Gnome 2.6 has a whole section on speeding up performance in the help manual. Just open the app up (the one with the life preserver icon) and find the system administration document (I think it's under the "desktop" category) and it'll be in there.
Remember kids, the only thing that separates the experts from the idiots is that the experts actually RTFM.
Do you really think that Microsoft would have played such hardball for Munich if it wasn't important for them?
Do you really think that they would have bankrolled SCO's lawsuits against IBM if they weren't worried?
Do you really think they would have set up their stupid Shared Source Initiative if they didn't see the train coming?
Do you really think they would have started putting on a nice public face towards Linux after calling it a cancer if they weren't scared about it?
Do you really think that they wouldn't have put up so many flimsy counter arguments to using Linux and OpenOffice if they weren't a threat?
Do you really think that the general population of programmers out there won't continue to push the Linux desktop forward, making it more and more capable of competing or even surpassing Windows?
Do you really think they got to be such a rich company without being paranoid?
This raises the metaphor to a new level. JIF and its ilk has saturated fats added to it in order to prevent oil separation. This is generally considered less healthy, and also generally causes a worse taste. Natural peanut butters, on the other hand, require a bit of stirring when you first open them, but you get a healthier food that generally tastes better.
Now compare Windows to Linux. Windows doesn't require that initial investment of work that Linux does (although this is improving daily) but you pay in terms of health (security) and flavor (customizability and tools). Linux gives you what the other guy is missing, at the cost of some work in terms of learning, installing, and transitioning. I'd rather not extrapolate this further to IBM vs Microsoft, but you get the idea.
Oh, and for those of you who have a Trader Joe's around, you need to try their peanut butter. It's easily the best I've had anywhere.
Fair enough, then demonstrate to me that this other genetic material exists. I'm not saying it doesn't, but we're a long way from proving that it does. Also, the current claims about nanobacteria's existence are hinged on the presence of DNA, so if that isn't really present then these researchers will probably back down on the entire idea or else demonstrate the presence and function of this other genetic material.
It's not so much about petty definitions of life as actually determining what the hell is going on. Some groups say that these things are pathological and that this pathology is caused the same as in any other bacteria, because they have DNA, express genes, make proteins and reproduce to amplify the pathogenic effect. Other groups say that all of this is not true at all, and that these things are just aggregates of mineral and protein. Isn't this distinction important, especially if these things really do turn out to cause disease (the jury is still out on this) and you want to treat them?
Doubtful. Increased optical density usually is due the formation of aggregates, not their breakdown. Bacteria grown in culture increase the optical density of the solution (this is one way to measure their growth rate) and if you add a predatory phage to lyse them, the solution will clear again. All the particles are still there (conservation of matter), but they are now in smaller form, so the solution is clear.
Agreed on all counts. The New Scientist article that someone else linked goes in to the politics a bit, and is a good read. Interestingly, the major article by a sceptic was submitted to PNAS, track II (for those who don't know, track II means it gets published without review, since PNAS is a bit of an "old boys club") so it should be taken with a grain of salt. That said, the science in that paper looks decent to me on the whole, at least in terms of raising questions that still need to be answered, most notably about the genetic material of these things.
There's also an article by the original group claiming that the nanobacteria induce apoptosis in a variety of cells, including fibroblasts in cell culture. This doesn't make any sense to me, evolutionarily, pathogenically, or physiologically. I haven't read that paper yet, only the abstract, but I still feel like I need to see a whole lot more good data on these things before I'll be convinced.
I read the PNAS article that was written by one of the critics. His claims that it's due to the aggregation of some other mineral particle, such as apatite, and has evidence to suggest that this is possible. He also does DNA sequence comparisons to demonstrate that there is a very strong similarity between the putative nanobacterial sequence and a common bacterial sequence for a certain very old (and therefore very well conserved over time) gene, indicating that it may be due to contamination.
While this evidence isn't very compelling in itself, it does cast a lot of doubt in my mind on how these nanobacteria could work. The DNA staining is very inconclusive (no distinct chromosomes), and even the culture conditions for the nanobacteria (including an experiment where they further dilute the nanobacteria culture and get no growth at all) make things very tenuous.
I'd personally like to see a genetic sequence for these things. I'd also love to hear an answer as to how they can replicate with such a small amount of DNA. I certaintly don't want to say that it's impossible for these things to exist, but the data so far is definitely inconclusive. Multiplication in serum does not equal life.
