Re:A lot of stuff in Gtk is replacing Gnome widget
on
GTK 2.6.0 Released
·
· Score: 0, Flamebait
Sir, it is truly morons like you that give Slashdot a bad name. That you post anonymously only does a disservice to more intelligent and insightful GNFOS trolls who use that name. If you have a point to make, I'm waiting for it, but you'll really need to make it both coherently and logically, because thus far you have done little but pretend you know that which you must not.
I've written more millions of lines of code than you have lived years on this earth. I did general-purpose distributed computing before it was cool. I wrote a Lisp interpreter in 6 hours. In short, I know a thing or two about writing code.
And you, sir, are wrong. Gnome and KDE are desktop environments. That they include thousands of googaws that make building applications easier is a disadvantage, and it's just that disadvantage that moving such googaws into the appropriate location, such as the upstream widget toolkit (in this case, Gtk+) can alleviate.
Right now, to get even one googaw or trinket out of Gnome for use in an application (the GtkHTML widget comes to mind as a past example of this, but as I haven't done any GUI development in several yearsI won't specify precisely which widget I am presently referring to), you end up installing the entirety of the Gnome libraries in an effort to ensure you meet the dependencies. Oftentimes, application programmers don't even know what their applications depend upon to build and/or run.
The disadvantage is that the line between widget toolkit and desktop is blurred and was blurred from shortly after the beginning of it all. Gtk+ and glib are a good example of two libraries with a unidirectional dependency and fine line of separation. If it's part of the graphical toolkit, it is in Gtk+. If it is a data structure or other non-graphical widget, it's in glib.
With Gtk+ versus libgnome*, though, the line is blurred. So many different libraries provide graphical widgets at the exact same level as Gtk+ provides them that an application developer often must choose between omitting functionality made possible by the Gnome widgets, implementing his own widgets to replace them, and writing code that depends upon higher-level Gnome libraries than it needs to.
If all the widgets were either consolidated into Gtk+ (as is the current trend) or consolidated instead into separate libraries with very well-defined territories, then application developers could get away from the type of bloat that is most annoying to many users: dependency bloat. There is nothing more obnoxious than finding a new piece of software (take, for example, mail-notification the project) and finding out that you have to install an entire desktop environment (here, Gnome) because it provides some minuscule piece of code that it depends upon.
The Gnome libraries could be rededicated to providing a functional and useful desktop environment and high-level IPC, and the applications that do not need those features could do just fine depending only upon specific libraries from a set of granular, well-defined libraries.
Re:A lot of stuff in Gtk is replacing Gnome widget
on
GTK 2.6.0 Released
·
· Score: 1
Don't think in terms of reducing dependencies as far as packages to install. Think in terms of reducing dependencies as in eliminating Gnome altogether. I hate Gnome. I think it sucks, and I refuse to use it. Every time I try using it, I end up hating it more. Same goes for KDE. The more useful Gtk+ (isn't that the actual name, as opposed to the inaccurate-as-always story blurb?) becomes without any additional libraries, the more likely it is that applications will be written directly atop Gtk+ rather than on top of Gnome, and the more likely I am to use those applications as a result.
And when you do know what GTK is, you probably know about the release because your distro's package manager tells you or because you are tracking the source and saw the release announcement on gtk.org or via mailing list before it was posted on Slashdot.
There's no reason to post this kind of crap on the front page. There are partner sites dedicated to software releases that do a much better job of explaining them, and some people actually do come here for the news, which often gets lost due to tinfoil-hat stories and software release announcements clogging up the front page.
6.2 it was (in both cases). The novelty of the local exploit was that it took under a minute including research, it didn't require a compiler or network access, it didn't require the OS media, and it didn't require a reboot.
I havne't used my Indigo2 much at all since the rooting occurred several years ago. I was on dial-up at the time and couldn't really download the patches, even if I had known where to get them from at the time.
I've actually found Linux to run equivalently or slower. On your Sparc hardware, I recommend OpenBSD/SPARC. I ran it happily for a long time on my Sparcstation IPX, whereas Linux was quite slow on it. Too bad a harsh winter in storage killed the NVRAM battery on little Sparky.
