Slashdot Mirror
Feature:Linux and X-Ray Astronomy
by Kevin Remhof
On July 23, 1999, NASA deployed the Chandra X-Ray observatory from the Space Shuttle Columbia. This telescope is a major leap forward in X-Ray astronomy. It will enable scientists to study dark matter, black holes, and other objects with great detail. I was checking out the information on the Chandra website and started to wonder about the software behind this telescope. So, I did a little searching on NASA'as website and found some interesting information. What I found is that X-Ray Astronomy is depending more and more on Linux for analyzing the data.
Background
X-Ray astronomy has been around since the early days of the space program. X-Ray telescopes work much the same as a regular telescope, but rely on X-Rays to view objects. The earth?s atmosphere absorbs most radiation from space. So, in order to use these telescopes, they have to be lifted out of earth?s atmosphere. The first X-Ray telescope sent into space took pictures of the Sun. Since the Sun is so close to earth, it was not hard to focus this crude telescope. Modern X-Ray telescopes are much more refined. The Chandra telescope could read the letters on a stop sign from 12 miles away.
X-Ray telescope data needs to be decoded before it is useful. This is similar to the concepts behind the SETI@Home project. Astronomers are allocated time to use the telescope. Chandra has about a 55 minute viewing time during each or bit around the earth. The telescope first gathers data without interpreting it. That data is then transferred to earth. Scientists then pick up chunks of that data for analysis. It is then that the power of X-Ray telescopes can be seen. X-Ray pictures are a little strange. Although attractive pieces of art, they are not easily understood by the general public.
The Linux Connection
Astronomers have used UNIX as their standard platform for years. Sun, HP, DE C/OSF were all commonplace. That is until Linux came along. Linux is becoming more and more popular and is threatening to take over as the platform of choice. Linux is now the darling of NASA. It is easy to powerful, easy to use, and porting applications from UNIX is not too difficult. NASA even uses Beowulf clusters for various projects which need large amounts of computing power.
Linux is a great choice for X-Ray astronomy because of the wide available of the OS and the software for it. Two popular programs are XANADU and FTOOLS. Both of them began as UNIX programs and have now been ported to Linux.
One astronomer, Ken Ebisawa has even written a HOWTO on "Linux for X-Ray Astronomers ". He details how to get Linux up and running and some of the various tools useful in X-Ray astronomy. According to the Goddard Space Center?s website, & quot;Ebisawa hopes to resolve the mystery of the Milky Way's so-called diffuse X-ray emission, a ubiquitous cloud of X-rays that blankets our galactic plane and whose origin is unknown. To do so, Dr. Ebisawa will look at a region of space where this cloud varies in X-ray brightness but where there seems to be no distinct energy sources to produce it." Ebisawa will use his Dell Inspiron laptop to analyze his data.
Open Source Software
The latest development in X-Ray astronomy is Open Source. Another program for analyzing X-Ray data is called SAOimage. Currently, it is being redesigned and updated using Open Source practices. Open Source is a natural for this because users often want to extend the software by combining it with other systems to come up with a research environment. It also helps to reduce some of the costs associated with development of such complex systems. Using the bazaar model of software development allows multiple developers to maintain and upgrade this complex code. It also allows the shared development of key pieces of code and functionality while satisfying individual needs of users.
Conclusion
Linux is rapidly becoming the OS of choice for X-Ray astronomers. This is yet another credit to the flexibility and power of Linux. More information on the Chandra scientists can be found here.
It is just as intelligent as most of /.'s comments. I've used Linux. I didn't like it. I prefer a real Unix. Solars, BSD and DEC UNIX are far better.
Linux
sucks
ass.
You'd be surprised how many people are still clinging to the VAXes over here!
You may have caught an earlier slashdot post about blackholes. That analysis was done on Linux.
IRAF may be open source, but don't expect to just run 'make' and have it work. It doesn't work that way. To install it takes hours of patience and is nearly impossible to understand. I speak from experience. I hear AIPS is worse.
The packages were designed for scientists, not for casual users. FTOOLS builds like a regular package (written by my users, I might add) and may offer the first step to something like SETI@Home. Until then, you'll have to leave it up to the scientists.