You raise a lot of really good points. It's all about tradeoffs again, but at least you know what you're talking about, and it's always good to discuss these things with someone who gets it, even if they don't like it. Too many people seem to argue against it without understanding.
Speaking of spatial stuff, one thing I'd like to see GNOME have, well Metacity really, is edge resistance when moving windows around. It certainly reinforces the illusion that windows are objects moving stuff around the screen when , for example, gaim doesn't want to move over another window without a little gentle persuasion.
Try holding down the shift key when you drag windows around. It has a sort of edge resistance/edge snapping. It's not perfect, and I'd rather have real edge resistance a la windowmaker (there was a huge debate with gnome dev's over this that I read) but it does get the job done well enough for me. I don't think I could stomach metacity without this.
The window manager I maintain for Debian is pwm, which is actually ion's predecessor (same author), so you could say I'm somewhat familiar with ion. I've tried to use ion, but just couldn't get it in to a shape where I could use it effectively. ion2 now has a pwm theme, so the old pwm is out of date, but the original pwm is still the lightest window manager you'll find, with maybe the one possible exception of twm. It's pretty basic, but has a few nice keybindings, the ability to bind others, a windowmaker dock, a couple of decent color schemes, and the same tabs that were later used in ion. Those happened to be just the features I want out of a bare-bones wm, no more no less, so it's a good fit for me.
I used to use windowmaker beforehand, and I liked it, although when I went back to it recently I found that it had somehow developed a bunch of annoying quirks while I was gone, like some apps not popping up icons when open. That was a shame, because I really used to love the wm, so I'll probably have to find a new alternate wm soon. I'll probably be checking out the *box derivatives for one.
No, you don't have to navigate two file trees. You start in the second browser in the file tree at exactly the same place that you were in the browser you invoked it from. Not that navigating two file trees is really a burden - they are always the same...
I'm perpetually annoyed by having to do it. "I was just here, why do I have to repeat this part of the process?" It's not really a big burden, but it's an annoyance and it really does feel like it slows me down. I have no hard numbers on this one, but then you have no hard numbers that it doesn't slow me down, so it's a draw.
As for chasing windows, yes I can take a guess at where they are going to show up, but that does NOT mean that the last place I left a window is close to where I am invoking it from, or isn't going to be covered up by my on-top GAIM or Tvtime window. I may know that a window is going to show up on the left edge of my screen, but that doesn't mean I enjoy traversing 1600 pixels to get there. That is lost time and effort, distracting me from my current task, which is moving files from A to B, not "scan across my desktop to my new working locus."
Well, if you want to keep your on-top window on top, that's your preference. No system is perfect, and that's just nitpicking. Of course, you could position every single window so that it never pops up over your gaim or tvtime window ever again, so this ceases to be a real problem. Your gaim and tvtime windows are also always in the same place, because you put them there. I mean, you wouldn't put your TV in the closet when you want to watch it, so why would you put it in a place where you've got windows popping up all the time? It's the same idea, and you have complete control over it.
Just because my machine has plenty of horsepower and resources doesn't mean I'm not using it at the moment for something other than seeing files in a directory I'm just passing through.
This is fair, but then you can always turn off the previewing option if it's such a burden to you (this is accessible in the nautilus 2.4 prefs at least). Again, tradeoffs happen, and this is just nitpicking. It also completely ignores the fact that spatial nautilus is actually much faster in practice than the old version.
As for files being where I left them, since I use "text" style file manager views, and tree browsers my files are ALWAYS where I left them - in alphabetical order in the same directory they were last time, unless I moved them. It is simplicity itself to jump down via my scrollwheel to the spot in the list where the files I want live.
When you add a new file to this directory, it'll shift all the other files that are organized after it alphabetically. This may be what you want, but it's not always ideal. If the real world worked this way, you'd be pretty freaked out. Imagine putting a book on the shelf and then the next time you turn around half the books on the shelf have moved! It's jarring. The same thing applies to files on computers. For large sets of files, like image directories, alphabetical organization is great (nautilus shouldn't remove the ability to sort alphabetically) but it's not always what you want. Spatial organization doesn't remove the organization scheme you speak of, it only adds another organizational level that works very well in practice.
The point I was trying to make is that no matter how many amazing things your GUI does for the computer illiterate if it is going to reduce the number of people who become literate, it is a bad thing.
