The biggest problem is of course the budget for NASA, which has been steadily decreasing over the last few years. I think it is ludicrous of us to expect NASA to launch missions like these without a proper budget for development, manufacturing, and testing of the equipment. In this case, apparently the money was simply not there to afford a complete test of the whole landing proceedure - which would have made it obvious that the engines were not going to work and that the censors for shutting them off would not work either
Too right. It was really really stupid on the Mars Polar Lander mission... for like 4 years we had so little funding, and the manager (Karen McBride) did an incredible job to secure the operations facility at UCLA, which was a first for a NASA operation to be run outside JPL.
Then, of course, landing comes closer, and we are glutted with money from NASA. But what's the point? We get 10 sysadmins one month before landing, and we have to spend hours training them up; much more preferable to have had one good one ten months before!
Compare this to the old days of Viking... billions of dollars spent, ten years of development, but they delivered an absolutely monumental amount of science back for years. Even if the MPL had been a huge success, it would still only have had three months of science, and that too not 24 hours a day! Instruments like the microphone only had about one hour a day or something stupid in which to transmit data back to earth.
Still, I have faith in Faster Better Cheaper; we cant afford mistakes like the Mars Orbiter again, where a simple case of bad luck (exploding fuel line) lost the entire mission. Painful as it was to lose the MPL/MCO/DS2 probes, they were cheap and cheerful, and we can throw more of them out to Mars.
To introduce myself, I worked on the Mars Polar Lander project as their Outreach Architect, and did some work with the Ground Data Systems crew.
That report is bull as far as I am concerned. We worked our ARSES off as launch came closer. I have never worked with a brighter, more intelligent, and more optimistic crew than the Mars Polar Lander team that was assembled at MVACS before that fateful landing.
After the Mars Climate Orbiter was lost, we had a huge influx of support from JPL; system admins, scientists, programmers, all drafted in at short notice simply to come in and fill in the budget shortage that we'd been suffering from before.
I somehow don't believe JPL would have thrown everything they had at us, to help us out, if they somehow knew this "secret" that the lander was doomed to fail.
It is possible that such a report could have kept secret by uppermost echelons of NASA, and kept from the team and the JPL management. I find that difficult to believe; I would count any such action as bordering criminal, after the incredible amount of hard work sunk by the staff of the mission before landing.
Really, the feeling of utter disbelief we had when it crashed said it all. Noone really expected it to happen, after all the effort, and it took a long time for some of the crew to come to terms with it.
Also, the Mars missions are underwent a fundamental and deep review of their future after this mission failed; it simply wasn't in NASA's interests to cover this up, since it really would have been a "worstcase" end to their Mars Surveyor 98 missions, with only ONE success (the global surveyor). If they did know about this, it would have made sense for them to come out and say it, and attempt a fix, rather than keep it secret, and throw (useless) resources at the team!
As with all other management, NASA executives can be dumb, but I don't believe they could be that dumb.
Why is this technique so useful compared to session-id tracking? I don't believe that it really is, or why such a fuss should be kicked up about it :
If I understand it correctly, it simply replaces the session ID normally stored in as a cookie/get-var in the hostname.
This would lead to extremely user-unfriendly domain names, and surely it would really bugger up users trying to bookmark the site (stale sessions could stay in bookmarks for a LONG time).
Also, its simply not as efficient as session IDs, which after one unfriendly GET, tend to store their results in a cookie which is transparently passed around. Surely dynamic DNS would all have to have really low TTLs and generally slow down site access if you have to do a large number of DNS lookups (which can be the slowest stage in an http access cycle?)
As I see it, the only problem with the session-id method is that it complicates serverside scripting, but with simply superb tools like PHPLIB all those details are abstracted away from the user. And also PHP4 has built in session handling to simplify things further. IIS has similar modules for ASP developers, and I'm sure others exist forother scripting languages (mod_perl? dunno...)
