Slashdot Mirror
Space Station's LAN
Muad was the first, but this
story was submitted many times: CNN is reporting on the
space station's LAN .
Check out those specs- 10base2? WinNT? Win95? Outlook?
At least the stations vitals are controlled by
Thinkpads running Solaris.
What the hell. 10 base 2(our favorite coax standard) is a single point of failure medium.
What bozo thought of this?
I'm a *NIX bigot, but if all they're going to
do up there is edit spread sheets, send email, and
play solitaire then why not use Windows?
That's all it's really good for anyway.
I can't wait to hear the news.. jan 1 2000 comes.. international space station LAN shuts down, all e-mail is lost.
heh, is this first post? nifty.
... Don't forget your install disks!
--
Jason Eric Pierce
It's pretty clear that for what the station system needs to do, Windows based OSes are the way to go. Instead of having to staff a unix sysadmin, we can send up another scientist to do valuable research. One of the great benefits of Windows is the UI, which is intuitive enough so any competant person can administer the network in their spare time. Maybe when unix gets a little easier to use, it may pose a threat for mission critical applications, but until then I'll take ease of use combined with equal performance and stability any day.
Hmmmm .... Well it might be arrogance but a LAN in space doesn't seem like rocket science to me. ;-)
Can't wait to see how well the closedness of winshit does them when their NT machine is bluescreening every few days and their 95 computers can't go for longer than a couple hours without being rebooted thanks to memory leaks in outlook and general 95 networking stupidity.
Good point. If there were going to be one or two fashionably-anti-MS *nix propellerheads on board, then it might make sense to be using non-MS software. But scientists don't have an infinite amount of time to spend obsessively dicking around with the internals of their OS. They actually have real work to do.
They should have used laser transceivers at the bulkheads/hatches.
NT can easily scale to 5 users.
As long as all they're doing is email,
they shouldn't have too many problems.
why not linux?
make a working image.. then dont update.
I'll tell you why they are using microsloth for the "entertainment" lan system.
A - Microsloth probably gave them all the software for free
B - the laptops were also probably given to them for free
C - if you have a pile of free stuff, there is no challenge to building the lan. the article makes it sound like he's doing a massive job. Hell a 15 year old geek can do what he is doing in 10 minutes.
Actually most of the scientists will be using Dauphin Orasis pen computers for their data collection and then wireless lan it to the central data collection server. It's already happening that way on the shuttle (and Dauphin DTR-1's were used on the first shuttle missions) I would not pay any heed to the CNN drivel.. they found some guy with a lowly and super simple job that would actually talk about the project.
Thinkpads... good god.. why not something that is rugged and designed for scientific use?
as for the linux troll, yes linux could be used successfully and easily... no worries about gnome crap if you run a simple x windows or even use Metro-X. and there are several thousand linux systems still in use that are happily running 1.xx kernels and haven been updated for years.
NOTE: you dont have to upgrade if you dont want to.
NASA keeps asking for more money for the space station, so I hope Washington knows that they are investing billions of dollars in a space station that will be running certain crappy OS's that are prone to failure. The could have been running Linux with WordPerfect8.
Forget solitare, Freecell is the way to go. Try and do that on your crappy linux.
I am not sure when the last time you used windows was, but i have nt and win98 and win98 running at home, and my nt box has not crashed in 6 months (would be more but i had to power it off to MOVE) and the 98 and 95 boxes have crashed a total of 3 times in the 6 mnths or so they have been up. As much as it pains me to admit it, m$ is getting stabler. now before you flame me, i am not a big fan of m$, I also run BeOS on the 98 box, and redhat linux 5.2 on the 95 box... and FyI Linux and Be are not exempt from crashing either ;) (even if noone will admit it damnit)
Jeff Cameron
jcameron@geocities.com
(anonymous because i am not creating an acct here hehe)
Outlook + Exchange server is the ONLY system that has ever caused me to lose email. It seems to WANT you to lose your email by putting it into proprietary BLOB databases and being just buggy enough to not be able to get them back. In addition if you don't have special backup/restore software you won't have the granularity to fix individual Exchange mailboxes without restoring the entire mailbox database on another NT server!
One thing was surprising though, and that was how quickly an Outlook install can deteriorate. When it works, it is fine.. but I saw 3 machines turn into GPF boxes within one week because of some corrupt calendar event that got circulated. I think they had to recreate those mailboxes just to recover.
I would rather; OPEN, STAT, LIST, RETR, QUIT than use another Outlook client. At least I'd know that *my* mail would be in an MBOX or mhdir that I could fix myself.
Daaaisy...
Daaaisy...
As soon as he said synchronizing email is a problem, I find him totally incompetent for his task and position. Did it occur to him that he could set up sendmail to compensated the up-and-down long delay connection? Talking about using off the shelf software ... Yeah, adjusting one self to adopt to the software is a easy solution, at least the astornaut are easier to train to adapt to their environment than the software.
Dave... I have upmost confidence in our mission.
Dave... this application has performed an illegal
operation and will be shut down. Please contact
the software vendor for further assistance.
At least most of the ones who seem to
do the real kinds of work. It's not
a surprise that most of the terminal
master's students from my graduate
school ended up working as systems
people.
Someone help me out here. Most hard drives use
aerodynamic foils to float over the platters.
How do they work in reduced gravity and low
air pressure? I always thought they used flash
disk.
I didn't think of the reinstalls... This means that they're going to have to send Win 98 and NT CDs into space for doing the, er, "regular maintenance".
Hmm.. Maybe we can arrange to send *all* the Windows 98 and NT CDs into space, like people periodically propose for radioactive waste.
If their software turns out to have a bug, I'll bet they start "obsessing" over whether it is open.
Good Windows is not used..else aliens may upload viruses just like we did to them in "Independence day".
I'm sorry Dave, but I'm afraid I can't do that...
Sorry about the anon. post.
The environment is not pure O2. It is probably the 79/20/1 mixture (some call it air) that we are breathing now. Pure oxygen has many drawbacks, including flamability and corrosion.
Ben
They don't use a pure oxygen atmosphere, but basically the same combination of gases we breathe here on earth. Anyway, even in pure oxygen, you need some fuel to do "kaboom"!
I agree, 10 base 2 works fine should it not be
moved or vibrated, or breathed upon, but once
it is its flaky from that point on. They could have picked a better medium...then again they could have picked a better os too...solaris is slow and msw well is crap.
What makes you think that they are breathing pure O2 up there?
Did it occur to you that Unix admin can do remote admin? I mean real remote admin, not things like running PCAnywhere ... or some tools that can crash the OS (or the OS crash the app.)
... Oh, besides, most scientist is also a competent Unix user (not necessay admin), and that make them a competent Windows user too (not the other way around)
What a moron
Again, you are moron, and I am really angry!@!@!#!
if had been used in place of solaris thinkpads :()
Inside the harddrive is no air - it's some inert gas and therefore the drive platters already are isolated from the outside. Zero-G should be no problem. The start with its huge acceleration will be much more challenging to the drive mechanics.
Obviously he is an idiot designing this thing, with all the linux systems and users at nasa, you know that only this asshole had any say so...he will show up to be a moron eventually, if they let the little problems show in the media that is.
The air pressure should be fine otherwise the astronauts would go poof. The disk heads should float nicely above the spinning platters. As for gravity, well think of a hard drive that is vertically mounted. Those work just fine, hence gravity is not really a factor.
Slashdot should sponsor the following event ... ... and there favour email program ...
1) Send each astronaut a Linux distro. (make sure at least a copy can get throught any inspection), so that they could re-install their system with Linux
2) Send each astronaut with a different Linux distro., because I am sure all scientist got style and has their own opinion which distro is better, but we will not worry, because all distro. will work great and co-exist (they even play nice to the evil OS, but not sure if the other way around is true though, poor Linux)
Long exposures to O2 above partial pressures of 0.6 bar causes chronic Oxygen toxicity. Pressures
above 1.2 bar can cause seizures and fits . Hence divers use mixtures such as heliox or trimix (1-4% O2, 0-5% N2 and the rest He)
O2 isn't at all good for you.
What about fiber? Its lower in power consumption, even if you are INCLUDING a hub. And what exactly is wrong with a hub topology?
Didn't the Navy have one of their new, improved and nearly bugfree cruisers play "sitting duck, dead in the water" with a blue screen of death recently? So much to viability for government work...
I'm not sure about the language support for all the various UNIX variants, but perhaps the ease of adding and using Russian, Japanese and Chinese character sets on top of an English Win9x environment was an issue for end-user interfaces in an international space station?
