The original creator of the mouse, Douglas Engelbart, always assumed you would use a one-handed "chord" keyboard with one hand, and use the mouse with the other hand. From what I have heard, if you invested the time to learn this, you could really rock.
I just updated my Debian computer, and now my Firefox is still called Firefox, but it has a much different logo (looks like a globe). If you compare the new logo to the official one, it's the same logo but with the fox removed.
I searched the Debian email archive web pages but could not find any message saying what's going on. I don't know if this is the new official compromise or not.
For one thing, it's probably pretty expensive to make an atomic clock, let alone one that can survive launch and 10 years of radiation. It's going to be big and heavy because it has to have a decade worth of fuel, batteries, solar panels, etc.
I'll have to agree there. Never mind my cluster idea, the general idea stands: right now, they expend huge amounts of effort on making the things fail-proof, because there is nothing you can do once it's in orbit. If you can assume cheap and frequent manned spaceflight, you might be able to relax a bit, and thus save some money.
GPS satellites have already been desinged, so it might make more sense to keep making them to the same design anyway even if there is frequent cheap manned space flight. Maybe a better example is orbiting telescopes. You might just design them with a short planned life, and keep replacing them ever few years with better and better ones. Compare with Hubble which was expensive (and still managed to launch with bad optics, big scandal).
As long as it would require multiple stages, the fuel cost wouldn't go down, you're still throwing part away, and you're going to be have to pay for recovery and refurb on your reusable vehicle.
What if you used one flight of a reusable spacecraft to launch a "space taxi" that can change orbits, and another flight to bring up a load of fuel for it, and another flight to bring up the satellite? The "space taxi" would not be discarded, it would stay in orbit indefinitely. It would be relatively cheap to just keep bringing more fuel for it.
Such a "taxi" should be straightforward; it doesn't have to survive in atmosphere, so just a crew compartment, a cargo rack, some fuel tanks and some jets should do. (Disclaimer: I'm not any sort of expert.)
If we can get space flight to the point where fuel is a major part of the costs, we will have made a huge stride forward. Right now, salaries for all the legions of people are the highest cost, and expensive hardware is most of the rest.
maybe the shuttle taught us that it is not possible to make such a thing
What the Shuttle taught us is that it is not possible for NASA to make such a thing.
No one has done it yet, but it actually looks reasonable. It's not clear that we can build a single-stage-to-orbit (SSTO) reusable spacecraft with our current knowledge, but it's quite clear that we could build at least a two-stage design (TSTO) where both stages are truly reusable. And most of the experts think SSTO is doable. It won't have much capacity, but you will be able to send things up in pieces. And it will be great for moving people and supplies.
One reusable spacecraft will cost more to build than one disposable rocket, but you will save big money if you can fly the reusable one a lot.
Manned spaceflight is way more expensive then sending up another sattelite.
It is now. True reusable spacecraft will change the game very much.
Currently, you build your space objects as large single pieces, with maximum size depending on which rocket you will use to do the launch. In the future, you will build your space objects as modules that dock together, with module size chosen to be convenient for the cargo capacity of your true reusable spacecraft. It will be much cheaper to send up several launches of a reusable spacecraft than to send one giant rocket that is completely consumed in the process.
How much would it cost to fly from Seattle to Los Angeles if the plane was somehow destroyed in the process? Air travel is pretty cheap because most of the cost is the fuel. If we could build a spacecraft that is fully reusable and easy to maintain, most of the cost would then be the fuel. You could put things in orbit for not much more than the cost of flying them all the way around the world; expensive, but not insanely so.
Good grief! The Shuttle is a horrible boondoggle. It is reusable in name only. It takes months of full-time work by a standing army of hundreds of people to refurbish a shuttle to make it ready to fly again.
We need the space equivalent of a 747 airplane: something that spends more time flying than being refurbished. That will bring down launch costs a great deal.
A space elevator would be great, but I don't want to wait for one. That's much harder engineering than a truly reusable spacecraft. And you will want the reusable spacecraft to help build the elevator anyway.
Sure, the government collects a lot in taxes. But I don't think you can really figure that in to the price tag of a project like this. You could mention taxes anytime you talk about the government spending money.
IIRC, there are only 26 operating satellites, give or take.
I Googled and found that there are currently 28. Thanks for the correction!
That means $105m * 50 over the entire life of the GPS project.
We shouldn't assume we can multiply 50 by $105 million, because the earlier GPS satellites were different and probably cost more. And I don't have any data at all on R&D costs, which you could add to the price tag. I was just interested in the costs of keeping the current system going, and how many launches per year.
So, 28 satellites not 50. But that's not all. I also found that the current generation of GPS satellites (the "Block IIA" and Block IIR" satellites) have a design life of less than 8 years. The "Block II" satellites must have a longer design life, since the oldest working satellite (SVN 13) has a launch date of 1989-06-10.