So while this might be of interest to some specific applications, I can't see it revolutionising the whole ecommerce industry with its cunning "new" user tracking system.
But then again, I might be talking bull:) Do correct my bullshit if it is that, please.
First of all, you should investigate using Cascading Style Sheets instead of tag for modern browsers. They are one simple tag and also degenerate gracefully for older browsers (which ignore them).
E.g., if I remember right, you'd use
BODY: { font-family: Arial,Helvetica,sans-serif }
And that overrides the body tag with the font you picked.
There are lots of cross-browser problems with using these unfortunately, but check out some decent sites like WebReference for info on them.
Incidentally, you often have to use lots of font tags if you use tables. A font tag will apply to multiple paragraphs, but not to the cell contents inside tables! You need extra font tags in there for that.
Heh, why not try to use one of the HTML compression tools which have a short javascript header which unpacks html on the fly? Problem is finding a small enough bit of "header JS", but it might be a cool way to get around the 5k limit while sticking to it.
Look at this freshmeat entry for AlgART HTML Picker to see what I mean:)
I followed the incredibly interesting link from this article regarding the "Really Bad Attitude" newsgroups that Netscape had setup, and that Microsoft subpeonaed (at http://www.jwz.org/gruntle/rbarip.html).
I noticed this quote :
In hindsight, complying with the company's Document Retention Policy (which at Netscape was basically, ``shred anything within 90 days unless you can't get your job done without it'') might have been a good idea.
How many major companies actually have a policy ilke this for electronic information? Most backups are tape/DLTs which last eternity, and is the only purpose of this policy to prevent liability with stuff lying around?
This sounds like it worked with paper-based archiving systems, where space simply doesn't exist to archive forever, and non-essential documents are destroyed, but none of the people I've done work for have had a similar policy at all.
So the question is... how many companies out there do this to avoid liability, or is there a different reason for it?
Yeah, the 5% comment came from a random statistic thrown out by one of the major UC colleges, for whom that is a hell of a lot... smaller colleges will suffer more obviously.
Quote: Think about guns - guns are made to kill people. They are used in crimes every day all over the nation. But still, they are legal, because we recognize that they have legitimate uses. You can restrict the use of guns, and you might even reduce crime, (I don't actually believe that, but that's an entirely different story)
Well, I'm from Ireland, guns are banned here in Europe, crime is lower, but you're right, that's another story. Not a good analogy with MP3s I believe though.
but you'd be giving up a portion of your freedom to do that. Similarly, you can ban mp3s, and you may reduce the amount of IP theft, or lower your bandwidth utilization, but you're giving something up. Namely, the positive aspects of downloading mp3s, and also, the students freedom to be in an environment that allows them to expand themselves as they see fit, not as the university sees fit.
You cannot condone freedom to commit illegal acts. You've already given up your freedom to kill your fellow human beings, and indeed, be protected againts such harm, by living under law. Are you the worse for it?
I can kinda sympathize with the bandwidth argument, but I really hate it when people change things midstream. If you come onto the university network, and sign an agreement saying "by signing this you agree not to do x, y, or z" then you have a choice, and you can go elsewhere. On the other hand, if you invest in a NIC for your student network, and then have regulations piled on you never agreed to, that's different, because you weren't given a choice.
The agreements haven't been changed at all. When I got net access from my University (Imperial College, London), there was a very very clear AUP that stated that illegal activities on their network are, you guessed it, ILLEGAL. Of course, uni's do overlook a number of illegal activities, but when those activities start to compromise your bandwidth flow, how can you expect them NOT to take action?!
If the trade of pirated/illegal MP3's is so rampant on a college campus, why not monitor the traffic, collect data, and then bust the offenders? You do this enough times, and people will start to realize they are being watched and will be caught.