What kind of scientists are these, chemists? All of the non-chemist scientists I know are more familiar with UNIX. That includes physicists and biologists, btw.
Last time I used windows (at work) W95 got so screwed up that even pressing the f**king start (what a girlie name for
;)
a button) crashed the machine.
Our (at work again) NT servers have to be rebooted every 3-4 days (especially the mail server).
Well, I hate to point fingers, but it sounds to me like your NT admin dosn't know his ass from a hole in the floor. Either your running your servers on flakey 'off brand' hardware or he installed it wrong. We run Sompaq servers here and our PDC has been up over 23 MONTHS. get a grip your experiances with winblowz are not indicitave of everyone elses
Why have BeOS and 98 on the same box, and linux and 95 on another, you can only run one OS (term used loosely to
allow for M$ Windoze) at a time.
Well, the 95/linux box is my roomates, 95 for games, linux for about everything else (even uses my WinNT proxy for internet access NICE!)
the 98/BeOS box is mine, 98 again for games and for ICQ, BeOS for everything else, including programming (it is one of the best c++ environments i've used)
Yes, Linux has been known to crash, twice for me in 2 years, and one of those was an X problem with the StarOffice
5.0 installation.
Like i said right now NO os is crash proof, and i'll agree that M$'s os' (win9x is a shell over dos really) crash far more than they should, but they don't crash every hour. If you have that problem, its either a. hardware b.corruption in a dll c. some software on it causing it to bsOd.
Jeff Cameron
jcameron@geocities.com
Windows 98 = Windows 95 + Ie 4
Windows95 = Dos + GUI Dosshell
DOS = Bad idea to begin with.
muhahhahahaha
They ran 2 of the 760 on Unix, it would be a drop in the bucket to run sendmail in terms of resource and stability.
First of all, to clear up all the talk about Skylab and such: the article did NOT say the LAN will be in ANY part used for station maintainence. Even that trajectory plotting program they mentioned most likely would not be actually connected to station controls in any way. It's probably just a time keeper based on a certain orbit.
I am not saying that using an MS based LAN is a good idea, but hey, they did say there will be spare parts lying around. I assume that means install CDs also. I do recommend they at least use Outlook Express, but in the big picture it probably doesn't matter that much. The LAN will be mostly used for personal stuff and data keeping. As long as they backup regularly (as anyone should anyways) they'll be okay.
Remember: the aerospace field is still FORTRAN dominant. Therefore the station itself is most likely running on fortran based programs for orbit management, life support, etc. This unlike the report I heard recently about the Navy actually installing NT as their primary OS, including on the ships and subs and stuff. Not just a LAN, but actually RUNNING the damn ships with it. They claim it will help out with real time battle management. As soon as I find the link again I'll submit it. It was like in the last month or so.
This is my personal observation. Most scientists have a working ability to use some operating system; unix, windows, macos, vms, etc. However, the vast majority aren't proficient at all wrt to sysadmin (okay, maybe on a Mac :-)). Their ability to write programs widely varies. Furthermore, their programming skills tend to degenerate from the time they were a graduate student or post-doc.
What you see more of is the use of scientific analysis software such a matlab. Even in this case, the programs are developed by someone else (grad student or post-doc), and then used by others. Canned software packages are heavily used especially for statistical analysis and signal processing.
When I was in grad school, the students in the group that I was in were required to write their own programs. Essentially, write or die. AFAIK, I'm the only one who is still hacking away.
"The next flight after crew arrival will bring the U.S. Laboratory Module that will house the file and print server unit and a hard-line printer."
Can you say ribon and 9-pin?
I thought you could...
Again, once you set things up, it take zero afford to maintain a Linux box. Besides, the time to wait for WinNT to reboot could add up too ...
Honestly, the way you guys are responding to NASA using 95/NT on a few laptops for *email*, you'd think it was a mission critical app. It's not. And it makes a lot of sense to use common, inexpensive, easy-to-use software that the people on board are familiar with, for such things as email/word processing. That way you have no training costs, and no custom software to build. If Outlook crashes, whoopee. So you reboot. A hassle but not a terribly big deal. Nothing critical is being run on these machines! The ones that will be doing anything important will be running Solaris. And the embedded stuff is probably embedded Unix anyway. So what's the fuss? It's not like the station is gonna crash to the earth because they're using Outlook for email to family members back home! Chill...
-- someone who actually *read* the article (gasp!)
You know, this person's first language may not be English. Not everyone can be an American.
Guess it's time to conjure up a new TLD like .spc, It would be fun the first time you analyzed your httpd logs and found win95-laptop-1.station.spc...
Hell, if all they're going to do is that stuff, why be on the Space Station at all?
How about getting some real work done, which means using real software.
Were talking laptops, right? Which means PCMCIA Ethernet cards. I've never seen a coax connector for such a card, just 10/100 base T connectors.
Do they really exist?
Or are the astronauts gonna get up there with their Thinkpads, walk up to the cable and go "oops!".
Who gives a rat's ass about the LAN technology...
/. it. Anyone figured out how to do a remote bootstrap install of linux over an active NT server? That would be the killer app... something that goes though NT servers, duplicating all functionality, but when the admin goes to log in, finds that it's really linux.
...what WOULD be interesting is a detailed article on the WAN technology. Hacking something in orbit... that's an interesting prospect. I can see it now :
Uhhh.... Huston, we have a problem... the CD-ROMs keep opening and closing.
Last one to hack the NT box's web page is a rotten egg. Hack it, then
Linux, the worm... coming to an NT partiton near you.
The bad thing about coax is that a failure anywhere along the line brings down the whole LAN, even if it's just somebody disconnecting their laptop and not putting a terminator on the T. If it's a cable defect its worse.
With 10baseT, if one cable goes out you know which one. If the hub goes out you know that too. Failure point determination (and fix) is much quicker.
> "E-mail is a bit of a challenge since we do not have a constant connection with the ground."
I about laughed myself silly at that. Another obvious failing of MS software. Geez, I remember when email was only transferred every hour or so (depending on configuration) via dial-up UUCP.
Bozos. I think all the good engineers left NASA/contractors a long time ago.
Geez, a bit harsh on the "windows programmers", eh? Guess what I program for a living? - Windows. Doesn't mean I'm stupid or that I like it, that's just what I can make money at right now.
I especially like the part where you say I have no ingenuity...
--
Jason Eric Pierce
Know anything other than plain old English? Know what is ohayo means? Know Spanish? Know Cantonese? Know ... What is your SAT score on English? How about TOFEL score?
How many 100BASE-FX/10BASE-FL PCMCIA cards have you seen? None.
If the LAN was all fibre, they'd need additional eqipment per station to convert the fibre into something the laptops could plug in to. This means more weight (remember, it costs $10k to get each pount into space), more points of failure, and more power being drawn by the additional equpment.
Try thinking this out before firing your mouth off like that. Moron.
"Aside from the trip up there, this hardware is going to be in THE ULTIMATE of places with regards
to G forces and shock"
Hmmm, I think the trip up there is going to be most of it for g-forces and shock. Most components for desktp computers can probably take hundreds of G's as long as they're not running at the time. So, if the G-forces on the trip up were too great for the components, they'd have to _drain_ the human beings out of the spacecraft.
As for actually being in freefall... I thought the whole point of microgravity was that the G-forces were, well, micro. So, except for components being thrown around, and I'm sure they have all sorts of clamps, clasps and nets to prevent that, the finished network shouldn't have to deal with too much physical stress. G-forces would get too extreme for the stations structure well before they got too extreme for the computer equipment.
The real thing to worry about is the radiation. The fact of the matter is that most computers are probably going to have an operational life of a few months up there. Very modern computers aren't really the best choice anyway. The technology of the Mir computer could be considered antiquated, for example. This has nothing to do with the budget, it has to do with the fact that, in a high radiation environment, tinier electronics are not better. The computers up there are going to be constantly malfunctioning due to the radiation.
I don't know why NASA really wants to do such a thing, if there are other, more viable alternatives around. E-mail? Pine is already very easy to use, and training an astronaut (which is by default a very intelligent person, or so they say) would cost nearly nothing. Or they could use Netscape on X from an extra Thinkpad laptop running Solaris, if they want a graphical interface. Wordprocessing? WordPerfect fits the bill. User friendly GUI? What the heck, the UNIX UI can be MADE user-friendly, so why use Windows? To risk having viruses in the ISS? To later have to bring it down because Windows didn't pass some FIPS test and is supposed to be banned from all government computers? Whoever wrote those specs needs a slap with the clue stick.