Assuming all new GPS satellites are Block IIRs with a useful life of 7.8 years, and assuming we need to keep 26 GPS satellites in orbit, thats 3 to 4 launches per year to replace aging satellites, about $350 million per year.
Honestly though, do you believe the gov pays $105m for each satellite in orbit? There are plenty of ways for them to get their money back.
I don't understand this comment. "Get their money back"?
The article says the satellite costs $45 million. I Googled a bit and found that the launch cost for a Delta 2 is around $50 to $60 million. The article also said the satellite being replaced is 11 years old, and at the end of its useful life, and that there are 50 GPS satellites.
Crunching the numbers, we have about $105 million to put up a GPS satellite, with about 11 useful years; call it $10 million per year. Multiplying by 50 satellites, we have $500 million per year cost for GPS. I never knew. Also, on average, each year 4 or 5 launches must happen to replace aging GPS satellites.
Note that the launch costs are actually higher than the cost of the satellite. Also, the satellite could probably be made more cheaply if launch costs were lower (instead of over-engineering it to never break, they might just launch a cluster of two in the same orbit, or just design it to be easily repaired). If and when private companies build reusable spacecraft that can carry a GPS satellite, the cost of GPS will go down a lot. A Boeing Delta 2 is completely used up in each GPS launch right now, so truly reusable spacecraft should be able to dramatically cut launch costs and still make money.
Before you can advance the state of the art, you have to reach the state of the art.
Rhythmbox is shaping up nicely, but don't forget that it really hasn't been aroud all that long. The Rhythmbox developers might do amazing, crazy things with it, but that will have to wait until they lay the foundation by adding the features people need, first. iPod owners need iPod support.
Consider the GNOME desktop itself. At the time it was started, KDE was already working and useful, and Windows already had years of evolution. GNOME has been playing catch-up for years!
I'd say the GNOME desktop is now in many ways state-of-the-art, the major exception being the File Open and File Save dialogs. I personally also think it is essential to have some kind of "device manager" that lets you browse your hardware (see what IRQs are in use, see whether the system thinks you have USB 1.1 ports or 2.0, etc.); that's coming very soon (HAL plus DBUS plus an application and boom, you have it). So GNOME is a few short steps away from the state of the art, and will soon be able to push it forward. GNOME Storage looks interesting, for example.
Despite the efforts of Microsoft and Apple, the desktop really isn't a swiftly moving target. Most innovations (e.g. ActiveDesktop) weren't useful or popular, and have been dropped; the ones that were kept are all easy to do. Within a short time, both GNOME and KDE will be caught up to the state of the art. And that is when advances become possible.
Note, however, that sometimes the state of the art is adequate, and there is no reason to push beyond it. Cars still have a steering wheel, a gas pedal, and a brake pedal, after how many years? Why not a gamepad interface with little thumb joysticks? Answer: people are used to what we have; people like what we have; it ain't broken, so don't fix it. The current desktop model, multiple overlapping windows with some sort of panel where you can see what you have running, is well-established and popular.
Still, if you want to do something completely different, it's easier than ever now. You don't have to build a whole desktop, you can focus on just changing the behavior of one piece of an already-built desktop. You want something shockingly new? Build it and see if anyone likes it. If it really is cool, people will help you. Even if you aren't a coder, mock up some screenshots and show them around.
I won't be helping you though, sorry. I'm pretty pleased with GNOME and the way it's going already.
I dunno. I personally would be happy if the text-entry box were there all the time.
I figure I can learn Ctrl+L -- especially since Mozilla-family web browsers use Ctrl+L to type in the location bar, which to me at least is kinda the same thing.
I hear that Mac OS X lets you just start typing and a text entry box appears. Maybe a future GNOME will do that.
Meanwhile this is way better than what we have now.
GUIs are about maximising DISCOVERABILITY. GNOME really fails at this.
I don't agree. GNOME has pared away all the wacky stuff, and what's left is easy to explore, so it's easy to discover things. Perhaps in some cases they pared away a little too much, but generally you can fix it with GConf.
It's not perfect, but I'd still rather run GNOME 2.x than anything else.
Given the things you say, Debian not being able to use the Firefox branding is by design, then. So why didn't the Debian guys receive a prompt answer to the effect of "You can't ship with the logo and the name; here's your consolation logo and the name you can use." This can't have been unexpected!
And I still think it's a mistake to force a fork in the very name of the product.
I'd like to be able to see that cool Firefox logo as my Firefox icon, but I can live without it.
Because it will cause confusion. The same browser, that looks the same and works the same, will have different names depending on where you get it.
I think debian would be complaining if Lindows had just called their distro "debian". It is, after all, a modified debian.
