Please create an account to participate in the Slashdot moderation system


Forgot your password?
The Internet Science

Folding@Home - Yet Another Distributed Client 44

braind writes: "The Stanford group has developed a new way to simulate protein folding ("distributed dynamics") which should remove the previous barriers to simulating protein folding. However, this method is extremely computationally demanding and we need your help. You can read more on the site." It's interesting seeing all these projects coming out - just a reminder that distributed is still around and we can always use more on our team. *grin* [addendum from timothy:] Note that the SDK used for this project was discussed here a few days ago, so you can even roll -- err, fold -- your own.
This discussion has been archived. No new comments can be posted.

Folding@Home - Yet Another Distributed Client

Comments Filter:
  • The leader of the committee running this project is, of course, the folding chair.
  • So are we folding or rolling?

    A. Fold proteins and get nothing
    B. Roll weed and get high

    A & B: Fold proteins that make up cannabis!

    Weird Pics []

  • to bring back the IMHO.

  • by andyh1978 ( 173377 ) on Tuesday September 26, 2000 @01:57PM (#752142) Homepage
    Whilst projects like [] and Seti@Home [] have clocked up shocking amounts of processor time (410497.11 years on Seti@Home), they're still running on the 'cool factor' of having your machine break codes or search for aliens.

    Sites such as ProcessTree [], and others, have been talking of paying for your computer time [] with micropayments, but so far nothing seems to have got off the ground.

    Presumably with the added incentive of cash, the number of computers taking part will rocket. Does anyone have any firm information on the progress of these schemes?
  • You /. readers are getting lazier and lazier! Now you even have computer programs written to do your laundry? Seriously, use your own two hands to put the clothing in the washer, turn the knob, put clothes in the dryer, hit the button, then fold your own clothes. This convergence of technology and household chores must end!
    (Humorous, folks. Really.)
  • Does anyone have any firm information on the progress of these schemes?

    I'm eagerly wondering about these, as well. I've got a collection of 'scrounged' computers at home on my network, a copy of Mosix [], and the ability to write download scripts for packets...I'm just itching to find out how big my home cluster will have to get to finance a broadband connection for me... :-)

    Joe Sixpack is dead!
  • by myc ( 105406 ) on Tuesday September 26, 2000 @02:22PM (#752145)

    (1) proteins are not static structures, they tend to change conformations in response to stimuli like binding to a ligand, or changes in the electrostatic microenvironment around them.

    (2) many proteins don't like to fold in isolation, they require the presence of other proteins that they naturally interact with.

    (3) protein sequence is linear (so-called primary structure); while local structural details may be predictable with some reliability (the so-called secondary structure, things like alpha helices and beta sheets), ultimately it is the final 3D fold with long range interactions (tertiary and higher structures) that form the final structure. You can imagine that the longer the protein, the harder it is to fold, due to the increased number of potential tertiary interactions.

    determination of the structure of a protein, and even relatively large protein complexes is not as technically challenging as it used to be for biophysicists these days. Tom Steitz's []group at Yale has managed to crystalize and solve the structure of the large ribosomal subunit (a **HUGE** molecule as far as the average biological molecular complex goes) at 2.4 angstrom resolution, which in itself is a monumental feat. I would not be surprised if Steitz is in contention for the Nobel prize for this work.

    The holy grail is eventually being able to reverse engineer a protein or ligand that is able to bind to part of a particular protein, using rational design. This is much harder than solving a structure. Pharmaceutical companies would love to be able to design this type of molecule for use as designer drugs, since it would take away much of the cost of R&D through trial and error. Big companies such as Merck basically screen for drugs the way Thomas Edison used to test materials; by having a warehouse full of stuff and testing it all.

    That being said, it's still a cool project :)

  • I like protien folding, but when I downloaded the client there was no source code, just a binary. How do I know I'm not running a trojan, or am I too paranoid.
  • by sydney094 ( 153190 ) on Tuesday September 26, 2000 @02:39PM (#752147)
    that there aren't any good ... and I mean really really good algorithms that do this type of work yet. It isn't going to be the panacea that protoemic researchers are looking for.

    Does anyone know exactly what models they will be using? Because there are only a few ways to actually go about this:

    1 - Use a known protein structure that is similar to the one under study, but silghtly different. You can also look for common motifs in a structure / sequence to compare the two. Basically you look at the sequences, and say, "Hey, those two proteins have similar sequences, so they probably look the same too."