Linux was not the first choice for the ground system that is being used for Chandra. Several years ago, NASA, actually the Marshall Space Flight Center (MSFC), that is the NASA center responsible for Hubble and Chandra, decided to use their in-house people to scale up a ground system that they had been using for the microgravity experiments that they flew on the Shuttle. Pretty much the ground system could send the command "Feed the rat" in so many words, and was not designed to support and control a complex science satellite.
Well, a couple of hundred million dollars later, the ground system was not scaling to meet the demands of Chandra: not just collecting science data, but running day to day operations. About 9-12 months ago, an engineer from the prime contractor decided to start working on a "backdoor" into the control system so that he could get his work done, even if the ground system was overloaded, working extremely slow, or just plain crashed. (This individual had done a similar thing while working on the Compton GRO program, so that he could get his work done using some scripting tools and some code on the UNIX boxes they were working on.)
Six months ago, this individual was slowly convincing others around him that a Linux dual boot on their PC's was a good thing, and could potentially be the only way that they would be able to check their work, review their telemetry, build their command loads, etc. Telecons start mentioning the installation of Linux "workstations" (i.e. PC's running Linux) at the prime contractor's software support facility, too.
That brings us to today, where information is becoming available showing that this individual's idea to run Linux, and the programs he created, was economically feasible (i.e. he had to do this for just about nothing, as NASA was not going to fund this effort overtly having blown millions of dollars on the Chandra ground system), and the best way to solve this mission critical issue.
I was ecstatic to see Chandra make to orbit safely without nary a problem (although it will always be AXAF to me).
It plots data well, and can manipulate data pretty well too, but it's not free in either sense (freedom or price). As far as I know the source code is unavailable and unmodifiable, and on top of that, extremely expensive.
In my office here at Columbia some people use IDL for its graphing capabilities. Others don't use it because of its awkward programming language, and some don't use it because it's not free software. We've ended up rewriting a number of its routines, which we imagine other people in the community have done as well.
Is anyone aware of an effort to replace IDL with a free software equivalent? For a community of scientists who program alot, programming scientific software seems like a natural thing to do.
Only a very small proportion of the software is _not_ open source (eg IDL). Most of the packages are reasonably straightforward to install, being a little more involved than the typical GNU package (ie configure, make, make install etc) but you'll get there in the end if you bother to RTFM. IRAF, though, certainly needs a deeper understanding and AIPS usually involves hitting your head against a wall for several days before it co-operates (it carrys a lot of historical baggage around with it).
My only problem with linux is that I can't keep up with the demand to install it (people want it on their desk, on their laptop and at home)!
JPL is using Linux to host the instrument sequencer for both the Keck Interferometer and Space Interferometer Mission. Although the sequencer will eventually move to an on-board system for flight, the Keck Interferometer is likely to keep it on the Linux box indefinitely (and the Keck Interferometer is funded through 2027 at least). http://huey.jpl.nasa.gov/
Although I wouldn't be willing to undertake the project by myself, I could get easily get obsessed with building my own radio telescope if I had several other like-minded people to help out... maybe its time to start a hacking group.
Linux being cheaper than a Sun is one plus. Its WAY more than just cheapness, though. With a CPU on your desktop running Linux, you suddenly OWN that magical ROOT password and don't have to depend on half-assed system administration policies/politics/personalities. Its a LIBERATING experience, I tell ya...
http://www.randomfactory.com/lfa.html The Linux for Astronomy cdrom is now in its third (2 cd set) release.
First :)
This really isn't anything new. Linux has been supplanting old unix workstations in astronomy for years. In my research group, we started making the switch about 3 years ago, and today we're almost completely a linux shop (we still have an old sparc that I'm keeping around just for grins).
In reponse to the question of what can the community do, I would give two suggestions:
Ciao
Shop Smart, Shop S-mart!
Aug 97. But don't they use a BSD-derivative on the primary systems?
...are nothing like one another, except at the most basic, functional level. (ie they have a hole in one end, some sort of focussing device and some way of recording what's going on). They both frequently contain big cylindrical bits. That's about it, though.
then use the term: "Roentgen-Rays"
;-)
named after the man who discoverd the X-Rays.
(he himself named the rays X-Rays because of
his understatement)
scnr.
Meme of the day: I browse "Disable Sigs: Checked". So should you.
Thank you for that wonderful piece of flamebait. This is a flame that I'm sure you won't read, given the level of maturity you've already expressed.