My Indigo2 runs Irix 6.2 quite well, and given that Linux/SGI tends to only work on the Challenge S and Indy machines, and barely does anything with graphics, I recommend you stick to Irix. It's actually a fun OS, and had one of the most usable early X11 interfaces. It can be made secure, but you're probably better off starting out behind a firewall and playing with it as a workstation. Heck, BZFlag is probably installed by default!:)
I had an Indigo2 get remote-rooted once. Oops. Then we had an Indy in the ACM office for a while. The President and I decided on a root password that, within 2 days, neither of us could remember. It took me nearly 50 seconds to root it without a compiler or network connection, and 30 seconds of that was spent waiting for the guy at the winterm next to me to let me Google for hints.
Keep it behind a firewall and you'll be fine. The Indy is a nice little box and lots of fun. I suggest keeping Irix on it, as half of the SGI experience is running Irix. I don't get people who buy every esoteric piece of hardware they can find and run the same OS on it as they do on their PC.
Has anyone done a TrackPoint/NippleStick/eClit hack for the iBook or Powerbook? I really want one but I absolutely despise the carpal-tunnel-inducing pointing device that Apple insists on providing.
The problem here is that spamming is easily modeled by game theory, and the spammers have a dominant strategy.
Your move: optimize how long you need to shut down your e-mail in order to minimize spam. Their move: check one day longer than your precaution allows for.
They can keep pushing it back until it is no longer useful for you to even have e-mail in the first place (i.e., you have more downtime than uptime), and either you end up not using e-mail at all or you end up receiving lots of spam.
Previously, the data was technically available to the public, but in a format that's not easily deciphered.
Presuming that this is a METAR replacement, then the format that was "not easily deciphered" is not really that bad at all. For the stuff that anyone reading Slashdot from under FL180 cares about, it's downright human-readable.
Of course, if my presumption is wrong (the article didn't appear clear at first glance) and this is for predictions of future weather rather than reports of current weather, then... neat.:)
Three things:
1. Speculation is, of course, required to determine $Y('04) in terms of future technology, but after you speculate, you can still use 2004 dollars to compare the approaches. As an armchair voodoo economist, I'm of the belief that everything can be at least equated to dollars.
2. Your motto...you'll forgive me for having a motto that differs from the status quo.
3. You proved me wrong about excluding enclosed habitats, but had I not excluded them categorically I (and anyone else silly enough to read this thread) might still exclude them silently from consideration.
As a side note, remember that the original story was about environmental conservation on Mars. As idiotic as it is, it indicates that there would likely be tree-huggers protesting against using Mars as a vast nuclear dump, so fission is probably not the popular option, no matter how much sense it makes. And so far there have been so few developments in fusion power that it's hard to believe it will be practical within 75 years.
Anything truly is possible, and it's the job of science to divine out the subclass of "anything" which is possible at any given time. It's the job of technology, though--and technology, not science, is the key to colonization--to implement what's currently possible and to work with science toward what's probable.
Right, because Texas is:
A) The only place in the world with the death penalty
B) Always electrocuting innocent people
C) At all relevant to this discussion
I was thinking that this is about South Korea, whereas North Korea's primary means of domestic communication are still starvation and threats of summary execution.
Not that it matters, but you're still dragging up the unnecessary comparison of 1974 dollars to 2004 dollars. If you stick with 2004 dollars, it's constant within the bounds necessary for this discussion.
Nowhere on the Wikipedia page for NYC does it claim a population of 25 million or a land area of 1200km^2. It does claim a metropolitan area population of 21 million, and a land area of 831km^2. That's actually denser than you wanted it to be, so I won't contradict your reasoning regarding those numbers.
However, the self-containment issue is still at large. I will address this later.
I'm not entirely anti-enclosed-habitat (by the way, "biodome" is a colloquial term that's easier to type and certainly communicates the same thing as "enclosed habitat," as our discussion of different shapes and my persistence in using the word "biodome" should indicate). In fact, in a lot of ways I'm playing Devil's Advocate here. Your initial argument against terraforming was sound. It's both cost- and time-prohibitive. It fails to prepare for colonization in time to really matter to us and it prepares too much of the planet for our uses.
So, let's be constructive and discuss how to make a long-term-survivable, self-contained Martian colony. We've established that an enclosed habitat of some form is the best known method. It will be very costly, compared to terraforming without constructing many habitats, but that cost can be paid incrementally as I discussed in another subthread. It can even be maintained profitably as it grows. (While we're at it, Wikipedia isn't thorough on the methodology of James J. Hill, but his railroad management is how a Martian colony should be run if it's to be housed in enclosed habitats.)