The problem, of course, is that no matter what is done to Windoze (NT/95/98), it will still cause problems for the whole station LAN. Microsoft Networking is usually a bit hard on the bandwidth, not to mention NetBIOS and so on. And Exchange is even harder on bandwidth than Windoze alone.
With that in mind, plus concerns about 10b-2's SPOF, here is a new plan that will work even better, and be cheaper in the long run (while admittedly more expensive at the get-go):
Start with one of Apple's new 400MHz PowerMac units, with as much RAM as possible (they'll take up to 1GB), and as much hard drive space as possible (up to 100GB inside, with an internal RAID). Then, install LinuxPPC (is DR4 the latest?) on it, completely removing the MacOS. Also install 100b-T with shielded TP cable to a wireless LAN transciever, also with 100b-T capability (if wireless 100Mbps radio networking exists; if not, use 10Mbps for all systems).
Now for the client units. Pick up a few PowerBook G3 units (292MHz; they run a bit faster than 300MHz units do, because of different bus speeds). Put in as much RAM and hard drive space as needed, then install wireless LAN transcievers (probably in PC Card form) on each system. On each of these, install LinuxPPC as on the server Mac.
Advantages: Linux-need I say more? I do? OK! Server is one computer, handling email and other communications, file services, and print services. Linux is essentially UNIX, with the reliability inherent in UNIX. NASA already knows UNIX, so the learning curve is extremely shallow. Configure each system once, and you won't have to do anything else until something goes seriously wrong, which is not nearly as often as Windoze. Essential applications exist on Linux, and shareware developers are coming out with new ones all the time. WordPerfect is available, so complex documents can be created.
Disadvantages: Linux is essentially UNIX, so it is not as easy to use as the MacOS or Windoze for the average user. The learning curve for the astronauts will be a bit steeper, but with a good X manager, even that will be pretty shallow. There are fewer applications available for office-type functions, like spreadsheets and the like, but even these differences will disappear as Linux becomes even more popular than it already is.
If system management is a concern, then use MacOS X. It is UNIX, based on the Mach kernel (which is about the best microkernel in existence) and NeXTStep, which is pretty reliable in its own right. Administration is easier than on Windoze, and you often don't have to reboot to activate a change in operating parameters, as you do with Windoze.
I chose Macintosh hardware because it is truly easier to use than PC hardware. Plug and play has always worked the right way, ever since 1984, and very little goes wrong with Macintosh hardware. Even so, it will probably be smart to have two similarly-equipped servers installed, to have a hot standby ready, in the rare event that the server would go down.
SendMail is a proven technology that works better on a bad day than Exchange does on a good day; I have used both, and SendMail is much easier to set up than Exchange, not to mention easier to keep functioning.
Anyway, that's my idea for the ISS LAN, not that NASA would do any of it.
The problem, of course, is that no matter what is done to Windoze (NT/95/98), it will still cause problems for the whole station LAN. Microsoft Networking is usually a bit hard on the bandwidth, not to mention NetBIOS and so on. And Exchange is even harder on bandwidth than Windoze alone.
With that in mind, plus concerns about 10b-2's SPOF, here is a new plan that will work even better, and be cheaper in the long run (while admittedly more expensive at the get-go):
Start with one of Apple's new 400MHz PowerMac units, with as much RAM as possible (they'll take up to 1GB), and as much hard drive space as possible (up to 100GB inside, with an internal RAID). Then, install LinuxPPC (is DR4 the latest?) on it, completely removing the MacOS. Also install 100b-T with shielded TP cable to a wireless LAN transciever, also with 100b-T capability (if wireless 100Mbps radio networking exists; if not, use 10Mbps for all systems).
Now for the client units. Pick up a few PowerBook G3 units (292MHz; they run a bit faster than 300MHz units do, because of different bus speeds). Put in as much RAM and hard drive space as needed, then install wireless LAN transcievers (probably in PC Card form) on each system. On each of these, install LinuxPPC as on the server Mac.
Advantages: Linux-need I say more? I do? OK! Server is one computer, handling email and other communications, file services, and print services. Linux is essentially UNIX, with the reliability inherent in UNIX. NASA already knows UNIX, so the learning curve is extremely shallow. Configure each system once, and you won't have to do anything else until something goes seriously wrong, which is not nearly as often as Windoze. Essential applications exist on Linux, and shareware developers are coming out with new ones all the time. WordPerfect is available, so complex documents can be created.
Disadvantages: Linux is essentially UNIX, so it is not as easy to use as the MacOS or Windoze for the average user. The learning curve for the astronauts will be a bit steeper, but with a good X manager, even that will be pretty shallow. There are fewer applications available for office-type functions, like spreadsheets and the like, but even these differences will disappear as Linux becomes even more popular than it already is.
If system management is a concern, then use MacOS X. It is UNIX, based on the Mach kernel (which is about the best microkernel in existence) and NeXTStep, which is pretty reliable in its own right. Administration is easier than on Windoze, and you often don't have to reboot to activate a change in operating parameters, as you do with Windoze.
I chose Macintosh hardware because it is truly easier to use than PC hardware. Plug and play has always worked the right way, ever since 1984, and very little goes wrong with Macintosh hardware. Even so, it will probably be smart to have two similarly-equipped servers installed, to have a hot standby ready, in the rare event that the server would go down.
SendMail is a proven technology that works better on a bad day than Exchange does on a good day; I have used both, and SendMail is much easier to set up than Exchange, not to mention easier to keep functioning.
Anyway, that's my idea for the ISS LAN, not that NASA would do any of it.
BTW: there are games on Linux.
1. They're letting M$ off the planet?
2. This could make for an interesting commercial... Show the occupants of the station playing solitare (obviously Win-whatever), then pull back over the shoulder (while the Win machine gets a BSOD?) to the machine running visibly something with the Solaris (whatever) logo and LIFE SUPPORT flashing or some other gimmicky graphic ad device (it has to get the point across, not be overly realistic). Over-caption: "Which would you trust to run YOUR business?"
--
AC. Not a puppet of either company, but it would be a neat ad to see.
"That's one small crash for Windows, one giant blue screen for mankind."
-D
dcross@cryogen.com
In my experience, ThinkPads are some of the best laptops. Of course there's probably something better - but I'm sure weight is critical
vunder bar! all wireless no more network cable..but wait...the spec called for cable didn't it?
As to these guys being well paid, I think that has more to do with seniority in a government position than actual knowledge/ability. This is the government we're talking about after all.
If you must know, I do keep in touch with these guys from time to time. Even they will tell you that most of the design decisions that are made have more to do with politics than actual engineering.
I agree with you that fibre is not the way to go. But I will disagree on the 10Base2. Having had the "pleasure" of having to track down a flaky connection once was enough to make me swear off them forever. They're just like the old christmas tree lights. One goes, they all go. You're pretty much stuck going from node to node to trace the fault.
As the station gets used more and more, the load on this network is going to grow. It will be used for data download from the various experiments. This network may need to act as a backup network for mission critical stuff. You want a network that degrades gracefully, where faults are easy to trace, and where repairs can be made with a minimum of fuss or loss of bandwidth.
[In space, no one can hear you scream... when your network crashes!]
The point is that when it fails, it fails hard.
One loose connector, and the whole network on the whole station goes out. Not a good plan.
The only place terminators will fit is on the tee. They have the same "gender" as the end of the cable.
Not necessarily. Keep in mind we are talking about a very small network covering a very small area. Manually diagnosing a faulty cable second won't take very long. But, if your hub goes out and you have to order a new one from Planet Earth you'll be waiting a while for a replacement. In a situation like this I can see a couple benefits to 10base2:
1) It has a bus topology (runs directly from one computer to the next) as opposed to a star bus topology (a seperate cable needs to be run to each computer.)
2) Less points of failure. 7 computers on 10base2 require 6 cable segments. On 10baseT or any other star topology you're looking at 7 cables plus a hub.
3) It doesn't require a whole lot of skill or equipment to repair a faulty 10base2 network. If I was stuck in outer space and needed to improvise something quickly I'd much rather have 10base2.
BTW, only the two end stations will require terminators. An additional terminator is not required when a machine is unplugged.
At first I thought it was a wierd choice, but the more I think about it, the more obvious it becomes why they chose 10base2.
KN
Got a URL for me. I miss my freecell.
If I were to claim to be a genius, I'd say. You really smart person! Your intellect is nothing compared to mine! Not "you're a moron.".