Rather than trying to express your pitiful self in such an annoying way, here are a couple other ideas for you.
1. Learn to use it. Linux is great, but not for the ignorant (which this sort of behavior proves you are).
2. How about surfing elsewhere? If you have used linux, and don't like it, use something else. Freedom of choice is a wonderful thing. And since so much stuff here is about linux, and you don't like linux, you have no reason to be here. Attracting flames in this way serves absolutely no purpose, other than possibly inflating your poor misguided ego. (Oh look! I can click on "Reply" and add to the stuff people read! I'm "3>Tr3/\/\3lY 3le3T" now for finding this "Reply" button! I think I'll be ANNOYING and send a message that does absolutely nothing but prove my own trollness and get sentient life-forms to yell at me! Yeah cool!)
--
Paranoid
Paranoid
Bwaahahahahaa.
For those of you that care, I just thought I'd let you know that all of the Chandra data analysis software is being developed on Sun/Solaris machines and ported to redhat and slackware. I've tested (and demo'd) the linux ports, and they're coming along nicely. It's definitely a step in the right direction to have this software available for linux boxes. All of this software is in the public domain, and covered with a license similar to the GPL, with the expected NASA spin on it.
/. Obviously y'all don't work in calibration or data systems, huh? :)
It's been amazing to watch the transition in the past few years at AAS (American Astronomical Society) and other professional meetings - now everyone has laptops running linux and loaded with their favorite analysis software.
I was shocked to see so many Chandra-related folks posting on
BTW, if any of you folks can show me how to get IRAF to build properly on LinuxPPC, reply to this!
-zeno
Hmm, this is an interesting article. Having given up on SETI after all the hoohaw about repeated packets and the overload of users, I thought this might provide some pointers for getting into some *real science*. However, it only really seems to be useful for those who already know what they are doing.
:-)
The HOWTO in particular caught my attention, but it is more of a "HOW TO set up a Linux system" than "HOW TO get involved in X-Ray astronomy".
I guess that if one doesn't have the appropriate background, this isn't a suitable subject for beginners to get involved in.
Maybe I'll stick to my home gene-splicing experiments
A little planning goes a long way...
It's cool that NASA is supporting Linux. But it is certainly true that the field of astronomy has been very Unix oriented for a long time (that is, for those who have given up their VAX systems already. :) )
Anotheuinr interesting thing about astronomy is the emphasis on open source software. Two of the most popular astronomical data analysis packages in the USA have been open source from the start.
The IRAF project (supported by the NSF) has been open source since it started in the early-mid 80's. It's primarily used for optical and infrared imaging and spectroscopy for ground-based and space-based observation, but it is also used for X-ray astronomy. Nearly all of the Hubble data is processed in IRAF.
Radio astronomers use the AIPS software, which is also open source and has been since its origin in 1978.
Both have supported Linux since about the time RedHat first appeared on the scene. Neither of these packages are recommended for non-specialists, but they are examples of free software which dominate a discipline.
First comment? ;)
LL
"If you are falling, dive." -Joseph Campbell
X-ray astronomy specifically and astronomy in general is a perfect place for LINUX (and UNIX previously) to grow. First, it's pretty much a skill training sourcing field. That is, the number of job opportunities is such that there are a lot of good people who just can't work in the field after they reach qualification and find work elsewhere, like in the computer industry. Second, astronomy is not a luxuriously supported discipline (double the funding and thats still true.) hence, astronomers learn to do things on the cheap (like not buying an MS setup, getting an open source one.) Third, it's a moderate sized field, small enough to have lots of personal contacts, large enough to make technology specialization at least plausible. Fourth, the amount of computing power needed in certain disciplines , particularly X-ray, is pretty much . . . well, astronomical. Hence, the people in the discipline become pretty handy with computers and analysis. All in all, astronomical research provides one of the best returns on the federal dollar invested in academic research. Of course, don't tell the government this.
So long and thanks for all the fish . . . !!!
As an X-ray astronomer myself I'd like to add that one of the main reasons for Linux being popular in our field is CHEAPNESS. I'm originally from the UK - where the computer systems are governed by a central organisation called Starlink (http://star-www.rl.ac.uk). They've been the subject of many budget cut backs that means that the previously supported platforms Solaris and Digital Unix are fast becoming too expensive to maintain.
