In spring, I went through the installation of all the major Linux distributions. They all sucked, because of various reasons.
Unfortunately, I didn't write the problems up. Of the RPM distributions, I liked Mandrake best, except that I had to install it about 5 times because some programs simply wouldn't work. After a few hours I found out that the system didn't work properly because of the Finnish language setting given in the first screen of the installation. Well, got that fixed, but after failing to install newest version of KDE properly, I kicked Mandrake out. I think it had some other critical problems too, but can't remember what anymore.
I installed SuSe Linux on my employer's two notebooks. It was hell, because SuSe didn't have a rescue CD. I was frustrated like hell and just couldn't believe this. Yeah, it had 3,5" rescue disk, but the notebooks didn't have such drives. Finally I managed to get it installed by using the Mandrake installation CD's rescue mode... Why did I need a rescue disk? I had accidentally written the LILO to wrong partition (not MBR), and I didn't want to go through the package selection again. Also setting up the network was hell, and fought with it for hours. I was unable to setup my audio card (SB AWE32) on my home machine, even with hours of fighting. The "YAST2" setup tool is sh*t, and slow as hell (and has annoyingly stupid name - why not just call it "System Installation Tool" or something???). In the second installation I didn't select ALL packages during the initial installation. Thus, I was forced to select 600 packages one-by-one using a notebook nanny because the package manager didn't support keyboard properly. It took about half an hour to select them all. On the plus side, SuSe has the prettiest boot screen I've ever seen.;-)
Corel Linux (based on Debian) had definitely the easiest installation! Everything went smooth, as it detected my AWE32 sound card and got even my exotic i740 video card and rare Viewpoint monitor working straight properly! Unfortunately, the support from Corel was non-existant, and there weren't any updates. I wanted to install new KDE 2.x, so I added some regular Debian sites to the update program. It went hell, as the Debian packages broke the system entirely.
I installed Debian on my home server/router/firewall. It was also hell, because I don't have a 3,5" disk drive, just CD, and Debian didn't have a bootable installation CD. It also didn't have a way to install it from RedHat. It took about two days to install it from RedHat, using chroot tweaks to run the installation scripts, etc. But that's nothing, then I had to fight with the package installations. I have to say that dselect has the most horrible user interface that I've ever seen in any program. Come on guys! Why can't you use "Esc" key for returning, but "Shift-Q"????? Aaaarghhh. Anyhow, using dselect and configuring the apt-get sources.list was horribly difficult, and I've been using Linux (RH) for 6 years and computers for 18 years. I can't even imagine how difficult it would be for a beginner. Choosing between the "stable" and "unstable" packages is also a hell to configure, and I still don't know how to do that in the source.list file, after using Debian for a few months. For some time the "unstable" worked, but then I started getting weird errors such as:
Sorry, but the following packages have unmet dependencies: apache: Depends: libdb2 (>= 2:2.7.7-4) but 2:2.7.7-8 is to be installed E: Sorry, broken packages
Then I switched back to the stable - which means having outdated packages all the time. Configuring the APT system is really confusing and non-intuitive for even an experienced Linux (RH) user (and I'm not really a computer beginner after 18 years). Everything requires reading huge amounts of documentation and help files. That's just NOT the way modern software should be done. All programs should be tested on beginners. But Debian developers clearly have some tendency to make the system so complex that they can feel superior. That's just idiotic.
Finally, I installed RedHat 7.1 to my personal workstation. Can't remember all the problems anymore, but they were plenty, as always with RedHat installations. Why on earth RH still doesn't have ReiserFS??? They say "it's not stable enough", but I switched from ext2 exactly because of non-fixable filesystem corruption. And I don't want to wait 30 minutes for the chkfs for the 60GB drive. The new "up2date" updater is a joke. On the first update, it tried to update the up2date itself, but ran into depency problems with Python (it's written with Python I guess). Yeah, this was the out-of-the-box installation. No, the updater didn't inform which package conflicted with which, but just that there was "a conflict" with SOME of the 20 packages it tried to install. Of course it didn't tell which package conflicted and how. I had to trace and update them manually with rpm, after a lot of trouble. After resolving that problem, I was able to make up2dates for a month, and now there's again a conflict. No, I'm not going to use up2date again before the GUI is fixed PROPERLY.