Nor does everyone want to be one.
"poorly written software" is the stuff those poor souls in orbit are going to use - Office, remember ? I am definitely not a Linux-for-everything person, but you have to acknowledge that Office is the embodyment of memory leaks. Get a grip - we don't need a flame war, and everyone can understand shielding problems, but you should also consider that the interior of the station is VERY well shielded, people need to live in there ! this thread started as a "wow - that is weird" thing and people like you turned it into a flame-war. Can't you guys realize that windows has very specific flaws, even if microsoft keeps stuffing your brain with the opposite? Linux people are not the narrow minded ones- in case you did not notice, Linux is learning the only good thing windows has, ease of use. Now, when NT will learn to be stable without me pulling books from O'reilly to fix it every minute, I will think that you MS-guys are learning too.
BTW - I use windows all the time, and for what it is worth, ease of use. I am not a mad-kernel-hacker, in fact, the only OS I seriously mess around with its internals is Windows, to fix it every time YOUR safe, strong , reliable piece of #@$^% blows on me. There are "badly written" programs for every OS, the only difference is that _some_ os is good enough not to blow on your face if someone can't write good code.
time in space is more expensive than per-pound costs. An hour of an astronaut's work is way more expensive than a couple of pounds. Moron.
The guys at NASA are smart, but I'd bet against a nobel prize that there are better solutions than this one.
I totally agree with the power argument of 10b2 drawing 0 power. THAT is a really good reason.But then... I am not sure it is enough of an advantage. Windows has crashed in space ALREADY, and kept astronauts fumbling for hours on "how do i do this" "it won't work" "reboot" etc etc. I am sure you can find the transcript of the mission or even the footage from NASAtv if you look hard enough.
I don't think any other OS has crashed in space, other than Windows, mind you. True, Ariane blew up last year for a software problem..... but, still, is anyone aware of a "software problems in space" site ?
Having done a lot of network installation starting way back in the bad old days of 10base5 thicknet, I'd say this is as good as NASA can get.
10base2 is easy to repair. Its big enough to be repaired with a pocket knife and fingers.
Most people with 10b2 failures went out and bought the cheapest cables they could find. NASA is probably assembling extremely hi-reliability cables with lots of strain relief.
Fiber itself is pretty noise resistant, but the electronics/laser bits are not. And an alpha particle strike on a fiber can cause a flash so bright it might burn out the IR receiver diode. It HAS happened on some satellites.
10b2 is shielded and pretty noise resistant. Important with no atmosphere between you and the sun. I bet they are using double shielding as well.
This entire story comes almost verbatim from a microsoft press release, a copy is sitting here on my desk. The PR people are just beating the drums because this is the first time NASA has allowed win/NT onto the ISS. It is still an amalgam of many different systems, and yes!, much of it is written in FORTRAN, or assembler. These win machines are just going to be there for the GUI and email, not for running any part of the station. So lighten up.
And the OCA is an amazing piece of work, using a huge amount of bandwidth up in the 24 GHz range, with a bunch of supergroups (old telco microwave term) each containing various data, video, and audio streams. It does NOT run any M$ code, NASA has the entire source and it is hand checked by QA teams after every update. And I do mean checked.
FACT: Windows95/NT will be responsible for a much greater power drain than a fiber converter. You're talking about a couple of volts DC here, per unit. And only when data is being transmitted or recieved. Open up a Centronics 10BaseT/F media converter sometime. It's practically nothing. Win95/NT, however, use up a HUGE amount of CPU idle time and RAM/Swap space (which it often flushes) doing.. umm.. well, we don't really know what it is doing, but I'm sure it's important.
FACT: 10Base2 is so incredibly inferior to 10BaseT in terms of data throughput, on a LAN you will suffer geometrically more from retransmits on 10Base2 than 10BaseT, and thus, the power differences tend to go away in terms of total power used per packet.
FACT: You can shield CAT5 if you want to. It's just a piece of tinfoil wrapped around the wiring and grounded on one side. So even if you -were- concerned about interference, you could still shield against it.
FACT: If they REALLY cared about power consumption, and weight they would have used an Alpha StrongARM chip and ANYthing but Windows.
I'm sorry folks. I have no religious attachment to Linux, or Unix, or any other OS. Whatever is right for the job at hand is what I live by. However, this is so incredibly ridiculous, I had to speak up. NASA is so blatantly NOT following technical responsibility here I cannot believe that any of you would try to defend them.
This decision is -completely- politically fueled. End of statement. Thank you for playing. There is no chance in hell that a self-respecting engineer would recommend this sort of setup and say it was the 'best in terms of performance and power' etc unless he had gotten the 16 ton weight From Above (tm) and management had set him up for it.
My god.. there must be dozens of ways to solve this problem, all of which would use less power than the proposed solution, and take up the same amount of space.
I'm not going to tell you what my credentials are to be saying this stuff.. it seems that people whom do that are crucified for being elitist.. so those of you whom research the problem will understand that what I am saying has basis in truth. Approach the problem yourself, in your own mind. You'll see it.
And cut out the 'they're smarter than all of you' bit folks. I've corrected NASA level people on technical issues before. They're human.. just like you and me. And they have managers.
Ok fine. But consider that today /. also linked to an article telling of a Navy warship that was dead in the water due to a stupid NT crash (one program attempts divide by zero and crashed the network...typical Mickeysoft OS)
I would think that if NT let us down with the naval firepower then maybe we shouldn't be using it anywhere near our space station.
There is however one group of employees at NASA who currently have no chance at being astronauts...the computer scientists. I mean, who needs a programmer or a sysadmin in space, when they could just upload any neccessary changes remotely.
The answer is Window! Now NASA will finally give us computer guys a chance, as they need somebody up there making sure the computers work.
Got my astronaut application in. How about you?
Who's going to tow the space station back to port?
Having worked with several shuttle astronauts, they have all reported "alpha particle" strikes in their eyes. Bright flashes in one eye when an alpha particle manages to get through all the external material but finally hits an atom somewhere near their retina. Result is a flash of photons, leaving them distracted for a minute, or waking them from a sound sleep.
Some optics sent into space on satellites have suffered from alpha flashes, leaving either a dead sensor or a patch of damaged cells. Don't know how the hubble gets around this, though.
And besides, I'd rather splice a 10b2 than a 10bF cable in weightlessness. Fiber is just too darn picky about getting it right, and requires training at least one astronaut.
I liked the comment someone made here about "they can fix it in their spare time". These guys aren't going to have any spare time, except for the allowed 1 hour per day of personal time. Every minute of their waking time during their shift is taken by experiments and other jobs. No rest for the lucky bastards.
10base2 has shielding, and you don't need a hub.
:)
if you had _read_ the article you would know that.
Does this mean there will be no Quake on the ISS?
Don't you think a hub IS a single point of
failure ? Should they have gone using multiple
switches with spanning tree and that whole
stuff ? Nah... 10base2 is just nice, and draws
0 power. That's the REAL point !
I was under the impression that most hard-core scientists had a working knowledge of Unix and it's derivatives and that a large amount of scientific software was based in the Unix world.
:)
Am I missing something? Luckily that no critical systems will use Windows... that would be truly scarey, but I don't think I would use WindowNT for anything... those little solaris thinkpads have more than enough power to handle 8 people's email and won't require maintainance every 2 weeks
The world is neither black nor white nor good nor evil, only many shades of CowboyNeal.
It's not a mission-critical environment. Just word processing and whatnot. Windows seems like the best choice in this environment...the astronauts have more important things to do that learning how to use Linux.
10 PRINT CHR$(205.5+RND(1)); : GOTO 10
Assuming the astronauts have a home computer (which is a reasonable assumption), they already know how to use Windows, which means that Windows has no learning curve whatsoever. Very few, if any, are likely to have Linux running on their home computers, so they will be less familiar with it. Even if the learning curve is fairly low, it's still more than no learning curve, and the astronauts have better things to be doing that learning how to use Linux/X/StarOffice.
10 PRINT CHR$(205.5+RND(1)); : GOTO 10
Read the article first, please.
1) these laptops are not going to be collecting data, so they can't "lose data that was important" if they crash.
2) I've never reinstalled Windows in the last few years. Reboots, sure, but not reinstalls. I install a lot of stuff too...if the astronauts are just doing word processing, the chances of really messing something up seriously aren't very good.
10 PRINT CHR$(205.5+RND(1)); : GOTO 10
So what do you suggest they use instead? Linux has not even applied for the test, and would fail if it did (it fails the requirement that all security-related software be binary-only).