Effort. It takes a lot more effort. It's simple enough just to add a line into a portblocking table to stop it. And you'd have to monitor continuously, keep the tools uptodate with the protocol, and if you slack off, word would get around and mp3 trade would start again. These are students we're talking about:-)
Quote: And another point. Banning Napster doesn't prevent people from pirating MP3's. Banning illegal MP3's will ban illegal MP3's. Funny how that works. It takes vigilance to stop crime, not another line in the old port filter.
There's principle, and there's being practical.
Me standing up in University and screaming "I BAN MP3S!" wouldn't do shit. Me blocking the tool that a lot of students use, which would cut down the usage drastically, would. And remember, almost all acceptable usage policies ban copyright infringments and other illegal activities using that network, so illegal MP3s are banned from the start.
I just can't see a line of conscientious students poring through their AUP's and then saying "Sorry Jim, I can't accept those files, they're illegal!"
Ban the popular tool. Easy enough to get around of course, but it would stop the 95% non-tech-savvy ones anyway.
Well, Napster is clearly a software product that is used almost exclusively to pirate MP3s, it leeches up a lot of bandwidth, and has some pretty bad security holes that means that potentially the client can be forced to serve up any file to a malicious Napster server.
What's wrong with universities banning it? Clearly, they cannot condone the flagrant ripping of MP3's on campus. I'm sure they'd do the same if they found 10000 warez sites running on their students' boxes.
I'm not saying that censorship is a good thing, just that in this case I fail to see how this is construed as censorship, given that using the application for anything other than illegal activities is fairly hard. 5% of university bandwidth is a hell of a lot as well.
No, it's not absolutely neccessary, since they send IF_MOD_SINCE requests, and don't always retrieve the data, etc.
But still, it's a good matter of principle to do so, to guarantee behaviour of clients.
Also, this only extends to private caches, and public caches really like to see http/1.1 headers before holding on to them (again, depends on the exact cache obviously)
Why compare it to a ridiculously expensive Thinkpad? My Sony Vaio F360 cost around $2600 and has a :
PII-400 128mb RAM 12gb Drive DVD 4x Firewire, and USB Built in modem 14.1 visible screen .. etc
and runs Windows and Linux like a dream. I must say I can't think of any killer apps that the Mac can run that those two OS's cant (even Photoshop is nicer on Windows now than the Mac, and the GIMP rules Linux)
Maybe mass-marketing the Macs would drive prices down... bit chicken and egg though.
Well, the expat library already exists and seems to be quite defacto under Unix.
At least, PHP and Apache use it, and well. And a couple of other utilities and CPAN modules. I've come across. It's fast, small, and not full of unneccessary crap.
Anyway, utilities like Apache and ProFTPD already have meta XML config scripts, and a fair few perl scripts that make use of XML::* as well.
I don't think I like the idea of having a dynamic XML library, and your entire/etc filesystem depending on it. What do you do when it breaks? It's like screwing up your LD.so.1... ow:)
Still, a standard, easy to use DOM for utilities would be nice, perhaps some kind of extension of the GetOpt stuff would work? It's succeeded largly because it's there and easy to use, and an XML standard would have to do the same.
For those of you interested in caching and how it can help large scale sites, I've helped co-author a technical report with Network Appliance, which was our experiences at accelerating the Mars Polar Lander website. That site used NetCache boxes, simple HTTP/1.1 cache-control headers, and a bit of cunningness to allow user-level tracking without letting the track requests filter through. As traditional, the site had a couple of problems which we've also included in the appendix after we fixed them, to hopefully save other people the same hassles in the future.
We would all save a scary amount of bandwidth if more sites were designed with public caches such as (the awesome) squid in mind, and it's a really simple use of headers that make it possible.
For those who use Apache and are interested in making your own sites more cache-friendly, I recommend you look at mod_expires, which is part of the default distribution of Apache, although not compiled in by default. If you have large, static images that rarely change, then go ahead and put week-month-year long expiry headers on them, and watch the hits for those redundant images drop right down on your web server. And if you suddenly need to change them, then it's no real problem, as all you have to do is change the images URL and it will become a "new" entity for purposes of caching.