I agree, but at the same time I realize that most people have absolutely no desire to become computer literate, and I don't want to force them in to such a thing. Further, I don't think spatial organization harms them from learning. If they want to learn, they will learn. And even further, if it helps them do their work better, then
Ok, I personally don't think it would be such a big deal to provide a desktop preference that's easily accessible for this kind of thing. I can understand why the gnome people didn't do it (it's a forcing device, a trick used by the original Mac people/Steve Jobs to force people to get used to the GUI paradigm) but there are plenty of good reasons to have this as an easily accessible option. Maybe for 2.8.
But on the other hand, most users really won't care about how the thing works. They generally just will use it. Aunt Tilly and challenged users really won't care.
Average users generally won't care either, but they'll care more about little things like desktop themes and sound effects and whatnot. What these users do care about is whether or not they can download mp3's, browse the web, and chat than their windowing metaphor. These people are the ones who will adopt some new program when they're told how cool it is, but they never mess around with the underlying OS unless they have to.
Power users should be able to handle this stuff with ease, and we shouldn't be concerned that they may have to google and run gconf-edit once. Besides, these people are probably using some other window manager anyway, or are at least happy to use one if they don't like what gnome's doing.
You make this sound like it's a bad thing. But it turns out that's how I organize a lot of things in my life. Each shelf on my bookshelf is a different topic (novels, technical, calvin and hobbes, etc). That's hierarchical. It REALLY gets hierarchical if you look in my filing cabinets! Bills are filled under bills, insurance policies under insurance policies, etc.
Ok, now how are those categories organized on the bookshelf? If you swapped your technical stuff with your Calvin and Hobbes comics, you'd be confused momentarily. Or if you moved the bookshelf all together across the room. That's the idea behind spatial organization, that such things don't happen. If you add a new book to your bookshelf, it doesn't cause the other books to move on its own. However, with an alphabetical organization, every time you add or remove file it will shift the position of other files in the window. Same is true for other methods.
The difference between the two kinds of organization is simple, and one that the proponents of spatial filesystems refuse to recognize, is that hierachical organization is for long term storage and spatial organization for short term stacks. Most things in my computer are for long terms storage (more than a week). Short term stuff (in my case) is organized into groups if icons on my desktop.
You've still got a heirarchical organization with the spatial metaphor. That's what nested directories are for, after all. But you've got that extra layer of metadata on top of it. No one is preventing you from organizing your files alphabetically in a spatial system, but you have that freedom if that's how you remember things best. Even for long-term information, knowing its physical location helps a lot. I know that my banking information is in the front of the file box while in the back are reference papers. There's still a nested heirarchy of data (bank info subdivided, etc) but I can get to it more quickly because I don't have to scan through the names of every folder in the box. This becomes even more useful when you have lots of folders in the box.
1. This guy is not an average user. He's an editor at a major computer magazine and has covered Linux stories for some time now.
2. The Gnome team did actual usability tests with plenty of real people. Apple did the same before this. The spatial concepts are neither new nor surprising.
3. There's plenty of average users who use linux, they usually just think that they're well above average. Hang around in an IRC linux support channel and you'll see what I mean. Their feedback is generally ignored for good reason.
4. The average user you're describing won't give two shits about spatial vs browser metaphor. They'll just learn it and use it if they have to because that's what's presented to them. They won't bother to go digging around for some stupid hidden option in a control panel or registry. That sort of thing defines a power user.
I never said that they did, but they do impede it.
Maybe there is a small conceptual hit, but its not a big one, and if it the "cost" of losing that conception is learning that files are abstract data that can be handled in a number of ways, well, your user is one step closer to grokking the computer using experience and thus no longer being a terrible burden because they are being kept in a state of ignorance.
I'm sorry, but practical experience disproves this. If you were correct, we'd have a lot of Windows users out there who understood how file systems worked a lot better.
Invoke a file browser. Drill down in tree to where you need to get files from.
Fire off a second file browser from the branch point in drilling down, as you would in the spatial file manager scenario above.
Continue navigating in the spawned browser via the tree (faster especially when the tree navigation allows you to skip loading a bunch of file attributes between here and there that you don't need to see) and select in the tree your destination.
You now have two file browsers open, one to your source directory, one to your destination directory. Select files from source, drag/drop/cut/paste/whatever to destination.
Done. You've only opened two windows, you've skipped viewing a ton of extraneous files that would be shown in the spatial navigator (which can be a HUGE delay if the machine is slow or under load, or large directories, or snapshots, or one of the many other situations), and as a bonus, you haven't had to chase fifty windows across the screen - not a big deal on a small screen, but on my 1600x1200 screen a REAL pain.
First off, you've re-proven my point about having to navigate two whole file trees, instead of one and a part. This is inefficient.