Debian isn't proposing to change the way the software works. They need two things: the ability to distribute the software under their own free software guidelines, and the ability to make slight tweaks to the source code--for example, if the code as written doesn't compile on 68000, they need to be able to patch it so it works.
mozilla.org's plan will work; consider that you can buy "Pink Tie" Linux from cheapbytes.com (which is Red Hat Linux without any Red Hat logos or Red Hat proprietary software). The world didn't end.
But the situations aren't identical: Pink Tie Linux isn't quite the same as Red Hat (as I said, anything proprietary is gone), and Red Hat doesn't want people who buy the cheapbytes.com product calling them for tech support. Debian's Firefox will be the same as mozilla.org's Firefox, and Debian users tend to post bugs on the Debian bug database so there shouldn't be much pressure on mozilla.org's bug database.
I don't think the mozilla.org guys are trying to insult the Debian guys, claiming they won't do a good job or something. It's just that the current mozilla.org policy has no flexibility: if you change anything, all bets are off.
And very possibly, even if Debian got special permission to change things and build with the Firefox logo, the result would not be "free enough" under Debian's own free software guidelines, so Debian still wouldn't be able to ship it. (Remember, Debian is obsessive about license details so we don't have to be.)
When I first read about the trademark on the Firefox logo, and the plan that if you see that logo you know it's a quality build, I just assumed that mozilla.org had thought it through completely.
Looks like I was wrong.
Debian asked about how the logo works, and from the mozilla.org answers, it appears there is no fallback plan yet! They don't have an alternate logo available. Worse, you can't even call a modified version "Firefox" anymore? That's a problem!
Given the mozilla.org plans for trademarks, I really don't think Debian can build with the official logo and the official name. That's a shame.
If mozilla.org lets Debian use the name and logo, Debian will build Firefox for about a dozen different architectures (Power PC, 68000, Alpha, etc.) and mozilla.org won't have to do it. mozilla.org would be crazy to keep this from happening.
I suggest a compromise plan: allow the artwork and the name for any version of Firefox, but add an official "seal" logo to the about: dialog, and add "official build" to the name in several places.
This new dialog is not only much more confusing looking, but seems bloated
I disagree. It has many new features compared to the old dialog, and they are cleanly laid out. You have bookmarks now, to quickly go to a folder you use often; there is a preview available now; and there are many different ways to quickly get to the folder you want (e.g. you can go up two folders with one click; you can go to your home directory with one click; etc.)
rather ugly
Matter of taste. That screenshot is using a theme I don't personally like, but in a more soothing theme, the new dialogs look just fine.
and doesn't have the text entry box -- i.e, they removed the one great feature they used to have!
Calm down. The text-entry box is still there if you want it. If it's not showing, as in that screenshot, Ctrl+L will make it appear. If you are a keyboard fan, you shouldn't have much trouble hitting one extra keystroke.
For a Save dialog, you don't even have to hit Ctrl+L; it's only the Open dialog that defaults to mouse-only operation.
I know they're attempting to appeal to inexperienced users, but they always seem to (1) do so in a way that pisses off experienced users, and (2) botch things up in the inexperienced-user case anyway.
During the months of discussion and testing before this release, did you provide any feedback to help them? If not, then perhaps you might want to hold back a bit on the abuse directed towards the GTK developers.
Hopefully someone will come up with a less crappy file-selector and all the major distros (at least debian) will use it.
Hey, it's free software. Fire up your favorite image editor, and start mocking up how it should look. I'm sure OSNews would publish an article about your new design, and I'm sure that someone, somewhere in the world, would code up a prototype for you. Or you could even code it yourself!
As for me, I am content with the new dialog and I'm looking forward to its arrival in Debian Unstable.
You have SCO, planning to sue everyone on the face of the Earth until they can collect a "license fee" on every *NIX system, including Linux and BSD. You have patents being granted on new inventions like "use the Internet to sell things". And you have vendors of proprietary software becoming increasingly nervous about the competition from free software; they might decide to play the lawsuit card.
It's not unthinkable that a company would sue end-users directly to "make an example" out of them; SCO already did just that, to AutoZone and DaimlerChrysler.
There are legal threats out there. Insurance against them isn't silly.
What I mean is: suppose he had donated the initial Linux kernel to the world, and then walked away from it? Without his management, we would still have some sort of Linux today. But I believe it would not be as good as it is.
If something took Linus away from Linux now, it would be sad but the Linux kernel would continue. I'm not trying to claim he's essential. But he is extrememly valuable.
[Linus] has coordinated a great deal of it, but linux is no longer his.
I disagree. He has done an outstanding job of managing the dev process for Linux. One of the most valuable things he does is to say "no": he won't let Linux bloat up needlessly and he won't accept badly-written patches. He has a vision for the kernel that has turned out to be excellent; he ignored the prevailing wisdom of the day that microkernel was the way to go, and that worked out (look where HURD is compared to Linux), he pushes at all times for simplicity (consider his interactivity boost, consider his plans to replace numeric IOCTLs with file-like semantics).