Linux definitely is not ready for desktop, or for beginners. I believe the main obstacle is the attitude of the developers, who require that users use huge amount of time studying the documentations.
Even as a heavy user, and as a software developer, I don't want to spend months of my time on studying useless things. Most people don't, and if Linux developers require them to do that, everybody loses.
I hope RedHat 7.x binary RPMs will be available SOON after the release this time. We didn't get the RPMs for the latest beta release at all (or at least I didn't find them).
RedHat doesn't seem to take packaging too seriously, unlike Mandrake, Suse and Debian, which typically provide the packages in just a few days.
I'm not quite sure why I haven't yet switched from using RedHat. I guess there was some reason, I always seem to forget what.
Weapons
a blessed +1 butterfly knife (alternate weapon; not wielded)
an uncursed +0 mini-axe
an uncursed +0 laser pointer
Armor
an uncursed waterproof +0 Goretex jacket
Comestibles
an uncursed candybar
Tools
an uncursed very expensive digital camera [0:340]
4 uncursed rechargeable AA batteries
an uncursed mini tripod
an uncursed +3 rechargeable flashlight
an uncursed Palm IIIxe PDA
an uncursed Palm III keyboard
2 uncursed AAA batteries
an uncursed cellphone
an uncursed wallet
an uncursed 0.3mm pencil
an uncursed 0.5mm pencil
an uncursed mini-magnifying glass
(Yeah, I really carry all that stuff every day in my backpack. Well, ok, not usually the axe.)
Brits have an excellent taste for satire, and make excellent comedy series, I must say. They are also known for not taking religion too seriously, unlike americans. From an unknown source:
SAN FRANCISCO MAN BECOMES FIRST AMERICAN TO GRASP SIGNIFICANCE OF IRONY
SAN FRANCISCO - We spoke to Jay Fullmer, 38, who became the first American
to get to grips with the concept of irony yesterday.
"It was weird," Fullmer said, "I was in London and, like, talking to
this guy and it was raining and shit and he said, like, great weather,
or something like that."
Said Fullmer: "And I thought - wait a minute, it's like, no way is it
great weather."
Fullmer soon realised that the other man's 'mistake' was deliberate.
"This guy was pretty cool about it," Fullmer said.
Fullmer, who is 39 next month and married with two children, aged 8 and
3, planned to use irony himself in future.
"I'm like saying it all the time." he said. "Weekend last I was like
grilling steaks and I like burned them to shit and I said 'great
weather'."
I guess the last paragraph is the most illustrative.;-)
The Sloan Digital Sky Survey is using CCDs to map one quarter of the entire sky, in five passbands. Its main camera uses a mosaic of 30 2048x2048 CCDs to cover an area about 2.5 degrees across (although there are gaps between the chips).
Ok, that sounds nice.
There seems to be a Digitized Sky Survey project, which digitizes the Palomar and some other plates. They are scanned from 6.5x6.5 degree plates to 1.7 or 1.0 arc seconds per pixel. That makes about...23400x23400 pixels per plate, which is a lot less than I guessed (the plates are rather grainy after all), only 4.6 times more pixels than with the SDSS camera.
Alpha Centauri's distance is 1.3 parsecs. It means that a planet at one AU (Earth's distance from Sun) would be a bit less than one arc second from the star. That's well within the limits of VLT and most middle-sized (>2m) earth-based telescopes.
The problem is that angular resolutions of telescopes are given for separating two objects of the same maginitude. A star is so damn bright compared to any orbiting planet (probably more than 1,000,000,000 times brighter) that, if the image is taken with an exposure time that would make the planet visible, the atmosphere will spread the star to many arc secs, even with the best adaptive and active optics. I think OWL would also have this problem.
I don't know exactly why Hubble can't see such planets. Probably its angular resolution is still not good enough. Also, CCDs don't always behave nicely when they are grossly overexposed, as the CCD would be if there's a planet in one pixel and its star two pixels away.