You do have a point about the file attributes, but with any sufficiently fast machine this should not be an issue, because the machine is usually just waiting on the human. Remember, this stuff worked just great circa 1984.
The number of total windows opened is inconsequential, unless you think double clicking is so much more of a pain. An important point is that while you're "chasing windows across the screen" is that you're not chasing them, because you know exactly where they'll pop up. This is a key point that people who haven't used spatial concepts extensively just don't understand. You know exactly where the next window will pop up because you left it there last time you closed it. Further, you know exactly where every file is in there, because they are exactly where you placed them before.
It's a lot like the concept behind radial menus, you don't have to read through everything to know where you're going. Think of it as driving home when you know your route really well versus navigating to some new place and having to slow down and read every sign along the way. This makes a huge difference in the way we work in practice. It's not about how many windows you open, it's about being able to map the data on the hard drive more efficiently in your own head.
When I can trivially turn off the sidebar in just about every file manager that has one, then it really doesn't have any bearing on why a spatial interface is better or worse than a non-spatial interface. It's like asserting chicken nuggets to be better than hamburgers because they don't have mustard on them.
The ability to turn it off and on has no bearing at all. You still turn it on to use it, and when you do use it, it still has the exact same flaws in that you have tons of extra directories that don't concern what you're doing visible on the screen. Further, the file tree concept has no concept of all of space, which is what we're arguing after all. The tree is always organized heirarchically first, and alphabetically/date/type-oriented second. Not based on where you "move" the file to. If you still don't understand what I mean by this, you really need to go and read the Ars Technica article.
For the past year I have been subjected to trolls about Gnome's usability campaign on a daily basis. This is merely my attempt to show that KDE's configurability isn't the horrible evil it's been made out to be.
Once again, stick to the topic at hand. Leave your personal vendettas at the door.
No one, of course. But what does this have to do with the spatial file interface? Nothing!
Sure it does. The whole tree view thing is totally-nonspatial. The fact that you have all this extra data hanging around due to the non-spatial metaphor demonstrates that the spatial method slims down the amount of crap the user has to mentally wade through when using the UI.
I'm not familiar with the 2.6 Nautilus, but I am very familiar with the Konqueror file manager. You get a choice, which is good. You choice is the default sidebar plus view, or just the view (my preference), or a split view (mc like), or even sidebar plus split view.
Konqueror will ALSO give you a spatial interface of a sort. Just configure "open folders in separate windows" (and without "open links in tabs") and you get most of what everyone is talking about.
Most of is not the same as the real thing. I use gnome 2.4's nautilus like you describe, and I've used Konq this way as well. It's not the same. Work with the old MacOS for a while or go read the Ars Technica article on spatial concepts. Most of the idea isn't the same thing because that's when you get a Win95-style UI.
That's the beauty of the KDE way. The user gets to decide how the user will work instead of settling for the default the developer chose (or trying to figure out from outdated and incomplete docs how to change it).
That's a whole other argument I don't want to go near. Trolling for one desktop or the other is not the point.
When people ask me why I use Linux, one of the things I always say is "I never have to pirate software anymore." Everyone ignores it, but this story demonstrates why I always mention it. When you don't have to pirate software, you don't have to worry that some program that you need but can't afford or don't want to pay for is going to destroy your system. All my stuff comes from a much more trusted source than Limewire.
Everyone I know who uses Windows and pirates software like this has to put up with this shit. It's just not worth it, especially when you just want to get your work done. Of course, in these days where you plug your machine in and you get a host of infections automatically within a 24 hour timespan perhaps no one really worries as much about these things anymore.
So I have to DOUBLE, click with a scrollwheel (not a nice experience)
I haven't switched from 2.4 to 2.6 yet, but I've been told that you can hold down shift key as you left click as well (which mirrors the old MacOS style).
Apple spends more than anyone on UI research and they have abandoned spacial..... are we to believe some hacker, former BeOS lover, is somehow more skilled than Apples UI teams?????? NO.
There's still a lot of UI wisdom from the pre-OSX days, and to simply dismiss it is foolish.
This is the same as saying we need "spacial web-browing" remove the back and forward buttons. Remove all buttons, the address bar EVERYTHING. And people can just navigate by "surfing the links" because it is more "natural".
Spatial webbrowsing is impractical because of the nature of HTML and the infinite space and chaotic organization of the network. In addition, you don't manipulate the web, you view it and interact with it. Spatial concepts become very useful when interacting with files (i.e. drag it to the trash to delete it) but isn't so necessary when all you're doing is looking at data. File browsing has a very different set of requirements and constraints.
