If I want to actually ship a working NeWS, I'll start with the original code. If it's too unusable, so be it
Your choice. But when you consider that you have no experienced individuals who know the code, it practically guarantees project failure.
Calling it "reference code", and making that a differentiating point, is dishonest. Rewrite it, use it as is, look at it and throw it out in disgust--all of those things are valid responses in what I'm stating, which is that I'll use the existing code.
It is not dishonest. Is Mozilla the original Netscape code? No! They rewrote it to decruftify the entire thing! The stability, poor performance, and other issues with the original, all pointed to a collapsing codebase. Thus it was a rewrite, just not a rewrite from scratch. Calling Mozilla the Netscape codebase is what would be dishonest.
Good God you are thick. Okay, have it your way. Linux is perfect, Linux is right, Linux is GOD. Well, your god at least. Some of us recognize that it could still use improvements. Especially since your idea of choice seems to be to not have any distributions that deviate from the perfection of your GOD. (i.e. fix the issues)
FWIW, I'm not saying that ALL Linux distros should move in a userfriendly direction, you are. I'm only suggesting that Home Desktop-focused move in that direction.
But Linux is your god. All hail Linux. Linux is perfect. Linux is wise. Linux is a stupid Operating System, not anything to worship. Oops, did I say that out loud?
You are DAMNED STRAIGHT I'd start with the 15 year old source code. You're not a programmer, are you?
Yes, I am a programmer. And if you think starting with 15 year old cruftified code is a GOOD idea, do yourself a favor and take a look at SAPDB sometime. Even better, try building it.
Old, unmaintained code can make a good reference and can even have pieces that you can rip wholesale. But the hundreds to thousands of little assumptions may not be correct anymore. This is particularly bad in a lot of business logic I've worked on, where there are thousands of little code patches in for situations that were transient and no longer apply. Rewriting the code (not from scratch) was the best method for decruftifying it.
if my goal is to release a working copy of NeWS (or whatever), I'd start with the original code first.
So, you're absolutely sure you're not going to get bit by assumptions made about the number of colors (Black and White), the resolutions available, the tools available, the processor instructions available, the cache design, the layout of the frame buffer, so on and so forth. You *really* think that 15 year old piece of code would be easier to get working on modern hardware than it would be to rewrite it using the code as a reference? Good luck.
DUH! because distributions are all different. They provide different things for different people.
Okay, good. It's *starting* to percolate through your skull a bit. Let's see if we can help it along.
I use ubuntu because I love the way it looks and feels but I cannot just install kwrite
And why not? Because...
because I do not want to install all the extra kde crap that goes with it. Is this the fault of the folks who make k3b or the folks who support ubuntu?
Ubuntu. If Ubuntu wanted a distro useful to all, they would have a minimum version of all common Linux libs, including the KDE classes. Or they could decide to just say "KDE programs will not be supported in our spec." At which point programs intended for KDE would need to be statically linked or distributed with the KDE libs if they are to work on Ubuntu.
But the common Linux choice is "none of the above." The user can install any mishmash of libraries and versions that he chooses, making "Ubuntu" a useless definition for a base of standard APIs. The fact that the user can destroy their system by installing a new version of GLib as required by program XYZ that they really want, is not a good thing either.
On what version of windows did you install videolan? XP? 2000? Does it work in windows 98?
Windows 2000. Not that it matters. Does Gnome 2.5 work on a vanilla RedHat 5.0 distribution? No? That's because you're confusing versions (i.e. a minimum platform level) with a base of standards for a given platform. APIs should move with the platform. As a result, the application developer can then say, "This program requires Ubuntu v5.2 and up" as opposed to, "This program requires a Linux distribution with Kernel 2.4, KDE 3.4, GLib 2.2, GCC 2.95, etc, etc, etc." At which point the user is supposed to track down each and every dependency. NO! If a dependency isn't offered by the OS, it needs to be included in the install.
Different distributions need support. If videolan doesn't have the time to do this for you and you refuse to learn to DIY, then look around.
I'm sorry, but what home user has time to mess around with trying to track down every last dependency? I know what I'm doing, and yet it was a PAIN in the ass trying to find every dependency that VLC required. Dozens of sites, one for each RPM. Not to mention that some of them where binary, some were source, and some I found in one form and had to use the other either because the package manager rejected it, or the source wouldn't compile. That's just stupid. A home user wants to click, click, done. If he runs into the problem like you have with Alice, he's going to do what anyone should do:
RETURN IT TO THE STORE
Pain and anguish can happen on all systems. But on the Linux Desktop it is treated as the norm.
Seriously, if you can muster a greenhouse that simple and effective, it could also be used to grow chickens
Again the pressure could be a problem. If you can help it, you don't want to be placing 10 times the inner pressure on the wall as is received by the outer wall. That also places extra stresses on the buried ring stiffener. But if necessary, it is doable.
Once that greenhouse has been in business for a few months, start hatching chickens and watch the perpetual protein (and fertilizer) machine rev up.
Not a bad idea, actually. It may make more sense to have a fully pressurized coop, then move the materials back and forth (as it were). Especially since loose chickens could potentially damage the tarp.
BTW, sorry dude, but you're going to have a hard time getting your C10H14N2 fix on Mars. The potatos might have a little, but the prime crops are unlikely to grow.;-)
A lot of engineering and a lot of failures went into creating the Saturn V, and disregarding all of it without even a glance seems like hubris to me.
That's sort of my point, though. The Saturn V design is still very interesting to engineers as a reference, but rebuilding the Saturn V itself is a bad idea. A bit like NeWS still being intersting for historical purposes, but not directly useful today.:-)
Not so long. I'll still be alive in two decades.:)
Oh good, you'll be just in time to see it being constructed.;-)
Hey, it's Hisham! Glad to have you in on the discussion!
Mod at will
Which is to say, mod parent up for being an expert with an on-topic opinon.;-)
I've looked at some of the work you guys have been doing, and I think you're on the right track. I have a few more thoughts that are going to be in my next blog (which I'm not going to replicate here, because that's a LOT of typing), but you guys are at least getting close from a source perspective.
For those who don't know what GoboLinux does, they break up programs into/Programs/$NAME_OF_PROGRAM structure. After that, symlinks are used in various places to maintain system wide compatibility. It's very interesting work, and is explained in more detail in the FAQ.
