Dell warms to Linux and open source

vnunet ^ | 22 Mar 2006 | Iain Thomson

[Golden Eagle]

Beowulf...wouldn't exist if not for free software

Bull manure. You freeware fanatics have definitely overdosed on you elixir

[antiRepublicrat]

Bull manure. You freeware fanatics have definitely overdosed on you elixir

Okay, what should Beowulf's creators have done? Where could they have gotten a closed operating system for free to freely hack into being able to run a supercomputer?

Maybe they could have begged Microsoft to give them all the soucce code for the then newly-released NT 3.1 with freedom to modify it at will, with a promise from Microsoft to extend that free license to other institutions that wanted to do a Beowulf. Now that's getting pretty rediculous.

[Golden Eagle]

They should have went to IBM or Sun or Cray, who someone similar who had the most expertise in such things. Do you really think Cray wouldn't have jumped through their shorts to help NASA modify UNICOS? All it would have taken is contract dollars. Now instead, we have Cray on the verge of bankruptcy, and releasing products called "Red Storm" just to try to keep up with the free copies.

[Stellar Dendrite]

LOL

[AFreeBird]

Do you really think Cray wouldn't have jumped through their shorts to help NASA modify UNICOS?

LOL.. sure as hell they would... for a price. I kinda like the fact that NASA used good old American ingenuity to build their own, using parts and systems no longer in production and taking up space in a warehouse somewhere.

More bang for the buck. OUR tax bucks to be exact.

And hey what are you gonna do.... those whacky scientists and engineers like to improvise, invent, adapt and overcome.

I know, I know, were' not paying them for that kind of thinking by gum. They're supposed to be exploring space and throw our tax money around like its candy.

Oh, and only buy their software from Micro$oft.

Right.

I don't know there GE, in computer years, M$ is getting a little long in the tooth, and all they seem able to do is put a new dress on an aging dog and call it new again. If they're not careful, they may go the way of VMS.

Oh, wait, VMS was a rock solid OS (did clustering too). Bad example...

[Golden Eagle]

LOL.. sure as hell they would... for a price. I kinda like the fact that NASA used good old American ingenuity to build their own

So you like the government serving the role of industry, verses the free enterprise. Next thing you'll tell me is you really dig that groovy idea by Stallman to just go ahead and charge taxes on everyone to provide it.

That software should have been contracted for, then kept private for US government use only. At some point, they may release the contractor to resell to other US businesses. Eventually, they might in rare circumstances allow other governments or foreign companies access. But generally that technology is the property of the US taxpaeyers, and should be safeguarded for it's worth, which obviously should not include giving the Chinese and Russians free copies.

[AFreeBird]

So you like the government serving the role of industry, verses the free enterprise.

You have got to be kidding me! Just what the hell do you think Free Enterprise is? NASA uses its in-house brain power to help design a supercomputer with commodity hardware; OLD hardware they no longer use; paid for with YOUR tax dollars, and you're giving them $#!t about it!

Making use with what you got is VERY American, and I sure as hell applaud them for it.

BTW: Here's little on beowulf origins.

From the book:

Engineering a Beowulf-style Compute Cluster

Robert G. Brown
Duke University Physics Department
Durham, NC 27708-0305
rgb@phy.duke.edu

Historical Perspective and Religious Homage

The concept underlying the beowulf-style compute cluster is not new, and was not invented by any one group at any one time (including the NASA group headed by Sterling and Becker that coined the name ``beowulf''). Rather it was an idea that was developed over a long period and that grew along with a set of open source tools capable of supporting it (primarily PVM at first, and later MPI). Note that this is not an attempt to devalue the contributions of Sterling and Becker in any way, it is simply a fact.

However, Thomas Sterling and Don Becker at NASA-Goddard (CESDIS) were, as far as I know, the first group to conceive of making a dedicated function supercomputer out of commodity components running entirely open source software and Don Becker, especially, has devoted a huge fraction of his life to the development of the open source software drivers required to make such a vision reality. Don actually wrote most of the ethernet device drivers in use in Linux today, which are the sine qua non of any kind of networked parallel computing1.23. The NASA group also made specific modifications to the Linux kernel to support beowulf design (like channel bonding) that are worthy of mention. Most recently Don Becker and Erik Hendricks and others from the original NASA-Goddard beowulf team have formed Scyld.com1.24, which both maintains the beowulf list and beowulf website and has produced a ``true beowulf in a box'' - the Scyld Beowulf CD - that can be used to transform any pile of PC's into a beowulf in literally minutes.

By providing the sexy name, a useful website, and the related mailing list they formed a nucleation point for all the users of PVM and MPI who were tired of programming in parallel on networks of expensive hardware with proprietary and expensive operating systems (like those offered at the time by IBM, DEC, SGI, Sun Microsystems, and Hewlett-Packard) only to have to buy the whole thing over and over again at very high cost as the hardware evolved. Once Linux had a reasonably reliable network and Intel finally managed to produce a mass-market processor with decent and cost-beneficial numerical performance (the P6), those PVM/MPI users, including myself, rejected those expensive, proprietary systems like radioactive waste and joined with others of a like mind on the beowulf list. This began an open source development/user support cycle that persists and is amazingly effective today.