Ultimately, what's interesting about the above is that a spatial metaphor encourages the user to interact with their data, where the portal viewing method that Windows and OSX uses is meant more for viewing, like on the web. Perhaps the reason why Apple switched to the portal viewer metaphor for OSX is that there's so much in UNIX that you're not supposed to manipulate, where in the old MacOS you could manipulate anything really.
And one thing that's very much in favor of spatial organization is that it's actually much faster to move files around than the porthole metaphor, at least if you know what you're doing. With the porthole method, you open up windows explorer (which by default is a totally separate icon/interface than starting from My Computer or whatever) and navigate to your file. Then you have to navigate to where you want to move it to on the sidebar. That's the most efficient way to do it, and you still have to bring up the sidebar, which may not expand far enough over to easily see as deep in the file tree as you need.
In contrast, with the one-window per file method, you open up each folder, holding down the shift key on each open so that the previous one is closed, and where the file tree branches off from the current file location and where it's going, you leave that window open, and keep drilling down the file tree. Then, once you get to your file, you go back to that branch point you left open and drill down the other half of the tree. This is much quicker because you rarely have to navigate two full trees, and you don't have to deal with a sidebar that's too small due to the fact that you have all this extra data hanging around. Who needs to see the whole damn file tree at a time on the sidebar? Once again, this sort of thing has no bearing when all you're doing is browsing the file tree and seeing what's there, but when actually manipulating it, it's of huge benefit.
I understand your frustration, but have you tried the new installer yet? We just released beta4, and it's gotten a lot of positive reviews and install reports so far. It's a ground-up rewrite, and while it's still in beta and is some features (like PPPoE) need to be fleshed out, it's a vast improvement over the old boot-floppies.
Our goal has been to build a much better installer, and so far we're succeeding. If you give it a try, I think you'll be pleased with the improvements.
The new installer will support PPPoE, the problem has been a lack of manpower to do so more than it being too obscure. There is basic support for it in the newest beta of the installer, but it's very basic and I'm not sure if it even works right now. Once again, the problem is lack of manpower to work on the damn thing. If you want to help us out, it'd be much appreciated, we could certaintly use it.
Great post. Really fantastic. Not 100% true, of course, since there are always the kids who do want to excel (me and my high school friends being good examples) but you're very much on target.
Interestingly enough, this effect carries over beyond high school and the like. When I tell people I'm getting my PhD in molecular bio, they stare at me like I just threw up on their shoes. They don't tease, mainly because they're stunned that anyone in real life would actually want to do something like science, but there's still that disbelief that such a thing is possible. I get this from even college educated people, which still shocks me.
Sure, they say that until they find out that they've got cancer or high blood preassure. But by then it's too late. Even then, most people never make the connection between paying taxes and living longer.
Slashdot Mirror
Gnome 2.6 has a whole section on speeding up performance in the help manual. Just open the app up (the one with the life preserver icon) and find the system administration document (I think it's under the "desktop" category) and it'll be in there.
Remember kids, the only thing that separates the experts from the idiots is that the experts actually RTFM.
Do you really think that Microsoft would have played such hardball for Munich if it wasn't important for them?
Do you really think that they would have bankrolled SCO's lawsuits against IBM if they weren't worried?
Do you really think they would have set up their stupid Shared Source Initiative if they didn't see the train coming?
Do you really think they would have started putting on a nice public face towards Linux after calling it a cancer if they weren't scared about it?
Do you really think that they wouldn't have put up so many flimsy counter arguments to using Linux and OpenOffice if they weren't a threat?
Do you really think that the general population of programmers out there won't continue to push the Linux desktop forward, making it more and more capable of competing or even surpassing Windows?
Do you really think they got to be such a rich company without being paranoid?
Microsoft is scared. And they should be.
This raises the metaphor to a new level. JIF and its ilk has saturated fats added to it in order to prevent oil separation. This is generally considered less healthy, and also generally causes a worse taste. Natural peanut butters, on the other hand, require a bit of stirring when you first open them, but you get a healthier food that generally tastes better.
Now compare Windows to Linux. Windows doesn't require that initial investment of work that Linux does (although this is improving daily) but you pay in terms of health (security) and flavor (customizability and tools). Linux gives you what the other guy is missing, at the cost of some work in terms of learning, installing, and transitioning. I'd rather not extrapolate this further to IBM vs Microsoft, but you get the idea.