In my experience, potatos will happily grow even with little to no light. My family used to keep large bags of potatos in the storage room in the basement, and pull them out as needed. Oddly, the potatos would occasionally sprout some pretty impressive roots.
Anecdotal, I know, but it seems that potatos hold up pretty well in low light.:-)
If the folks at ubuntu or mandrake or wherever would take the time to package it then it wouldn't be this mess it is.
And this is as poor of an argument as ever. Why should the user be restricted to only software officially supported by a Linux distro? VLC isn't supported by Microsoft or Apple. VideoLan provides it. Why should I have to ask Ubuntu, Mandrake, Fedora, SuSE, Uncle Bob's Linux, or any other distro to compile and repackage every piece of software in existance?
And what if I *do* want to run a commercial piece of software? Wait, we're supposed to attack them for not giving it to us at no cost, then make them release the source for Football Linux, PenguiHeads Linux, and Turbine Linux to repackage.
Wouldn't it be a smidge easier to allow the program writer to target a given set of APIs, and place his program in a place where it's unlikely to interfere with other programs?
It took me about a month to get used to using linux after several years of sucking on the windows teat. Now every time one of my friends asks me for help I just chuckle about how easy it would be to fix their problem if only they had a decent command line or a decent set of tools installed by default.
I'm long past that stage. I just tell everyone to get a Mac. That way they can manage their own computer, and if something *does* go wrong (which is pretty rare on Macs), I have a command line in front of me.
Most of the public doesn't care enough to go through the effort of "upgrading?" to a less OS.
Of course they don't. The thing about illegal competition is that it's illegal because it's an abuse of your position. The customer feels that he's getting a deal, but that's not the point.
Taking the common gas station example, customers get really happy when one gas station underprices another by $.50. The fact that this is hurting the consumer in the long run (less competition) escapes that customer. He just wants cheap gas.
The same is true of Microsoft. The fact that they effectively put Real and Netscape out of business is the real point, not what the consumer feels. As a result, this EU decision is weak at best.
Go install VLC on say... SuSE. FEEL the pain and anguish of trying to install something and having the system reject it. Then FEEL the pain of trying to do all the updates through source compiles. FEEL the pain of tracking down every dependency. FEEL the pain of having the program not work even after you compile it. FEEL the frustration of having no menu icon for the program, and having no easy method for setting up a menu icon. FEEL the angst as you have to browse through thousands of little/bin programs looking for VLC to make an association.
That is the pain I've gone through (and every user goes through) every time they try to install something outside of what the distribution provides. You know how I install VLC on Windows? 1)Run Installer. 2)Run Program. You know how I run it on my Mac? 1)Download and Run program.
So don't give me any fanatic crap about "it's all just the same LALALALALALA!" Pull your fingers out of your ears and PAY ATTENTION. Issues exist that are blocking Linux from being a viable desktop platform. If you'd rather ignore them, fine. But don't even TRY to tell me that it's all in my head.
If you really wanted Linux adopted in the home, then you'd recognize where issues exist, and realize that good solutions are being worked on.
Unfortunetly, all plants grown on Mars will still be freeze-dried before eaten.
*chuckle* No, they really are thinking about growing the food.:-)
The article was interesting (despite the fact that the ESA seems to be already picking out foods for a mission they don't yet have), but I would have liked to know more about how they planned to grow food on Mars. For example, the soil samples seem to tell us all kinds of different things about the actual composition of Martian soil. Have we found a concensus on what materials we'll need to bring to grow plants. Last I heard, nitrogen was going to be the biggest issue.
The other thing I'd like to see is someone actually developing a Martian greenhouse design. However you make it, you'll want the greenhouse to be light, portable, and easy to setup. My current thoughts are that a transparent, inflatable tarp would do the trick. We'd first need to know what the minimum pressure is that the plants require before we design the tarp. Hopefully, they can survive in pressures similar to Mars's surface. That knowledge could then be used to develop a greenhouse that works like this:
1. The tarp would be planted into the ground. Depending on the pressure required, it could either be nailed in with stakes (how primitive, but effective) or a stiff ring could be buried into the ground, thus creating an airtight seal.
2. CO2 could then be pumped from the surrounding atomosphere into the greenhouse. Depending on the plant, a certain amount of oxygen may need to be initially pumped in.
3. The pump system should move air in and out of the tarp area. Oxygen would be separated out, and replacement CO2 would be pumped from outside.
So far, so good. But then what about solar energy? Does enough energy reach Mars' surface to support these plants? Does artifical lighting need to be added? (I guess that's why they went with potatos. Little to no light necessary.)
Your post, I'm sorry, shows lousy reading comprehension.
While software and computer hardware have changed in the past 15 years, the laws of physics haven't.
What's your point? Computer hardware is subject to the same laws of physics as hardware. Are you saying that just because the laws of physics are the same that it's just as easy now to build a SPARCStation-1 now as it was back then? Nevermind that Sun doesn't produce the chips, the motherboard, the memory, the connectors, the keyboards, the monitors, the disks, or the cases any more. That's all irrelivant. We've got the software!
We have on one hand, the blueprints for a tested launch vehicle, essentially a turnkey space program that meets our needs.
No, we have the absolute equivalent of crufty code. None of the parts are made any longer, the original Saturn V team is retired or dead, and we have zero remaining experience with the electronics and mechanical design used in the original rocket. Even if we could just throw another one together, we have no idea whether it would work right or not. Many of the design parameters took issues into account that noone knows about/understands any longer.
For example, did you know that the wiring of the Saturn V control electronics was designed to take into account the delays introduced by the primitive wiring? Even if we could recreate the larger transistors and wiring used back then, many of our processes have improved to the degree to where it would throw off the entire timing of the rocket! i.e. Our Saturn V 1/2 would be more likely to fly up and tip over than it would be to follow a straight path. All because the engines on one side prefired according to a timing that's no longer valid!
How many years and how many needless billions will it take this time to reach a solution which may only be 10% more efficient?
One hell of a lot less than it would cost to start over with the base Saturn V concept in mind. Do you like starting with old code that you feel like you could rewrite in far less time than it takes to update it? e.g. I tried getting the original adventure shell to run on a modern SH shell a short time ago. Wouldn't work. As I tried to fix it, I realized that the syntax had changed to the degree to where it was easier to rewrite it using the original as a reference than it was to debug it.