He's not perfect. His continuing refusal to accept kernel debugger hooks in the mainline kernel is silly (he has claimed that kernel debuggers are a crutch, for those who don't fully grok the kernel).
But without Linus, the Linux kernel would not be as amazingly great as it is today.
AFIK the 2.4.x kernel with the premptive hack in the kernel oly supported premption of opeation at certain predefiend points and not at arbitrarily.
Read the article I linked. The kernel can preempt anytime it is outside an SMP spinlock. This is true for 2.4 and 2.6. On a uniprocessor system, there aren't actually SMP spinlocks, but the preemption code took adavantage of the SMP spinlock code.
It sure looks like a troll. You toss around technical-sounding terms and make vague assertions. I've followed the kernel preemption patch as it appears in the news, and I don't remember ever seeing the words "preemption points" so I don't know what you are talking about. Linux kernel preemption allows preemption except when an SMP spinlock would have been invoked. What are these "preemption points"? Can you provide a URL to a page that explains them?
anyone familair with the way something like Solaris works knows how primitive the thread support in linux is.
If you don't want people to think you are a troll, don't say things like this without pointing to some kind of reference that backs this up. Since Linux kernel 2.6 just got the NPTL, threads are way better than they used to be. Are you claiming that Solaris is way better than NPTL on kernel 2.6? Or did you not know about NPTL? Or are you just trolling?
the new scheduler is quite controversial, i wouldnt be surprised if they return to the previous one later
References for this statement, please? All the articles I have been reading are very enthusiastic about the new scheduler, especially with the interactivity boost.
Are you sure about the X nice level? Your symptoms sound exactly like what happened to me when I ran 2.6 for the first time, and my problem was the X nice level.
For 2.6, you want X to run at nice 0. Many Linux distros set X to nice -10 for kernel 2.4 and older, but for 2.6 that gums up the works.
Debian users can fix it like so:
dpkg-reconfigure xserver-common
Then, when it asks you what X nice level you want, set it to zero.
I have always had good luck with HP printers. And HP gets a gold star for their Linux driver support: the Linux drivers are fully free software and have all the features of the Windows drivers.
HP has made a few cheap, junky printer models, and I don't recommend those. You should try getting her an HP DeskJet 895C, 970C, or 990C off eBay, and see if that doesn't sort out her issues. Or if she doesn't care about color, you could even get her a LaserJet 1200 series; those just print PostScript (well, a compatible clone) so there aren't any driver issues.
My Dad clings stubbornly to his HP DeskJet and HP DeskJet Plus! They are slow, but built like tanks, and his ancient DOS word processor can talk to them.
One thing I really like about HP DeskJet printers: when you swap the ink cartridge, you are also swapping the print head. My DeskJet printers over the years have had few problems, and usually swapping out the ink cartridge fixed the problem. (If the problem is clogged nozzles, use the printhead cleaning features from the "toolbox". Alas, the toolbox hasn't been ported to Linux yet.)
If I could, I would mod the parent up as +1, Interesting. But I'll just have to reply instead.
First of all, I did some digging, and discovered a charming utility called "gnome-open". It does the same thing as "start". I just tried gnome-open on one of my.ogg music files, and it's playing now in the default.ogg player.
I am intrigued by the idea of a pipe-oriented application to handle File Open and File Save. I don't think it would actually be as big a win as you think. It's possible to use shared libraries, so there is no reason to statically link the libraries. And with the GNOME 2.6 dialogs, it is possible for an app to add extra controls to the dialogs; see the infamous screen shot with "Frob the file" (or the newer "Lart the next user who asks about this checkbox"). You propose switches and pipes for this, but calling a shared library means an application can simply register callbacks--easy and fast.
There are already command-line tools that pop up GUI dialogs, for use in your shell scripts. For GNOME 2.x, the tool of choice is Zenity, which can do what you wanted: for example, it can put up a file selector dialog and return the chosen file on the standard output.
Zenity doesn't currently offer a printer chooser, color chooser, or font chooser. You could probably work around these lacks; for example, you could call gtklp instead of Zenity to deal with printing. Or you could add these features to Zenity.
You could also script in Python instead of shell, and you can do whatever you want (including design nontrivial dialogs in Glade).
I can't believe anyone is actually trying to argue that Debian is easier to install/upgrade than Mac OS X!
Install -- no. But Debian is so easy to upgrade, and you don't even have to reboot.
When the Progeny guys finish getting the Red Hat installer working with Debian, it will be a lot easier to install than it currently is, too. I haven't installed OS X so I cannot compare.
Slashdot Mirror
The original creator of the mouse, Douglas Engelbart, always assumed you would use a one-handed "chord" keyboard with one hand, and use the mouse with the other hand. From what I have heard, if you invested the time to learn this, you could really rock.
http://sloan.stanford.edu/mousesite/1968Demo.html
steveha
I just updated my Debian computer, and now my Firefox is still called Firefox, but it has a much different logo (looks like a globe). If you compare the new logo to the official one, it's the same logo but with the fox removed.