    2 - Good old ab initio methods where you reduce the conformational energy to the optimal folding pattern. This is basically looking only at the sequence and saying "If I were a protein, what would I look like."

    Both are relatively time consuming, but I'm not sure how suited distribution is to this task. The first method requires a great deal of database lookups, and the second requires a lot of computing power under the hood. With distribution, you don't have the database backend to work with, so it must be the brute force method. But I have yet to see any studies where ab initio have been anywhere near a 95% level of accuracy (compared to x-ray crystal structures). The best I've seen is around 75%. This isn't quite as helpful as it might sound. You can get some good results and working models this way, but you can't do a great deal with drug design with an inaccurate model.

    They had links to the papers citing their algorithms, but they links were not yet active... If they have a better way to do this, I'll be quite impressed, but for now, I think that a machine like IBM's Blue Gene [] has a better chance.

    And neither of these methods really takes into account post-translational modifications, phosphorylations, cleavage, activation, etc... (basically all the extra stuff your cells do to proteins before they are "activated").

  • by Cramer ( 69040 ) on Tuesday September 26, 2000 @02:42PM (#752148) Homepage
    Umm, I'd suspect they are likely to follow in the footsteps of a lot of the "dot-com"s. While some will argue "it sounds good on paper", that's where it should stay. I won't bore you with the details. But, this won't scale and simply cannot work without a great deal of costly planning and infrastructure which is ultimately unprofitable. But then, who cares about profitable *smirk*

    Just think about it... you owe how many people three cents? Is that 0.03$US or 0.03$CDN? What about the inscrupulous people SETI and DCTI already have to deal with? These problems (and many others) aren't simple and a handful of MBA's with fists full of seed-money aren't competent to deal with them.

    Most of the clones are the ideas of business types. They have little or no computer science or engineering background. To these people, all numbers are preceeded by a dollar sign. Most of them point to SETI as the basis for their business: SETI has zillions of... blah, blah, freakin' blah. They don't understand what SETI is, how it works, or why thousands of people contribute entire offices of machines to the cause. They see that big number and want to plant a `$'!

    A few years ago everyone wanted to be an ISP. A year ago everyone wanted to be a "dot-com". A few months ago everyone was chanting IPO -- Redhat stock is where now? Now everyone wants to be an "ASP" and "distributed network"s are all the rage. (Technically, they are all client-server not distributed. They form an easily splintered tree; the clients do not talk to each other. However, like profitability, no one cares.)
  • Isnt it about time we started doing medicine on a quantum level, as everyone who does medicine thinks that they're so damm smart, but all the fuck they do is learn outdated techniques that suck basically. If they want to advance they have to think on a new level
  • by SIGFPE ( 97527 ) on Tuesday September 26, 2000 @03:01PM (#752150) Homepage
    A few years ago I worked in computational chemistry for a pharmaceutical company. Determining the conformation of a molecule is a *hard* problem. We're dealing with quantum mechanics (QM) rather than classical mechanics and many-body QM problems are notoriously difficult. For example if you have just *two* particles the space of possible configurations is 6 dimensional (in this simple example you can use symmetry to simplify things). The wavefunction is a function on a six dimensional space. For a protein you might want to deal with hundreds or thousands of nuclei and many more electrons. You might be determining a wavefunction on a 100,000 dimensional space. Let me give a taste of how big that is. Imagine we discretise this space so that we only have *ten* steps along each dimension. Then we have 1 with 100,000 zeros after it discrete points in the space. That's *big*. So clearly any attempt to solve this problem on a classical (ie. non-quantum) computer is a gross approximation. I have serious doubts about our ability to solve this problem today - even with a billionfold increase in the power of computers. When I worked in this computational chemistry department all of the molecular modelling packages had parameters you could tune. A computational chemist would run a simulation. If the result wasn't to their taste they'd tweak the parameters and run it again. Then they'd run it a few more times. As X-ray data came in they'd fine tune their parameters to make their simulated model match. Eventually they'd give a seminar showing how their simulation matched the real results - when in fact all they'd done is find the set of simulation parameters that matched reality. These parameters were purely hacks tweaked to make things look like the experimental results. They had no a priori worth. If you took these tweaked parameters and tried them on the next simulation with a different parameter guess what! They wouldn't work. And this was for relatively simple biologicaly active compounds - not entire proteins. This is a problem that grows exponentially with the number of bodies. Thankfully some of these people realised that what they were doing was no better than Voodoo. So I hope someone can convince me that there have been big improvements before we collectively build the world's biggest waster of CPU time. Keep your cycles for SETI@home - at least then they might be useful.
  • How do I know I'm not running a trojan

    The risk is the same when running ANY program that is downloaded, regardless of purpose.