A far better solution would be to use the proven Saturn V (or even the Energia) boosters while we invest seed money in space elevator related research.
The Energia *is* a realistic possibility for now. As long as we don't wait too much longer, Russia is still capable of retooling to build the rockets. Plus, much of their experience is still intact.
As for the Space Elevator, don't hold your breath. We've got a LONG way to go before such a beast can be constructed. It's not that it's a bad idea, it just isn't here quite yet.
Once we have commercial semi-orbital flight, it won't be much of a leap to orbital commerce.
*sigh* There's a HUGE difference between sub-orbital and orbital. Orbital flight has a Delta-V that is extremely difficult for all but the most powerful rockets to achieve. Sub-orbital has a Delta-V that is insignificant in comparison.
Go read Rei's writeup on this, and I think you'll understand the issue.
You have not heard of Lycoris before because they are indeed not a major Linux player. I used them for a while, but the hardware support was lacking.
I don't know if anyone else here remembers it, but I loved the LinuxWorld "review" of Lycrois. It was two or three paragraphs long and went something like this:
"I installed Lycoris, a Linux billed as an XP clone. My screen tunred a weird shade of green when using the Desktop. That sucked. Fin."
After reading that, I was just standing there with a horrified look on my face, then started muttering "but, but, he didn't even TRY to fix the issue, or even tell us anything ABOUT the OS!"
Ah, here it is. Gotta love journalistic integrity.:-)
Re:I'm all for science/technology/astronomy but...
on
Back to Moon in 2015?
·
· Score: 1
Ok I read the wiki article, and it was interesting, although for me it raised more questions than it answered.
Most of your questions seem to be related to a misunderstanding of how reactors work. Very quickly:
1. The reaction is initiated as soon as fuel is ready to be pumped. The fuel is not heated as a whole, but rather heated as it passes through the throat of the exhaust channel.
2. Nuclear rockets are not magical. They can only run for as long as they have fuel. Existing chemical rockets are sufficient for interplanetary travel, but nuclear rockets produce more power. That power == faster travel and more fuel reserves.
3. You can, and do, simply turn off the reaction. Any reactor that lacks the ability to control the reaction would be likely to enter a prompt-critical stage, followed by a melt-down or super-critical reaction. As for slowing down, you do it like any other rocket does it. Turn and burn.
4. This is only a concern if you're using nuclear rockets for a launch. In the case of a launch, your best bet is a flight profile over the ocean (as is done today), plus some protective material (which is required by the engine anyway). Should something go wrong, the entire assembly should fall into the ocean, cool, and become harmless until it can be retrieved. (We've got a few sunken nuclear subs to attest to the fact that reactors in the ocean don't seem to be much of a problem.) In the case of Spiral Two, however, the engines would be likely to stay on orbital trajectories at all times. Which means that they *can't* reenter since they don't have the Delta-V.
FYI, RTGs are currently protected in a case that can survive reentry with no issues. (And it has been tested.) Before this policy was implmented, a couple of RTGs were intentionally (US) and unintentionally (Russia) burned up in the atmosphere with no real effect.
Linux doesn't need consolidation as much as it just needs to address the existing usablility issues. I touched upon this in the Symphony OS story, and I hope to get a new blog entry up on it in the next day or two.
The long and short of it is:
1. The packaging system is user-unfriendly. 2. The locations of programs are user-unfriendly. 3. The folder layout of Linux systems is user-unfriendly. 4. The lack of a standard base of installed libraries is application (and thus user) unfriendly.
If this can at least be solved at the distribution level, then we'll be good to go. But right now a given distribution means different things to different people depending on what packages are installed.
(P.S. Speaking of my blog, I get a kick out of the fact that the story I submitted on my last entry is still pending. Since Saturday. Guess the editors just think it's cool to look at or something.:-P)
Realistically you'll need to move small loads into and out of space to swap out personnel, replacement parts and tools, etc. You can't launch a Saturn V, or even a shuttle, just to bring Bob the plumber and the length of pipe you forgot up to the construction site.
1. Most lunar base plans call for mining and construction facilities. Self-sufficiency is the key.
2. CEV Spiral One is intended to be a (relatively) inexpensive method for launching personnel. Leave the supplies to the big rockets, and the people to the little ones.
3. I have little doubt that a workable space-plane could be built with sufficient R&D. Such a space plane could lower the cost of space access for people, even if the cost of shipping materials remains high. Sufficient R&D will materialize after sufficient economic incentive exists. Or perhaps Rutan will build it while we're all not looking.:-)
Re:Infrastructure/Building material
on
Back to Moon in 2015?
·
· Score: 3, Interesting
See that little blue dot next to my name? I was into nuclear engines long before I ever found slashdot. However, I don't trust NASA to complete what it starts. I've lived near the Cape for too long.
I realize that we both want the same thing, we just disagree on the approach. Most of the NASA issues I've seen go way back to the moment that Nixon effectively shut down the space program. Any and all money was channelled into the Shuttle program, which was subsequently screwed up by Nixon's demands for a "jack of all trades" vehicle. The engineers did an amazing job on the shuttle, and by all rights its technology and power outstrips that of the Saturn V. The problem is the amount of crap hard-bolted to that technology, and the Carter administration's demands to "stretch out" the costs. The result was a half-rate craft that was unable to complete any mission effectively.
Regean tried to build on the existing investment by supporting NASA's plans for Space Station Freedom, a Lunar Transit, and eventually a Lunar Base. He also pushed NASA to complete the National Space Plane to provide for cheap access to space. But between the Shuttle's downtime after Challenger and Clinton's cutbacks to the space program (which resulted in that piece of shit in the sky known as the ISS), all the original goals of Regean's program have been missed.
Now we're scrapping all of our super-fancy technology and doing it the way we know it can be done. I see this as a *good* thing because it's the only way our space program will progress. Remember, NASA's current budget is being sucked dry by flying and maintaining the space shuttle. At $500 million per launch, it's anything but and effective method for getting to space!
The CEV program (even if only Spiral One is completed) will free up NASA's coffers to do more interesting stuff than sending 104 metric tonnes up and down the gravity well. (Yes, the Space Shuttle weighs that much.) We don't have figures yet, but even at $100 million per launch the CEV will be 5 times as cost effective as the shuttle is today. My guess is that realistic launch costs will settle somewhere closer to $50-$75 million per launch. A significant savings.