10 PRINT CHR$(205.5+RND(1)); : GOTO 10
You haven't looked hard enough. I've always used LinkSys Combo cards on my laptops, mainly 'just in case' I need the coax somewhere. They're the main brand sold at Staples last time I was there.
This
I doubt anyone would have to learn anything from scratch.
If the platform they're familiar with is Unix, of course they'll be famialiar with Unix apps such as WordPerfect, LaTeX (they _are_ scientists, after all), Pine and the like.
Posted by HolyMackeralAndy:
BSOD will take on a whole new meaning.
Posted by dcantey:
Hey, Keyboard Boy, Ever terminate STP. It can't be fun in a space suit. Anyway there is no such thing as Shielded Unshielded Twisted Pair.
Sending an IT staff member on-site is going to cost a tad bit more than an airplane ticket in this case, I think. Given that, the 'cost' of learning a Unix over Windows is well worth it. Besides, it's not like we're dealing with idiots who can't program their VCRs, which is MS's target audience. These are astronauts for Crissakes. Learning Unix is nothing compared to the rest of the training they end up going through in preparation for a mission. These are guys who think nothing of going through complicated checklists just to take a leak.
Using Windows for this, even for something as simple as word processing and e-mail, is a major mistake because it cannot be fixed remotely. And why on earth (or not on earth) do they plan on making the communications link to the ground be some Windows proprietary crap too. That locks them in for purely artificial reasons. The comm link shouldn't give a rat's ass what the OS of the clients is, IMO.
Don't label something "offtopic" unless you know the topic well enough to tell what's on topic.
No, they aren't. They're using Windows. Windows and networking and minimum complexity don't belong in the same sentence.
Don't label something "offtopic" unless you know the topic well enough to tell what's on topic.
I doubt they will use printers of *any* kind up there. Where would they throw away the old paper? stuff on paper is a bad idea in general I would think.
Don't label something "offtopic" unless you know the topic well enough to tell what's on topic.
To summarize, a number of people have claimed that Windows is a good choice because the astronauts won't have time to muck about with the extra time it takes to get unix working. With Windows, the story goes, the astronauts can just go about their business without worrying about the OS.
Then, in response to this, others point out that Windows is horrible for administration and crashing and so on, and can't be administrated remotely very well.
In response, the Winvocate says that this is not a problem becasue, as the article points out, the laptops are being sent up preconfigured by competent personnel on the ground. If anything breaks they won't try to fix it remotely, just send up a new machine on the next run and take down the old one.
Am I the only one that sees the contradiction here? Hellloooo, Mcfly! If the ground personnel are the only ones configuring the machines, and they spend the time to get them all working perfectly before they get sent up, then this knocks the legs out from under the argument that Unix would be bad because the astronauts would waste time fiddling with it.
If all the admin happens on the ground, then Unix can be set up even better than Windows can be, and you have no reason to claim the astronauts have to set things up themselves.
Remember that there are two Solaris laptops going up too. Are *they* going to be adminstrated by the astronauts? Of course not.
Don't label something "offtopic" unless you know the topic well enough to tell what's on topic.
Uhm, okay. Then they can start by using an OS that makes proper use of the HLT instruction.
Don't label something "offtopic" unless you know the topic well enough to tell what's on topic.
Skylab! Aaaaaeeeeiiiii!
It's got great shielding - and I doubt their going to be running a Webserver or sending video around the LAN up there... ;>
...it wasn't pure O2. I figured they'd run a mixture similar to the "air" down here.
idiot - scientists don't care about OS, they just want to write fortran (or matlab or whatever) and run batch jobs. it's only geeks who spend all their time playing and no time working that actually get upset over it.
...a "Service Pack 5.2" upgrade. You know SP5.2, it's the one made by RedHat.
The next Cmdr Taco duplicate will be ready soon, but subscribers can beat the rush and see it early!
Fiber is faster, and doesn't need to be sheilded from EM interference, and is way sexy, technologically speaking. what were they thinking, indeed?
These are my friends, See how they glisten. See this one shine, how he smiles in the light.
Fiber is faster.
Hm.
I wonder if all the air can leak out the improperly sealed hatch quicker than the proper locking protocol can be transmitted from the gateway machine to the hatch control mechanism over 10base2.
I guess we'll just have to wait and see. . .
These are my friends, See how they glisten. See this one shine, how he smiles in the light.
Tin cans and string then! Definately no need for EM sheilding there!
These are my friends, See how they glisten. See this one shine, how he smiles in the light.
Just because you lack the NT experience and finesse required to keep your NT machine up for any length of time does NOT mean there aren't others that can.
These machines are surely not going to be shipped up there empty with a box of 95/NT CDROMs. They are going to be installed and CONFIGURED by ground personnel. Unnecessary/unstable drivers and applications are going to be removed and the systems are going to be placed in a state to ensure optimum reliability.
Even if a Win95 machine crashes, what's lost? They're only using these things for word processing, spreadsheets, e-mail, and your typical office/laboratory data entry sort of stuff. Perhaps a game of solitaire. They're not using these things to adjust the space station's orientation or oxygen mixture.
I'm a Linux person, but I'm really getting kind of sick of all these stupid anti-Windows posts people are making.
Do you honestly think the people working at NASA managing and planning these things are STUPID? They're a fuckload smarter than you are and if I were you, I'd give them the benefit of the doubt every time.
Linux is not the best solution for this situation. The astronauts are not going to have the time nor desire to re-learn an OS (especially one with the learning curve of Linux/Unix).
A - Probably not, but if they did, so what? The ground crew is going to be spending hours upon hours reconfiguring it and setting it up so that it's as reliable as possible (which can be pretty reliable, especially considering the resourcefulness of NASA engineers)
B - Probably not, but even if they are, so what?
C - I doubt that. They aren't building your typical SOHO LAN here. They need to ensure the equipment is protected and will function without fail for several years in orbit around earth. Network problems and equipment failures must be avoided.
they found some guy with a lowly and super simple job that would actually talk about the project.
That lowly simple worker has ten times the brains you do and probably makes ten times what you do. NASA doesn't hire stupid or shortsighted people.
Thinkpads... good god.. why not something that is rugged and designed for scientific use?
Why design and build proprietary hardware (at ten times the cost) that can unexpectedly fail when you have commercial, high-quality, high-reliability solutions right off the shelf? Aside from the trip up there, this hardware is going to be in THE ULTIMATE of places with regards to G forces and shock. Also, the astronauts are not going to have the time or desire to re-learn an operating system or hardware tool. They use Windows on the ground and are all probably very familiar with it. Why force them to change?
as for the linux troll, yes linux could be used successfully and easily... no worries about gnome crap if you run a simple x windows or even use Metro-X. and there are several thousand linux systems still in use that are happily running 1.xx kernels and haven been updated for years.
But why force the astronauts to spend time learning to use a brand new, complex operating system? Why force them to use a different word processor, spreadsheet and e-mail program than what they've used before?
NASA engineers are not stupid.
While I still think they made a good decision using Windows PC's for the astronauts themselves, the back-end stuff could have been done a lot more reliably and easily if they used a Unix server for it...
:)
Though there might have been reasons they couldn't have. What if the data collection software they're using was written for NT? Since mass is a big thing, they would only want to take up the bare minimum of hardware. If they absolutely need an NT server for one or two tasks, a compromise with the e-mail system seems appropriate.
There's probably a lot more to the story than what we're reading on the CNN article. Just put your faith in the fact that these guys are NASA engineers and they've probably already thought of this.
Think about this. The space station is orbiting several hundred miles up. Computer links with the ground are very lossy and prone to errors and they are NOT in 100% contact. There are times when the space station is completely out of contact with ground personnel.
These things makes remote administration somewhat difficult (including SSH to a Unix machine).
Besides, these systems are going to be pre-configured for optimal reliability by ground crew. They're not sending systems up with buggy/unstable drivers or loaded down with unnecessary and unpredicable applications. Problems like the ones you're describing are going to be nearly non-existant.
What makes you think they NEED training?
Chances are, these guys are already familiar with Outlook. It's not exactly difficult to use.
Did anyone ever consider the fact that maybe these astronauts WERE consultated with respect to their operating systems of choice?
If I were going to use a PC to just check e-mail and do some occasional word processing, it pains me to say it, but I think I would choose Win95 over Linux. Linux is great for the things I use it for, but really, it's a bit overkill for what the astronauts need.
I believe they're using them for e-mail, word processing, data entry/spreadsheets, graphical information display (they mentioned a ground-control-style application suite for tracking the station's position).
