when you customise a good interface you invariably make it less productive and just don't realise it. Secondly, it is extremely poor user interface design to have two modes - one for new users and one for power users.
Very funny. Tell me another good one -- how about that joke where it's faster to pull down menus than to hit accelerator keys?
(Some Mac guys claim this, seriously. They say that it is slower to hit Ctrl+S or whatever than to take my hand off the keyboard, find the mouse, move the pointer, pull down a menu, find the item I want, click on it, and put my hand back on the keyboard.)
You know, I've got both Galeon and Mozilla customized so that when I click my middle mouse button on a link, that link opens in a new tab. This isn't the default behavior. According to you, I am suffering reduced productivity because I now spin up new tabs with a single button click.
The beauty of a customizable interface is that it can adapt to the way you want to work. I frequently want to open links in a new tab, so it saves me time and effort to turn a menu operation into a single middle-click. You cannot convince me that I am not better off.
I will meet you halfway: it's not a good idea to take really fundamental things and make them odd. I don't think the ability to customize the single-click behavior of your mouse would be a very good thing; it would ruin you for using other computers, and ruin your computer for others to use.
As for having two modes, I think it is a very good idea. The only problem I see with customizations is when Joe customizes his setup his way, and then I want to use Joe's computer. If I can kick it into non-customized mode for a moment, I can use his computer very well, and both of us are happier.
I looked into using a Palm for a bike computer, and concluded that I wasn't very interested. I live in the Seattle area, and I ride in the rain. The BikeBrain solution comes with a plastic protector for your Palm, but it isn't really waterproof.
The good thing of course is that a Palm can capture a lot of data. But just capturing wheel spin data to show speed and distance isn't enough to make me buy either of these solutions.
Last autumn I bought myself a Specialized P.Brain computer. I love it; it collects wheel turn data (like the two Palm solutions) and also altitude and heart rate data. A PC interface lets you capture your data and make pretty charts. You can get a graph showing your speed, altitude, and heart rate plotted against either time or distance. Read more about it here.
The PC download software is for Windows; I'm planning to try to get it working under WINE if I can. The data is stored in some opaque binary format, but you can get the data out with Dan Connelly's Perl script (get it here.
The P.Brain isn't the only data-collecting bike computer. There are other brands. I have heard good things about the Polar XTrainer. There are even computer systems that directly measure your power output; you have a wheel built with a power-measuring hub, and the computer keeps track of power. Pro riders (including Lance Armstrong) use these. For example, the Power-Tap.
steveha
Re:Very, very worthwhile for cyclists.
on
Palm on a Bicycle
·
· Score: 2
It looks like BikeBrain has an altimeter in it, so it can also give you data like what your best time on a hill was, adjusted by the grade of the hill.
It looks like it has an altimeter, but it doesn't.
It keeps track of how far you have gone on your ride, and if it has a route plan that includes altitude, it shows what your altitude ought to be if you are following the route plan. The altitude graph, alas, will not work right if you alter the route slightly, or ride some completely new route.
There are bike computer systems with altimeters built in, and one of those would be better.
What I would love to see is everyone who is working on anything remotely redundant to drop what they are doing, put their collective heads together and come up with a real competitor for Microsoft
To some extent, redundant projects are a good thing. If two projects are competing, they can spur each other on to better results. Look at GNOME and KDE, for example.
Also, when there are two projects, the potential downside is reduced for trying something new. If GNOME bets on Mono, and KDE steers clear of it, then if Mono turns out to be a bad idea, we can all switch to KDE. (But note that it might be easier just to fork GNOME and switch to the non-mono fork, especially for those of us already using GNOME.)
But it's moot anyway. People work on whatever they want to work on. I think the world has enough text editors, but no one cares what I think; if some guy wants to write a text editor, he's going to do it. Free software, freedom. Nothing you can do about it, so why worry about it?
I can use netcat and ExtractStream to (usually) get a TV show onto one of my computers for editing and reencoding.
Whoa, could you tell me more? You are just about to sell me a TiVo.
What quality does the TiVo capture at? Can it capture full 640x480 TV pictures? (Sometimes called 720x480, if you are talking about entire scan lines and not just the square-pixels viewable part.)
I have no problem with paying money to TiVo for their box. I have no problem paying monthly for TV listings. I have a problem with doing the above and then not having any way to burn my own DVDs; I know you can hook a VCR up to a TiVo but I'm not interested. I have been saying for years that I would buy a TiVo when I can put one on my home network and grab video for DVD burning. Maybe that day has come?
the guy does a lot of great work and all, but most influential INNOVATORS??? We're talking a Unix clone here.
Linux is not an exact clone of UNIX. The first version of Linux was a clone of the most important bits of UNIX, but it has since grown in many different directions.
For one example, consider the HTTP daemon that runs inside the kernel, serving up static pages without leaving kernel mode and seamlessly handing dynamic pages off to Apache. There had never been anything like it before (although Microsoft has copied that idea for their web server).
Read Linux Weekly News. There is always new stuff shaking in the kernel development. An O(1) scheduler! Copy-on-write to make fork() really fast! Reverse-mapping in the VM subsystem! I don't know enough about the OS landscape to tell you which of these exist in which OSes, but I'm pretty sure that many of them are brand-new.
It's like some company made the most advanced car ever, and it had hundreds of cool little features that make it safer or nicer or whatever. If you are standing next to it, you just say "It's a car." But an automotive engineer can look under the hood and know better.