Unfortunately, it seems like the capital, either political or monetary, to build them isn't easy to find. To me, that says that the designs aren't really that great. Building prototypes is the most important step of R&D - and I was saying that we need to build better engines before spending $umpteen billion in space.
Now slow down a moment here. Quite a few prototypes *have* been built. The NERVA program was considered successful, and was ready to fly prior to the cancellation of pretty much all space programs. The Orion was prototyped in many forms (you've seen the Put-Put video I assume?) but finally died in the cancellation of the Saturn V program. The linear aerospike engines and hyrdrogen slush technologies were key to the X-33. That program was underfunded and undercommitted to by NASA (as was the Delta Clipper). Most of these engines are developed enough to take a risk on, but the only one you're likely to see in the near term is the Nuclear Thermal Rockets. The rest will wait until we again have aerospace companies fighting to create the best hardware. (They were pretty disillusioned after their treatment in the 90's.)
Other engine concepts:
- Nuclear Salt Water Rocket: Must be tested in space due to the radioactivity of the fuel.
- Gas-Core Nuclear Rocket: Research is progressing, but no working prototype yet exists.
- Antimatter catalyzed engines: This is related to the Orion engine, and cannot be used on Earth due to the nuclear test ban treaty. A mission is already planned, however.
- Ion Engines: These are already used.
- Solar Sails: These have been used on a few test missions.
- M2P2 Solar Sails: Under development. Could be useful for a more powerful Orion design.
- Deadalus: Excellent solution for travel beyond our sol
Re:I'm all for science/technology/astronomy but...
on
Back to Moon in 2015?
·
· Score: 1
Also look up ION engine.
You dont need thrust all of the time but if you have fuel to burn so to speak. A constant low output can do the job also. Would provide for a simulated gravity for the ship.
You're not going to get much gravity out of an Ion engine. About the thrust equal to the weight of a small stack of paper is about the max you're going to get, even with a nuclear power source.
Nuclear engines *could* do constant thrust, but most aren't designed that way. Running the reactor hot enough for that long tends to wear out the components rather quickly.
I think you'll see Spiral Two initially consisting of nuclear rockets (since the astronauts will need to get to the moon quickly), then you'll later see Ion rockets used for cargo-only runs.:-)
When you reach orbit. You half way to any where in space.
Robert Heinlein. He was more or less correct, too.:-)
While the semiconductor and transistor created a massive new market, the real boon was the ability to miniturize the transistor on a chip. That didn't come until the microprocessor was invented by Intel some 25 years later. And even then, it still took awhile for the technology to catch on in popular computing circles. Mainframes were still built with larger semiconductor technology until the 1980s.
The same thing is true of space access. You'll see existing efficient chemical engines, nuclear thermal engines, and Ion engines used in the CEV program. That will bootstrap space access and lead to the more powerful (but yet undeveloped) Orion, Nuclear Salt Water, Plasma, Daedalus, Antimatter, and other very useful and powerful engines. Don't make the industry wait until the proverbial microprocessor is perfected. The space rush needs to start NOW, or there will be no "hardware boom" like in the 1980s.
I'm not certain you understand the issues. It's not just that the parts don't exist anymore, it's that the factories to make those parts don't exist anymore. To make a nut of the type required by the Saturn V, we'd need to retool an entire factory. Across all the nuts, bolts, sheet metal, piping, and other heavy industry components, it would take an excessive amount of time and money to rebuild the program.
The problem is that US industry has moved on to more advanced materials than those used in the Saturn V program. The result is that it would actually be *cheaper* to build a new super-booster from scratch. (Or pay the Russians to retool for the Energia.)
Think of it this way: NeWS was a powerful windowing system back in the early 90's that put X-Windows to shame. It has since disappeared due to Sun's attempt to keep the technology proprietary. Now if you wanted to revisit the concepts of NeWS, would you start with unmaintained 15 year old code that makes assumptions about machines that don't even exist any longer, or would you reuse the original concept in a new design that would take advantage of the far simpler programming model of today's languages, environments, and hardware?
Cruftiness is a problem in all areas of engineering, not just programming.:-)
Re:I'm all for science/technology/astronomy but...
on
Back to Moon in 2015?
·
· Score: 1
Becoming more of a "do all disposable" similar to the ever reliable Appolo hardware that was quickly adapted to 3 totaly different sets of missions with little effort.
Actually, most of the proposed CEV hardware is reasonably reusable. The booster itself is still lost (not that we've gotten much milage out of the "reusable" SRBs and non-reusable LHOx tank), but the crew vehicle will probably be capable of quick turnarounds. The primary difference between the shuttle and the CEVs is that the CEVs are only a life support and reentry capsule. All the tricky parts of the mechanics (e.g. engines that guzzle a swimming pool per second) are left to the disposable boosters.
Of course, this is only true of Spiral One. Spirals two and up will most likely be 100% reusable, since they will live their entire lifetimes up in space.
I do hear what you're saying, though. Idealogically, a new super-technology (the X-33, Delta Clipper, Space Plane?) should come along and be the Next Big Thing(TM). The problem that we've faced is that the Shuttle has been proven to be a step in the wrong direction. It was a jack of all trades, and a master of none. As a result, ideology must take a back seat to pragmatism.:-)
Re:I'm all for science/technology/astronomy but...
on
Back to Moon in 2015?
·
· Score: 1
In other words, yes you can derive thrust from a neutron stream without using a transfer mass a la NERVA-type vehicles. No clue on efficiency, or ease, but it is definitely within the possibility of physics to channel the neutron output of a reactor into a specific direction and use that output to provide thrust.
As I understand it, a neutron powered rocket is highly infeasible due to the need to maintain a chain reaction, and the inability to control neutrons (no charge). If all the neutrons were channelled out of the rocket, then there would be nothing to maintain the fission reaction. The neutrons that are used to maintain the reaction dump their energy by boucing off of atoms, thus creating massive amounts of thermal energy. Since we don't want to be using the reactor core as a propellant, we have to transfer that thermal energy to some form of working fluid.
At least that's my understanding of why such a rocket wouldn't work. (It would be damn efficient if it could, though.) But now we're off on a tangent.:-)
I have some very fun ideas involving high Isp "excimer plasma rockets" which I'd love to try out.