Since they already use Windows for a great deal (most? all?) of that stuff on the ground, it only makes sense to keep the astronauts using the same systems once they're in orbit.
They may also have some other software not specifically mentioned in the article (like, say, data gathering software) that was written for NT/95.
A great deal (most? all?) of the data collected is actually going to be analyzed on the ground. The scientists up there are in most cases just performing the experiments and gathering the data. They're doing no analysis, which means the scientific software isn't really needed in orbit.
Everyone whines about how much the ISS costs. Do you think it's going to be cheap to rewrite NT software for Unix? Do you think it's going to be cheap to train the astronauts to use a new operating system? A new e-mail system?
Do you think they hired some high-school VB programmer to design this thing? This guy is a LOT smarter than you are and I think you should give him a little more credit.
Just because he's not of the "use Linux for everything" mentality doesn't mean he's an idiot. He's probably got a whole staff of NASA engineers that are all approaching these problems and have decided upon the specs they have for reasons you (or I) have absolutely no knowledge of.
These guys aren't going to be using their 95 machines to locate another workstation connected via PPTP. They're doing your typical e-mail reading, word processing and generic SOHO tasks.
We have a "cluster" of NT workstations here serving e-mail to anywhere from 30-50 thousand users. The only time they've had to be rebooted was due to software upgrades, hardware upgrades or hardware failures. We have competant people running our NT servers. Perhaps you don't.
I'm not saying their applications are efficient in the least. I think it's considerably easier for a Unix admin to build and maintain a stable Unix server doing the same things that the NT server does.
But then again you don't know the whole story here, do you? There could be any number of reasons why they need NT up there. Everyone here is going by simply what was presented in the article. I promise you there's information we don't know.
With regards to Word/Outlook locking up the PC because of font problems, remember: The systems are going to be configured for reliability and optimal usability. They're not going to come loaded with a thousand different fonts. They're going to be working on these things, not building complicated proposals. With that in mind, it shouldn't be difficult to strip a Windows system down so that it works reliably.
you guys realize that windows has very specific flaws, even if microsoft keeps stuffing your brain with the opposite?
Oh, I agree, 100%. I'm not saying these systems will be *as* stable or problem-free as their Unix counterparts. All I'm saying is that the astronauts certainly have experience configuring Windows systems (like any other moderately educated computer user), and that's what they've been using to date on the ground.
All I was saying was that there are valid reasons for NASA choosing to have the astronauts use Windows for their "personal" computers. The silly claims about them wasting hours upon hours rebooting the machines every other minute and having to reinstall Windows every few weeks are just stupid.
I will think that you MS-guys are learning too.
Hah! Nobody's ever called me an "MS-guy" before. Usually I'm arguing in Linux's/Unix's favor, but I'm afraid for this article I have to step up to Microsoft's defense, simply because the Linux bigots feel they have to come out of the woodwork with every conceivable Windows cliche. It's pathetic. If you have a reasonable, educated argument, FEEL FREE to post it. Some have. But for those that feel they just have to jump on the "Linux r00lZ! WindoZe SuX!!" bandwagon without ANY real knowledge of the circumstances of the situation need to get a life.
time in space is more expensive than per-pound costs. An hour of an astronaut's work is way more expensive than a couple of pounds. Moron.
Ahh.. I thought this was going to be one of the first "real" dialogs, but I see we've resorted to name-calling. *sigh*.. It was going so well, too.
If the single two factors we had to consider was the astronaut's time and the payload mass, I'd be tempted to agree. However, neither you nor I have even a MINIMAL idea of what tasks these astronauts are going to be doing and what exactly they'll be doing with their '95 machines and the NT server aside from what's already been discussed.
Again, these are NASA engineers. You can be brutally certain that they have considered using non-Windows operating systems to do the task, but for some mixture of reasons, they opted instead to use Windows. Why? *I* don't know. *I* didn't make the decisions. It's not me you should be calling a moron. You have absolutely NO idea what factors they're considering. Again, they're not stupid. Give them some credit.
Hell, if you're really that concerned, write NASA and see if you can get more information about their decision.
don't think any other OS has crashed in space, other than Windows, mind you. True, Ariane blew up last year for a software problem..... but, still, is anyone aware of a "software problems in space" site ?
Again, I agree that Windows is not the most stable OS choice, but again, this is not the only factor to consider. You don't have all of the information. Don't be so quick to judge.
Besides, these are NOT CRITICAL MACHINES. A crash or reboot is not going to cause the loss of millions of dollars of astronaut time. I'm not saying that it's NOT going to crash. Let's say that an astronaut's machine requires a reboot once every week (an easy goal for NASA -- my Win98 PC at home, running continuously, used nearly continuously by various people, needs to be rebooted an average of once every two weeks). That's what, a maximum of 3 minutes? 3 minutes/week is nothing. He could easily spend 30 seconds a day brushing his teeth a bit longer.
The guys at NASA are smart, but I'd bet against a nobel prize that there are better solutions than this one.
Why don't you write them and enlighten them? Say, "Hi guys, I know you've been studying this problem for months if not years, and I know you have degrees in all of these types of things, and I know I have but an inkling of knowledge regarding what you're going to be needing these machines to be doing and the environment it's all going to be working in, but really, you're all morons! You could be doing it SO much better if you used Linux!"
Please...
It's to stop the space-station being invaded by aliens! As soon as the alien invasion fleet realises that the space-station is using Windows, there won't be enough gold-plated latinum in the galaxy to coax them to step one foot inside!
It's a small world and it smells funny; I'd buy another if it wasn't for the money; Take back what I paid (SoM)
You're right, NASA is really trying to save a buck here and there on networking hardware. Give me a break. I guarantee you that a spool of fiber and a concentrator is lighter than a big roll of shielded 10Base-2.
--
Sheilding in space is VERY important. The sun gives out tremendous amounts of electromagnetic energy.. as well as x-rays and other forms of radiation. An unshielded system would simply NOT work because of an extremely large amount of data error caused by these fields/rays.
On earth we don;t have this problem because the earths magnetic field and atmosphere protect us from this. You won't have that in space.
-Ex-Nt-User
Nope Hard Drives are filled with just regular air. It's very low particle..clean room level air.. but it's not some inert gas or something like that. (I know I work at a HD company)
As far as takeoff.. there will be no problem as long as the drives are not spoinning at the moment. A lached drive is capable of withstanding upwards of 100G forces.
-Ex-Nt-User
They point out in the article that 10base2 uses less power than 10baseT.
That's kind of important when you're running everything off large solar cells and batteries.
Not to mention a 10base2 LAN will require less cable than a 10baseT LAN, as well as a hub. This saves a few ounces of weight...
This is important when you are flying around 250 miles above the planet.
Guys -- this is actually good news!
:)
What it means is that NASA is actually breaking away from the mystique of custom, over-priced solutions. Which greatly increasess our chances of living on the moon in our old age.
Does Windows suck -- YES. Does it matter? I mean, we're talking about email here. With any luck, we'll have another USS Yorktown style incident.
I have to admit though that the 10Base-2 is a bit harder to swallow.
-- Slashdot sucks.
besides the obvious (email) can anyone (constructively) confirm to what use they might use MS-Win95/NT based software for?
I've had a look at a lot of the sw at nasa and most of it's number crunching for calcualting position etc...
peterrenshaw ~ Another Scrappy Startup
As long as you have air pressure, hard-drives have no problem, they use laptops on the space shuttle
/.-ers will tell you, aerodynamic lift is not very dependent on gravity,
all the time.
Despite what some
which you can verify by spinning-up your hard-drive in various orientations (but not while
it's spinning, angular momentum can do bad things).
No gravity, small problem, but not for the reason you might think. The biggest problem with any
electronic or mechanical system in space is heat dissipation. On earth, hot air rises, causing
the hot air to move away from the heat source (convection). In zero G, heat doesn't rise so
you have to have fans or heat pipes.
Hard-drives probably go through many more thermal recalibration cycles in space, but this is minor.
Freecell has worked perfectly under Wine for a year.
But hardcore microgamers play reversi (othello) anyways.
Isn't MS software only licensed for use in one time zone?
This could give rise to a new phenomenon: Windows re-entry day.
---- "If we have to go on with these damned quantum jumps, then I'm sorry that I ever got involved" - Erwin Schrodinger
Hey, I for one, agree that Windows can be set up to be fairly stable. It's why I insist on doing fresh installs when computers get shifted around at work. No extra garbage from app uninstalls and crap like that. BUT, in one of your posts you stated that w/out installing more software, Win will stay as stable as the day it was installed. Phooey. I've set up my own system with 2 high end apps, and nothing more (except device drivers.) Six months that install lasted. A record.