I searched the Debian email archive web pages but could not find any message saying what's going on. I don't know if this is the new official compromise or not.
steveha
For one thing, it's probably pretty expensive to make an atomic clock, let alone one that can survive launch and 10 years of radiation. It's going to be big and heavy because it has to have a decade worth of fuel, batteries, solar panels, etc.
I'll have to agree there. Never mind my cluster idea, the general idea stands: right now, they expend huge amounts of effort on making the things fail-proof, because there is nothing you can do once it's in orbit. If you can assume cheap and frequent manned spaceflight, you might be able to relax a bit, and thus save some money.
GPS satellites have already been desinged, so it might make more sense to keep making them to the same design anyway even if there is frequent cheap manned space flight. Maybe a better example is orbiting telescopes. You might just design them with a short planned life, and keep replacing them ever few years with better and better ones. Compare with Hubble which was expensive (and still managed to launch with bad optics, big scandal).
As long as it would require multiple stages, the fuel cost wouldn't go down, you're still throwing part away, and you're going to be have to pay for recovery and refurb on your reusable vehicle.
What if you used one flight of a reusable spacecraft to launch a "space taxi" that can change orbits, and another flight to bring up a load of fuel for it, and another flight to bring up the satellite? The "space taxi" would not be discarded, it would stay in orbit indefinitely. It would be relatively cheap to just keep bringing more fuel for it.
Such a "taxi" should be straightforward; it doesn't have to survive in atmosphere, so just a crew compartment, a cargo rack, some fuel tanks and some jets should do. (Disclaimer: I'm not any sort of expert.)
If we can get space flight to the point where fuel is a major part of the costs, we will have made a huge stride forward. Right now, salaries for all the legions of people are the highest cost, and expensive hardware is most of the rest.
steveha
maybe the shuttle taught us that it is not possible to make such a thing
What the Shuttle taught us is that it is not possible for NASA to make such a thing.
No one has done it yet, but it actually looks reasonable. It's not clear that we can build a single-stage-to-orbit (SSTO) reusable spacecraft with our current knowledge, but it's quite clear that we could build at least a two-stage design (TSTO) where both stages are truly reusable. And most of the experts think SSTO is doable. It won't have much capacity, but you will be able to send things up in pieces. And it will be great for moving people and supplies.
One reusable spacecraft will cost more to build than one disposable rocket, but you will save big money if you can fly the reusable one a lot.
steveha
Manned spaceflight is way more expensive then sending up another sattelite.
It is now. True reusable spacecraft will change the game very much.
Currently, you build your space objects as large single pieces, with maximum size depending on which rocket you will use to do the launch. In the future, you will build your space objects as modules that dock together, with module size chosen to be convenient for the cargo capacity of your true reusable spacecraft. It will be much cheaper to send up several launches of a reusable spacecraft than to send one giant rocket that is completely consumed in the process.
How much would it cost to fly from Seattle to Los Angeles if the plane was somehow destroyed in the process? Air travel is pretty cheap because most of the cost is the fuel. If we could build a spacecraft that is fully reusable and easy to maintain, most of the cost would then be the fuel. You could put things in orbit for not much more than the cost of flying them all the way around the world; expensive, but not insanely so.
steveha
Good grief! The Shuttle is a horrible boondoggle. It is reusable in name only. It takes months of full-time work by a standing army of hundreds of people to refurbish a shuttle to make it ready to fly again.
We need the space equivalent of a 747 airplane: something that spends more time flying than being refurbished. That will bring down launch costs a great deal.
A space elevator would be great, but I don't want to wait for one. That's much harder engineering than a truly reusable spacecraft. And you will want the reusable spacecraft to help build the elevator anyway.
steveha
Sure, the government collects a lot in taxes. But I don't think you can really figure that in to the price tag of a project like this. You could mention taxes anytime you talk about the government spending money.
steveha
IIRC, there are only 26 operating satellites, give or take.
I Googled and found that there are currently 28. Thanks for the correction!
That means $105m * 50 over the entire life of the GPS project.
We shouldn't assume we can multiply 50 by $105 million, because the earlier GPS satellites were different and probably cost more. And I don't have any data at all on R&D costs, which you could add to the price tag. I was just interested in the costs of keeping the current system going, and how many launches per year.
So, 28 satellites not 50. But that's not all. I also found that the current generation of GPS satellites (the "Block IIA" and Block IIR" satellites) have a design life of less than 8 years. The "Block II" satellites must have a longer design life, since the oldest working satellite (SVN 13) has a launch date of 1989-06-10.
Assuming all new GPS satellites are Block IIRs with a useful life of 7.8 years, and assuming we need to keep 26 GPS satellites in orbit, thats 3 to 4 launches per year to replace aging satellites, about $350 million per year.