Btw, Alpha Centauri is a multiple star, and the close component stars probably can't have stable planets. Proxima Centauri (one of the components) is farther away from the others and might have planets (I guess).
Space based and ground based telescopes compliment each other. Right now, the Hubble's primary mission is that of a scout...
;-) And for some reason, Hubble seems to get the 99% of the media time of all telescopes. Oh, and Arecibo. I don't read newspapers much (mostly just Excite and some local newspapers occasionally), but I've never seen an article about VLT in any news source. There probably are many, but compared to Hubble, they are quite rare. Well, I guess space is much sexier.
One thing which I've been wondering about VLT is the usage of just digital imaging (FORS1, etc)...or at least I haven't noticed any VLT cameras using traditional film (correct me if I'm wrong). I know that CCDs are great for making photometric measurements because of their linearity, but their resolution is nothing compared to large photographic plates used in older cameras. For example, FORS1 is just 2048x2048.
Well, ok, camera resolution might not be so important in most research, but I would imagine that doing the Palomar sky survey (hundreds of huge plates) with CCDs would be impossible (it would probably require trillions of pictures)....And the best space poster pictures are still the ones taken with the Palomar 5m.;-)
So, how are surveys made now or in future, with CCDs or plates? Are surveys or other hi-res imaging still relevant?
[I'm eagerly waiting for a job decision from ESO at Garching, should come next week. It would be great to get to mess up^H^H^H^H^H^H^Hdevelop the computer systems there.;-)]
The case seems silly. MySQL database is double-licensed with GNU GPL and a proprietary license. Now MySQL has terminated the proprietary license, so if NuSphere distributes the product, it may break both against MySQL's copyright and GPL.
I'm just not so sure that they can say that NuSphere breaks against both licenses, but just the proprietary one, which it has bought in the first place.
Thus this seems just a case of creating a GPL violation from a normal license violation (if that really is the case). Probably the motivation is to damage the opponent more than reasonably, and to get the anger of the Free Software community, and perhaps also help in legal matters.
There's been an editor's note added to the article which says:
"We checked with our lawyers and they categorically deny having said anything that could have been interpreted as the comment that is attributed to them."
What on earth is happening? According to the Netcraft survey, IIS has just gained +5.49% from last month (to 25.88% market share), and Apache has lost -4.29% (down to 58.73% share). It's more than 2 million new servers in one month, from last month's 6 million servers, a +34% jump!
In "active sites", the jump is just +1.77% (Apache -1.89% down to 60.53%).
People must have heard that IIS is unsecure, and immediately installed it just to be one of the worm spreaders?
The jump is biggest ever. I guess it must be because of some new bundled IIS server. Perhaps the one in the new JesusWindows?
The BSA is so concerned about software piracy that they've decided to use only Free Software:
The site www.bsa.org is running Apache/1.3.12 OpenSSL/0.9.5a (Unix) AuthMySQL/2.20 PHP/3.0.16 on FreeBSD.
FreeBSD is not Free Software. Open Source, shure. Even Microsoft likes the FreeBSD license, as it is allowed to steal that code, remember?
Also Apache is not Free Software; it doesn't use GPL but the Apache Group License, which is basically a BSD license (you don't have to distribute the source code). I guess same applies to PHP as well.
MySQL is Free Software, double-licensed with GPL and proprietary licenses.
Yeah. But, the problem may be that non-ecommerce-people (such like me) don't know about the other existing systems. With Microsoft's visibility, Passport will probably be the first such system they learn about. And since it's new and buzzy, and hey, "Microsoft is the standard", that's what they want, and therefore businesses have to provide it. Possibly not, but I'm a bit worried about this.
I think that most of the current authentications and epayment systems are rather fragmented, and thus not very well known, and some may see Passport a solution to that. A bad solution, we know.
Either my brain doesn't function (no morning tea yet), or there's something really sick in the document.
The section "Wait a minute! Are you saying that there are no drive letters in Linux?" (page 14) says that Linux doesn't have drive letters but mount points. Correct so far.