Yeah, granted, bandwidth is getting cheaper now, but for us poor Europeans, it's still a scarce commodity and we need to worry about these things:-)
Perhaps a special case could be worked out for developers not intending to make a profit from their development resources... free-for-non-commercial-use license perhaps?
Although it smacks of corporate greediness, and it is apparent that Linux developers who work for free can't join the members list, why can't the corporate Linux companies, such as Red Hat, Caldera or VA Linux systems join, and make their copies of the specifications available to developers. There may be copyright issues, but if the USB developers work "for" one of these companies, then surely there is no problem. The end-result would still be GPLed, and available to all, it's just a nominal shift in who you are working for.
Linux has a number of high-profile companies now, and it's in their interests to ensure the best hardware compatibility for their distributions. $2500 is a small price to pay for that from their point of view, surely.
If you are concerned about security and reliability, then BIND may not be the best solution for you. It has a history of security exploits, and in general the codebase is a complete mess.
Everyone's favourite daemon-revamper Dan Bernstein has an excellent DNS suite of tools called DNSCache which you can obtain from http://cr.yp.to/dnscache.html. This looks pretty much ultrasecure (as most of his stuff is), and also very robust and modular. In addition, you don't need to muck around with zone files and so on, and the interface for adding information is ridiculously simple thanks to some scripts in the package.
For example, to add a new alias to an already existing domain, all I have to do is :
./add-alias new.recoil.org 1.2.3.4 make
and thats it! Serial numbers etc all taken care of automatically for you.
I've been running and playing with DNSCache and can recommend to anyone looking for an alternative to BIND.
I used to work for the Mars Polar Lander project, on its Ground Data Systems. One thing about this announcement is that its only making official what already happens a lot in NASA.
Internally, we used a lot of Linux boxes, and even more GNU tools on Solaris boxes. In fact, the whole image processing pipeline was built using a series of BASH scripts by the NASA AMES team, as part of their operations deployment.
The web servers for the MPL were Solaris x86 boxes driven by Apache and PHP, while the support boxes for DNS and mail were Linux (red hat, stripped down). All ultra-reliable, getting them secure was the hardest bit of the whole operation really (we considered OpenBSD for a while, but didnt have enough inhouse skills with it)
And, to reassure you, no Windows servers were used for anything other than looking pretty. Oh, and Quake III. None of the Linux boxes had 3D cards:)
Hi, I'm part of the MPL team, and thought you guys might like to know that we are taking the flight-spare MPL to a secret location in Death Valley, and running a complete simulation of operations as they might have been had it landed on Mars.
The point? To test the instruments and make sure they would have done what they are meant to, as a number will be reused in future missions (the Stereo Surface Imager in particular).
You can follow all the images and simulations if you're interested in what actual science was being conducted, at our official site www.marspolarlander.com
Cheers, Anil Madhavapeddy - Ground Data Systems - Outreach Mars Volatiles and Climate Surveyor
We are using a combination of Solaris x86 boxes and Linux boxes to run the website, along with NetApp cache boxes and load balancing DNS around various hub points to handle the majority of hits.
The actual Ground Data System uses Sun (sparc, ultra, etc) boxes, a lot of heavy SGIs and so on... no hint of x86 procs there:-)
Anil Madhavapeddy - anil@mars.ucla.edu Mars Polar Lander Project http://www.marspolarlander.com/
Slashdot Mirror
Quote:
... for like 4 years we had so little funding, and the manager (Karen McBride) did an incredible job to secure the operations facility at UCLA, which was a first for a NASA operation to be run outside JPL.
... billions of dollars spent, ten years of development, but they delivered an absolutely monumental amount of science back for years. Even if the MPL had been a huge success, it would still only have had three months of science, and that too not 24 hours a day! Instruments like the microphone only had about one hour a day or something stupid in which to transmit data back to earth.