Honestly though, do you believe the gov pays $105m for each satellite in orbit? There are plenty of ways for them to get their money back.
I don't understand this comment. "Get their money back"?
steveha
The article says the satellite costs $45 million. I Googled a bit and found that the launch cost for a Delta 2 is around $50 to $60 million. The article also said the satellite being replaced is 11 years old, and at the end of its useful life, and that there are 50 GPS satellites.
Crunching the numbers, we have about $105 million to put up a GPS satellite, with about 11 useful years; call it $10 million per year. Multiplying by 50 satellites, we have $500 million per year cost for GPS. I never knew. Also, on average, each year 4 or 5 launches must happen to replace aging GPS satellites.
Note that the launch costs are actually higher than the cost of the satellite. Also, the satellite could probably be made more cheaply if launch costs were lower (instead of over-engineering it to never break, they might just launch a cluster of two in the same orbit, or just design it to be easily repaired). If and when private companies build reusable spacecraft that can carry a GPS satellite, the cost of GPS will go down a lot. A Boeing Delta 2 is completely used up in each GPS launch right now, so truly reusable spacecraft should be able to dramatically cut launch costs and still make money.
steveha
Don't forget Hastings's Law:
Before you can advance the state of the art, you have to reach the state of the art.
Rhythmbox is shaping up nicely, but don't forget that it really hasn't been aroud all that long. The Rhythmbox developers might do amazing, crazy things with it, but that will have to wait until they lay the foundation by adding the features people need, first. iPod owners need iPod support.
Consider the GNOME desktop itself. At the time it was started, KDE was already working and useful, and Windows already had years of evolution. GNOME has been playing catch-up for years!
I'd say the GNOME desktop is now in many ways state-of-the-art, the major exception being the File Open and File Save dialogs. I personally also think it is essential to have some kind of "device manager" that lets you browse your hardware (see what IRQs are in use, see whether the system thinks you have USB 1.1 ports or 2.0, etc.); that's coming very soon (HAL plus DBUS plus an application and boom, you have it). So GNOME is a few short steps away from the state of the art, and will soon be able to push it forward. GNOME Storage looks interesting, for example.
Despite the efforts of Microsoft and Apple, the desktop really isn't a swiftly moving target. Most innovations (e.g. ActiveDesktop) weren't useful or popular, and have been dropped; the ones that were kept are all easy to do. Within a short time, both GNOME and KDE will be caught up to the state of the art. And that is when advances become possible.
Note, however, that sometimes the state of the art is adequate, and there is no reason to push beyond it. Cars still have a steering wheel, a gas pedal, and a brake pedal, after how many years? Why not a gamepad interface with little thumb joysticks? Answer: people are used to what we have; people like what we have; it ain't broken, so don't fix it. The current desktop model, multiple overlapping windows with some sort of panel where you can see what you have running, is well-established and popular.
Still, if you want to do something completely different, it's easier than ever now. You don't have to build a whole desktop, you can focus on just changing the behavior of one piece of an already-built desktop. You want something shockingly new? Build it and see if anyone likes it. If it really is cool, people will help you. Even if you aren't a coder, mock up some screenshots and show them around.
I won't be helping you though, sorry. I'm pretty pleased with GNOME and the way it's going already.
steveha
There's a big database on there
What's more, Apple didn't share any of the details needed for free software to support the iPod. People had to figure it out on their own.
steveha
I dunno. I personally would be happy if the text-entry box were there all the time.
I figure I can learn Ctrl+L -- especially since Mozilla-family web browsers use Ctrl+L to type in the location bar, which to me at least is kinda the same thing.
I hear that Mac OS X lets you just start typing and a text entry box appears. Maybe a future GNOME will do that.
Meanwhile this is way better than what we have now.
GUIs are about maximising DISCOVERABILITY. GNOME really fails at this.
I don't agree. GNOME has pared away all the wacky stuff, and what's left is easy to explore, so it's easy to discover things. Perhaps in some cases they pared away a little too much, but generally you can fix it with GConf.
It's not perfect, but I'd still rather run GNOME 2.x than anything else.
steveha
steveha
Given the things you say, Debian not being able to use the Firefox branding is by design, then. So why didn't the Debian guys receive a prompt answer to the effect of "You can't ship with the logo and the name; here's your consolation logo and the name you can use." This can't have been unexpected!
And I still think it's a mistake to force a fork in the very name of the product.
I'd like to be able to see that cool Firefox logo as my Firefox icon, but I can live without it.
steveha
How is that a problem?
Because it will cause confusion. The same browser, that looks the same and works the same, will have different names depending on where you get it.
I think debian would be complaining if Lindows had just called their distro "debian". It is, after all, a modified debian.