But then it implies that ''/dev/fd0 (or/dev/floppy or/mnt/floppy)...)'' are such mount points. Eeeeehhhhhh?
More: ''...There can be many SCSI drives, on one or more controllers. The mount point designations just keep incrementing up through/dev/sdz....''
Gods! The writer doesn't know the difference between a device file and a mount point!
Let's see further: ''
Unlike the primary floppy drive (/dev/fd0) Linux doesn't automatically map a second diskette drive (/dev/fd1) to something like/dev/floppy or/mnt/floppy. However, if desired you can set up your own alias, such as/mnt/floppy2. The mapping entries that are used for this are located in the/etc directory, in a file called fstab. To add an alias for fd1, use an editor to open/etc/fstab and...'' (emphasis mine)
Ah, so the/mnt-files are just ''aliases''! Rrright. It's always nice to learn something new!
This came up on...July 29th on Slashdot. I guess the real technical writers are on summer vacation, and the article was submitted by the summer trainee. Am I wrong?
The IBM site seems to have a huge number of other interesting Linux articles too, at their Linux Zone.
The site seems to have lots of introductionary (such as the new FAQ) and general technical articles (hardware stability, journaling filesystems, embedded Linux, firewalls, etc). The Linux Zone seems to be also an index to interesting documents on other sites.
Unfortunately some of the articles (at least the tutorials) require annoying registration.
I doubt MS is able to destroy all Free Software and Open Source just by a single protocol change. It would be just paranoid to think that..NET isn't everything, and won't be.
However, Microsoft's vision seems to be making the.NET as popular and central in people's lives as Internet. In such a case, FS||OSS has to provide a solution, just like MS had to revolt their Internet strategy a few years ago.
I guess other parts of.NET will be ''easy'' to duplicate, except the Passport. So, create an alternative, based on an OPEN protocol, and make it preferably an ISO standard. Then, make the Free.NET support both.
Of course, it may be difficult for an alternative authentication&payment system to compete with Passport. But I'd believe some such systems exist already and many banks and other e-business companies will be competing with Microsoft anyhow. They'll just need co-operation from the FS||OSS community.
If USA is too tough a battleground, start in Europe. It's much easier to get an authentication/e-payment system working in a small country than in USA. An European system might have some taxation advantages too, as MS is a foreign company. It might also be possible to join with some mobile phone payment systems, which are already quite common.
That "license" agreement is for the web site, not LinuxDA.
Perhaps, but in that case it's rather misleading. Read it carefully. They write about ''software found on [this site]''.
But yes, probably that's just a mistake and the Linux kernel will be free and GPL as usual, but all the softwares on top of it will not. Probably you can't even develop software for it without their cossstly SDK. Much use, not.
There is a free DEMO (their caps) version to download, says the release, but the final version will be payware.
Pay...ware? Linux? Me not understand?
The site www.linuxda.com says:
SINGLE COPY LICENSE The materials at this Site are copyrighted and any unauthorized use or reproduction of any materials at this Site is prohibited. You may download one copy of the information or software ("Materials") found on Empower sites on a single computer for your personal, non-commercial internal use only unless specifically licensed to do otherwise by Empower Technologies Inc. in writing or as allowed by any license terms which accompany or are provided with individual Materials. This is a license, not a transfer of title, and is subject to the following restrictions: you may not (a) modify the Materials or use them for any commercial purpose or for any public display, performance, sale, or rental; (b) decompile, reverse engineer, or disassemble software Materials except and only to the extent permitted by applicable law; (c) remove any copyright or other proprietary notices from the Materials; (d) transfer the Materials to another person. You agree to prevent any unauthorized use or copying of the Materials.
Excuse me, but I thought Linux is under the GPL license or something?
"IRC+10216 is a carbon-rich star in which the concentration of carbon exceeds that of oxygen," Melnick said. "In such stars, we expect all the oxygen atoms to be bound up in the form of carbon monoxide (an oxygen atom and a carbon atom bound together), with almost nothing left over to form water (one oxygen atom bound to two hydrogen atoms).