Look under the hood of Linux development.
steveha
Re:Linux on the desktop is a matter of ambition
on
Alan Cox Interview
·
· Score: 2
So, why aren't we using more gui installers?
Maybe you should start up a project to make a standard GUI installer that everyone can use.
The main benefit to the viewer is that you will see the same movie on opening night, and four weeks later. Currently, film prints can get scratched; they can break and get (possibly poorly) patched; and you can often see a visual glitch when one reel of film finishes and another reel starts up. The problems with film can get worse, much worse, if you have an idiot or a klutz as a projectionist, or if the movie house is really cheap and doesn't take proper care of the film and/or the projection equipment.
The number one fear I have about digital is that a projector might get one or more stuck pixels, and then there will be wrong pixels in the same places for the whole movie. So we are getting rid of all the little annoying errors that accrete over time with film, but possibly gaining annoying errors that last all movie (not changing until the projector is repaired!).
Another thing: this is the wild-and-wooly early days of digital projection. We will no doubt go through several rounds of standards changes and upgrades. (I haven't seen 1920x1200 on a big screen--maybe it's really gorgous--but it is hard for me to imagine that this level of quality will still be in use 20 years from now.) So the theater owners may invest in expensive projectors, and then find that they have to buy newer and even more expensive projectors to keep up.
One benefit to theater owners, however: they won't have to pay as much for the actual distribution media the film comes on. If they have multiple theaters that want to show the film, currently they need multiple prints. With digital, I don't know for sure how they will do it, but I imagine they will ship out the movie on magnetic tape, and load up one or more RAID arrays at the theater to drive the projectors. (Anyone who knows for sure, please speak up.)
"updatedb" was running, eating plenty of system ressources
I have an old box that is running an older version of Mandrake Linux. The "logrotate" command runs at odd intervals (frequently during the work day, not in the middle of the night) and the computer is extremely slow when that happens. (I'm looking forward to installing Debian on that computer. I understand Debian, and I've never had a problem with logrotate on Debian anyway.)
The worst was when the hard drive filled up. The logrotate command was running continuously; the hard drive was rattling nonstop. I discovered that I had files like this:
I managed to kill the logrotate process. Then I ran a find command to find all files that had a ".gz" in the name, and delete them. It took over 10 minutes to find and delete them all! There were thousands, many whose filenames were over 80 characters from ".gz.gz.gz.gz..." extensions.
Once/var/log was clean, I nuked some junk I didn't need, and the computer was decently fast again.
(Now I understand well why/var should be on its own partition! When/tmp is full, bad things happen. If you only have one partition, when/var is full,/tmp is full... and when you have too many.MP3 files, you can set off an unpleasant chain reaction.)
The only problem that a CLR supposedly solves is the maintanence of the bindings. Instead of binding Gtk to perl and python and ada and C and C++, etc., you bind it in a library in the CLR. Except that to access that new CLR binding, you need to have perl and python and ada and C and C++ compilers that target the CLR, which is certainly a more glamorous job than maintaining bindings for every language under the sun, but is *WAY MORE* complex.
Nice summary. The big win with a CLR is that you need to write Perl/CLR once, and Python/CLR once, and so on; then you write each class library and interface for Mono once and everything just works.
Compare with the current system: for each new interface you add, you need to write the Python bindings, the Perl bindings, etc. etc. So you need to write an M*N matrix of bindings, where M is the number of interfaces and N is the number of supported languages.
Assuming the CLR works as Miguel thinks it will, you go from M*N to M+N, a huge win, even if writing the CLR interfaces for each language is much harder than writing some bindings. And the larger M and N become, the bigger the win becomes.
HP talks a good Linux, but does approximately nil to further the cause, or even make their products work with Linux.
I like HP inkjet printers. I strongly believe they're more reliable than Epson's. But I've bought three Epson printers in the last two years, only because they worked better (or at all) with Linux.
HP devlopers wrote a printer driver for Linux. It was initially released with a license similar to the BSD license, with an extra clause saying that to use it you needed to own an HP printer. They promised to have their lawyers make sure they were clear on patent issues, and then drop the extra clause if possible. Guess what, they did drop it, and now it is available as a purely free piece of software. (Get it here.)
This makes HP the only company to have released a free software driver for its products. Linux has excellent support for Epson printers, but Epson didn't do the work.
If you want to print photos, you are still better off under Linux with an Epson, because none of the Linux drivers for HP DeskJet printers support the 2400x1200 DPI photo printing mode yet. But it is just a matter of time. I'm hoping that HP will add that themselves to their own free driver.
If 600x600 DPI color printing is enough for you, you can use the HP DeskJet with Linux.
I claimed that the KDE developers, when given a choice between Qt and a free clone of Qt, chose Qt... despite the non-free status of Qt at the time. Then I introduced that article to support my claim.
The 1998 copyright date doesn't matter, all that matters is whether the reference is accurate and supports my claim.
HTML spam is much worse than plain text spam. I always set my mail client to prefer plain text, because you can't embed cookies in URLs to images in plain text. Doing so in an HTML mail gives a clear indication that the mail address is valid (as the image has been requested, the mail has been recieved and read).
This is why the Evolution mail client is now my favorite. It lets you read HTML mail, but by default it will never put any hits on any servers; it only renders the passive tags (bold, italics, link, header, etc. but nothing downloaded from a server).
KDE/GNOME is a bit off-topic for a discussion of the license of Mono's class libraries, but what the heck, I'm game.
the KDE project was based on Qt due to it being a nice toolkit and due to the naivete of the core team. They just weren't aware of the implications, and couldn't understand it because, let's face it, programmers aren't lawyers.