Debian isn't proposing to change the way the software works. They need two things: the ability to distribute the software under their own free software guidelines, and the ability to make slight tweaks to the source code--for example, if the code as written doesn't compile on 68000, they need to be able to patch it so it works.
mozilla.org's plan will work; consider that you can buy "Pink Tie" Linux from cheapbytes.com (which is Red Hat Linux without any Red Hat logos or Red Hat proprietary software). The world didn't end.
But the situations aren't identical: Pink Tie Linux isn't quite the same as Red Hat (as I said, anything proprietary is gone), and Red Hat doesn't want people who buy the cheapbytes.com product calling them for tech support. Debian's Firefox will be the same as mozilla.org's Firefox, and Debian users tend to post bugs on the Debian bug database so there shouldn't be much pressure on mozilla.org's bug database.
I don't think the mozilla.org guys are trying to insult the Debian guys, claiming they won't do a good job or something. It's just that the current mozilla.org policy has no flexibility: if you change anything, all bets are off.
And very possibly, even if Debian got special permission to change things and build with the Firefox logo, the result would not be "free enough" under Debian's own free software guidelines, so Debian still wouldn't be able to ship it. (Remember, Debian is obsessive about license details so we don't have to be.)
steveha
When I first read about the trademark on the Firefox logo, and the plan that if you see that logo you know it's a quality build, I just assumed that mozilla.org had thought it through completely.
Looks like I was wrong.
Debian asked about how the logo works, and from the mozilla.org answers, it appears there is no fallback plan yet! They don't have an alternate logo available. Worse, you can't even call a modified version "Firefox" anymore? That's a problem!
Given the mozilla.org plans for trademarks, I really don't think Debian can build with the official logo and the official name. That's a shame.
If mozilla.org lets Debian use the name and logo, Debian will build Firefox for about a dozen different architectures (Power PC, 68000, Alpha, etc.) and mozilla.org won't have to do it. mozilla.org would be crazy to keep this from happening.
I suggest a compromise plan: allow the artwork and the name for any version of Firefox, but add an official "seal" logo to the about: dialog, and add "official build" to the name in several places.
steveha
This new dialog is not only much more confusing looking, but seems bloated
I disagree. It has many new features compared to the old dialog, and they are cleanly laid out. You have bookmarks now, to quickly go to a folder you use often; there is a preview available now; and there are many different ways to quickly get to the folder you want (e.g. you can go up two folders with one click; you can go to your home directory with one click; etc.)
rather ugly
Matter of taste. That screenshot is using a theme I don't personally like, but in a more soothing theme, the new dialogs look just fine.
and doesn't have the text entry box -- i.e, they removed the one great feature they used to have!
Calm down. The text-entry box is still there if you want it. If it's not showing, as in that screenshot, Ctrl+L will make it appear. If you are a keyboard fan, you shouldn't have much trouble hitting one extra keystroke.
For a Save dialog, you don't even have to hit Ctrl+L; it's only the Open dialog that defaults to mouse-only operation.
I know they're attempting to appeal to inexperienced users, but they always seem to (1) do so in a way that pisses off experienced users, and (2) botch things up in the inexperienced-user case anyway.
During the months of discussion and testing before this release, did you provide any feedback to help them? If not, then perhaps you might want to hold back a bit on the abuse directed towards the GTK developers.
Hopefully someone will come up with a less crappy file-selector and all the major distros (at least debian) will use it.
Hey, it's free software. Fire up your favorite image editor, and start mocking up how it should look. I'm sure OSNews would publish an article about your new design, and I'm sure that someone, somewhere in the world, would code up a prototype for you. Or you could even code it yourself!
As for me, I am content with the new dialog and I'm looking forward to its arrival in Debian Unstable.
steveha
...when they really are out to get you.
You have SCO, planning to sue everyone on the face of the Earth until they can collect a "license fee" on every *NIX system, including Linux and BSD. You have patents being granted on new inventions like "use the Internet to sell things". And you have vendors of proprietary software becoming increasingly nervous about the competition from free software; they might decide to play the lawsuit card.
It's not unthinkable that a company would sue end-users directly to "make an example" out of them; SCO already did just that, to AutoZone and DaimlerChrysler.
There are legal threats out there. Insurance against them isn't silly.
steveha
What I mean is: suppose he had donated the initial Linux kernel to the world, and then walked away from it? Without his management, we would still have some sort of Linux today. But I believe it would not be as good as it is.
If something took Linus away from Linux now, it would be sad but the Linux kernel would continue. I'm not trying to claim he's essential. But he is extrememly valuable.
steveha
[Linus] has coordinated a great deal of it, but linux is no longer his.
I disagree. He has done an outstanding job of managing the dev process for Linux. One of the most valuable things he does is to say "no": he won't let Linux bloat up needlessly and he won't accept badly-written patches. He has a vision for the kernel that has turned out to be excellent; he ignored the prevailing wisdom of the day that microkernel was the way to go, and that worked out (look where HURD is compared to Linux), he pushes at all times for simplicity (consider his interactivity boost, consider his plans to replace numeric IOCTLs with file-like semantics).