The challenges that I can identify are:
Getting the initial enclosed habitat set up
Getting a genetically diverse and procreationally willing population into the habitat
Making the colony self-contained
Giving the colony the capability of expanding its habitat
The down payment, so to speak, is the first two items. #3 would be a significant cost, but it could be amortized over some time. For example, shipments of a year's supply of everything could be sent along with that year's self-containment additions. Hydroponics and genetically engineered plants are probably the keys, here. The genetically engineered plants could theoretically produce all the proteins and amino acids that you need to thrive (which regular plants do not).
Self-containment is more than food, though. Energy production, mining, and manufacturing are the primary ingredients beyond the basic food, water, and shelter required for human survival. Life or no life, Mars almost certainly has no fossil fuels. A source of energy is needed. As hydrogen is probably not abundant on Mars, I doubt fuel cells are the answer. Wind power might be useful, depending on the reliability of Martian windstorms. Solar energy is always available, but costly in terms of resources.
One thing Mars has plenty of is rocks. I don't know much about the composition, but I'm sure enough iron and silicon could be harvested to be useful for all construction needs, presuming that the energy cost of refinement wouldn't outweigh the benefits of the refined products. What other building materials are available there?
Self-containment plus manufacturing naturally leads the way to self-propelled expansion. And it's when a colony can grow itself without any external input whatsoever that it truly becomes viable.
What are your thoughts on the challenges I've identified here?
The problem with USE flags lies in the sheer amount of effort and "Oh, shit, I have to start over from scratch!" they are to get right. The cause of this is that, through trial and error, I often found that I should have set such-and-such flag before I did the first emerge six days of compiling ago, and now I could just as well start completely from scratch. I have never found a truly well-documented list of USE flags and how they interact with each other and with certain applications.
The sheer amount of trivia that needs to be remembered in order to actually install and maintain a perfectly-crafted Gentoo system doesn't seem to be worth it.
I liked Gentoo, don't get me wrong. I just found it lacking in a few areas. Right now my PC runs Arch, which tracks versions pretty well, has a good wealth of desktop-productivity software (it was quick on the ball with Gaim 1.0, Firefox 1.0, etc.), and isn't nearly as bloated as Gentoo tends to be. I'd be amenable to installing Gentoo again, but would want a definite battle plan going in.
On a side note, I still haven't found anything remotely as good as aptitude for managing installed packages. I like package menus that clearly show me what's available, what I already have, and what will happen if I add, remove, or upgrade a particular package. Is there anything similar for Gentoo? That was another killjoy for me.
PS: Thanks for keeping up with this thread so far. I really do appreciate your recommendation of e17 and indication that Gentoo's got it.:)
I've had serious issues with Gentoo. After spending a day installing it, it won't run. And even when it does run, it tends not to be stable. Moreover, there's no way at all that I can run the same distro on my laptop and PC if I do that, as, quite frankly, my 650MHz laptop with 128MB of RAM would likely run it fine, if the hard drive weren't only 12GB. Speaking of which, Gentoo is bloated in the first place. It frustrates me.:(
Are there any other X window managers that do that? Specifically, any that have had an actual release of the version that does that, or even that have had any release at all since Clinton was in office? As fun as it is to deal with source all the time, my lifestyle demands that I not invest time in manually tracking source code and dependencies.
Where did you get your figure for the area of NYC? Are you certain that's the area that houses the residences and workplaces of 25 million people?
Also, you operate upon the mistaken assumption that NYC is self-contained. It's nowhere near it, particularly in terms of the basic necessities of life. NYC doesn't produce its own food or energy, and those two form the heaviest requirements for human life, particularly on Mars. But mining and manufacturing are other areas where NYC's production falls below its demand, and those are the requirements for a growing society on Mars.
And dollars would have been just fine, as long as you stick to today's dollars, because 2004 dollars are static. I don't know why you're afraid of using them.
I run XFCE (on Arch on my PC and on Debian sarge on my laptop for the time being) and only have a few problems with it. Everything else is perfect, but...
First and foremost, the file manager sucks. It's exceedingly slow, difficult to navigate, and easy to click into things like "SMB Network" by mistake. The top-level hierarchy is also heterogeneous - "Book" (wtf?), home directory, "SMB Network", "Applications", "Find", "Trash", and "Fstab". But mainly it's slow as hell.