This excuse does not fly. Perhaps it could have been true for a short time, but lots of people pointed out that KDE was depending on non-free software. KDE could have adopted the Harmony project (free library compatible with Qt) but they chose not to do so.
(There was an essay about this, with a title something like "Why KDE still isn't a good idea". I've searched the web for a long time and I couldn't find it. I think I read it on Slashdot, but I can't find it here either. Can anyone help me out?)
It seems clear to me that we have GNOME to thank for Troll Tech freeing up their license on the Qt library.
To admit that both sides wanted the same thing would be to admit that the GNUbites were a bit wrong to spew so much venom at the KDE crowd.
I'd prefer that no one spew venom at anyone, least of all at fellow free-software developers. But I must say that I have seen about as much anti-GNOME venom as anti-KDE venom. Both desktops have lots of rabid supporters.
Doggone it, news items like this just show how harmful having GNOME around is for the Free desktop. Nowadays, to be free-from-cost in the KDE world, one has to release their code under a Free license. To do otherwise is to pay a princely sum to Troll Tech, which most people don't want to do. The GNOME project, however, has wanted to get into bed with commercial projects since the beginning, and this is a great example. Such a license is bad for the Free world, though they'll not admit that their darling environment would be so.
Let me get this straight. It is better for the free software world to depend on a commercial product, one you must sometimes pay for the right to use, than to depend on a completely free library? Nonsense. To believe that, you would have to believe that contributions by business to free software are tainted and bad. I don't believe that. If IBM wants to donate a journaling file system, big iron patches, or anything else to the free community, I'm all for it.
I also note your implication: if we didn't have GNOME we would have more free apps. Let's think about this. If some company wanted to make and sell a product to run under Linux, and they didn't want to pay for a Qt license, you think that they would then go ahead and write the product and give it away for free? Not in this world. They would just forget about writing the product. The people who want to release their software free will do it, whether they are developing for GNOME or KDE. Lack of a free alternative will not cause more people to give stuff away.
If anyone's been harmful to the Free desktop, it's GNOME, not KDE.
Utter nonsense. The competition between the two desktops has made both of them better. The KDE guys are doing a great job, but so are the GNOME guys. The big difference is that companies like Sun and HP are going to ship GNOME on their computers, not KDE. But you have totally failed to make any case that it's a bad thing when more computers are running a free desktop.
How do the next few months look in terms of the ability of either Intel or AMD to improve upon these products?
The Pentium 4 is now being made with a.13 micron process, vs. the.18 micron process still being used to make the Athlon. When AMD starts selling Athlon chips made with the.13 micron process, they will once again be beating the Pentium 4 in benchmarks.
That is, assuming AMD gets the.13 micron process version out in a reasonable amount of time. If AMD takes too long getting their act together, Intel may be able to push the Pentium 4 to some ridiculous level (5 GHz or something) and win.
However, the Athlon will continue to rule in price/performance. Those of us who pay for their own computers will likely keep buying Athlons.
The latest I have seen on the AMD website is that the.13 micron chips should be out in "1H 2002", which presumably means June this year. That shouldn't be too late by any means; I doubt that Intel can do much with the Pentium 4 by that time.
P.S. Who among us really needs more performance than current Athlons? Even when Doom III comes out, current CPUs will have adequate performance (it's the 3D graphics card you will need to drop money on, not your CPU). I'm looking forward to buying a.13 micron Duron chip, to use in low-heat/low-noise computers.
Thank you very much for the info. I have downloaded the low latency patches for 2.4.17, and I'll give them a try. Since I am not maintaining either the LL patches or the preempt patches, I don't actually care which one is smaller or more "elegant", I just want my low latency.
I'm building a PC based on this theory. A case I got from Coolcases.com, modified with two 120mm fans; one, under the power supply, blows out, and the other is on the bottom front of the case to blow in. I'm not done testing it, but so far I haven't bothered to hook up the one on the front; the one on the back pulls enough air through the case to keep things cool enough, and it's quiet.
I am using the Zalman "flower" copper cooler, a CNPSCU 6000, with a 92mm cooling fan that has a variable-speed control. On its lowest speed, it is keeping my Athlon XP 1700+ (1.46 GHz) CPU cool enough. (I don't think 60 degrees C is "cool", but it is well under the 90 degrees C max for an Athlon XP, so it is "cool enough".)
I got the Zalman at Coolerguys.com. The guy there asked me what kind of motherboard I am using. I told him: an Asus A7V266-E. He looked it up, and said that some of the larger coolers would not fit well, but that the Zalman should. It does fit, although it is a bit tight: I had to bend the cooling fins slightly on one side, away from the memory slots, in order to fit two memory DIMMs in. The motherboard has three slots for memory, and slot 1, the closest to the CPU, cannot be used at the same time that you are using a Zalman flower cooler. That's okay, I'll survive somehow with just the 512MB of RAM I've got in the other two slots.
If I feel the need, I can hook up the front case fan too, so I have some safety margin. But I haven't needed that second fan so far. Maybe I should hook it up to a switch on the front panel, in case it might come in handy during hotter summer days.
But only if you actually compile it in to the kernel, right? Including the patches, which are not enabled unless you choose the option while configuring the kernel, wouldn't mess up kernel debugging, would it?
It is small, but it's not really elegant
I am not a kernel hacker, and I was not commenting on the code; I feel that the idea of leveraging the SMP locks is elegant.
I am one of the large pool of people who don't need realtime features, but do not want our movies or music playback to stutter when the system gets busy. The preemption patch seems to be doing that for me, but I might be interested in the full low latency patch.