The biggest problem is of course the budget for NASA, which has been steadily decreasing over the last few years. I think it is ludicrous of us to expect NASA to launch missions like these without a proper budget for development, manufacturing, and testing of the equipment. In this case, apparently the money was simply not there to afford a complete test of the whole landing proceedure - which would have made it obvious that the engines were not going to work and that the censors for shutting them off would not work either
Too right. It was really really stupid on the Mars Polar Lander mission
Then, of course, landing comes closer, and we are glutted with money from NASA. But what's the point? We get 10 sysadmins one month before landing, and we have to spend hours training them up; much more preferable to have had one good one ten months before!
Compare this to the old days of Viking
Still, I have faith in Faster Better Cheaper; we cant afford mistakes like the Mars Orbiter again, where a simple case of bad luck (exploding fuel line) lost the entire mission. Painful as it was to lose the MPL/MCO/DS2 probes, they were cheap and cheerful, and we can throw more of them out to Mars.
To introduce myself, I worked on the Mars Polar Lander project as their Outreach Architect, and did some work with the Ground Data Systems crew.
That report is bull as far as I am concerned. We worked our ARSES off as launch came closer. I have never worked with a brighter, more intelligent, and more optimistic crew than the Mars Polar Lander team that was assembled at MVACS before that fateful landing.
After the Mars Climate Orbiter was lost, we had a huge influx of support from JPL; system admins, scientists, programmers, all drafted in at short notice simply to come in and fill in the budget shortage that we'd been suffering from before.
I somehow don't believe JPL would have thrown everything they had at us, to help us out, if they somehow knew this "secret" that the lander was doomed to fail.
It is possible that such a report could have kept secret by uppermost echelons of NASA, and kept from the team and the JPL management. I find that difficult to believe; I would count any such action as bordering criminal, after the incredible amount of hard work sunk by the staff of the mission before landing.
Really, the feeling of utter disbelief we had when it crashed said it all. Noone really expected it to happen, after all the effort, and it took a long time for some of the crew to come to terms with it.
Also, the Mars missions are underwent a fundamental and deep review of their future after this mission failed; it simply wasn't in NASA's interests to cover this up, since it really would have been a "worstcase" end to their Mars Surveyor 98 missions, with only ONE success (the global surveyor). If they did know about this, it would have made sense for them to come out and say it, and attempt a fix, rather than keep it secret, and throw (useless) resources at the team!
As with all other management, NASA executives can be dumb, but I don't believe they could be that dumb.
Why is this technique so useful compared to session-id tracking? I don't believe that it really is, or why such a fuss should be kicked up about it :
...)
:) Do correct my bullshit if it is that, please.
If I understand it correctly, it simply replaces the session ID normally stored in as a cookie/get-var in the hostname.
This would lead to extremely user-unfriendly domain names, and surely it would really bugger up users trying to bookmark the site (stale sessions could stay in bookmarks for a LONG time).
Also, its simply not as efficient as session IDs, which after one unfriendly GET, tend to store their results in a cookie which is transparently passed around. Surely dynamic DNS would all have to have really low TTLs and generally slow down site access if you have to do a large number of DNS lookups (which can be the slowest stage in an http access cycle?)
As I see it, the only problem with the session-id method is that it complicates serverside scripting, but with simply superb tools like PHPLIB all those details are abstracted away from the user. And also PHP4 has built in session handling to simplify things further. IIS has similar modules for ASP developers, and I'm sure others exist forother scripting languages (mod_perl? dunno
So while this might be of interest to some specific applications, I can't see it revolutionising the whole ecommerce industry with its cunning "new" user tracking system.
But then again, I might be talking bull
First of all, you should investigate using Cascading Style Sheets instead of tag for modern browsers. They are one simple tag and also degenerate gracefully for older browsers (which ignore them).
E.g., if I remember right, you'd use
BODY: { font-family: Arial,Helvetica,sans-serif }
And that overrides the body tag with the font you picked.