Next up, I'd love for the main launcher bar thingy leaves a little to be desired, as does the task bar. Also, the only really good configuration I've found is to have them at opposite ends of the screen, which is annoying. A system more like the MacOS X dock would be nice - application launchers that tell you when the application is already open and give you a small control menu for it.
Fixing the previous issue gets rid of the entire taskbar except for the small system tray area. Making this into a launcher bar applet would solve that problem and get rid of the taskbar entirely.
That's all I can think of right now. There are other issues that preclude me from declaring Linux "ready for the desktop" (even though I use it as such 80% or more of the time and for 95% or better of my serious work), such as Firefox launch time, Thunderbird new mail notification while iconified, and communication between those two; but those are all basically unrelated to Xfce4.
No. Those are the apps you require in order to make your system usable, regardless of desktop environment. They would all work essentially the same regardless of window manager, even if it's twm or none at all.
The way you were saying how you could build one bit at a time just made it click in my head - if you turn each segment of track profitable before you build the next segment, you can use the profits to build the entire railway and not owe a dime when you're done. Obviously, the cost of getting the first segment of habitat to Mars is insanely high, but if you can turn it profitable (not necessarily in terms of cashflow, but in terms of production; and ignoring the cost of getting it there in the first place; it just has to produce more than it consumes) then you can build up the rest of the colony from there.
Now...let's design us a Martian colony that works that way.:)
As I point out in a (heated) sibling thread, I was talking about "significant" human population meaning in the hundreds of millions or preferably the billions. Biodomes to support ten people in Arizona for a single year are impractical right now.
Supporting 10 million times that for 100 times as long would, at first glance, require 10 million times the construction costs and 1 billion times the operating costs (and that's just for 100 million people), and that's neglecting the cost of getting materials to Mars or, alternatively, large-scale mining and manufacturing set up on Mars.
Obligatory Futurama reference: Pauly Shore, on the fact that he can, in the year 3000, live in an actual biodome: "An unattractive prospect. While researching for the role, I ran computer simulations demonstrating incontrovertibly that the whole bio-enclosure concept is fundamentally flawed, be it expressed via dome, sphere, cube, or even the stately tetrahedron, buuuuuuuuuddy."
Slashdot Mirror
Sir, it is truly morons like you that give Slashdot a bad name. That you post anonymously only does a disservice to more intelligent and insightful GNFOS trolls who use that name. If you have a point to make, I'm waiting for it, but you'll really need to make it both coherently and logically, because thus far you have done little but pretend you know that which you must not.
I've written more millions of lines of code than you have lived years on this earth. I did general-purpose distributed computing before it was cool. I wrote a Lisp interpreter in 6 hours. In short, I know a thing or two about writing code.
And you, sir, are wrong. Gnome and KDE are desktop environments. That they include thousands of googaws that make building applications easier is a disadvantage, and it's just that disadvantage that moving such googaws into the appropriate location, such as the upstream widget toolkit (in this case, Gtk+) can alleviate.
Right now, to get even one googaw or trinket out of Gnome for use in an application (the GtkHTML widget comes to mind as a past example of this, but as I haven't done any GUI development in several yearsI won't specify precisely which widget I am presently referring to), you end up installing the entirety of the Gnome libraries in an effort to ensure you meet the dependencies. Oftentimes, application programmers don't even know what their applications depend upon to build and/or run.
The disadvantage is that the line between widget toolkit and desktop is blurred and was blurred from shortly after the beginning of it all. Gtk+ and glib are a good example of two libraries with a unidirectional dependency and fine line of separation. If it's part of the graphical toolkit, it is in Gtk+. If it is a data structure or other non-graphical widget, it's in glib.
With Gtk+ versus libgnome*, though, the line is blurred. So many different libraries provide graphical widgets at the exact same level as Gtk+ provides them that an application developer often must choose between omitting functionality made possible by the Gnome widgets, implementing his own widgets to replace them, and writing code that depends upon higher-level Gnome libraries than it needs to.
If all the widgets were either consolidated into Gtk+ (as is the current trend) or consolidated instead into separate libraries with very well-defined territories, then application developers could get away from the type of bloat that is most annoying to many users: dependency bloat. There is nothing more obnoxious than finding a new piece of software (take, for example, mail-notification the project) and finding out that you have to install an entire desktop environment (here, Gnome) because it provides some minuscule piece of code that it depends upon.