Oh, and for those of you who have a Trader Joe's around, you need to try their peanut butter. It's easily the best I've had anywhere.
Fair enough, then demonstrate to me that this other genetic material exists. I'm not saying it doesn't, but we're a long way from proving that it does. Also, the current claims about nanobacteria's existence are hinged on the presence of DNA, so if that isn't really present then these researchers will probably back down on the entire idea or else demonstrate the presence and function of this other genetic material.
It's not so much about petty definitions of life as actually determining what the hell is going on. Some groups say that these things are pathological and that this pathology is caused the same as in any other bacteria, because they have DNA, express genes, make proteins and reproduce to amplify the pathogenic effect. Other groups say that all of this is not true at all, and that these things are just aggregates of mineral and protein. Isn't this distinction important, especially if these things really do turn out to cause disease (the jury is still out on this) and you want to treat them?
Doubtful. Increased optical density usually is due the formation of aggregates, not their breakdown. Bacteria grown in culture increase the optical density of the solution (this is one way to measure their growth rate) and if you add a predatory phage to lyse them, the solution will clear again. All the particles are still there (conservation of matter), but they are now in smaller form, so the solution is clear.
Agreed on all counts. The New Scientist article that someone else linked goes in to the politics a bit, and is a good read. Interestingly, the major article by a sceptic was submitted to PNAS, track II (for those who don't know, track II means it gets published without review, since PNAS is a bit of an "old boys club") so it should be taken with a grain of salt. That said, the science in that paper looks decent to me on the whole, at least in terms of raising questions that still need to be answered, most notably about the genetic material of these things.
There's also an article by the original group claiming that the nanobacteria induce apoptosis in a variety of cells, including fibroblasts in cell culture. This doesn't make any sense to me, evolutionarily, pathogenically, or physiologically. I haven't read that paper yet, only the abstract, but I still feel like I need to see a whole lot more good data on these things before I'll be convinced.
I read the PNAS article that was written by one of the critics. His claims that it's due to the aggregation of some other mineral particle, such as apatite, and has evidence to suggest that this is possible. He also does DNA sequence comparisons to demonstrate that there is a very strong similarity between the putative nanobacterial sequence and a common bacterial sequence for a certain very old (and therefore very well conserved over time) gene, indicating that it may be due to contamination.
While this evidence isn't very compelling in itself, it does cast a lot of doubt in my mind on how these nanobacteria could work. The DNA staining is very inconclusive (no distinct chromosomes), and even the culture conditions for the nanobacteria (including an experiment where they further dilute the nanobacteria culture and get no growth at all) make things very tenuous.
I'd personally like to see a genetic sequence for these things. I'd also love to hear an answer as to how they can replicate with such a small amount of DNA. I certaintly don't want to say that it's impossible for these things to exist, but the data so far is definitely inconclusive. Multiplication in serum does not equal life.
Yeah, Mos Def appears to be the only black guy on the cast and crew. That's irony for you. Go Mos!
Try holding down the shift key when you drag windows around. It has a sort of edge resistance/edge snapping. It's not perfect, and I'd rather have real edge resistance a la windowmaker (there was a huge debate with gnome dev's over this that I read) but it does get the job done well enough for me. I don't think I could stomach metacity without this.
The window manager I maintain for Debian is pwm, which is actually ion's predecessor (same author), so you could say I'm somewhat familiar with ion. I've tried to use ion, but just couldn't get it in to a shape where I could use it effectively. ion2 now has a pwm theme, so the old pwm is out of date, but the original pwm is still the lightest window manager you'll find, with maybe the one possible exception of twm. It's pretty basic, but has a few nice keybindings, the ability to bind others, a windowmaker dock, a couple of decent color schemes, and the same tabs that were later used in ion. Those happened to be just the features I want out of a bare-bones wm, no more no less, so it's a good fit for me.
I used to use windowmaker beforehand, and I liked it, although when I went back to it recently I found that it had somehow developed a bunch of annoying quirks while I was gone, like some apps not popping up icons when open. That was a shame, because I really used to love the wm, so I'll probably have to find a new alternate wm soon. I'll probably be checking out the *box derivatives for one.
I'm perpetually annoyed by having to do it. "I was just here, why do I have to repeat this part of the process?" It's not really a big burden, but it's an annoyance and it really does feel like it slows me down. I have no hard numbers on this one, but then you have no hard numbers that it doesn't slow me down, so it's a draw.