    The only way to be mostly sure, is to audit the source of every program you run, then compile and build yourself. Even then, you have to worry about the compiler, as the infamous compiler trojan [] illustrates, as described by Ken Thompson.
  • About the ribosome group...

    This took them a long time to do... and this isn't exactly a good model to work from. True, this is a huge model, and is great work. I personally think that the odds are good that a Nobel prize is in the picture. Simply obtaining the protein took a long time, and the crystallization was the hard part.

    But this ribosome is not the same as that of a human. I'm not sure which one this is, probably related to T. aquaticus or something like that (some thermophile), so it is going to be very different than anything coming out of the Human Genome project.

  • Source will be available soon, the only reason it's not already is because one of the FORTRAN libs needs to come from another site.
  • I agree it's good to be sceptical of everything. However, IMHO the situation is *much* better than SIGFPE's comments. Computation chemistry has been useful in quantitative analysis of many different areas in molecular biology and chemistry. We (the Folding@home team) have been able to in fact run folding simulations which agree *quantitatively* with experiment, in terms of rates, thermodynamics, structure, etc.
  • Why do you think Blue Gene has a better chance? With enough users, we'll have more CPU power. What they do have is fast communication between processors. However, for many applications (including our method of simulating folding dynamics), this is not needed. I think that both methods will be useful and complementary. No doubt, computational biology is definitely getting very exciting ...
  • A USEFUL distributed project. I have been interested in distributed computing since I first heard of the project but I couldn't help but feel it was a waste of time to crack encryption for cash. SETI didn't intrigue me - I leave the X-files on the TV where they belong. OGR was a bit better, but still kind of pointless. But THIS actually has some use. Count me in!
  • Are your quantitative agreements true *predictions*? I've seen some interesting ways results that are already known can feedback into the simulation without the developers realising. For example if you have a closely related family of molecules 1 to N with known properties and N+1 is unknown and you use 1 to N to calibrate your simulation then you have a good chance of getting N+1 right simply because your complex looking simulator is doing nothing but simple curve fitting and is simply interpolating from the properties on 1 to N. Does your simulator work in entirely new domains? Does it really simulate a priori or do you need to tune the laws of physics on a per-molecule basis? I know of great successes with simple inorganic compounds but I'm yet to be convinced with complex organic molecules. On the other hand my scepticism could be completely out of place which means that your work is very very cool! But then I'll launch into my tirade about how the knowing conformations doesn't help as much as chemists claim. I have great scepticism about the whole rational drug design thing and think combinatorial chemistry is the way to go. But that's another discussion...
  • by cosmicaug ( 150534 ) on Tuesday September 26, 2000 @03:52PM (#752158)
    Looks cool. Is it open source? I'm concerned that clients like SETI (and this) could just be an NSA setup to have the public decrypt its own communications on the government's behalf.

    From their site: []

    Why no Mac/Solaris/etc version?

    We're looking for good programmers to help with the ports to Mac, Solaris, etc. In general, the Cosm libraries should be easy to port and thus (with some help), we should be able to whip out these versions. Interested in volunteering? Please email [mailto].

    Presumably if you volunteer to port to system x they'll have to let you see the source code. They might even let you see it if you ask nicely for all I know.

    As for SETI, I don't know if their code is available at all (I think not --at least officially); but I know they do not want any unofficial versions around and that they've even refused assistance to produce versions optimized for the 3DNow extensions in AMD chips (none exist now AFAIK).

  • With the upcoming election what some people really would like to hear is a simpler way of exchanging long protein strings. That'll help avoid embarrasing situations of Gore and Bush walking down the street holding hands.
  • by Foldinghome ( 235676 ) on Tuesday September 26, 2000 @04:04PM (#752160)
    > I hope they come out with a version that can work without the screensaver.

    yea, you're not alone and we do have one (for linux and windows): check out the Folding@home site [] and go to the download page [], sign up, and then download.
  • Their windows client sucks. I installed it on 3 different machines, 2 of them locked up after 10 minutes and required a reboot, and the 3rd one rebooted itself after 15 minutes. Yes, I know windows normally does that, but it does it consitently with this screensaver thingy installed.