Could you explain what "live locks" are, please? I'm not familiar with the term.
I have been following these two patches a bit, because I want my desktop to be as snappy and responsive as possible.
The current low-latency patches work by finding "hot spots" in the kernel code where something is taking a long time, and then putting in a hack to make the code yield. The good part is that you can get a really low-latency kernel; the bad part is that you have to touch the kernel code in hundreds of places, and the kernel code gets really ugly. I remember reading that Ingo Molnar, who wrote a giant low-latency patch that worked this way, agreed with Linus that his low-latency patch was just too ugly and huge and should not be included in the main source tree.
The preemption patch is comparatively small and elegant. It leverages the work that has already been done to make SMP work correctly. I'm using it on my Linux desktops, and I like it.
On one of the mailing lists, Linus said that he wants the Linux kernel to gain low latency the cleanest way: find all parts that are slow, and instead of hacking them to yield, re-write them so they are faster (but still clean code that is easy to understand). This is of course the ideal, but when will it be finished? The preemption patch is available now, and works now, and I am using it now.
In my humble opinion, the XOR cursor patent is worse than any of his top four. That patent totally flunks the "obviousness" test: any first-year computer science student, taking a graphics class, could write code that infringes on that patent.
"Write code to move a cursor around." Hmm, what are the operations I can use to set and clear pixels? AND, OR, and XOR. Hey, wait a minute...
At least they didn't grant a patent on "A Technique Whereby Raster Graphic Image Fragments Are Made to Appear in a Blank Frame Buffer by the Use of the OR Operator". Hmmm, wait, maybe I should file that one before someone beats me to it.
I'm sure his comment had something to do with the recent horrible security breach found in Windows XP, one that lets an attacker totally take over an XP box.
I'm also sure that Microsoft's patch has fixed that breach, although there will probably be others.
If you like XP and want to use it, that's fine by me. It just seems odd to me that some guy got moderated up to 5 pointing out an "error" that wasn't. Oh well, it's no big deal.
Actually, Windows XP does include this exact feature.
Re-read the part where he says "and they only have access when you do that CLICK!" Roblimo, the author of the article, equates XP with "a permanent backdoor into your computer."
Hey, maybe next time you can do your homework instead of complaining about things you don't understand.
Debian "stable" is famous for being out-of-date. It is also famous for being stable. The two somewhat go together, since Debian doesn't have paid full-time people hammering together updates.
Potato was frozen when the kernel was at 2.2 and Xfree86 was at 3.x. If you want to run Potato, you can get packages for Xfree86 4.1.x and kernel 2.4.x; if you want the latest cutting-edge stuff, all you have to do is update your system.
To update your system:
edit your sources.list to point to a mirror of "unstable", then run "apt-get update", then run the command "apt-get dist-upgrade" over and over until it reports that all packages were installed.
I hope this helps, and maybe next time you won't shoot your mouth off so obnoxiously.
For the love of god... upgrade stable.
Oh gee, what a great idea. Debian wasn't planning to upgrade stable, but now that you suggested it, I'm sure they will get right on it.
Sarcasm aside, if you had taken even a little while to read the debian.org web page, you might have found out that "Woody" is in the middle of a "freeze" process, which will take time... but when it is done, it will become the new "stable" branch.
And apt... what a joke
Actually, it's not a joke. It's one of the best things about Debian, and if you don't like it, maybe you should be running Red Hat. (Red Carpet does some of the same things as APT, but you may wind up having to pay money every month to use it. APT is always free.)
If you want a menu front-end to APT, you have many choices. I like aptitude ("apt-get install aptitude" if you want to try it) but there is also gnome-apt and others.
Your attitude sucks. You ought to work on that.
steveha
Re:And I just put 2.2r4 on yesterday....
on
Debian 2.2r5 Released
·
· Score: 4, Informative
Let me just ask a general question then - what's the best way to install a fresh Woody system? My current system is a bastardized Potato with some ugly Sid stuff thrown in to make it complicated. I want a Woody system, upgraded to a 2.4 kernel with ipfilter, and I'm willing to start from scratch.
Do I remove all non-Woody sources from the apt-get sources file? Do I manually remove all potato and unstable packages? Or is it best to format and start over?
First of all, if the only thing you really want is a 2.4.x kernel and ipfilter, you don't need to upgrade everything; you could just get kernel sources, build a 2.4.x kernel, and go from there.
But if you want to update your system, here is what to do:
Edit your sources.list file to point to a Debian mirror for the "unstable" packages. (Or "testing" if you want to try that, but I'm perfectly happy with unstable.)
Run "apt-get update", which fetches the list of new packages.
Run "apt-get dist-upgrade", which downloads the new packages and installs them.
The Debian APT system is really cool, but it isn't absolutely perfect. It will try to install packages like libc first, and then later on install packages that depend on the other packages; but sometimes it fails. Sometimes it will try to install a package, only to have the install fail because some needed package wasn't there. This especially happens when upgrading from Potato to unstable.
The solution is simple: you just keep running "apt-get dist-upgrade", over and over, until it reports that all packages installed. Each time more packages will install, as the dependencies get installed.
I've done this about twice, and that's what worked for me.
Slashdot Mirror
when you customise a good interface you invariably make it less productive and just don't realise it. Secondly, it is extremely poor user interface design to have two modes - one for new users and one for power users.
Very funny. Tell me another good one -- how about that joke where it's faster to pull down menus than to hit accelerator keys?