There are lots of cross-browser problems with using these unfortunately, but check out some decent sites like WebReference for info on them.
Incidentally, you often have to use lots of font tags if you use tables. A font tag will apply to multiple paragraphs, but not to the cell contents inside tables! You need extra font tags in there for that.
HTH
Heh, why not try to use one of the HTML compression tools which have a short javascript header which unpacks html on the fly? Problem is finding a small enough bit of "header JS", but it might be a cool way to get around the 5k limit while sticking to it.
:)
Look at this freshmeat entry for AlgART HTML Picker to see what I mean
So is the only purpose of the policy to protect against possible future liability then?
Economically, it doesn't make a huge amount of sense, given the cheap mass storage costs and admin costs associated with the policy.
I followed the incredibly interesting link from this article regarding the "Really Bad Attitude" newsgroups that Netscape had setup, and that Microsoft subpeonaed (at http://www.jwz.org/gruntle/rbarip.html).
... how many companies out there do this to avoid liability, or is there a different reason for it?
I noticed this quote :
In hindsight, complying with the company's Document Retention Policy (which at Netscape was basically, ``shred anything within 90 days unless you can't get your job done without it'') might have been a good idea.
How many major companies actually have a policy ilke this for electronic information? Most backups are tape/DLTs which last eternity, and is the only purpose of this policy to prevent liability with stuff lying around?
This sounds like it worked with paper-based archiving systems, where space simply doesn't exist to archive forever, and non-essential documents are destroyed, but none of the people I've done work for have had a similar policy at all.
So the question is
Yeah, the 5% comment came from a random statistic thrown out by one of the major UC colleges, for whom that is a hell of a lot ... smaller colleges will suffer more obviously.
Quote: Think about guns - guns are made to kill people. They are used in crimes every day all over the nation. But still, they are legal, because we recognize that they have legitimate uses. You can restrict the use of guns, and you might even reduce crime, (I don't actually believe that, but that's an entirely different story)
Well, I'm from Ireland, guns are banned here in Europe, crime is lower, but you're right, that's another story. Not a good analogy with MP3s I believe though.
but you'd be giving up a portion of your freedom to do that. Similarly, you can ban mp3s, and you may reduce the amount of IP theft, or lower your bandwidth utilization, but you're giving something up. Namely, the positive aspects of downloading mp3s, and also, the students freedom to be in an environment that allows them to expand themselves as they see fit, not as the university sees fit.
You cannot condone freedom to commit illegal acts. You've already given up your freedom to kill your fellow human beings, and indeed, be protected againts such harm, by living under law. Are you the worse for it?
I can kinda sympathize with the bandwidth argument, but I really hate it when people change things midstream. If you come onto the university network, and sign an agreement saying "by signing this you agree not to do x, y, or z" then you have a choice, and you can go elsewhere. On the other hand, if you invest in a NIC for your student network, and then have regulations piled on you never agreed to, that's different, because you weren't given a choice.
The agreements haven't been changed at all. When I got net access from my University (Imperial College, London), there was a very very clear AUP that stated that illegal activities on their network are, you guessed it, ILLEGAL. Of course, uni's do overlook a number of illegal activities, but when those activities start to compromise your bandwidth flow, how can you expect them NOT to take action?!
If the trade of pirated/illegal MP3's is so rampant on a college campus, why not monitor the traffic, collect data, and then bust the offenders? You do this enough times, and people will start to realize they are being watched and will be caught.
:-)
Effort. It takes a lot more effort. It's simple enough just to add a line into a portblocking table to stop it. And you'd have to monitor continuously, keep the tools uptodate with the protocol, and if you slack off, word would get around and mp3 trade would start again. These are students we're talking about
Quote: And another point. Banning Napster doesn't prevent people from pirating MP3's. Banning illegal MP3's will ban illegal MP3's. Funny how that works. It takes vigilance to stop crime, not another line in the old port filter.
There's principle, and there's being practical.