  • IBM has an protein folding intiative called Blue Gene that was reported on back in Dec. 1999.
    CNet's article is here [], and IBM's is here [].
  • Ha! Obscure Simpsons reference in tha house.
  • I ran it for 24 hours on w2k. no issues.
  • Isnt it about time we started doing medicine on a quantum level, as everyone who does medicine thinks that they're so damm smart, but all the fuck they do is learn outdated techniques that suck basically. If they want to advance they have to think on a new level

    Right. It's imperative that we go straight to the source -- and heal those sick, neglected elementary particles! There's nothing more dangerous than an electron with an advanced case of malaria.
  • FreeBSD users will be pleased to learn that the linux-redhat client works fine under emulation. brandelf -t Linux client and run as per the instructions.
  • What if there was a program that took your spare cpu cycles and used them for whatever was needed at the moment? That way EVERYBODY was sharing EVERYBODY ELSES spare CPU time?

    There should be a way to messure an estimate the amount of time it would take to do a calculation, or the number of cycles it takes to complete. If it's say, 10 million cycles or more, send off a request for spare cycles!

    Obviously there are some real problems with my dream world idea here: Network latency and bandwith problems. Imagine if everycomputer in the world was this way, and they used the internet to chatter the information... Wow, the bandwith would be sucked up pritty quickly!

    Okay, enough of a weird idea. I've heard of selling cpu cycles, but I'm more intrested in a common pool.
  • ...or at least start working on problems for the benefit of humanity such as this project rather than goofing around with random cipher texts.

    well, it would be kinda cool if the military turned to to simulate their new weapons (and see the images as they develop)...

    i have no morals... :p
  • This project sounds awesome, but my concern is how do I know that they are not wasting my CPU cycles for nothing?

    Having a lot of computer power is not going to help them solve anything,if theory underlying their simulation or algorithm being used for folding is incorrect.

    • Where can we find description of algorithm?
    • Has method been benchmarked against known set of protein folds?
    • How accurate are protein models being produced?
  • So, this means then that you are simulating folding dynamics for your algorithm. Any citation available?

    I usually see this process as requiring a great deal of feedback before proceeding to the next step. For example, if R23 moves by 2.5 angstroms, how this affects to torsional strain placed upon Q102. This seems more problematic since you are essentially trying to optimize an entire model from scratch.

    I guess the big concern is how you modularize the algorithm. The work of one process is directly related to the work of another, and I find it difficult to see how you can proceed without direct feedback from one process to another. Again, I think a citation would be helpful, especially to help understand how the algorithm is compartmentalized.

    I can see how you could have one process calculating the thermodynamic energy of one model versus another, but then the question is, how do you choose to modify the model? Is it a matter of checking all possibilities and choosing the one with the lowest energy? This could be distributed easily, but how efficient is it?
  • Does anyone know the name of the computational biology test that is done every year (or so I think) where people put their algorithms to work on a protein that has not had a structure published? Last I knew, the best methods were tested against an unpublished (soon to be) x-ray crystal structure. This makes for a nice blinded approach to the algorithm. The algorithms have gotten better, but unfortunately, aren't too good yet. And really, the only way to verify any result is with a crystal structure (ignoring mutagensis with activity assays, etc...)
  • I've been participating in a similar distributed computing project called Folderol []. The graphics aren't as pretty, but they seem to be using a genetic algorithm of some sort.
  • With a static logo that is displayed all the time in the lower left corner of the screen this is actually not a screensaver.
  • sdk that is used for this project is cosm. cosm was developed by former [] developer. you can find more info on cosm on [http]
  • protein sequence is linear; while local structural details may be predictable with some reliability, ultimately it is the final 3D fold with long range interactions that form the final structure. You can imagine that the longer the protein, the harder it is to fold, due to the increased number of potential tertiary interactions.
    It definitely is a hard problem, but it's the next logical thing to attack now that the human genome is more or less sequenced. The use of seti@home-style distributed computing seems like a good idea, except for those long range electrostatic and van der Waals interactions you mention. For a distributed system that relies on a central server, those are the killer. They represent an enormous amount of global communication on each time step of the simulation, and therefore a big bottleneck if they all have to pass thru the central server. This is a strong argument in favor of allowing client-to-client communication. That would allow the thing to scale much better.