I think you're still missing my point. You don't want to develop *new* technologies until AFTER you've developed a market using *proven* space technologies. All the engines we could ever want are already on the drawing boards. Even engines for interstellar flight. The key is to get into space first, THEN there will be an economic incentive for R&D.
Slashdot Mirror
If I want to actually ship a working NeWS, I'll start with the original code. If it's too unusable, so be it
Your choice. But when you consider that you have no experienced individuals who know the code, it practically guarantees project failure.
Calling it "reference code", and making that a differentiating point, is dishonest. Rewrite it, use it as is, look at it and throw it out in disgust--all of those things are valid responses in what I'm stating, which is that I'll use the existing code.
It is not dishonest. Is Mozilla the original Netscape code? No! They rewrote it to decruftify the entire thing! The stability, poor performance, and other issues with the original, all pointed to a collapsing codebase. Thus it was a rewrite, just not a rewrite from scratch. Calling Mozilla the Netscape codebase is what would be dishonest.
Good God you are thick. Okay, have it your way. Linux is perfect, Linux is right, Linux is GOD. Well, your god at least. Some of us recognize that it could still use improvements. Especially since your idea of choice seems to be to not have any distributions that deviate from the perfection of your GOD. (i.e. fix the issues)
FWIW, I'm not saying that ALL Linux distros should move in a userfriendly direction, you are. I'm only suggesting that Home Desktop-focused move in that direction.
But Linux is your god. All hail Linux. Linux is perfect. Linux is wise. Linux is a stupid Operating System, not anything to worship. Oops, did I say that out loud?
You are DAMNED STRAIGHT I'd start with the 15 year old source code. You're not a programmer, are you?
Yes, I am a programmer. And if you think starting with 15 year old cruftified code is a GOOD idea, do yourself a favor and take a look at SAPDB sometime. Even better, try building it.
Old, unmaintained code can make a good reference and can even have pieces that you can rip wholesale. But the hundreds to thousands of little assumptions may not be correct anymore. This is particularly bad in a lot of business logic I've worked on, where there are thousands of little code patches in for situations that were transient and no longer apply. Rewriting the code (not from scratch) was the best method for decruftifying it.
if my goal is to release a working copy of NeWS (or whatever), I'd start with the original code first.
So, you're absolutely sure you're not going to get bit by assumptions made about the number of colors (Black and White), the resolutions available, the tools available, the processor instructions available, the cache design, the layout of the frame buffer, so on and so forth. You *really* think that 15 year old piece of code would be easier to get working on modern hardware than it would be to rewrite it using the code as a reference? Good luck.
DUH! because distributions are all different. They provide different things for different people.
Okay, good. It's *starting* to percolate through your skull a bit. Let's see if we can help it along.
I use ubuntu because I love the way it looks and feels but I cannot just install kwrite
And why not? Because...
because I do not want to install all the extra kde crap that goes with it. Is this the fault of the folks who make k3b or the folks who support ubuntu?
Ubuntu. If Ubuntu wanted a distro useful to all, they would have a minimum version of all common Linux libs, including the KDE classes. Or they could decide to just say "KDE programs will not be supported in our spec." At which point programs intended for KDE would need to be statically linked or distributed with the KDE libs if they are to work on Ubuntu.
But the common Linux choice is "none of the above." The user can install any mishmash of libraries and versions that he chooses, making "Ubuntu" a useless definition for a base of standard APIs. The fact that the user can destroy their system by installing a new version of GLib as required by program XYZ that they really want, is not a good thing either.
On what version of windows did you install videolan? XP? 2000? Does it work in windows 98?
Windows 2000. Not that it matters. Does Gnome 2.5 work on a vanilla RedHat 5.0 distribution? No? That's because you're confusing versions (i.e. a minimum platform level) with a base of standards for a given platform. APIs should move with the platform. As a result, the application developer can then say, "This program requires Ubuntu v5.2 and up" as opposed to, "This program requires a Linux distribution with Kernel 2.4, KDE 3.4, GLib 2.2, GCC 2.95, etc, etc, etc." At which point the user is supposed to track down each and every dependency. NO! If a dependency isn't offered by the OS, it needs to be included in the install.
Different distributions need support. If videolan doesn't have the time to do this for you and you refuse to learn to DIY, then look around.
I'm sorry, but what home user has time to mess around with trying to track down every last dependency? I know what I'm doing, and yet it was a PAIN in the ass trying to find every dependency that VLC required. Dozens of sites, one for each RPM. Not to mention that some of them where binary, some were source, and some I found in one form and had to use the other either because the package manager rejected it, or the source wouldn't compile. That's just stupid. A home user wants to click, click, done. If he runs into the problem like you have with Alice, he's going to do what anyone should do:
RETURN IT TO THE STORE
Pain and anguish can happen on all systems. But on the Linux Desktop it is treated as the norm.
Seriously, if you can muster a greenhouse that simple and effective, it could also be used to grow chickens
;-)
Again the pressure could be a problem. If you can help it, you don't want to be placing 10 times the inner pressure on the wall as is received by the outer wall. That also places extra stresses on the buried ring stiffener. But if necessary, it is doable.
Once that greenhouse has been in business for a few months, start hatching chickens and watch the perpetual protein (and fertilizer) machine rev up.
Not a bad idea, actually. It may make more sense to have a fully pressurized coop, then move the materials back and forth (as it were). Especially since loose chickens could potentially damage the tarp.
BTW, sorry dude, but you're going to have a hard time getting your C10H14N2 fix on Mars. The potatos might have a little, but the prime crops are unlikely to grow.
A lot of engineering and a lot of failures went into creating the Saturn V, and disregarding all of it without even a glance seems like hubris to me.
:-)
:)
;-)
That's sort of my point, though. The Saturn V design is still very interesting to engineers as a reference, but rebuilding the Saturn V itself is a bad idea. A bit like NeWS still being intersting for historical purposes, but not directly useful today.
Not so long. I'll still be alive in two decades.
Oh good, you'll be just in time to see it being constructed.
Hey, it's Hisham! Glad to have you in on the discussion!
;-)
/Programs/$NAME_OF_PROGRAM structure. After that, symlinks are used in various places to maintain system wide compatibility. It's very interesting work, and is explained in more detail in the FAQ.
Mod at will
Which is to say, mod parent up for being an expert with an on-topic opinon.