Well, if you want to keep your on-top window on top, that's your preference. No system is perfect, and that's just nitpicking. Of course, you could position every single window so that it never pops up over your gaim or tvtime window ever again, so this ceases to be a real problem. Your gaim and tvtime windows are also always in the same place, because you put them there. I mean, you wouldn't put your TV in the closet when you want to watch it, so why would you put it in a place where you've got windows popping up all the time? It's the same idea, and you have complete control over it.
This is fair, but then you can always turn off the previewing option if it's such a burden to you (this is accessible in the nautilus 2.4 prefs at least). Again, tradeoffs happen, and this is just nitpicking. It also completely ignores the fact that spatial nautilus is actually much faster in practice than the old version.
When you add a new file to this directory, it'll shift all the other files that are organized after it alphabetically. This may be what you want, but it's not always ideal. If the real world worked this way, you'd be pretty freaked out. Imagine putting a book on the shelf and then the next time you turn around half the books on the shelf have moved! It's jarring. The same thing applies to files on computers. For large sets of files, like image directories, alphabetical organization is great (nautilus shouldn't remove the ability to sort alphabetically) but it's not always what you want. Spatial organization doesn't remove the organization scheme you speak of, it only adds another organizational level that works very well in practice.
I agree, but at the same time I realize that most people have absolutely no desire to become computer literate, and I don't want to force them in to such a thing. Further, I don't think spatial organization harms them from learning. If they want to learn, they will learn. And even further, if it helps them do their work better, then
Ok, I personally don't think it would be such a big deal to provide a desktop preference that's easily accessible for this kind of thing. I can understand why the gnome people didn't do it (it's a forcing device, a trick used by the original Mac people/Steve Jobs to force people to get used to the GUI paradigm) but there are plenty of good reasons to have this as an easily accessible option. Maybe for 2.8.
But on the other hand, most users really won't care about how the thing works. They generally just will use it. Aunt Tilly and challenged users really won't care.
Average users generally won't care either, but they'll care more about little things like desktop themes and sound effects and whatnot. What these users do care about is whether or not they can download mp3's, browse the web, and chat than their windowing metaphor. These people are the ones who will adopt some new program when they're told how cool it is, but they never mess around with the underlying OS unless they have to.
Power users should be able to handle this stuff with ease, and we shouldn't be concerned that they may have to google and run gconf-edit once. Besides, these people are probably using some other window manager anyway, or are at least happy to use one if they don't like what gnome's doing.
You've still got a heirarchical organization with the spatial metaphor. That's what nested directories are for, after all. But you've got that extra layer of metadata on top of it. No one is preventing you from organizing your files alphabetically in a spatial system, but you have that freedom if that's how you remember things best. Even for long-term information, knowing its physical location helps a lot. I know that my banking information is in the front of the file box while in the back are reference papers. There's still a nested heirarchy of data (bank info subdivided, etc) but I can get to it more quickly because I don't have to scan through the names of every folder in the box. This becomes even more useful when you have lots of folders in the box.
1. This guy is not an average user. He's an editor at a major computer magazine and has covered Linux stories for some time now.
2. The Gnome team did actual usability tests with plenty of real people. Apple did the same before this. The spatial concepts are neither new nor surprising.
3. There's plenty of average users who use linux, they usually just think that they're well above average. Hang around in an IRC linux support channel and you'll see what I mean. Their feedback is generally ignored for good reason.
4. The average user you're describing won't give two shits about spatial vs browser metaphor. They'll just learn it and use it if they have to because that's what's presented to them. They won't bother to go digging around for some stupid hidden option in a control panel or registry. That sort of thing defines a power user.
I'm sorry, but practical experience disproves this. If you were correct, we'd have a lot of Windows users out there who understood how file systems worked a lot better.
First off, you've re-proven my point about having to navigate two whole file trees, instead of one and a part. This is inefficient.
You do have a point about the file attributes, but with any sufficiently fast machine this should not be an issue, because the machine is usually just waiting on the human. Remember, this stuff worked just great circa 1984.
The number of total windows opened is inconsequential, unless you think double clicking is so much more of a pain. An important point is that while you're "chasing windows across the screen" is that you're not chasing them, because you know exactly where they'll pop up. This is a key point that people who haven't used spatial concepts extensively just don't understand. You know exactly where the next window will pop up because you left it there last time you closed it. Further, you know exactly where every file is in there, because they are exactly where you placed them before.
It's a lot like the concept behind radial menus, you don't have to read through everything to know where you're going. Think of it as driving home when you know your route really well versus navigating to some new place and having to slow down and read every sign along the way. This makes a huge difference in the way we work in practice. It's not about how many windows you open, it's about being able to map the data on the hard drive more efficiently in your own head.