Me standing up in University and screaming "I BAN MP3S!" wouldn't do shit. Me blocking the tool that a lot of students use, which would cut down the usage drastically, would. And remember, almost all acceptable usage policies ban copyright infringments and other illegal activities using that network, so illegal MP3s are banned from the start.
I just can't see a line of conscientious students poring through their AUP's and then saying "Sorry Jim, I can't accept those files, they're illegal!"
Ban the popular tool. Easy enough to get around of course, but it would stop the 95% non-tech-savvy ones anyway.
Well, Napster is clearly a software product that is used almost exclusively to pirate MP3s, it leeches up a lot of bandwidth, and has some pretty bad security holes that means that potentially the client can be forced to serve up any file to a malicious Napster server.
What's wrong with universities banning it? Clearly, they cannot condone the flagrant ripping of MP3's on campus. I'm sure they'd do the same if they found 10000 warez sites running on their students' boxes.
I'm not saying that censorship is a good thing, just that in this case I fail to see how this is construed as censorship, given that using the application for anything other than illegal activities is fairly hard. 5% of university bandwidth is a hell of a lot as well.
No, it's not absolutely neccessary, since they send IF_MOD_SINCE requests, and don't always retrieve the data, etc.
But still, it's a good matter of principle to do so, to guarantee behaviour of clients.
Also, this only extends to private caches, and public caches really like to see http/1.1 headers before holding on to them (again, depends on the exact cache obviously)
Why compare it to a ridiculously expensive Thinkpad? My Sony Vaio F360 cost around $2600 and has a :
... bit chicken and egg though.
PII-400
128mb RAM
12gb Drive
DVD 4x
Firewire, and USB
Built in modem
14.1 visible screen
.. etc
and runs Windows and Linux like a dream. I must say I can't think of any killer apps that the Mac can run that those two OS's cant (even Photoshop is nicer on Windows now than the Mac, and the GIMP rules Linux)
Maybe mass-marketing the Macs would drive prices down
Well, the expat library already exists and seems to be quite defacto under Unix.
/etc filesystem depending on it. What do you do when it breaks? It's like screwing up your LD.so.1 ... ow :)
At least, PHP and Apache use it, and well. And a couple of other utilities and CPAN modules. I've come across. It's fast, small, and not full of unneccessary crap.
Anyway, utilities like Apache and ProFTPD already have meta XML config scripts, and a fair few perl scripts that make use of XML::* as well.
I don't think I like the idea of having a dynamic XML library, and your entire
Still, a standard, easy to use DOM for utilities would be nice, perhaps some kind of extension of the GetOpt stuff would work? It's succeeded largly because it's there and easy to use, and an XML standard would have to do the same.
For those of you interested in caching and how it can help large scale sites, I've helped co-author a technical report with Network Appliance, which was our experiences at accelerating the Mars Polar Lander website. That site used NetCache boxes, simple HTTP/1.1 cache-control headers, and a bit of cunningness to allow user-level tracking without letting the track requests filter through. As traditional, the site had a couple of problems which we've also included in the appendix after we fixed them, to hopefully save other people the same hassles in the future.
:-)
The technical report can be found at http://www.netapp.com/tech_library/307 1.html
We would all save a scary amount of bandwidth if more sites were designed with public caches such as (the awesome) squid in mind, and it's a really simple use of headers that make it possible.
For those who use Apache and are interested in making your own sites more cache-friendly, I recommend you look at mod_expires, which is part of the default distribution of Apache, although not compiled in by default. If you have large, static images that rarely change, then go ahead and put week-month-year long expiry headers on them, and watch the hits for those redundant images drop right down on your web server. And if you suddenly need to change them, then it's no real problem, as all you have to do is change the images URL and it will become a "new" entity for purposes of caching.
Yeah, granted, bandwidth is getting cheaper now, but for us poor Europeans, it's still a scarce commodity and we need to worry about these things
-anil-
Agreed, my comment was posted in haste ...