    There is hope in some algorithms (such as DPMTA []) which intelligently partition large groups of particles to simplify the computation of long-range forces:

    ...the classical N-Body problem involves computing the net effect of the interactions of each pair of particles out of a set of N... the amount of computation grows as the square of the number of particles, for the naive implementation... The FMA process, however, uses a Multipole Expansion (MPE) to represent the effects of a group of particles as a single entity. By using the MPE when computing forces on a particle, and doing operations to combine multipole expansions, the overall amount of computation can be reduced to an almost linear relationship with the number of particles.
    Hopefully the folding@home folks are aware of such algorithms, and are using them to reduce the need for inter-client communication. By farming out as much of that computation as possible to the clients, they minimize the reliance on their non-scalable server CPU, and they also effectively slow down the clients a little, postponing the day when they find themselves hopelessly bandwidth-bound.
  • I don't think these people [] are wasting CPU time.

    Doing this [] distributed over the Internet, however, is unlikely.

  • I think there are good reasons to believe that a reasonable description of protein dynamics is possible without quantum calculations.

    Proteins are built up out of twenty standard amino acids, and it turns out that if you can make a model which describes the behaviour of individual amino acids well (with reference to quantum calculations, or experimental data), then you can describe their collective behaviour in proteins quite well too.

    The Amber [], CHARMM [], and GROMOS [] parameter sets for doing this are quite refined, and simulations using these parameters appear to agree pretty well with reality.

    The big problem is that, as the project pages mention, computer simulations of proteins have only recently hit the 1 microsecond range. What they don't tell you is that many common-or-garden proteins fold on a millisecond, second, or longer timescale. That's a factor of a million you have to brute-force your way through. A simulation also deals with one protein molecule at a time, while nature tends to fold a couple of billion of them at once, so it doesn't matter if a few don't quite make it to the correct fold in a reasonable time.