[AFreeBird]

Sorry, hit post instead of preview. I meant to include the source for the book quoted.

Engineering a Beowulf-style Compute Cluster

[Golden Eagle]

Thomas Sterling and Don Becker at NASA-Goddard (CESDIS) were, as far as I know, the first group to conceive of making a dedicated function supercomputer out of commodity components

Why was this technology not privately kept as property of the United States Government? Why was it given away, completely free of charge, to anyone in the world who wanted a copy? What exactly did we receive in return? Nothing much, that I know of.

[rzeznikj at stout]

And, IIRC, the NSA releases SELinux patches, software, and kernels for users of other distros.

Not to mention that various individuals have repackaged them for major distributions (e.g. Fedora/Red Hat, Debian, Slackware, Ubuntu, Gentoo, SuSE.)

Even Fermilab puts out its own distribution of Linux based on RHEL source RPMs.

[AFreeBird]

What exactly did we receive in return? Nothing much, that I know of.

Therein lies the problem.

First off as was written in the chapter quoted, the cluster (parrallel computing), is a very old idea, Think IBM and Amdahl back in the sixties.

Anyway, what did we get in return? Okay, we took obsolecent spare computers lying around and turned them into a supercomputer to serve the needs of a civilian scientific R&D agency of USGOV. Saved taxpayer money and contributed to scientific research. BTW: Aren't the taxpayers supposed to get some tangible bennefits from federal expenditures? Why can't a small start up business make use of such computing power?

Knowledge, learned either as a result of building and programming such a system, and/or the results it yielded, at a fraction of the cost of a shiney new IBM or Cray and the millions in costs that went along with them.

Shared knowledge, tends to create even more inovations, not to mention educating and expanding the knowledge of sudents AND teachers of the computing sciences. And basically contributing to the future of systems and artificial intelligence. Scary and exciting all the same.

I know you think that all ideas should be contained and controlled, espcially by the USGOV or MightyBig Corp, but I'd like to think that even if there's a seemingly level playing field, we'll make better use if it. In any event, people aren't stupid, some even have a tendency to come up with similar ideas independently - accross oceans even.

And no, we're not locking the internet. BTW: The DoD basically gave away all its research and infrastructure on their little invention. If you're really that concerned with USGOV giving away such technologies; you're too late.

[Golden Eagle]

Making something available to US Citizens obviously does not mean you have to make it available to every citizen of the world, unless you think it is the responsibility of our government to provide for the world, which it appears you do.

Make it no different than any other form of federal aid, they apply, they might receive, but not if they're an illegal alien, much less living and working in Iran or North Korea.

Why do you keep insisting on giving them equal access to our own citizens, they certainly didn't pay any of our taxes, and might use the technology against us or our allies militarily.

[antiRepublicrat]

They should have went to IBM or Sun or Cray, who someone similar who had the most expertise in such things. ... All it would have taken is contract dollars.

That statement clearly shows you know nothing about Beowulf and the concepts behind it, despite me having told you more than once. Shut up on the subject and live in your ignorance.

and releasing products called "Red Storm" just to try to keep up with the free copies.

That's called "competition" in case you've never heard of it. BTW, it is highly proprietary, and runs a combination of UNICOS and Linux. As usual, what really sets Red Storm apart is the proprietary scalable system software and extremely high-speed Cray interconnect and I/O technology.

[Golden Eagle]

Shut up on the subject and live in your ignorance.

LMAO is that the best you can do? There's no reason that software couldn't have been contracted from Cray for probably less than it cost Nasa to do it internally.

That's called "competition"

So Cray has to now compete with our government, who is releasing free products that destroy their business? Yes according to you, following Stallman's plan to a T, with software taxes and credits soon to follow.

[antiRepublicrat]

There's no reason that software couldn't have been contracted from Cray for probably less than it cost Nasa to do it internally.

You still don't get it. Read up on what a Beowulf is before you spout further nonsense.

So Cray has to now compete with our government, who is releasing free products that destroy their business?

Cray uses free products in what it sells.

Contracting for a full product when there is a free version available that will do the job is a subsidy. I don't like subsidies. If the government feels that $100,000 worth of contributions to Linux will do the same as a million dollar contract, that's $900,000 of my tax dollars that were just saved. Plus the government can redistribute that $100,000 worth of programming if it wants to, meaning I actually get a tangible return on my tax dollars.

Can't beat that situation. Only people who think the sole benificiaries of government IT spending should be the corporate beltway bandits have a problem with that.

[AFreeBird]

You're incredible, you know that? Can't see the forrest for the trees.

Listen up; the technology genie is long out of the bottle, and by sharing little ideas like Beowulf, they can grow bigger and better. I know you want to lock up all information here in the US, cause we all know that the poor dumb basterds in the rest of the world couldn't scratch their ass' unless we showed them how to do it.

I suppose NASA shouldn't work with and share ideas with Russia's space agency. I mean what could we possibly learn from the folks that first put a satellite, dog, and a man in orbit, and built and operated the first permanantly manned space station, and set endurance records for a man in space.

