMS says, GNU/Linux.

This story is important because in the Western society something that is not "mainstream" is not considered a success. You don't see Linux on a lot of consumer devices where people graphically interact with the OS and hence Linux as an OS is deemed "not mainstream" and hence deemed "not a success". These stories will be unnecessary only when society moves past this shallowness and I just don't see that happening any time soon.

BTW, my favorite "Linux is awesome" moment? Watching an inflight system in front of my seat in a Delta flight reboot..

Computational physics platforms are not "mainstream" products in the context of western society. That being said within the scientific community Linux is certainly "mainstream."

Scientific community is not "mainstream" either because it projects "nerdiness" and "geekiness", both of which are associated with negative connotations and don't fit into the anti-intellectual narrative that is broadly perpetuated by the media.

I'd tend to agree with you in that we could see it as 'every little helps'. Positive publicity never hurt, especially in view of the desktop market which remains stubbornly proprietary. I wonder when Linux will go mainstream there.

That's more like a "Linux is mainstream" moment. If it was truly a "Linux is awesome" moment, Linux would have given you more legroom and a free beer.

Linux is indeed mainstream but as rogerchucker many measure the success of Linux from the success of desktop Linux.

> You don't see Linux on a lot of consumer devices where people graphically interact with the OS

Umm, Android?

Nobody in the mainstream identifies it with Linux.

Does that matter?

Yes it does

Going to the moon was a huge success that didn't need cheerleading from a faction of operating system fanatics. The success here is the science. The machinery is irrelevant.

Calling Linux "not mainstream" is fair enough. Calling it "not a success" is downright wrong and misguided. Linux is successful beyond the wildest dreams of anyone. No-one would have believed back in the early 1990's that an open source Unix system would one day become as big as Linux is.

Linux powers Google, Facebook, Twitter, and a lot many other popular web sites. Linux is used on Android and a few lesser known mobile systems. Many if not most of the supercomputers in the Top 500 list are powered by Linux.

Hell, measured by the number of CPU's it runs on, Linux may even be the most widely used operating system today and it's been steadily increasing. (HN: anyone have any stats on this one?)

Go check your facts.

> BTW, my favorite "Linux is awesome" moment? Watching an inflight system in front of my seat in a Delta flight reboot..

We've all seen a public display with a rebooted Linux setup or stuck in bootloader. However, Windows blue screen is a lot more common.

What part of Linux "is not deemed" a success did you not understand?

So did oxygen ... and food ... and beverages.

Reply on news.ycombinator.com