(Some Mac guys claim this, seriously. They say that it is slower to hit Ctrl+S or whatever than to take my hand off the keyboard, find the mouse, move the pointer, pull down a menu, find the item I want, click on it, and put my hand back on the keyboard.)
You know, I've got both Galeon and Mozilla customized so that when I click my middle mouse button on a link, that link opens in a new tab. This isn't the default behavior. According to you, I am suffering reduced productivity because I now spin up new tabs with a single button click.
The beauty of a customizable interface is that it can adapt to the way you want to work. I frequently want to open links in a new tab, so it saves me time and effort to turn a menu operation into a single middle-click. You cannot convince me that I am not better off.
I will meet you halfway: it's not a good idea to take really fundamental things and make them odd. I don't think the ability to customize the single-click behavior of your mouse would be a very good thing; it would ruin you for using other computers, and ruin your computer for others to use.
As for having two modes, I think it is a very good idea. The only problem I see with customizations is when Joe customizes his setup his way, and then I want to use Joe's computer. If I can kick it into non-customized mode for a moment, I can use his computer very well, and both of us are happier.
steveha
I looked into using a Palm for a bike computer, and concluded that I wasn't very interested. I live in the Seattle area, and I ride in the rain. The BikeBrain solution comes with a plastic protector for your Palm, but it isn't really waterproof.
The good thing of course is that a Palm can capture a lot of data. But just capturing wheel spin data to show speed and distance isn't enough to make me buy either of these solutions.
Last autumn I bought myself a Specialized P.Brain computer. I love it; it collects wheel turn data (like the two Palm solutions) and also altitude and heart rate data. A PC interface lets you capture your data and make pretty charts. You can get a graph showing your speed, altitude, and heart rate plotted against either time or distance. Read more about it here.
The PC download software is for Windows; I'm planning to try to get it working under WINE if I can. The data is stored in some opaque binary format, but you can get the data out with Dan Connelly's Perl script (get it here.
The P.Brain isn't the only data-collecting bike computer. There are other brands. I have heard good things about the Polar XTrainer. There are even computer systems that directly measure your power output; you have a wheel built with a power-measuring hub, and the computer keeps track of power. Pro riders (including Lance Armstrong) use these. For example, the Power-Tap.
steveha
It looks like BikeBrain has an altimeter in it, so it can also give you data like what your best time on a hill was, adjusted by the grade of the hill.
It looks like it has an altimeter, but it doesn't.
It keeps track of how far you have gone on your ride, and if it has a route plan that includes altitude, it shows what your altitude ought to be if you are following the route plan. The altitude graph, alas, will not work right if you alter the route slightly, or ride some completely new route.
There are bike computer systems with altimeters built in, and one of those would be better.
steveha
What I would love to see is everyone who is working on anything remotely redundant to drop what they are doing, put their collective heads together and come up with a real competitor for Microsoft
To some extent, redundant projects are a good thing. If two projects are competing, they can spur each other on to better results. Look at GNOME and KDE, for example.
Also, when there are two projects, the potential downside is reduced for trying something new. If GNOME bets on Mono, and KDE steers clear of it, then if Mono turns out to be a bad idea, we can all switch to KDE. (But note that it might be easier just to fork GNOME and switch to the non-mono fork, especially for those of us already using GNOME.)
But it's moot anyway. People work on whatever they want to work on. I think the world has enough text editors, but no one cares what I think; if some guy wants to write a text editor, he's going to do it. Free software, freedom. Nothing you can do about it, so why worry about it?
steveha
I can use netcat and ExtractStream to (usually) get a TV show onto one of my computers for editing and reencoding.
Whoa, could you tell me more? You are just about to sell me a TiVo.
What quality does the TiVo capture at? Can it capture full 640x480 TV pictures? (Sometimes called 720x480, if you are talking about entire scan lines and not just the square-pixels viewable part.)
I have no problem with paying money to TiVo for their box. I have no problem paying monthly for TV listings. I have a problem with doing the above and then not having any way to burn my own DVDs; I know you can hook a VCR up to a TiVo but I'm not interested. I have been saying for years that I would buy a TiVo when I can put one on my home network and grab video for DVD burning. Maybe that day has come?
steveha
the guy does a lot of great work and all, but most influential INNOVATORS??? We're talking a Unix clone here.
Linux is not an exact clone of UNIX. The first version of Linux was a clone of the most important bits of UNIX, but it has since grown in many different directions.
For one example, consider the HTTP daemon that runs inside the kernel, serving up static pages without leaving kernel mode and seamlessly handing dynamic pages off to Apache. There had never been anything like it before (although Microsoft has copied that idea for their web server).
Read Linux Weekly News. There is always new stuff shaking in the kernel development. An O(1) scheduler! Copy-on-write to make fork() really fast! Reverse-mapping in the VM subsystem! I don't know enough about the OS landscape to tell you which of these exist in which OSes, but I'm pretty sure that many of them are brand-new.
It's like some company made the most advanced car ever, and it had hundreds of cool little features that make it safer or nicer or whatever. If you are standing next to it, you just say "It's a car." But an automotive engineer can look under the hood and know better.
Look under the hood of Linux development.
steveha
So, why aren't we using more gui installers?
Maybe you should start up a project to make a standard GUI installer that everyone can use.
I just use apt-get, and I love how easy it is.
steveha
what is the benefit to the theater or the viewer?