I've looked at some of the work you guys have been doing, and I think you're on the right track. I have a few more thoughts that are going to be in my next blog (which I'm not going to replicate here, because that's a LOT of typing), but you guys are at least getting close from a source perspective.
For those who don't know what GoboLinux does, they break up programs into
In my experience, potatos will happily grow even with little to no light. My family used to keep large bags of potatos in the storage room in the basement, and pull them out as needed. Oddly, the potatos would occasionally sprout some pretty impressive roots.
:-)
Anecdotal, I know, but it seems that potatos hold up pretty well in low light.
If the folks at ubuntu or mandrake or wherever would take the time to package it then it wouldn't be this mess it is.
And this is as poor of an argument as ever. Why should the user be restricted to only software officially supported by a Linux distro? VLC isn't supported by Microsoft or Apple. VideoLan provides it. Why should I have to ask Ubuntu, Mandrake, Fedora, SuSE, Uncle Bob's Linux, or any other distro to compile and repackage every piece of software in existance?
And what if I *do* want to run a commercial piece of software? Wait, we're supposed to attack them for not giving it to us at no cost, then make them release the source for Football Linux, PenguiHeads Linux, and Turbine Linux to repackage.
Wouldn't it be a smidge easier to allow the program writer to target a given set of APIs, and place his program in a place where it's unlikely to interfere with other programs?
It took me about a month to get used to using linux after several years of sucking on the windows teat. Now every time one of my friends asks me for help I just chuckle about how easy it would be to fix their problem if only they had a decent command line or a decent set of tools installed by default.
I'm long past that stage. I just tell everyone to get a Mac. That way they can manage their own computer, and if something *does* go wrong (which is pretty rare on Macs), I have a command line in front of me.
Most of the public doesn't care enough to go through the effort of "upgrading?" to a less OS.
Of course they don't. The thing about illegal competition is that it's illegal because it's an abuse of your position. The customer feels that he's getting a deal, but that's not the point.
Taking the common gas station example, customers get really happy when one gas station underprices another by $.50. The fact that this is hurting the consumer in the long run (less competition) escapes that customer. He just wants cheap gas.
The same is true of Microsoft. The fact that they effectively put Real and Netscape out of business is the real point, not what the consumer feels. As a result, this EU decision is weak at best.
Go install VLC on say... SuSE. FEEL the pain and anguish of trying to install something and having the system reject it. Then FEEL the pain of trying to do all the updates through source compiles. FEEL the pain of tracking down every dependency. FEEL the pain of having the program not work even after you compile it. FEEL the frustration of having no menu icon for the program, and having no easy method for setting up a menu icon. FEEL the angst as you have to browse through thousands of little /bin programs looking for VLC to make an association.
That is the pain I've gone through (and every user goes through) every time they try to install something outside of what the distribution provides. You know how I install VLC on Windows? 1)Run Installer. 2)Run Program. You know how I run it on my Mac? 1)Download and Run program.
So don't give me any fanatic crap about "it's all just the same LALALALALALA!" Pull your fingers out of your ears and PAY ATTENTION. Issues exist that are blocking Linux from being a viable desktop platform. If you'd rather ignore them, fine. But don't even TRY to tell me that it's all in my head.
If you really wanted Linux adopted in the home, then you'd recognize where issues exist, and realize that good solutions are being worked on.
Unfortunetly, all plants grown on Mars will still be freeze-dried before eaten.
:-)
*chuckle* No, they really are thinking about growing the food.
The article was interesting (despite the fact that the ESA seems to be already picking out foods for a mission they don't yet have), but I would have liked to know more about how they planned to grow food on Mars. For example, the soil samples seem to tell us all kinds of different things about the actual composition of Martian soil. Have we found a concensus on what materials we'll need to bring to grow plants. Last I heard, nitrogen was going to be the biggest issue.
The other thing I'd like to see is someone actually developing a Martian greenhouse design. However you make it, you'll want the greenhouse to be light, portable, and easy to setup. My current thoughts are that a transparent, inflatable tarp would do the trick. We'd first need to know what the minimum pressure is that the plants require before we design the tarp. Hopefully, they can survive in pressures similar to Mars's surface. That knowledge could then be used to develop a greenhouse that works like this:
1. The tarp would be planted into the ground. Depending on the pressure required, it could either be nailed in with stakes (how primitive, but effective) or a stiff ring could be buried into the ground, thus creating an airtight seal.
2. CO2 could then be pumped from the surrounding atomosphere into the greenhouse. Depending on the plant, a certain amount of oxygen may need to be initially pumped in.
3. The pump system should move air in and out of the tarp area. Oxygen would be separated out, and replacement CO2 would be pumped from outside.
So far, so good. But then what about solar energy? Does enough energy reach Mars' surface to support these plants? Does artifical lighting need to be added? (I guess that's why they went with potatos. Little to no light necessary.)
This is, I'm sorry, a lousy analogy.
Your post, I'm sorry, shows lousy reading comprehension.
While software and computer hardware have changed in the past 15 years, the laws of physics haven't.
What's your point? Computer hardware is subject to the same laws of physics as hardware. Are you saying that just because the laws of physics are the same that it's just as easy now to build a SPARCStation-1 now as it was back then? Nevermind that Sun doesn't produce the chips, the motherboard, the memory, the connectors, the keyboards, the monitors, the disks, or the cases any more. That's all irrelivant. We've got the software!
We have on one hand, the blueprints for a tested launch vehicle, essentially a turnkey space program that meets our needs.
No, we have the absolute equivalent of crufty code. None of the parts are made any longer, the original Saturn V team is retired or dead, and we have zero remaining experience with the electronics and mechanical design used in the original rocket. Even if we could just throw another one together, we have no idea whether it would work right or not. Many of the design parameters took issues into account that noone knows about/understands any longer.
For example, did you know that the wiring of the Saturn V control electronics was designed to take into account the delays introduced by the primitive wiring? Even if we could recreate the larger transistors and wiring used back then, many of our processes have improved to the degree to where it would throw off the entire timing of the rocket! i.e. Our Saturn V 1/2 would be more likely to fly up and tip over than it would be to follow a straight path. All because the engines on one side prefired according to a timing that's no longer valid!
How many years and how many needless billions will it take this time to reach a solution which may only be 10% more efficient?