Nope you're right, we should be sharing anything with them. Nothing to see there, please move along.

And I guess any and all military, economic and political ties we have with friends and allies around the world should be out too. What do we have to learn or gain from such endevours. After all, we wouldn't want any of our ideas to find their way out of our country. No tangible bennefits from them at all.

Ummm.... in such a light; shouldn't your boy Bill; master of all that is holy, pull back from outsourcing code work to places like India? I mean, sure all the code work done there is covered by confidentiality agreements, and M$ patent lawyers are submitting applications for every line of code ever written (in the history of the world if they could) by their Indian development center, but the ideas and methodologies those coders learn from doing that work will be with them for the rest of their lives, and might - God forbid - spawn new ideas and methodologies that might find their way into (the horror) other products from other companies in other countries.

We're DOOMED!!

[AFreeBird]

Oh and since the subject of NASA and computing came up, well I hesitate to share this with you; I wouldn't want to to have an attack or anything.... This is from last year. It was an open conference. My God, do you know how many foreigners could have attended. I mean, who are these people to share our state secrets with others of like mind.

SEW-29

29^th Annual IEEE/NASA Software Engineering Workshop

Greenbelt Marriott Hotel, Greenbelt, MD, USA

6-7 April 2005

(Tutorials 3 & 8 April 2005)

Co-located with 12^th Annual IEEE International Conference on Engineering of Computer-Based Systems, http://sel.gsfc.nasa.gov/ECBS2005, as part of Systems and Software Week

Sponsored by NASA Goddard Space Flight Center, Software Engineering Laboratory

IEEE Computer Society, Technical Council on Software Engineering

Background

The 29^th Annual IEEE/NASA Software Engineering Workshop will be held at the Greenbelt Marriott Hotel, Greenbelt, Maryland, in Metropolitan Washington DC, 6-7 April 2005, as part of Systems and Software Week

Scope

[Emphasis added for GE] The workshop aims to bring together NASA technical staff, contractors, academics and industrial practitioners interested in the advancement of software engineering principles and techniques. The workshop provides a forum for reporting on past experiences for describing new and emerging results and techniques, and for exchanging ideas on best practice and future directions. Of particular importance is relevance to NASA’s mission and goals, and how techniques might be applied, or adapted for use, at NASA, or how NASA’s techniques might be used or adapted for more generic use. [ I'm sure you know that this is NOT the way scientists and engineers behave, it must be some secret conspiracy ]

Topics of interest include, but are not limited to:

Ø      Metrics and experience reports

Ø      Software quality assurance

Ø      Formal methods and formal approaches to software development

Ø      Software engineering processes and process improvement

Ø      CMM and CMMI

Ø      Requirements engineering

Ø      Software Architectures

Ø      Real-time Software Engineering

Ø      Software maintenance, reuse, and legacy systems

Ø      Agent-based software systems

Submissions

Both full papers (maximum 10 pages in IEEE format, or equivalent length in free format) and extended abstracts/ industrial experience reports (4 to 6 pages, free format) should be submitted electronically (plain-text, .doc, .pdf or .ps formats only, please) . A link to an electronic submission site will be available here shortly.

Deadlines

1 December, 2005	Full Papers and Extended Abstracts/Industrial Experience Reports Submissions
1 February, 2005	Notification to Authors
6-7 April, 2005	SEW-29 in Greenbelt
8 May, 2005	Camera Ready Copy due for Post Proceedings

Both full papers and abstracts will be reviewed by the Program Committee, and full versions of selected papers will be due one month after the workshop (8 May ,2005) to be included in the workshop proceedings that will be published by IEEE Computer Society Press. At least one author per paper will be required to register and present at the workshop.

Any inquiries should also be directed by email to Michael.G.Hinchey@nasa.gov; for faster responses, please include “SEW” in the subject line.

Tutorials

Tutorials will be held on 3^rd April and 8^th April; proposals for full-day and half-day tutorials should be sent electronically to Mike Stark, Michael.E.Stark@nasa.gov by 1 November, 2005.

Conference Organizers

Conference Chair: Mike Hinchey

Program Committee Chair: Margaret Caulfield

Finance Chair: Don Jamison

Registration Chair: John Cook

Tutorials Chair: Mike Stark

Publication Chair: Chris Rouff

[Golden Eagle]

You don't have to convince me how much you admire the Russians, it's been obvious from the beginning you want them and others to have our technology free of charge. Thankfully, most Americans are a lot smarter than you.

Russia gave Iraq U.S. war secrets

[Golden Eagle]

Contracting for a full product when there is a free version available that will do the job is a subsidy.

No it's not, it's buying a product produced by free enterprise, not replacing free enterprise with government produced products like socialists like to do.

Of course you're still parroting Stallman with every post of yours. You call him a kook, but then march along in perfect lockstep.

[Golden Eagle]

foreigners could have attended

Why should they? All they have to do is sit back and download the final product for free. They don't have to invest a single dime, but get the same rewards as those of us who actually paid for it. Which is of course an outrage to me, but perfectly fine with one worlders like yourself. Take from the rich, give to the poor, or those that want to bomb you, etc.