Once again, stick to the topic at hand. Leave your personal vendettas at the door.
Agreed. That was one of the few really nice things about the System 7-> System 8 switch.
Gnome doesn't take away your command prompt or ability to write scripts, last time I checked.
Most of is not the same as the real thing. I use gnome 2.4's nautilus like you describe, and I've used Konq this way as well. It's not the same. Work with the old MacOS for a while or go read the Ars Technica article on spatial concepts. Most of the idea isn't the same thing because that's when you get a Win95-style UI.
That's a whole other argument I don't want to go near. Trolling for one desktop or the other is not the point.
When people ask me why I use Linux, one of the things I always say is "I never have to pirate software anymore." Everyone ignores it, but this story demonstrates why I always mention it. When you don't have to pirate software, you don't have to worry that some program that you need but can't afford or don't want to pay for is going to destroy your system. All my stuff comes from a much more trusted source than Limewire.
Everyone I know who uses Windows and pirates software like this has to put up with this shit. It's just not worth it, especially when you just want to get your work done. Of course, in these days where you plug your machine in and you get a host of infections automatically within a 24 hour timespan perhaps no one really worries as much about these things anymore.
There's still a lot of UI wisdom from the pre-OSX days, and to simply dismiss it is foolish.
Spatial webbrowsing is impractical because of the nature of HTML and the infinite space and chaotic organization of the network. In addition, you don't manipulate the web, you view it and interact with it. Spatial concepts become very useful when interacting with files (i.e. drag it to the trash to delete it) but isn't so necessary when all you're doing is looking at data. File browsing has a very different set of requirements and constraints.
Ultimately, what's interesting about the above is that a spatial metaphor encourages the user to interact with their data, where the portal viewing method that Windows and OSX uses is meant more for viewing, like on the web. Perhaps the reason why Apple switched to the portal viewer metaphor for OSX is that there's so much in UNIX that you're not supposed to manipulate, where in the old MacOS you could manipulate anything really.
And one thing that's very much in favor of spatial organization is that it's actually much faster to move files around than the porthole metaphor, at least if you know what you're doing. With the porthole method, you open up windows explorer (which by default is a totally separate icon/interface than starting from My Computer or whatever) and navigate to your file. Then you have to navigate to where you want to move it to on the sidebar. That's the most efficient way to do it, and you still have to bring up the sidebar, which may not expand far enough over to easily see as deep in the file tree as you need.
In contrast, with the one-window per file method, you open up each folder, holding down the shift key on each open so that the previous one is closed, and where the file tree branches off from the current file location and where it's going, you leave that window open, and keep drilling down the file tree. Then, once you get to your file, you go back to that branch point you left open and drill down the other half of the tree. This is much quicker because you rarely have to navigate two full trees, and you don't have to deal with a sidebar that's too small due to the fact that you have all this extra data hanging around. Who needs to see the whole damn file tree at a time on the sidebar? Once again, this sort of thing has no bearing when all you're doing is browsing the file tree and seeing what's there, but when actually manipulating it, it's of huge benefit.
I understand your frustration, but have you tried the new installer yet? We just released beta4, and it's gotten a lot of positive reviews and install reports so far. It's a ground-up rewrite, and while it's still in beta and is some features (like PPPoE) need to be fleshed out, it's a vast improvement over the old boot-floppies.
Our goal has been to build a much better installer, and so far we're succeeding. If you give it a try, I think you'll be pleased with the improvements.
The new installer will support PPPoE, the problem has been a lack of manpower to do so more than it being too obscure. There is basic support for it in the newest beta of the installer, but it's very basic and I'm not sure if it even works right now. Once again, the problem is lack of manpower to work on the damn thing. If you want to help us out, it'd be much appreciated, we could certaintly use it.
Great post. Really fantastic. Not 100% true, of course, since there are always the kids who do want to excel (me and my high school friends being good examples) but you're very much on target.
Interestingly enough, this effect carries over beyond high school and the like. When I tell people I'm getting my PhD in molecular bio, they stare at me like I just threw up on their shoes. They don't tease, mainly because they're stunned that anyone in real life would actually want to do something like science, but there's still that disbelief that such a thing is possible. I get this from even college educated people, which still shocks me.
Sure, they say that until they find out that they've got cancer or high blood preassure. But by then it's too late. Even then, most people never make the connection between paying taxes and living longer.