One hell of a lot less than it would cost to start over with the base Saturn V concept in mind. Do you like starting with old code that you feel like you could rewrite in far less time than it takes to update it? e.g. I tried getting the original adventure shell to run on a modern SH shell a short time ago. Wouldn't work. As I tried to fix it, I realized that the syntax had changed to the degree to where it was easier to rewrite it using the original as a reference than it was to debug it.
A far better solution would be to use the proven Saturn V (or even the Energia) boosters while we invest seed money in space elevator related research.
The Energia *is* a realistic possibility for now. As long as we don't wait too much longer, Russia is still capable of retooling to build the rockets. Plus, much of their experience is still intact.
As for the Space Elevator, don't hold your breath. We've got a LONG way to go before such a beast can be constructed. It's not that it's a bad idea, it just isn't here quite yet.
Once we have commercial semi-orbital flight, it won't be much of a leap to orbital commerce.
*sigh* There's a HUGE difference between sub-orbital and orbital. Orbital flight has a Delta-V that is extremely difficult for all but the most powerful rockets to achieve. Sub-orbital has a Delta-V that is insignificant in comparison.
Go read Rei's writeup on this, and I think you'll understand the issue.
You have not heard of Lycoris before because they are indeed not a major Linux player. I used them for a while, but the hardware support was lacking.
:-)
I don't know if anyone else here remembers it, but I loved the LinuxWorld "review" of Lycrois. It was two or three paragraphs long and went something like this:
"I installed Lycoris, a Linux billed as an XP clone. My screen tunred a weird shade of green when using the Desktop. That sucked. Fin."
After reading that, I was just standing there with a horrified look on my face, then started muttering "but, but, he didn't even TRY to fix the issue, or even tell us anything ABOUT the OS!"
Ah, here it is. Gotta love journalistic integrity.
Ok I read the wiki article, and it was interesting, although for me it raised more questions than it answered.
Most of your questions seem to be related to a misunderstanding of how reactors work. Very quickly:
1. The reaction is initiated as soon as fuel is ready to be pumped. The fuel is not heated as a whole, but rather heated as it passes through the throat of the exhaust channel.
2. Nuclear rockets are not magical. They can only run for as long as they have fuel. Existing chemical rockets are sufficient for interplanetary travel, but nuclear rockets produce more power. That power == faster travel and more fuel reserves.
3. You can, and do, simply turn off the reaction. Any reactor that lacks the ability to control the reaction would be likely to enter a prompt-critical stage, followed by a melt-down or super-critical reaction. As for slowing down, you do it like any other rocket does it. Turn and burn.
4. This is only a concern if you're using nuclear rockets for a launch. In the case of a launch, your best bet is a flight profile over the ocean (as is done today), plus some protective material (which is required by the engine anyway). Should something go wrong, the entire assembly should fall into the ocean, cool, and become harmless until it can be retrieved. (We've got a few sunken nuclear subs to attest to the fact that reactors in the ocean don't seem to be much of a problem.) In the case of Spiral Two, however, the engines would be likely to stay on orbital trajectories at all times. Which means that they *can't* reenter since they don't have the Delta-V.
FYI, RTGs are currently protected in a case that can survive reentry with no issues. (And it has been tested.) Before this policy was implmented, a couple of RTGs were intentionally (US) and unintentionally (Russia) burned up in the atmosphere with no real effect.
Linux doesn't need consolidation as much as it just needs to address the existing usablility issues. I touched upon this in the Symphony OS story, and I hope to get a new blog entry up on it in the next day or two.
:-P)
The long and short of it is:
1. The packaging system is user-unfriendly.
2. The locations of programs are user-unfriendly.
3. The folder layout of Linux systems is user-unfriendly.
4. The lack of a standard base of installed libraries is application (and thus user) unfriendly.
If this can at least be solved at the distribution level, then we'll be good to go. But right now a given distribution means different things to different people depending on what packages are installed.
(P.S. Speaking of my blog, I get a kick out of the fact that the story I submitted on my last entry is still pending. Since Saturday. Guess the editors just think it's cool to look at or something.
Realistically you'll need to move small loads into and out of space to swap out personnel, replacement parts and tools, etc. You can't launch a Saturn V, or even a shuttle, just to bring Bob the plumber and the length of pipe you forgot up to the construction site.
:-)
1. Most lunar base plans call for mining and construction facilities. Self-sufficiency is the key.
2. CEV Spiral One is intended to be a (relatively) inexpensive method for launching personnel. Leave the supplies to the big rockets, and the people to the little ones.
3. I have little doubt that a workable space-plane could be built with sufficient R&D. Such a space plane could lower the cost of space access for people, even if the cost of shipping materials remains high. Sufficient R&D will materialize after sufficient economic incentive exists. Or perhaps Rutan will build it while we're all not looking.
See that little blue dot next to my name? I was into nuclear engines long before I ever found slashdot. However, I don't trust NASA to complete what it starts. I've lived near the Cape for too long.
I realize that we both want the same thing, we just disagree on the approach. Most of the NASA issues I've seen go way back to the moment that Nixon effectively shut down the space program. Any and all money was channelled into the Shuttle program, which was subsequently screwed up by Nixon's demands for a "jack of all trades" vehicle. The engineers did an amazing job on the shuttle, and by all rights its technology and power outstrips that of the Saturn V. The problem is the amount of crap hard-bolted to that technology, and the Carter administration's demands to "stretch out" the costs. The result was a half-rate craft that was unable to complete any mission effectively.
Regean tried to build on the existing investment by supporting NASA's plans for Space Station Freedom, a Lunar Transit, and eventually a Lunar Base. He also pushed NASA to complete the National Space Plane to provide for cheap access to space. But between the Shuttle's downtime after Challenger and Clinton's cutbacks to the space program (which resulted in that piece of shit in the sky known as the ISS), all the original goals of Regean's program have been missed.
Now we're scrapping all of our super-fancy technology and doing it the way we know it can be done. I see this as a *good* thing because it's the only way our space program will progress. Remember, NASA's current budget is being sucked dry by flying and maintaining the space shuttle. At $500 million per launch, it's anything but and effective method for getting to space!
The CEV program (even if only Spiral One is completed) will free up NASA's coffers to do more interesting stuff than sending 104 metric tonnes up and down the gravity well. (Yes, the Space Shuttle weighs that much.) We don't have figures yet, but even at $100 million per launch the CEV will be 5 times as cost effective as the shuttle is today. My guess is that realistic launch costs will settle somewhere closer to $50-$75 million per launch. A significant savings.