Ok, the star has a lot of oxygen. The astronomers assume that the oxygen is bound to the carbon, but what if this is not the case? IANAA, but what if there's enough free oxygen and hydrogen than can bind, and then perhaps escape from the star? (I'm not sure if its possible that just the oxygen comes from the star and then binds to interstellar hydrogen.)
Another, perhaps better possibility might that the water is from an earlier stage, where all the oxygen was not bound to carbon. My introductionary astronomy book says about small stars:
"When the core runs out of helium, the star will have two shells at different depths. Hydrogen burns at the outern shell [to helium] and helium in the inner [to carbon and oxygen]. This structure is unstable and the matter can mix, or escape to space and form a shell, much like a planetary nebulae." (translated roughly from a Finnish textbook).
Thus, if the escaping mixed matter is hydrogen and oxygen....-> huge cloud of water!
If this idea is valid, the astronomers probably have already thought about it and have taken it into account, and for some reason have judged it unsatisfactory.
I personally find it very difficult to imagine how a dying star could vaporize the comets around it, if it couldn't do it when it was still young and bright. Perhaps it could have done it during some middle-age crisis, for example if it was an inflated gas giant at some point.
It seems that many who have answered were germans. I guess this very biased.
Strangely, the typical OS developer is 21-22 years old male (98%). Well, perhaps 21-27. That means not much experience. 33% are students, so perhaps that explains it. But then, 37% are university graduates/masters, plus 4% PhDs.
78% are not paid for developing OS. This isn't nice. But 30% report having profited professionally.
3.5% do not work (I belong to that sad part). I just wonder which option all the students have reported...
The yearly income distribution is rather even in the entire range, perhaps about 30,000-40,000 EUR per year on average. I think US salaries are higher than european; it would be nice to have per-country statistics too.
They seem to be spending typically
Typically they use Debian, although SuSE and RedHat are not far behind. 33% favour KDE, and 23% Gnome. Umm, 11% Windows, 7% "pure tex"?
36% are masochists who use vi (can this really be true?), while emacs has 28%.
Hmm, the site seems to be slashdotted, so I can't summarize the last page, "computer abilities".
Testing the exoskeletons may produce many interesting results.
Just imagine what would happen if the pants suddenly "detect" that the muscleman test pilot really wants to perform movements suitable for a ballet dancer.
After a few pretty splits and bends, which all 15-year old ballerinas would envy, the pieces of the soldier's bones are collected from the suit with tweezers.
Slashdot Mirror
Unfortunately, I didn't write the problems up. Of the RPM distributions, I liked Mandrake best, except that I had to install it about 5 times because some programs simply wouldn't work. After a few hours I found out that the system didn't work properly because of the Finnish language setting given in the first screen of the installation. Well, got that fixed, but after failing to install newest version of KDE properly, I kicked Mandrake out. I think it had some other critical problems too, but can't remember what anymore.
I installed SuSe Linux on my employer's two notebooks. It was hell, because SuSe didn't have a rescue CD. I was frustrated like hell and just couldn't believe this. Yeah, it had 3,5" rescue disk, but the notebooks didn't have such drives. Finally I managed to get it installed by using the Mandrake installation CD's rescue mode... Why did I need a rescue disk? I had accidentally written the LILO to wrong partition (not MBR), and I didn't want to go through the package selection again. Also setting up the network was hell, and fought with it for hours. I was unable to setup my audio card (SB AWE32) on my home machine, even with hours of fighting. The "YAST2" setup tool is sh*t, and slow as hell (and has annoyingly stupid name - why not just call it "System Installation Tool" or something???). In the second installation I didn't select ALL packages during the initial installation. Thus, I was forced to select 600 packages one-by-one using a notebook nanny because the package manager didn't support keyboard properly. It took about half an hour to select them all. On the plus side, SuSe has the prettiest boot screen I've ever seen. ;-)
Corel Linux (based on Debian) had definitely the easiest installation! Everything went smooth, as it detected my AWE32 sound card and got even my exotic i740 video card and rare Viewpoint monitor working straight properly! Unfortunately, the support from Corel was non-existant, and there weren't any updates. I wanted to install new KDE 2.x, so I added some regular Debian sites to the update program. It went hell, as the Debian packages broke the system entirely.