The main benefit to the viewer is that you will see the same movie on opening night, and four weeks later. Currently, film prints can get scratched; they can break and get (possibly poorly) patched; and you can often see a visual glitch when one reel of film finishes and another reel starts up. The problems with film can get worse, much worse, if you have an idiot or a klutz as a projectionist, or if the movie house is really cheap and doesn't take proper care of the film and/or the projection equipment.
The number one fear I have about digital is that a projector might get one or more stuck pixels, and then there will be wrong pixels in the same places for the whole movie. So we are getting rid of all the little annoying errors that accrete over time with film, but possibly gaining annoying errors that last all movie (not changing until the projector is repaired!).
Another thing: this is the wild-and-wooly early days of digital projection. We will no doubt go through several rounds of standards changes and upgrades. (I haven't seen 1920x1200 on a big screen--maybe it's really gorgous--but it is hard for me to imagine that this level of quality will still be in use 20 years from now.) So the theater owners may invest in expensive projectors, and then find that they have to buy newer and even more expensive projectors to keep up.
One benefit to theater owners, however: they won't have to pay as much for the actual distribution media the film comes on. If they have multiple theaters that want to show the film, currently they need multiple prints. With digital, I don't know for sure how they will do it, but I imagine they will ship out the movie on magnetic tape, and load up one or more RAID arrays at the theater to drive the projectors. (Anyone who knows for sure, please speak up.)
steveha
"updatedb" was running, eating plenty of system ressources
/var/log was clean, I nuked some junk I didn't need, and the computer was decently fast again.
/var should be on its own partition! When /tmp is full, bad things happen. If you only have one partition, when /var is full, /tmp is full... and when you have too many .MP3 files, you can set off an unpleasant chain reaction.)
I have an old box that is running an older version of Mandrake Linux. The "logrotate" command runs at odd intervals (frequently during the work day, not in the middle of the night) and the computer is extremely slow when that happens. (I'm looking forward to installing Debian on that computer. I understand Debian, and I've never had a problem with logrotate on Debian anyway.)
The worst was when the hard drive filled up. The logrotate command was running continuously; the hard drive was rattling nonstop. I discovered that I had files like this:
auth.log.gz
auth.log.gz.gz
auth.log.gz.gz.gz
[...]
auth.log.gz.gz.gz.gz.gz.gz.gz.gz.gz.gz.gz.gz.gz
[...]
I managed to kill the logrotate process. Then I ran a find command to find all files that had a ".gz" in the name, and delete them. It took over 10 minutes to find and delete them all! There were thousands, many whose filenames were over 80 characters from ".gz.gz.gz.gz..." extensions.
Once
(Now I understand well why
steveha
The only problem that a CLR supposedly solves is the maintanence of the bindings. Instead of binding Gtk to perl and python and ada and C and C++, etc., you bind it in a library in the CLR. Except that to access that new CLR binding, you need to have perl and python and ada and C and C++ compilers that target the CLR, which is certainly a more glamorous job than maintaining bindings for every language under the sun, but is *WAY MORE* complex.
Nice summary. The big win with a CLR is that you need to write Perl/CLR once, and Python/CLR once, and so on; then you write each class library and interface for Mono once and everything just works.
Compare with the current system: for each new interface you add, you need to write the Python bindings, the Perl bindings, etc. etc. So you need to write an M*N matrix of bindings, where M is the number of interfaces and N is the number of supported languages.
Assuming the CLR works as Miguel thinks it will, you go from M*N to M+N, a huge win, even if writing the CLR interfaces for each language is much harder than writing some bindings. And the larger M and N become, the bigger the win becomes.
steveha
HP talks a good Linux, but does approximately nil to further the cause, or even make their products work with Linux.
I like HP inkjet printers. I strongly believe they're more reliable than Epson's. But I've bought three Epson printers in the last two years, only because they worked better (or at all) with Linux.
HP devlopers wrote a printer driver for Linux. It was initially released with a license similar to the BSD license, with an extra clause saying that to use it you needed to own an HP printer. They promised to have their lawyers make sure they were clear on patent issues, and then drop the extra clause if possible. Guess what, they did drop it, and now it is available as a purely free piece of software. (Get it here.)
This makes HP the only company to have released a free software driver for its products. Linux has excellent support for Epson printers, but Epson didn't do the work.
If you want to print photos, you are still better off under Linux with an Epson, because none of the Linux drivers for HP DeskJet printers support the 2400x1200 DPI photo printing mode yet. But it is just a matter of time. I'm hoping that HP will add that themselves to their own free driver.
If 600x600 DPI color printing is enough for you, you can use the HP DeskJet with Linux.
steveha
Yes, and it's copyright 1998.
Way too old to be of any discussion value.
I claimed that the KDE developers, when given a choice between Qt and a free clone of Qt, chose Qt... despite the non-free status of Qt at the time. Then I introduced that article to support my claim.
The 1998 copyright date doesn't matter, all that matters is whether the reference is accurate and supports my claim.
steveha
HTML spam is much worse than plain text spam. I always set my mail client to prefer plain text, because you can't embed cookies in URLs to images in plain text. Doing so in an HTML mail gives a clear indication that the mail address is valid (as the image has been requested, the mail has been recieved and read).
This is why the Evolution mail client is now my favorite. It lets you read HTML mail, but by default it will never put any hits on any servers; it only renders the passive tags (bold, italics, link, header, etc. but nothing downloaded from a server).
steveha
There was an essay about this... I couldn't find it.
Found it.
Why KDE is Still a Bad Idea
steveha
KDE/GNOME is a bit off-topic for a discussion of the license of Mono's class libraries, but what the heck, I'm game.
the KDE project was based on Qt due to it being a nice toolkit and due to the naivete of the core team. They just weren't aware of the implications, and couldn't understand it because, let's face it, programmers aren't lawyers.