You may not notice the difference in the software you run, but with multitrack digital audio, system degradation is very apparent. It demands real time, stable performance. Windows can do that for a while, but the registry seems to get "crusty." The best version of Windows isn't even available unless you bought a new computer between 8/96 and 5/98. Win95B was the only one worth spitting on. Go try and buy it. (This excludes NT, which is a whole other set of pros and cons... not a huge leap forward, IMHO.)
Also, buggy apps should be irrelavent to the OS. If the app crashes, restart it. But it shouldn't take the system down with it, increasing the chances that the registry will get corrupted. That's not something I want astronauts I'm paying to worry about.
On the other hand, I don't want them spending the time learning new apps, either. I'll take a guess that this is why NASA chose MS. Hmm, I think the article said something like that, huh?
Further, since Microsoft itself has admitted their OS's crash far more frequently than Solaris or Linux, it seems pretty silly to claim otherwise.
I've been involved with a highly competent Windows (3.x, 9x, etc.) support department (configuration, installation, etc.) for over three years at a good-sized university. My experience evidently diverges from yours strongly when it comes to OS stability.
Kythe
(Remove "x"'s from
Kythe
I wouldn't like to change a toner cartridge in zero-gravity...
We need an open-source OCA project!
"E-mail is a bit of a challenge since we do not have a constant connection with the ground."
I guess they won't be reading Slashdot.
On the bright side, at least the NASA people aren't using WinXX for any of the "real" work up there. But I'd hate to see them lose any important scientific data to a crashed computer.
...playing my son at home, over scrounged ethernet cable/NIC's. He on the Win95, Me on Linux. Too bad running on the better OS doesn't prevent me from getting my ars whooped by that 14 yr old kid every time we play... ;-)
Your Servant, B. Baggins
Given that they won't be able to establish peppy TCP/IP links to run something like, heck, PCAnywhere effectively to do remote configuration and maintenance of the systems, this looks like an pretty bad idea. There are probably going to be situations where their only practical choice will be to ship up replacement machines or ship up an MCSE to keep things working.
A Windows network can be a real bear to maintain remotely in situations where you don't have a WAN or realtime dialup access to fall back on. NASA is not staffed by Unix-phobic fools, though, so I have to cling to the hope that they've got redundant Tivoli or BackOffice servers in place, an ability to do network boot and a total wipe of the client machines, and the ability to wipe and rebuild the servers from scratch and then do data-only recovery from tape.
Using unix does not require a sys admin on the space station as unix machines can be administered remotely ( and I would imagine that they manage the Solaris machines they have up there in this manner .... )
I don'tknow how you substantiate your claims of "equal performance and stability". ROFL. I guess you are just trolling.
Cheers,
-- Elflord
Most controls in the station will be embedded systems. Its not like WinNT is going to be controlling the flow of oxygen.
If someone needs to use a word processor, Word is as good (or crappy) as any. No, AbiWord, Klyx or any of the other lame-o linux toys probably weren't on the list. People in orbit can't be bothered to obsess over the openness of the QT libraries, GNOME support, or any of the other minutae that linux nerds give themselves migraines over.
Yeah, that's right, no one at nasa knows what they are doing. They should have called you.
nuff said.
Are you in space already?
"intuitive enough so any competant person can administer the network in their spare time"
My 8 year old can admin a 5 person network remotly from school during lunch. And yet again, classic mistake number 1: When dealing with a network, a UI doesnt make your job easier. Most of the time it makes it worst, by standing between you and the job at hand. Or worst, it makes a incompetent think he/she knows what he/she is doing...
On the other hand, I thought the people going to the station would be highly trained scientists and technicians. Usually these people learned to use Unix when they got the tape from ATT for their PDP-11.
As for UI, if you think Unix has no UI, your case is lost and closed. Go play solitaire.
I was the primary subsystem engineer on the data systems network some years back at JSC. At that time, the network was spec'd as a Dual-redundant 100 Mbps FDDI ring with a 50-Mbps Ku-band uplink. The system was supposed to actually be linked directly to the internet. LynxOS was spec'd as the OS with X-Windows and a plethora of wonderful control and monitoring applications. I can't believe that after 10 years and how many billions of dollars, we end up with thinkpads that are going to have to be rebooted every hour.
All mine and others effort, flushed down the toilet with little to show in its place.
I guess going back to school was the right decision after all, I just wouldnt have been able to take this had I stayed.
Frank W. Miller
Yeah, the Windows environment is no doubt going to cause crashes and data loss and all that (unless NASA & gang are very lucky). But important scientific data being lost? C'mon, the ISS is about international politics, there is going to be very little in the way of science going on up there. There is only so much that can be learned by bouncing around in reduced gravitiy while feeding rats. Who cares if some of their "important data" is lost, it will just give them an excuse to shovel millions more in taxpayer dollars up there to upgrade their systems...
duck and cover!
-- To dream a dream is grand, but to live it is divine. -- Leto ][
I'm currently using a 3Com 589D, with a combo adatper, both twisted pair and coax.. You can't switch between the 2 easily (the various drivers seem to assume TP), but it can be done.. I've done it at home before I can a TP setup..
Slashdot is like Playboy: I read it for the articles
Blue Screen in the sky from Redmond.....
Slashdot is like Playboy: I read it for the articles
That lowly simple worker has ten times the brains you do and probably makes ten times what you do. NASA doesn't hire stupid or shortsighted people.
Nope. Salaries there suck. I can go there (have an offer), or I can find a job in industry in BA. Guess what I will do. Will not even think about that stupid goverment job.
NASA engineers are not stupid.
Indeed, but they are very conservative, and suffer a LOT from politically motivated managers. Goverment science is an horror show. Seen enough of it. Sheesh.
<^>_<(ô ô)>_<^>
Politically motivated or not, these guys know their stuff and I don't think they would ever put the safety or comfort of the astronauts in jeopardy because of stupid politically-based OS decisions.
You will be surprised, how many design and especially mission concept decision were dictated for political reasons. Of course we make sure that it does work. But usually there are other, and usually better, ways of doing things. NASA leadership is very politically motivated. They want cost reduction - grab off the shelf component - who cares that a specialized solution can be more reliable, and actually cheaper to maintain. They still to not count salaries of support pesonnel into the cost of the project - at least in our project (I am outside NASA, but do work on a project for them). I is really frustrating the way burocracy there work. This is the reason I refused to consider employment there, though the people are good indeed. Though the salaries are not high at all. Well, in comparison.
<^>_<(ô ô)>_<^>
Good chemists tend to be using IRIX now, on SGI Origin 2000 machines.
Are you nuts !? this is not about doing sysadmin or not (which unix can do remotely _anyways_), this is about an astronaut's time being *pretty expensive*. Just the cost of waiting for a reboot is astronomical, in space. How many times did you reinstall windows, huh ? think at doing that in space, at how darn expensive that is. Talk about zero-administration costs =)
Don't give me that crap - and keep counting how many times thay'll have to reinstall and (huh, huh) configure Windows crazy networking - you WILL need a sysadmin in space, after all.
I bet a pizza that within five years of operation they'll switch the stuff to linux or solaris - the only things that got windows in space were the applications, not the OS. Remember that.
(I'd rather run Wine - sheesh).
10b2 draws zero power. It does NOT draw 0 time. that is going to be the real problem. And Yes, I would tell those guys at NASA what to do as much as to any other average joe who needs advice.
This is going to be fun. Not for the poor soul on Eart who has to administer it, though.
--- "I didn't think anyone would understand it" -Prof. Bob Muller
But why force the astronauts to spend time learning to use a brand new, complex operating system? Why force them to use a different word processor, spreadsheet and e-mail program than what they've used before? Because can do that on earth, in the YEARS between missions. Windows reinstalls will happen in space instead wich is both uneconomical AND silly. And, yes, people know how to use windows. Did you read they are going to use Office ? How many times did that crash on you ? Of course they know how to use it - everyone does, I am posting from M$ IE - but what is convenient in Earth (rebbot - no big deal) is NOT in space. They'll either use those stations just for astronaut's e-mail and excel (in which case, unless you lose the data of an experiment in a crash, reboot is a good excuse for a coffee break) or they are headed for trouble. Hey, just wait - I am sure that with the current attitude toward computing, you'll hear CNN reports for the first two-or-three crashes that lose data that was important or ruin a million-bucks experiment. I don't care how many PHds they have at NASA - the guy knows what he's doing for sure. It will be nice to see astronauts replacing his 10b2 cable when anything happens, or cope with windows _oh so stable_ configurations and networkin'. "hey - I lost those drivers for the space telescope in the last crash" =) -- just kidding. Hey - You WILL SEE what good advertisement those crashes will be for the M$ platform. just wait and see.