I installed Debian on my home server/router/firewall. It was also hell, because I don't have a 3,5" disk drive, just CD, and Debian didn't have a bootable installation CD. It also didn't have a way to install it from RedHat. It took about two days to install it from RedHat, using chroot tweaks to run the installation scripts, etc. But that's nothing, then I had to fight with the package installations. I have to say that dselect has the most horrible user interface that I've ever seen in any program. Come on guys! Why can't you use "Esc" key for returning, but "Shift-Q"????? Aaaarghhh. Anyhow, using dselect and configuring the apt-get sources.list was horribly difficult, and I've been using Linux (RH) for 6 years and computers for 18 years. I can't even imagine how difficult it would be for a beginner. Choosing between the "stable" and "unstable" packages is also a hell to configure, and I still don't know how to do that in the source.list file, after using Debian for a few months. For some time the "unstable" worked, but then I started getting weird errors such as:
Sorry, but the following packages have unmet dependencies:
apache: Depends: libdb2 (>= 2:2.7.7-4) but 2:2.7.7-8 is to be installed
E: Sorry, broken packages
Then I switched back to the stable - which means having outdated packages all the time. Configuring the APT system is really confusing and non-intuitive for even an experienced Linux (RH) user (and I'm not really a computer beginner after 18 years). Everything requires reading huge amounts of documentation and help files. That's just NOT the way modern software should be done. All programs should be tested on beginners. But Debian developers clearly have some tendency to make the system so complex that they can feel superior. That's just idiotic.
Finally, I installed RedHat 7.1 to my personal workstation. Can't remember all the problems anymore, but they were plenty, as always with RedHat installations. Why on earth RH still doesn't have ReiserFS??? They say "it's not stable enough", but I switched from ext2 exactly because of non-fixable filesystem corruption. And I don't want to wait 30 minutes for the chkfs for the 60GB drive. The new "up2date" updater is a joke. On the first update, it tried to update the up2date itself, but ran into depency problems with Python (it's written with Python I guess). Yeah, this was the out-of-the-box installation. No, the updater didn't inform which package conflicted with which, but just that there was "a conflict" with SOME of the 20 packages it tried to install. Of course it didn't tell which package conflicted and how. I had to trace and update them manually with rpm, after a lot of trouble. After resolving that problem, I was able to make up2dates for a month, and now there's again a conflict. No, I'm not going to use up2date again before the GUI is fixed PROPERLY.
Linux definitely is not ready for desktop, or for beginners. I believe the main obstacle is the attitude of the developers, who require that users use huge amount of time studying the documentations.
Even as a heavy user, and as a software developer, I don't want to spend months of my time on studying useless things. Most people don't, and if Linux developers require them to do that, everybody loses.
RedHat doesn't seem to take packaging too seriously, unlike Mandrake, Suse and Debian, which typically provide the packages in just a few days.
I'm not quite sure why I haven't yet switched from using RedHat. I guess there was some reason, I always seem to forget what.
Weapons
a blessed +1 butterfly knife (alternate weapon; not wielded)
an uncursed +0 mini-axe
an uncursed +0 laser pointer
Armor
an uncursed waterproof +0 Goretex jacket
Comestibles
an uncursed candybar
Tools
an uncursed very expensive digital camera [0:340]
4 uncursed rechargeable AA batteries
an uncursed mini tripod
an uncursed +3 rechargeable flashlight
an uncursed Palm IIIxe PDA
an uncursed Palm III keyboard
2 uncursed AAA batteries
an uncursed cellphone
an uncursed wallet
an uncursed 0.3mm pencil
an uncursed 0.5mm pencil
an uncursed mini-magnifying glass
(Yeah, I really carry all that stuff every day in my backpack. Well, ok, not usually the axe.)
Ok, that sounds nice.