This excuse does not fly. Perhaps it could have been true for a short time, but lots of people pointed out that KDE was depending on non-free software. KDE could have adopted the Harmony project (free library compatible with Qt) but they chose not to do so.
(There was an essay about this, with a title something like "Why KDE still isn't a good idea". I've searched the web for a long time and I couldn't find it. I think I read it on Slashdot, but I can't find it here either. Can anyone help me out?)
It seems clear to me that we have GNOME to thank for Troll Tech freeing up their license on the Qt library.
By the way, read this:
http://www.ntlug.org/~cbbrowne/qtcontroversy.html
To admit that both sides wanted the same thing would be to admit that the GNUbites were a bit wrong to spew so much venom at the KDE crowd.
I'd prefer that no one spew venom at anyone, least of all at fellow free-software developers. But I must say that I have seen about as much anti-GNOME venom as anti-KDE venom. Both desktops have lots of rabid supporters.
Doggone it, news items like this just show how harmful having GNOME around is for the Free desktop. Nowadays, to be free-from-cost in the KDE world, one has to release their code under a Free license. To do otherwise is to pay a princely sum to Troll Tech, which most people don't want to do. The GNOME project, however, has wanted to get into bed with commercial projects since the beginning, and this is a great example. Such a license is bad for the Free world, though they'll not admit that their darling environment would be so.
Let me get this straight. It is better for the free software world to depend on a commercial product, one you must sometimes pay for the right to use, than to depend on a completely free library? Nonsense. To believe that, you would have to believe that contributions by business to free software are tainted and bad. I don't believe that. If IBM wants to donate a journaling file system, big iron patches, or anything else to the free community, I'm all for it.
I also note your implication: if we didn't have GNOME we would have more free apps. Let's think about this. If some company wanted to make and sell a product to run under Linux, and they didn't want to pay for a Qt license, you think that they would then go ahead and write the product and give it away for free? Not in this world. They would just forget about writing the product. The people who want to release their software free will do it, whether they are developing for GNOME or KDE. Lack of a free alternative will not cause more people to give stuff away.
If anyone's been harmful to the Free desktop, it's GNOME, not KDE.
Utter nonsense. The competition between the two desktops has made both of them better. The KDE guys are doing a great job, but so are the GNOME guys. The big difference is that companies like Sun and HP are going to ship GNOME on their computers, not KDE. But you have totally failed to make any case that it's a bad thing when more computers are running a free desktop.
steveha
How do the next few months look in terms of the ability of either Intel or AMD to improve upon these products?
.13 micron process, vs. the .18 micron process still being used to make the Athlon. When AMD starts selling Athlon chips made with the .13 micron process, they will once again be beating the Pentium 4 in benchmarks.
.13 micron process version out in a reasonable amount of time. If AMD takes too long getting their act together, Intel may be able to push the Pentium 4 to some ridiculous level (5 GHz or something) and win.
.13 micron chips should be out in "1H 2002", which presumably means June this year. That shouldn't be too late by any means; I doubt that Intel can do much with the Pentium 4 by that time.
.13 micron Duron chip, to use in low-heat/low-noise computers.
The Pentium 4 is now being made with a
That is, assuming AMD gets the
However, the Athlon will continue to rule in price/performance. Those of us who pay for their own computers will likely keep buying Athlons.
The latest I have seen on the AMD website is that the
P.S. Who among us really needs more performance than current Athlons? Even when Doom III comes out, current CPUs will have adequate performance (it's the 3D graphics card you will need to drop money on, not your CPU). I'm looking forward to buying a
steveha
Thank you very much for the info. I have downloaded the low latency patches for 2.4.17, and I'll give them a try. Since I am not maintaining either the LL patches or the preempt patches, I don't actually care which one is smaller or more "elegant", I just want my low latency.
Thanks again and see you around.
steveha
I'm building a PC based on this theory. A case I got from Coolcases.com, modified with two 120mm fans; one, under the power supply, blows out, and the other is on the bottom front of the case to blow in. I'm not done testing it, but so far I haven't bothered to hook up the one on the front; the one on the back pulls enough air through the case to keep things cool enough, and it's quiet.
I am using the Zalman "flower" copper cooler, a CNPSCU 6000, with a 92mm cooling fan that has a variable-speed control. On its lowest speed, it is keeping my Athlon XP 1700+ (1.46 GHz) CPU cool enough. (I don't think 60 degrees C is "cool", but it is well under the 90 degrees C max for an Athlon XP, so it is "cool enough".)
I got the Zalman at Coolerguys.com. The guy there asked me what kind of motherboard I am using. I told him: an Asus A7V266-E. He looked it up, and said that some of the larger coolers would not fit well, but that the Zalman should. It does fit, although it is a bit tight: I had to bend the cooling fins slightly on one side, away from the memory slots, in order to fit two memory DIMMs in. The motherboard has three slots for memory, and slot 1, the closest to the CPU, cannot be used at the same time that you are using a Zalman flower cooler. That's okay, I'll survive somehow with just the 512MB of RAM I've got in the other two slots.
If I feel the need, I can hook up the front case fan too, so I have some safety margin. But I haven't needed that second fan so far. Maybe I should hook it up to a switch on the front panel, in case it might come in handy during hotter summer days.
steveha
It makes kernel debugging considerably harder
But only if you actually compile it in to the kernel, right? Including the patches, which are not enabled unless you choose the option while configuring the kernel, wouldn't mess up kernel debugging, would it?