So Linux came along - cheap hardware - free OS - easy portability and it's Unix enough that for users of DU or Solaris it can be picked up quickly.
Now most of the work I'm doing is on Linux - alot of the software developement for XMM (ESA's companion to Chandra) is being done on Linux (including work I am doing myself).
So no - as someone else pointed out - Linux in astronomy is nothing new - I'm betting on Linux being the number one operating system used in Astronomy - simply because that area is populated by hackers and the computer savvy, exactly the kind of people who made Linux what it is today.
> you suddenly OWN that magical ROOT password
Hmm... just because it's Linux doesn't stop it being administered. I admit that I have the root passwords in my current X-ray astronomy job (for Solaris as well as Linux) - but in my last job the root passwords to the Linux boxes were just as secret as those for the Solaris and Digital Unix boxes.
For those of you who do not know, Tucson AZ is perhaps the Astronomy Captital of the World, with Kitt Peak National Observatory in our back-yard and perhaps the highest per capita population of Astronomers (especially Ph.D.s) in the World.
I'm mostly Familiar with Steward Observatory and the University of Arizona Department of Astronomy. There, we primarily use Sun Workstation and the IRAF project, while most of the software applications revolve around FORTRAN Language, due to the industry standard in the Astronomy Community.
I know from personal experience, that Linux Platforms like Red Hat and Caldera are often "passed" around from office to office and home to home so that the PC fleet can network to the Unix severs and scientists can do a majority of their Data reduction at the comfort of their home or Office PC and not some slow terminal connection.
Also on the horizon, the Astronomy community is eager to network everything together (with security and what-knot, so that some hacker kid can't get access to the HST) so that the Astronomer doing research can sit in his lab/office and do his research over a computer connection, rather then spend his precious time in an observatory maneuvering the Telescope and doing menial tasks of adaptive optics and taking dark images to adapt for errors in the telescope, that could easily be done remotely. All the while, while porting Unix software like qphot, emacs, super mongo, and IRAF to their Linux PC and working with their data instantly.
That ends this report from the field, this is BaronCarlos.
*Carlos: Exit Stage Right*
"Geeks, Where would you be without them?"
*Carlos: Exit Stage Right*
"Geeks, Where would you be without them?"
"Got Linux?"
Interesting, I must say... Hence, the people in the discipline become pretty handy with computers and analysis. All in all, astronomical research provides one of the best returns on the federal dollar invested in academic research. Of course, don't tell the government this. Of course, I'll add my $0.01 worth: I think it's great that such innovative, ingeniutive, inventive, etc. people exist. Now, if only the government'd start cutting budgets in other areas, maybe some of those slackers would be forced to actually start thinking a bit.
Then again, I realize that many of you may find that extremist, in which case I'd advise you, and the rest of the "population at large" to ignore me.
I'm finished now. Thank you for your time.
Insert mind here.
This is slightly off topic, but...
Not too long ago, I found a book in a used bookstore called "How to Build a Radio Telescope" - it was published in the late 50's-early 60's (IIRC), and a the schematics seem to be based on tube technology. However, it seems pretty in depth on everything else.
Are there any geeks out there that would be interested in such a book? I picked it up because of the title - I mean, how often do you find such a book - anywhere?
I am not saying I am going to sell it - but I might give thought to converting it to HTML or something if enough people are interested...
Reason is the Path to God - Anon
Linux is currenty being used for some operations at the Operational Control Center for Chandra. It is being used to process some data. It was an obvious choice, from what I'm told. The majority of the system runs on SGI's, though. Throw in a couple Suns for good measure and a couple NT's to screw things up (trust me they do!). Personnally, I'd like to switch to Linux exclusively, but these are decision that take YEARS to be made. Chandra has only a 5 years expectancy.
Sig
Appended to the end of comments you post. 120 chars
Stupid Dweeb.... use your time for something more productive than wastying a few mor bytes of space on /.'s Hard drives.
:-0
You probably just read here cause you still cant figure out what this whole 'Linux" thing is.
www.mp3.com/Undocumented
So how long until Linux gets used on a shuttle mission? It'll be the first Open Source operating system in orbit.
(I wonder how a penguin would react to zero-g)
Any sufficiently advanced civilization is indistinguishable from Gods.