There seems to be a Digitized Sky Survey project, which digitizes the Palomar and some other plates. They are scanned from 6.5x6.5 degree plates to 1.7 or 1.0 arc seconds per pixel. That makes about...23400x23400 pixels per plate, which is a lot less than I guessed (the plates are rather grainy after all), only 4.6 times more pixels than with the SDSS camera.
The problem is that angular resolutions of telescopes are given for separating two objects of the same maginitude. A star is so damn bright compared to any orbiting planet (probably more than 1,000,000,000 times brighter) that, if the image is taken with an exposure time that would make the planet visible, the atmosphere will spread the star to many arc secs, even with the best adaptive and active optics. I think OWL would also have this problem.
I don't know exactly why Hubble can't see such planets. Probably its angular resolution is still not good enough. Also, CCDs don't always behave nicely when they are grossly overexposed, as the CCD would be if there's a planet in one pixel and its star two pixels away.
Btw, Alpha Centauri is a multiple star, and the close component stars probably can't have stable planets. Proxima Centauri (one of the components) is farther away from the others and might have planets (I guess).
One thing which I've been wondering about VLT is the usage of just digital imaging (FORS1, etc)...or at least I haven't noticed any VLT cameras using traditional film (correct me if I'm wrong). I know that CCDs are great for making photometric measurements because of their linearity, but their resolution is nothing compared to large photographic plates used in older cameras. For example, FORS1 is just 2048x2048.
Well, ok, camera resolution might not be so important in most research, but I would imagine that doing the Palomar sky survey (hundreds of huge plates) with CCDs would be impossible (it would probably require trillions of pictures). ...And the best space poster pictures are still the ones taken with the Palomar 5m. ;-)
So, how are surveys made now or in future, with CCDs or plates? Are surveys or other hi-res imaging still relevant?
[I'm eagerly waiting for a job decision from ESO at Garching, should come next week. It would be great to get to mess up^H^H^H^H^H^H^Hdevelop the computer systems there. ;-)]
I'm just not so sure that they can say that NuSphere breaks against both licenses, but just the proprietary one, which it has bought in the first place.
Thus this seems just a case of creating a GPL violation from a normal license violation (if that really is the case). Probably the motivation is to damage the opponent more than reasonably, and to get the anger of the Free Software community, and perhaps also help in legal matters.
This smells bad.
"We checked with our lawyers and they categorically deny having said anything that could have been interpreted as the comment that is attributed to them."
I mean, life, even artificial, will find a way. ''Ooh'' and ''aah'', that's how it starts, and then comes the running and the screaming.
I'll get worried about nano-replicators after they build the first self-replicating machine factory even with human workers. It's hard.
But then, the certain Lexx episode made a nice demonstration what could happen with enough self-replicating robot arms... ;-)
In "active sites", the jump is just +1.77% (Apache -1.89% down to 60.53%).
People must have heard that IIS is unsecure, and immediately installed it just to be one of the worm spreaders?
The jump is biggest ever. I guess it must be because of some new bundled IIS server. Perhaps the one in the new JesusWindows?
I'm not amazed if some don't grasp the FS and OSS concepts, if I still make these mistakes after so many years...
With the given pricing, the marketing gap between this and a real notebook is very narrow.
Definitely not for pockets.
FreeBSD is not Free Software. Open Source, shure. Even Microsoft likes the FreeBSD license, as it is allowed to steal that code, remember?
Also Apache is not Free Software; it doesn't use GPL but the Apache Group License, which is basically a BSD license (you don't have to distribute the source code). I guess same applies to PHP as well.
MySQL is Free Software, double-licensed with GPL and proprietary licenses.
I think that most of the current authentications and epayment systems are rather fragmented, and thus not very well known, and some may see Passport a solution to that. A bad solution, we know.
The section "Wait a minute! Are you saying that there are no drive letters in Linux?" (page 14) says that Linux doesn't have drive letters but mount points. Correct so far.
But then it implies that '' /dev/fd0 (or /dev/floppy or /mnt/floppy)...)'' are such mount points. Eeeeehhhhhh?
More: ''...There can be many SCSI drives, on one or more controllers. The mount point designations just keep incrementing up through /dev/sdz. ...''