It is small, but it's not really elegant
I am not a kernel hacker, and I was not commenting on the code; I feel that the idea of leveraging the SMP locks is elegant.
I am one of the large pool of people who don't need realtime features, but do not want our movies or music playback to stutter when the system gets busy. The preemption patch seems to be doing that for me, but I might be interested in the full low latency patch.
Could you explain what "live locks" are, please? I'm not familiar with the term.
steveha
I have been following these two patches a bit, because I want my desktop to be as snappy and responsive as possible.
The current low-latency patches work by finding "hot spots" in the kernel code where something is taking a long time, and then putting in a hack to make the code yield. The good part is that you can get a really low-latency kernel; the bad part is that you have to touch the kernel code in hundreds of places, and the kernel code gets really ugly. I remember reading that Ingo Molnar, who wrote a giant low-latency patch that worked this way, agreed with Linus that his low-latency patch was just too ugly and huge and should not be included in the main source tree.
The preemption patch is comparatively small and elegant. It leverages the work that has already been done to make SMP work correctly. I'm using it on my Linux desktops, and I like it.
On one of the mailing lists, Linus said that he wants the Linux kernel to gain low latency the cleanest way: find all parts that are slow, and instead of hacking them to yield, re-write them so they are faster (but still clean code that is easy to understand). This is of course the ideal, but when will it be finished? The preemption patch is available now, and works now, and I am using it now.
steveha
In my humble opinion, the XOR cursor patent is worse than any of his top four. That patent totally flunks the "obviousness" test: any first-year computer science student, taking a graphics class, could write code that infringes on that patent.
"Write code to move a cursor around." Hmm, what are the operations I can use to set and clear pixels? AND, OR, and XOR. Hey, wait a minute...
At least they didn't grant a patent on "A Technique Whereby Raster Graphic Image Fragments Are Made to Appear in a Blank Frame Buffer by the Use of the OR Operator". Hmmm, wait, maybe I should file that one before someone beats me to it.
steveha
I'm sure his comment had something to do with the recent horrible security breach found in Windows XP, one that lets an attacker totally take over an XP box.
I'm also sure that Microsoft's patch has fixed that breach, although there will probably be others.
If you like XP and want to use it, that's fine by me. It just seems odd to me that some guy got moderated up to 5 pointing out an "error" that wasn't. Oh well, it's no big deal.
steveha
Actually, Windows XP does include this exact feature.
Re-read the part where he says "and they only have access when you do that CLICK!" Roblimo, the author of the article, equates XP with "a permanent backdoor into your computer."
steveha
Hey, maybe next time you can do your homework instead of complaining about things you don't understand.
Debian "stable" is famous for being out-of-date. It is also famous for being stable. The two somewhat go together, since Debian doesn't have paid full-time people hammering together updates.
Potato was frozen when the kernel was at 2.2 and Xfree86 was at 3.x. If you want to run Potato, you can get packages for Xfree86 4.1.x and kernel 2.4.x; if you want the latest cutting-edge stuff, all you have to do is update your system.
To update your system:
edit your sources.list to point to a mirror of "unstable", then run "apt-get update", then run the command "apt-get dist-upgrade" over and over until it reports that all packages were installed.
I hope this helps, and maybe next time you won't shoot your mouth off so obnoxiously.
For the love of god... upgrade stable.
Oh gee, what a great idea. Debian wasn't planning to upgrade stable, but now that you suggested it, I'm sure they will get right on it.
Sarcasm aside, if you had taken even a little while to read the debian.org web page, you might have found out that "Woody" is in the middle of a "freeze" process, which will take time... but when it is done, it will become the new "stable" branch.
And apt... what a joke
Actually, it's not a joke. It's one of the best things about Debian, and if you don't like it, maybe you should be running Red Hat. (Red Carpet does some of the same things as APT, but you may wind up having to pay money every month to use it. APT is always free.)
If you want a menu front-end to APT, you have many choices. I like aptitude ("apt-get install aptitude" if you want to try it) but there is also gnome-apt and others.
Your attitude sucks. You ought to work on that.
steveha
Let me just ask a general question then - what's the best way to install a fresh Woody system? My current system is a bastardized Potato with some ugly Sid stuff thrown in to make it complicated. I want a Woody system, upgraded to a 2.4 kernel with ipfilter, and I'm willing to start from scratch.
Do I remove all non-Woody sources from the apt-get sources file? Do I manually remove all potato and unstable packages? Or is it best to format and start over?
First of all, if the only thing you really want is a 2.4.x kernel and ipfilter, you don't need to upgrade everything; you could just get kernel sources, build a 2.4.x kernel, and go from there.
But if you want to update your system, here is what to do:
Edit your sources.list file to point to a Debian mirror for the "unstable" packages. (Or "testing" if you want to try that, but I'm perfectly happy with unstable.)
Run "apt-get update", which fetches the list of new packages.
Run "apt-get dist-upgrade", which downloads the new packages and installs them.
The Debian APT system is really cool, but it isn't absolutely perfect. It will try to install packages like libc first, and then later on install packages that depend on the other packages; but sometimes it fails. Sometimes it will try to install a package, only to have the install fail because some needed package wasn't there. This especially happens when upgrading from Potato to unstable.
The solution is simple: you just keep running "apt-get dist-upgrade", over and over, until it reports that all packages installed. Each time more packages will install, as the dependencies get installed.
I've done this about twice, and that's what worked for me.
steveha