He's not perfect. His continuing refusal to accept kernel debugger hooks in the mainline kernel is silly (he has claimed that kernel debuggers are a crutch, for those who don't fully grok the kernel).
But without Linus, the Linux kernel would not be as amazingly great as it is today.
steveha
AFIK the 2.4.x kernel with the premptive hack in the kernel oly supported premption of opeation at certain predefiend points and not at arbitrarily.
Read the article I linked. The kernel can preempt anytime it is outside an SMP spinlock. This is true for 2.4 and 2.6. On a uniprocessor system, there aren't actually SMP spinlocks, but the preemption code took adavantage of the SMP spinlock code.
steveha
WTF why is this troll.
It sure looks like a troll. You toss around technical-sounding terms and make vague assertions. I've followed the kernel preemption patch as it appears in the news, and I don't remember ever seeing the words "preemption points" so I don't know what you are talking about. Linux kernel preemption allows preemption except when an SMP spinlock would have been invoked. What are these "preemption points"? Can you provide a URL to a page that explains them?
anyone familair with the way something like Solaris works knows how primitive the thread support in linux is.
If you don't want people to think you are a troll, don't say things like this without pointing to some kind of reference that backs this up. Since Linux kernel 2.6 just got the NPTL, threads are way better than they used to be. Are you claiming that Solaris is way better than NPTL on kernel 2.6? Or did you not know about NPTL? Or are you just trolling?
the new scheduler is quite controversial, i wouldnt be surprised if they return to the previous one later
References for this statement, please? All the articles I have been reading are very enthusiastic about the new scheduler, especially with the interactivity boost.
steveha
Are you sure about the X nice level? Your symptoms sound exactly like what happened to me when I ran 2.6 for the first time, and my problem was the X nice level.
For 2.6, you want X to run at nice 0. Many Linux distros set X to nice -10 for kernel 2.4 and older, but for 2.6 that gums up the works.
Debian users can fix it like so:
dpkg-reconfigure xserver-common
Then, when it asks you what X nice level you want, set it to zero.
steveha
I have always had good luck with HP printers. And HP gets a gold star for their Linux driver support: the Linux drivers are fully free software and have all the features of the Windows drivers.
HP has made a few cheap, junky printer models, and I don't recommend those. You should try getting her an HP DeskJet 895C, 970C, or 990C off eBay, and see if that doesn't sort out her issues. Or if she doesn't care about color, you could even get her a LaserJet 1200 series; those just print PostScript (well, a compatible clone) so there aren't any driver issues.
My Dad clings stubbornly to his HP DeskJet and HP DeskJet Plus! They are slow, but built like tanks, and his ancient DOS word processor can talk to them.
One thing I really like about HP DeskJet printers: when you swap the ink cartridge, you are also swapping the print head. My DeskJet printers over the years have had few problems, and usually swapping out the ink cartridge fixed the problem. (If the problem is clogged nozzles, use the printhead cleaning features from the "toolbox". Alas, the toolbox hasn't been ported to Linux yet.)
steveha
If I could, I would mod the parent up as +1, Interesting. But I'll just have to reply instead.
.ogg music files, and it's playing now in the default .ogg player.
First of all, I did some digging, and discovered a charming utility called "gnome-open". It does the same thing as "start". I just tried gnome-open on one of my
I am intrigued by the idea of a pipe-oriented application to handle File Open and File Save. I don't think it would actually be as big a win as you think. It's possible to use shared libraries, so there is no reason to statically link the libraries. And with the GNOME 2.6 dialogs, it is possible for an app to add extra controls to the dialogs; see the infamous screen shot with "Frob the file" (or the newer "Lart the next user who asks about this checkbox"). You propose switches and pipes for this, but calling a shared library means an application can simply register callbacks--easy and fast.
There are already command-line tools that pop up GUI dialogs, for use in your shell scripts. For GNOME 2.x, the tool of choice is Zenity, which can do what you wanted: for example, it can put up a file selector dialog and return the chosen file on the standard output.
Zenity doesn't currently offer a printer chooser, color chooser, or font chooser. You could probably work around these lacks; for example, you could call gtklp instead of Zenity to deal with printing. Or you could add these features to Zenity.
You could also script in Python instead of shell, and you can do whatever you want (including design nontrivial dialogs in Glade).
steveha
I can't believe anyone is actually trying to argue that Debian is easier to install/upgrade than Mac OS X!
Install -- no. But Debian is so easy to upgrade, and you don't even have to reboot.
When the Progeny guys finish getting the Red Hat installer working with Debian, it will be a lot easier to install than it currently is, too. I haven't installed OS X so I cannot compare.
steveha