Gods! The writer doesn't know the difference between a device file and a mount point!
Ah, so the /mnt-files are just ''aliases''! Rrright. It's always nice to learn something new!
This came up on...July 29th on Slashdot. I guess the real technical writers are on summer vacation, and the article was submitted by the summer trainee. Am I wrong?
The site seems to have lots of introductionary (such as the new FAQ) and general technical articles (hardware stability, journaling filesystems, embedded Linux, firewalls, etc). The Linux Zone seems to be also an index to interesting documents on other sites.
Unfortunately some of the articles (at least the tutorials) require annoying registration.
However, Microsoft's vision seems to be making the .NET as popular and central in people's lives as Internet. In such a case, FS||OSS has to provide a solution, just like MS had to revolt their Internet strategy a few years ago.
I guess other parts of .NET will be ''easy'' to duplicate, except the Passport. So, create an alternative, based on an OPEN protocol, and make it preferably an ISO standard. Then, make the Free .NET support both.
Of course, it may be difficult for an alternative authentication&payment system to compete with Passport. But I'd believe some such systems exist already and many banks and other e-business companies will be competing with Microsoft anyhow. They'll just need co-operation from the FS||OSS community.
If USA is too tough a battleground, start in Europe. It's much easier to get an authentication/e-payment system working in a small country than in USA. An European system might have some taxation advantages too, as MS is a foreign company. It might also be possible to join with some mobile phone payment systems, which are already quite common.
20 lines now, about one coming every 15 mins.
Quite many seem to be coming Taiwanian or other Far-East countries such as Thailand.
Perhaps, but in that case it's rather misleading. Read it carefully. They write about ''software found on [this site]''.
But yes, probably that's just a mistake and the Linux kernel will be free and GPL as usual, but all the softwares on top of it will not. Probably you can't even develop software for it without their cossstly SDK. Much use, not.
Pay...ware? Linux? Me not understand?
The site www.linuxda.com says:
Excuse me, but I thought Linux is under the GPL license or something?
Dada.
Ok, the star has a lot of oxygen. The astronomers assume that the oxygen is bound to the carbon, but what if this is not the case? IANAA, but what if there's enough free oxygen and hydrogen than can bind, and then perhaps escape from the star? (I'm not sure if its possible that just the oxygen comes from the star and then binds to interstellar hydrogen.)
Another, perhaps better possibility might that the water is from an earlier stage, where all the oxygen was not bound to carbon. My introductionary astronomy book says about small stars:
Thus, if the escaping mixed matter is hydrogen and oxygen....-> huge cloud of water!
If this idea is valid, the astronomers probably have already thought about it and have taken it into account, and for some reason have judged it unsatisfactory.
I personally find it very difficult to imagine how a dying star could vaporize the comets around it, if it couldn't do it when it was still young and bright. Perhaps it could have done it during some middle-age crisis, for example if it was an inflated gas giant at some point.
*shrug*, IANAA, just an amateur.
Strangely, the typical OS developer is 21-22 years old male (98%). Well, perhaps 21-27. That means not much experience. 33% are students, so perhaps that explains it. But then, 37% are university graduates/masters, plus 4% PhDs.
78% are not paid for developing OS. This isn't nice. But 30% report having profited professionally.
3.5% do not work (I belong to that sad part). I just wonder which option all the students have reported...
The yearly income distribution is rather even in the entire range, perhaps about 30,000-40,000 EUR per year on average. I think US salaries are higher than european; it would be nice to have per-country statistics too.
They seem to be spending typically Typically they use Debian, although SuSE and RedHat are not far behind. 33% favour KDE, and 23% Gnome. Umm, 11% Windows, 7% "pure tex"? 36% are masochists who use vi (can this really be true?), while emacs has 28%.
Hmm, the site seems to be slashdotted, so I can't summarize the last page, "computer abilities".
Just imagine what would happen if the pants suddenly "detect" that the muscleman test pilot really wants to perform movements suitable for a ballet dancer.
After a few pretty splits and bends, which all 15-year old ballerinas would envy, the pieces of the soldier's bones are collected from the suit with tweezers.