... free-for-non-commercial-use license perhaps?
Perhaps a special case could be worked out for developers not intending to make a profit from their development resources
-anil-
Although it smacks of corporate greediness, and it is apparent that Linux developers who work for free can't join the members list, why can't the corporate Linux companies, such as Red Hat, Caldera or VA Linux systems join, and make their copies of the specifications available to developers. There may be copyright issues, but if the USB developers work "for" one of these companies, then surely there is no problem. The end-result would still be GPLed, and available to all, it's just a nominal shift in who you are working for.
Linux has a number of high-profile companies now, and it's in their interests to ensure the best hardware compatibility for their distributions. $2500 is a small price to pay for that from their point of view, surely.
-anil-
If you are concerned about security and reliability, then BIND may not be the best solution for you. It has a history of security exploits, and in general the codebase is a complete mess.
Everyone's favourite daemon-revamper Dan Bernstein has an excellent DNS suite of tools called DNSCache which you can obtain from http://cr.yp.to/dnscache.html. This looks pretty much ultrasecure (as most of his stuff is), and also very robust and modular. In addition, you don't need to muck around with zone files and so on, and the interface for adding information is ridiculously simple thanks to some scripts in the package.
For example, to add a new alias to an already existing domain, all I have to do is :
./add-alias new.recoil.org 1.2.3.4
make
and thats it! Serial numbers etc all taken care of automatically for you.
I've been running and playing with DNSCache and can recommend to anyone looking for an alternative to BIND.
Anil
I used to work for the Mars Polar Lander project, on its Ground Data Systems. One thing about this announcement is that its only making official what already happens a lot in NASA.
:)
Internally, we used a lot of Linux boxes, and even more GNU tools on Solaris boxes. In fact, the whole image processing pipeline was built using a series of BASH scripts by the NASA AMES team, as part of their operations deployment.
The web servers for the MPL were Solaris x86 boxes driven by Apache and PHP, while the support boxes for DNS and mail were Linux (red hat, stripped down). All ultra-reliable, getting them secure was the hardest bit of the whole operation really (we considered OpenBSD for a while, but didnt have enough inhouse skills with it)
And, to reassure you, no Windows servers were used for anything other than looking pretty.
Oh, and Quake III. None of the Linux boxes had 3D cards
Anil Madhavapeddy
Hi, I'm part of the MPL team, and thought you guys might like to know that we are taking the flight-spare MPL to a secret location in Death Valley, and running a complete simulation of operations as they might have been had it landed on Mars.
The point? To test the instruments and make sure they would have done what they are meant to, as a number will be reused in future missions (the Stereo Surface Imager in particular).
You can follow all the images and simulations if you're interested in what actual science was being conducted, at our official site www.marspolarlander.com
Cheers,
Anil Madhavapeddy - Ground Data Systems - Outreach
Mars Volatiles and Climate Surveyor
http://www.marspolarlander.c om/overview/finalsite.html
--
Anil Madhavapeddy, anil@mars.ucla.edu
Outreach Architect, Mars Polar Lander, UCLA
Um. NASA _is_ landing a space craft on Mars. These are just two spin offs that release during the start of the landing sequence.
Anil Madhavapeddy
Mars Volatiles and Climate Surveyor
We are using a combination of Solaris x86 boxes and Linux boxes to run the website, along with NetApp cache boxes and load balancing DNS around various hub points to handle the majority of hits.
... no hint of x86 procs there :-)
The actual Ground Data System uses Sun (sparc, ultra, etc) boxes, a lot of heavy SGIs and so on
Anil Madhavapeddy - anil@mars.ucla.edu
Mars Polar Lander Project
http://www.marspolarlander.com/
The official Mars Polar Lander website is at http://www.marspolarlander.com/
It's been pretty well publicised this last week in press releases and stuff.
Anil Madhavapeddy, anil@mars.ucla.edu