Unfortunately, it seems like the capital, either political or monetary, to build them isn't easy to find. To me, that says that the designs aren't really that great. Building prototypes is the most important step of R&D - and I was saying that we need to build better engines before spending $umpteen billion in space.
Now slow down a moment here. Quite a few prototypes *have* been built. The NERVA program was considered successful, and was ready to fly prior to the cancellation of pretty much all space programs. The Orion was prototyped in many forms (you've seen the Put-Put video I assume?) but finally died in the cancellation of the Saturn V program. The linear aerospike engines and hyrdrogen slush technologies were key to the X-33. That program was underfunded and undercommitted to by NASA (as was the Delta Clipper). Most of these engines are developed enough to take a risk on, but the only one you're likely to see in the near term is the Nuclear Thermal Rockets. The rest will wait until we again have aerospace companies fighting to create the best hardware. (They were pretty disillusioned after their treatment in the 90's.)
Other engine concepts:
- Nuclear Salt Water Rocket: Must be tested in space due to the radioactivity of the fuel.
- Gas-Core Nuclear Rocket: Research is progressing, but no working prototype yet exists.
- Antimatter catalyzed engines: This is related to the Orion engine, and cannot be used on Earth due to the nuclear test ban treaty. A mission is already planned, however.
- Ion Engines: These are already used.
- Solar Sails: These have been used on a few test missions.
- M2P2 Solar Sails: Under development. Could be useful for a more powerful Orion design.
- Deadalus: Excellent solution for travel beyond our sol
Also look up ION engine.
:-)
:-)
You dont need thrust all of the time but if you have fuel to burn so to speak. A constant low output can do the job also. Would provide for a simulated gravity for the ship.
You're not going to get much gravity out of an Ion engine. About the thrust equal to the weight of a small stack of paper is about the max you're going to get, even with a nuclear power source.
Nuclear engines *could* do constant thrust, but most aren't designed that way. Running the reactor hot enough for that long tends to wear out the components rather quickly.
I think you'll see Spiral Two initially consisting of nuclear rockets (since the astronauts will need to get to the moon quickly), then you'll later see Ion rockets used for cargo-only runs.
When you reach orbit. You half way to any where in space.
Robert Heinlein. He was more or less correct, too.
While the semiconductor and transistor created a massive new market, the real boon was the ability to miniturize the transistor on a chip. That didn't come until the microprocessor was invented by Intel some 25 years later. And even then, it still took awhile for the technology to catch on in popular computing circles. Mainframes were still built with larger semiconductor technology until the 1980s.
The same thing is true of space access. You'll see existing efficient chemical engines, nuclear thermal engines, and Ion engines used in the CEV program. That will bootstrap space access and lead to the more powerful (but yet undeveloped) Orion, Nuclear Salt Water, Plasma, Daedalus, Antimatter, and other very useful and powerful engines. Don't make the industry wait until the proverbial microprocessor is perfected. The space rush needs to start NOW, or there will be no "hardware boom" like in the 1980s.
I'm not certain you understand the issues. It's not just that the parts don't exist anymore, it's that the factories to make those parts don't exist anymore. To make a nut of the type required by the Saturn V, we'd need to retool an entire factory. Across all the nuts, bolts, sheet metal, piping, and other heavy industry components, it would take an excessive amount of time and money to rebuild the program.
:-)
The problem is that US industry has moved on to more advanced materials than those used in the Saturn V program. The result is that it would actually be *cheaper* to build a new super-booster from scratch. (Or pay the Russians to retool for the Energia.)
Think of it this way: NeWS was a powerful windowing system back in the early 90's that put X-Windows to shame. It has since disappeared due to Sun's attempt to keep the technology proprietary. Now if you wanted to revisit the concepts of NeWS, would you start with unmaintained 15 year old code that makes assumptions about machines that don't even exist any longer, or would you reuse the original concept in a new design that would take advantage of the far simpler programming model of today's languages, environments, and hardware?
Cruftiness is a problem in all areas of engineering, not just programming.
Becoming more of a "do all disposable" similar to the ever reliable Appolo hardware that was quickly adapted to 3 totaly different sets of missions with little effort.
:-)
Actually, most of the proposed CEV hardware is reasonably reusable. The booster itself is still lost (not that we've gotten much milage out of the "reusable" SRBs and non-reusable LHOx tank), but the crew vehicle will probably be capable of quick turnarounds. The primary difference between the shuttle and the CEVs is that the CEVs are only a life support and reentry capsule. All the tricky parts of the mechanics (e.g. engines that guzzle a swimming pool per second) are left to the disposable boosters.
Of course, this is only true of Spiral One. Spirals two and up will most likely be 100% reusable, since they will live their entire lifetimes up in space.
I do hear what you're saying, though. Idealogically, a new super-technology (the X-33, Delta Clipper, Space Plane?) should come along and be the Next Big Thing(TM). The problem that we've faced is that the Shuttle has been proven to be a step in the wrong direction. It was a jack of all trades, and a master of none. As a result, ideology must take a back seat to pragmatism.
In other words, yes you can derive thrust from a neutron stream without using a transfer mass a la NERVA-type vehicles. No clue on efficiency, or ease, but it is definitely within the possibility of physics to channel the neutron output of a reactor into a specific direction and use that output to provide thrust.
:-)
As I understand it, a neutron powered rocket is highly infeasible due to the need to maintain a chain reaction, and the inability to control neutrons (no charge). If all the neutrons were channelled out of the rocket, then there would be nothing to maintain the fission reaction. The neutrons that are used to maintain the reaction dump their energy by boucing off of atoms, thus creating massive amounts of thermal energy. Since we don't want to be using the reactor core as a propellant, we have to transfer that thermal energy to some form of working fluid.
At least that's my understanding of why such a rocket wouldn't work. (It would be damn efficient if it could, though.) But now we're off on a tangent.
I have some very fun ideas involving high Isp "excimer plasma rockets" which I'd love to try out.
:-)
I think you're still missing my point. You don't want to develop *new* technologies until AFTER you've developed a market using *proven* space technologies. All the engines we could ever want are already on the drawing boards. Even engines for interstellar flight. The key is to get into space first, THEN there will be an economic incentive for R&D.
Clear?