The Gnome libraries could be rededicated to providing a functional and useful desktop environment and high-level IPC, and the applications that do not need those features could do just fine depending only upon specific libraries from a set of granular, well-defined libraries.
Don't think in terms of reducing dependencies as far as packages to install. Think in terms of reducing dependencies as in eliminating Gnome altogether. I hate Gnome. I think it sucks, and I refuse to use it. Every time I try using it, I end up hating it more. Same goes for KDE. The more useful Gtk+ (isn't that the actual name, as opposed to the inaccurate-as-always story blurb?) becomes without any additional libraries, the more likely it is that applications will be written directly atop Gtk+ rather than on top of Gnome, and the more likely I am to use those applications as a result.
And when you do know what GTK is, you probably know about the release because your distro's package manager tells you or because you are tracking the source and saw the release announcement on gtk.org or via mailing list before it was posted on Slashdot.
There's no reason to post this kind of crap on the front page. There are partner sites dedicated to software releases that do a much better job of explaining them, and some people actually do come here for the news, which often gets lost due to tinfoil-hat stories and software release announcements clogging up the front page.
6.2 it was (in both cases). The novelty of the local exploit was that it took under a minute including research, it didn't require a compiler or network access, it didn't require the OS media, and it didn't require a reboot.
I havne't used my Indigo2 much at all since the rooting occurred several years ago. I was on dial-up at the time and couldn't really download the patches, even if I had known where to get them from at the time.
I've actually found Linux to run equivalently or slower. On your Sparc hardware, I recommend OpenBSD/SPARC. I ran it happily for a long time on my Sparcstation IPX, whereas Linux was quite slow on it. Too bad a harsh winter in storage killed the NVRAM battery on little Sparky.
:)
My Indigo2 runs Irix 6.2 quite well, and given that Linux/SGI tends to only work on the Challenge S and Indy machines, and barely does anything with graphics, I recommend you stick to Irix. It's actually a fun OS, and had one of the most usable early X11 interfaces. It can be made secure, but you're probably better off starting out behind a firewall and playing with it as a workstation. Heck, BZFlag is probably installed by default!
It's actually different music on each line, from what I've seen. I know my Indigo2 has a piano lick, versus the sound on the Indy.
I had an Indigo2 get remote-rooted once. Oops. Then we had an Indy in the ACM office for a while. The President and I decided on a root password that, within 2 days, neither of us could remember. It took me nearly 50 seconds to root it without a compiler or network connection, and 30 seconds of that was spent waiting for the guy at the winterm next to me to let me Google for hints.
Keep it behind a firewall and you'll be fine. The Indy is a nice little box and lots of fun. I suggest keeping Irix on it, as half of the SGI experience is running Irix. I don't get people who buy every esoteric piece of hardware they can find and run the same OS on it as they do on their PC.
Has anyone done a TrackPoint/NippleStick/eClit hack for the iBook or Powerbook? I really want one but I absolutely despise the carpal-tunnel-inducing pointing device that Apple insists on providing.
The problem here is that spamming is easily modeled by game theory, and the spammers have a dominant strategy.
Your move: optimize how long you need to shut down your e-mail in order to minimize spam. Their move: check one day longer than your precaution allows for.
They can keep pushing it back until it is no longer useful for you to even have e-mail in the first place (i.e., you have more downtime than uptime), and either you end up not using e-mail at all or you end up receiving lots of spam.
I tried to read the article, but I had too many computers early on and don't know how to read.
Sometimes you have to resort to looking at the next field to figure out through a process of elimination what the current field is.
;)
And that's why it's fun to code for.
Previously, the data was technically available to the public, but in a format that's not easily deciphered.
... neat. :)
Presuming that this is a METAR replacement, then the format that was "not easily deciphered" is not really that bad at all. For the stuff that anyone reading Slashdot from under FL180 cares about, it's downright human-readable.
Of course, if my presumption is wrong (the article didn't appear clear at first glance) and this is for predictions of future weather rather than reports of current weather, then
Three things:
1. Speculation is, of course, required to determine $Y('04) in terms of future technology, but after you speculate, you can still use 2004 dollars to compare the approaches. As an armchair voodoo economist, I'm of the belief that everything can be at least equated to dollars.