  • As I recall, one of the initial concerns (and I certainly think that this is a legitimate concern) that the Seti@Home folks had with releasing their source was that if/when these distributed computing projects started taking on the air of a competition (which they have), the availability of source code would make cheating that much more tempting. If you're depending on the return of data for serious scientific research, you don't want some script kiddie monkeying around with your results just so Team 31337 can pull into the lead...
  • FOR RELEASE: 27 SEPTEMBER 2000 AT 14:00 ET US University of Chicago Medical Center Prions, abnormally folded proteins associated with several bizarre human diseases, may hold the key to a major mystery in evolution: how survival skills that require multiple genetic changes arise all at once when each genetic change by itself would be unsuccessful and even harmful. In a study in the September 28, 2000, issue of Nature researchers at the Howard Hughes Institute at the University of Chicago describe a prion-dependent mechanism that seems perfectly suited to solving this dilemma, at least for yeast. It allows yeast to stockpile an arsenal of genetic variation and then release it to express a host of novel characteristics, including the ability to grow well in altered environments. "We found that a heritable genetic element based on protein folding, not encoded in DNA or RNA, allows yeast to acquire many silent changes in their genome and suddenly reveal them," said Susan Lindquist, Ph.D., professor of molecular genetics and cell biology at the University of Chicago, Howard Hughes Investigator and principal author of the study. There are thousands of proteins in every cell and each one has to fold into just the right shape in order to function. In prion diseases, which include mad cow disease and Creutzfeldt-Jakob disease, a normal cell protein, PrP, assumes an abnormal shape. Mis-folded proteins are usually just degraded, but the prion protein causes other PrP proteins to mis-fold, too, creating a protein-folding chain reaction. Thus, they act as infectious agents. As more and more of the proteins fold into the prion shape, they form inactive aggregates which lead to dysfunction and disease. A few years ago geneticists made the startling discovery that yeast, the organism found in bread and beer, has prions, too. Yeast prions are unrelated to the mammalian prions, and don't harm humans or yeast. They do, however, have the unusual property of mis-folding in the same peculiar way and spreading their change in shape from one protein to another. Mother cells pass these proteins to their daughters, so the change, once it occurs, is inherited from generation to generation. Because yeast prions act much like mammalian prions and are easier to study, scientists hope they will offer clues about how these mis-folding chain reactions get started and how they might be stopped. But the real puzzle is why these things exist in yeast cells in the first place. University of Chicago researchers appear to have found the answer, and it has broad and unexpected implications: the yeast prion seems to play an adaptive role and may greatly influence evolutionary processes. The prion protein they studied is called Sup35. It normally ensures that yeast faithfully translate the genetic code. Specifically, Sup35 recognizes special signals that tell the entire protein production machinery to stop when it is supposed to stop. Sup35 doesn't function in its prion state. As a result, the protein production machinery runs right through the "stop signs." This means that usually silent regions of the genetic code are suddenly expressed. Because these regions are normally not expressed, they don't face selective pressures that prevent mutations from accumulating. The prion therefore uncovers, all at once, a wealth of previously hidden genetic mutations and creates a completely new set of growth properties. Suddenly cells change the kind of food they eat, change their resistance to antibiotics and even grow colonies with completely different shapes. In some cases the prion may simply cause the protein production machinery to read through the "stop sign" at the end of a normal gene. This would create a protein whose function is altered by the addition of a new tail. In other cases the cell machinery may produce a completely new protein from a mutated gene that is not ordinarily translated because it contains a stop signal. The key to its effect is the stable inheritance of the prion state and the normal state. A spontaneous switch between the two states occurs approximately once in a million generations. Because a yeast colony produces a new generation every two hours, in a short time a colony will produce some members that have switched their state. "It's an 'all or nothing' switch, with the changes immediately inherited by all the progeny," said Lindquist. "But because the cell maintains the ability to switch back, the prion switch allows cells to occupy a new niche without losing their capacity to occupy the old." The researchers exposed seven distinct genetic strains of yeast in their prion and non-prion states to 150 different growth conditions. The prion-positive state had a substantial effect on the growth of the yeast in nearly half of the conditions tested. In more than 25 percent of these cases its effects were positive. The incredible diversity of the advantages conveyed by the prions indicated that each strain had different novel genes turned on in its prion-positive state. This prion switch is conserved in yeast across very distantly related genetic strains. Though the switch may have evolved as an accidental consequence of a shape change in an unimportant functioning part of the Sup35, its conservation suggests an evolutionary advantage. "It may be that the prion switch offers yeast a way to respond to commonly fluctuating environments," said Lindquist. "During its evolution S. cerevisiae (brewers' yeast) must have met with such erratic environments that it needed to maintain a global mechanism for exploiting genome-wide variation." By providing yeast with a way to respond to fluctuating environments, the prion switch may offer a significant evolutionary advantage. "Though we haven't shown it yet, selective pressure should operate to 'fix' the advantageous genes, which could then be read and translated at all times," said Lindquist. Prion mechanisms could be more common than previously suspected and exert an important influence on the rates and mechanisms of evolutionary change. "We need to expand our understanding of inheritance," said Lindquist. "It involves much more than a certain nucleic acid sequence of DNA." Susan L. Lindquist is the Albert D. Lasker Professor of Medical Sciences, Department of Molecular Genetics & Cell Biology at the University of Chicago and a Howard Hughes Medical Institute Investigator. Her co-author is Heather L. True, a Fellow in the Department of Molecular Genetics & Cell Biology at the University of Chicago. html
  • has had difficulty with people falsly submitting large blocks of work. Not only have people submitted large blocks of work that was not completed but I beleive that their have been falsefied reports of positive returns.

    It is for this reason that they only release the source to their core. And not to their network interface. This is an attempt to prevent a script kiddie from tweaking the source to boost their stats. In order to do this you actually have to reverse engineer the client/server communications protocol.
  • Hey, I said on a quantum level not molecular, bitch grrrrrrrrrrrrrr........and fuck andshit and stuff
  • From the FAQ [] (second question):

    Who "owns" the results? What will happen to them?
    Unlike other distributed computing projects, Folding@home is run by an academic institution (specifically the Pande Group, at Stanford University's Chemistry Department), which is a non-profit institution dedicated to science research and education. The results from Folding@home will be made available on several levels. First, we put movies and images of all folding runs on the web for everyone to see...

Slowly and surely the unix crept up on the Nintendo user ...