--- "I didn't think anyone would understand it" -Prof. Bob Muller
Apart from the flame war I seem to have started, lets look at the bright side: most of ms-sided folks say it is possible to make windows stable. Some other guy will rather say "sure thing, just do not run it". What about a Stable-windows howto?
It could _really_ spare a lot of people some serious headache and make life easier for power users. So... Should we ask NASA for the howto or... is someone up to the task ? Don't tell me "remove unstable *stuff* and it will work". too easy! Give us the procedures ! Do you think we don't want a stable windows installation just to be able to flame M$ ? Hey, most of us use windows anyways. So.. take it up the linux way! Show us what Windows can do. Or do we need to ask the NASA guys for this ?
Hey - never thought this morning I would spark such a flame !
(cool - I never replied to myself, before!)
--- "I didn't think anyone would understand it" -Prof. Bob Muller
Hold on... the specs I saw a year ago stated solaris on everything running Metro-X as the X window server with 100baseT. But this was the scientific and control system not the recreational system for the residents. I dont see NASA jumping on the win bandwagon when it's not on the shuttles, on anything in mission-control, and all their programmers are UNIX heads to begin with. NASA knows what mission critical means, and they would never put a microsoft product on an important system, maybe the entertainment system or the personal laptops for the crew to play doom,quake,etc.. but never for anything of importance. (the toilets on the shuttle run embedded unix! so NT isnt safe enough to flush the toilets!) remember this is CNN and they arent the most reliable, and this was a press release from a company I'd bet that is either doing the work or supplying the products.
I will bet a nice shiny new quarter that Unix is doing the important stuff on the station.
Do not look at laser with remaining good eye.
Hey folks please note:
Separate two thinkpads to run the station (with slowarez)...
Did anybody think about this. Isn't this a nice opinion about M$ capability for use in mission critical environment.
Once again note it is not even let close to the mission.
And a second note - it is very interesting that the austronauts are not considered smart enough to use solaris for mail...
Doesn't this cause thoughts...
Baker's Law: Misery no longer loves company. Nowadays it insists on it
http://www.sigsegv.cx/
1. Get yourself a modern unix box. I would not mind hearing your opinion again after that.
2. You did not read the whole article. Please note that UNIX is what is used for the mission. Windows is for toys...
Baker's Law: Misery no longer loves company. Nowadays it insists on it
http://www.sigsegv.cx/
Haha this is one of the funniest things I have read all day.. If the troll was actually serious I would be scared.
Seems to me there are lots of people making intellegent and thoughtful arguments for using a non-microsoft platform on the station, and then there is one or two people who can only come up with things like "fanboys" and "propeller-heads" to justify their pro-MS position.
Cassius your whole argument boils down to the assumption that for the stations crew of _scientists_ that using unix is somehow a "change". Now no one here really knows what the crew of a space station was "brought up on", but seing as they will all be educated scientists, it would not come as a surprise to me if they are already used to working in a unix environment.
I also happen to agree with the posters that mention the decision to use Windoze is more politically motivated than technically or logistically motivated. I worked for NASA for a time and I know how science and politics mixes there.
I hope their link is pretty fast. That way they can handle the bloat from the HTML, RTF, or WFT? file formats that piece of trash converts plain old text e-mail into. Christ, I've had to handle some email sent to one of our users in Word format.
I believe most scientists are *nix advocates. I have people here that won't touch a M$ OS when doing research work.
I'd love to see how much trouble they have when they need a patch right now from their precious COTS sources. If I had to operate the thing, I would be screaming GIVE ME THE CODE - GIVE ME THE SPECS - NOW!
Why is it we always end up with dinosaur brains designing "The Next Generation" platforms?
-- ultra1
- I fail to see how fiber needs all this weighty shielding - it pushes photons, and is usually the network media of choice in electromagnetically noisy environments
- Fiber isn't even close to being bleeding edge technology - I was splicing fiber with hand tools five years ago, no problem
- 10Base2 draws power - in fact, you can feed 12.5 volts to devices via a 10Base2 tap. I'm betting that laser diodes draw less, especially for multimode fiber
- Who said this was going to be assembled in space? The modules go up pre-constructed
- I can see the transceivers causing trouble - the damn things are so light, though, you could send up a few spares without feeling it
Yeah, I was ranting... I just wish I knew more about why they made the decisions they did. It would appear at first glance that they could have chosen better technology for space.-- ultra1
One thing that I could find quite important for the recreational computers is power management. What was the last laptop you saw running some flavor of unix with power management? Every volt is precious up there I'd assume so why not use something that won't suck up all the power of the station if you leave it on for a bit?
Where do you want to get sucked out to today?
Proxim's rangelan/2 isn't supported by Linux. In fact, Proxim won't disclose a damn thing about the cards -- My University uses it for the wireless network here and is giving up on them for Lucent WaveLAN for many reasons, Linux compatibility being a major issue. I've emailed them many times about getting specs, with only requests for $20,000 for their source and an NDA for my troubles.
Jim
If I'm not back again this time tomorrow...
Sure sendmail has it's problems, but it was designed to handle flakey network links....
Didn't Widndows failed some kind of a test (i think FIPS)?
Doesn't it mean that goverment controled computers cant run Windows?
NASA is govermental, no?
Anyway - I can still imagine the movie:
"Apollo 13 part 2: BSoD":
They get GPF when trying to use both the oxygen and temperature control software at the same time.
They try to escape but the spaceship wont boot!
Too bad they didnt keep these rescue floppies, huh?
---
I'm going to live forever, or die in the attempt.
These astronauts,
:-)
They're really stupid, right?
They're the ones who failed school.
They couldn't even get jobs at Burger King
So they became astronauts.
That makes sense.
And ofcourse,
Whats better than Windows for idiots?
I guess a mac.
I guess it doesnt matter the stability of these computers,
After all its just life support control and Millions of $$$ worth of research.
Nobody needs oxygen ALL THE TIME.
So go on astros, play solitaire.
I guess that's a better thing to do than "learn how to use UNIX"
---
I'm going to live forever, or die in the attempt.
These astronauts,
They're really stupid, right?
They're the ones who failed school.
They couldn't even get jobs at Burger King,
So they became astronauts.
That makes sense.
And ofcourse,
Whats better than Windows for idiots?
(A Mac?)
I guess it doesnt matter the stability of these computers,
After all its just life support control and Millions of $$$ worth of research.
Nobody needs oxygen ALL THE TIME.
So go on astros, play solitaire.
I guess that's a better thing to do than "learn how to use UNIX"
---
I'm going to live forever, or die in the attempt.
Astronauts aren;t briliant computer users, they may now nothing at all about computers
With Fiber and 10baseT you have to run a seperate line to where ever for a hub
If you have trouble with these specs, build your own space station.
Eh?
In one NT lab I sysadmined on co-op, the NT server was a tad flakey so if you moved the mouse too fast you got the BSOD, So I just left it alone except to do backups.
Eh?
For once, someone is dead on right. Windows 95 is a toy operating system that has a few haphazard add-ons (networking, PPP, DNS, FTP, etc.). These add-ons fall well within the specs of *their* mission (they need to f**k around with e-mail). So they put the evil, bastard computers from hell on a spindly little backbone and set them a chugging. Let's hope they don't run anything mission critical. Mebbe Windows works better in space :-).
aÍÍ©ÍÌÍ£Ì'̽ͩÌÍzÍYÌÍÌY
Just had a horrifying thought of them actually running the station on the 95 Laptop, with Back Orifice installed :) I can just picture the scene with the K33l dUdEz on earth "Dude lets make it hit Mir for extra points.."
Anyone wanna start a pool on how soon the network'll crash..?
Your whole argument is based on an assumption for which you have no foundation: That it is somehow a change of platform for the crew to use Unix or Linux. Do you know the crew? Do you KNOW what OSs they are most familiar with?
Have you ever tried searching for scientific software? Then you'd know that there are huge amounts of Unix software out there written by scientists with a non-computer science background. For people with a long scientific career, you'll find a lot of Unix users, and a lot of Unix programmers.