2. Your motto...you'll forgive me for having a motto that differs from the status quo.
3. You proved me wrong about excluding enclosed habitats, but had I not excluded them categorically I (and anyone else silly enough to read this thread) might still exclude them silently from consideration.
As a side note, remember that the original story was about environmental conservation on Mars. As idiotic as it is, it indicates that there would likely be tree-huggers protesting against using Mars as a vast nuclear dump, so fission is probably not the popular option, no matter how much sense it makes. And so far there have been so few developments in fusion power that it's hard to believe it will be practical within 75 years.
Anything truly is possible, and it's the job of science to divine out the subclass of "anything" which is possible at any given time. It's the job of technology, though--and technology, not science, is the key to colonization--to implement what's currently possible and to work with science toward what's probable.
Right, because Texas is:
A) The only place in the world with the death penalty
B) Always electrocuting innocent people
C) At all relevant to this discussion
I was thinking that this is about South Korea, whereas North Korea's primary means of domestic communication are still starvation and threats of summary execution.
Nowhere on the Wikipedia page for NYC does it claim a population of 25 million or a land area of 1200km^2. It does claim a metropolitan area population of 21 million, and a land area of 831km^2. That's actually denser than you wanted it to be, so I won't contradict your reasoning regarding those numbers.
However, the self-containment issue is still at large. I will address this later.
I'm not entirely anti-enclosed-habitat (by the way, "biodome" is a colloquial term that's easier to type and certainly communicates the same thing as "enclosed habitat," as our discussion of different shapes and my persistence in using the word "biodome" should indicate). In fact, in a lot of ways I'm playing Devil's Advocate here. Your initial argument against terraforming was sound. It's both cost- and time-prohibitive. It fails to prepare for colonization in time to really matter to us and it prepares too much of the planet for our uses.
So, let's be constructive and discuss how to make a long-term-survivable, self-contained Martian colony. We've established that an enclosed habitat of some form is the best known method. It will be very costly, compared to terraforming without constructing many habitats, but that cost can be paid incrementally as I discussed in another subthread. It can even be maintained profitably as it grows. (While we're at it, Wikipedia isn't thorough on the methodology of James J. Hill, but his railroad management is how a Martian colony should be run if it's to be housed in enclosed habitats.)
The challenges that I can identify are:
The down payment, so to speak, is the first two items. #3 would be a significant cost, but it could be amortized over some time. For example, shipments of a year's supply of everything could be sent along with that year's self-containment additions. Hydroponics and genetically engineered plants are probably the keys, here. The genetically engineered plants could theoretically produce all the proteins and amino acids that you need to thrive (which regular plants do not).
Self-containment is more than food, though. Energy production, mining, and manufacturing are the primary ingredients beyond the basic food, water, and shelter required for human survival. Life or no life, Mars almost certainly has no fossil fuels. A source of energy is needed. As hydrogen is probably not abundant on Mars, I doubt fuel cells are the answer. Wind power might be useful, depending on the reliability of Martian windstorms. Solar energy is always available, but costly in terms of resources.
One thing Mars has plenty of is rocks. I don't know much about the composition, but I'm sure enough iron and silicon could be harvested to be useful for all construction needs, presuming that the energy cost of refinement wouldn't outweigh the benefits of the refined products. What other building materials are available there?
Self-containment plus manufacturing naturally leads the way to self-propelled expansion. And it's when a colony can grow itself without any external input whatsoever that it truly becomes viable.
What are your thoughts on the challenges I've identified here?
PS: Good argument contra terraforming.
The problem with USE flags lies in the sheer amount of effort and "Oh, shit, I have to start over from scratch!" they are to get right. The cause of this is that, through trial and error, I often found that I should have set such-and-such flag before I did the first emerge six days of compiling ago, and now I could just as well start completely from scratch. I have never found a truly well-documented list of USE flags and how they interact with each other and with certain applications.
:)
The sheer amount of trivia that needs to be remembered in order to actually install and maintain a perfectly-crafted Gentoo system doesn't seem to be worth it.
I liked Gentoo, don't get me wrong. I just found it lacking in a few areas. Right now my PC runs Arch, which tracks versions pretty well, has a good wealth of desktop-productivity software (it was quick on the ball with Gaim 1.0, Firefox 1.0, etc.), and isn't nearly as bloated as Gentoo tends to be. I'd be amenable to installing Gentoo again, but would want a definite battle plan going in.
On a side note, I still haven't found anything remotely as good as aptitude for managing installed packages. I like package menus that clearly show me what's available, what I already have, and what will happen if I add, remove, or upgrade a particular package. Is there anything similar for Gentoo? That was another killjoy for me.
PS: Thanks for keeping up with this thread so far. I really do appreciate your recommendation of e17 and indication that Gentoo's got it.
Aren't there other concerns, such as window size? When you're piping that much down the line, an ACK every 48 bytes just isn't right.
I've had serious issues with Gentoo. After spending a day installing it, it won't run. And even when it does run, it tends not to be stable. Moreover, there's no way at all that I can run the same distro on my laptop and PC if I do that, as, quite frankly, my 650MHz laptop with 128MB of RAM would likely run it fine, if the hard drive weren't only 12GB. Speaking of which, Gentoo is bloated in the first place. It frustrates me. :(
Are there any other X window managers that do that? Specifically, any that have had an actual release of the version that does that, or even that have had any release at all since Clinton was in office? As fun as it is to deal with source all the time, my lifestyle demands that I not invest time in manually tracking source code and dependencies.
Where did you get your figure for the area of NYC? Are you certain that's the area that houses the residences and workplaces of 25 million people?
Also, you operate upon the mistaken assumption that NYC is self-contained. It's nowhere near it, particularly in terms of the basic necessities of life. NYC doesn't produce its own food or energy, and those two form the heaviest requirements for human life, particularly on Mars. But mining and manufacturing are other areas where NYC's production falls below its demand, and those are the requirements for a growing society on Mars.
And dollars would have been just fine, as long as you stick to today's dollars, because 2004 dollars are static. I don't know why you're afraid of using them.
I run XFCE (on Arch on my PC and on Debian sarge on my laptop for the time being) and only have a few problems with it. Everything else is perfect, but...
First and foremost, the file manager sucks. It's exceedingly slow, difficult to navigate, and easy to click into things like "SMB Network" by mistake. The top-level hierarchy is also heterogeneous - "Book" (wtf?), home directory, "SMB Network", "Applications", "Find", "Trash", and "Fstab". But mainly it's slow as hell.
Next up, I'd love for the main launcher bar thingy leaves a little to be desired, as does the task bar. Also, the only really good configuration I've found is to have them at opposite ends of the screen, which is annoying. A system more like the MacOS X dock would be nice - application launchers that tell you when the application is already open and give you a small control menu for it.
Fixing the previous issue gets rid of the entire taskbar except for the small system tray area. Making this into a launcher bar applet would solve that problem and get rid of the taskbar entirely.
That's all I can think of right now. There are other issues that preclude me from declaring Linux "ready for the desktop" (even though I use it as such 80% or more of the time and for 95% or better of my serious work), such as Firefox launch time, Thunderbird new mail notification while iconified, and communication between those two; but those are all basically unrelated to Xfce4.
No. Those are the apps you require in order to make your system usable, regardless of desktop environment. They would all work essentially the same regardless of window manager, even if it's twm or none at all.
The way you were saying how you could build one bit at a time just made it click in my head - if you turn each segment of track profitable before you build the next segment, you can use the profits to build the entire railway and not owe a dime when you're done. Obviously, the cost of getting the first segment of habitat to Mars is insanely high, but if you can turn it profitable (not necessarily in terms of cashflow, but in terms of production; and ignoring the cost of getting it there in the first place; it just has to produce more than it consumes) then you can build up the rest of the colony from there.
:)
Now...let's design us a Martian colony that works that way.
As I point out in a (heated) sibling thread, I was talking about "significant" human population meaning in the hundreds of millions or preferably the billions. Biodomes to support ten people in Arizona for a single year are impractical right now.
Supporting 10 million times that for 100 times as long would, at first glance, require 10 million times the construction costs and 1 billion times the operating costs (and that's just for 100 million people), and that's neglecting the cost of getting materials to Mars or, alternatively, large-scale mining and manufacturing set up on Mars.
Obligatory Futurama reference: Pauly Shore, on the fact that he can, in the year 3000, live in an actual biodome: "An unattractive prospect. While researching for the role, I ran computer simulations demonstrating incontrovertibly that the whole bio-enclosure concept is fundamentally flawed, be it expressed via dome, sphere, cube, or even the stately tetrahedron, buuuuuuuuuddy."