Then who is supposed to fund it? The average person?
The annoying thing is, that's what federal funds come from: Average people's tax. No government should be able to limit research just based on some arbitrary pseudo-religious/ethical argument.
When you can't answer, you can always phone a friend?
Slowly and surely, I'm teaching my colleauge sitting next to me with the arcane art of shells. He has improved quite a lot. These days he can ssh into a Linux box, do what he wants to do and exit without damaging anything, even if he logins as root.
I was just about to mention TORA as well. On the other hand Tora in sourceforge is GPL'ed code. If you can get it compiled on Windows I can'see where the problem would be (apart from QT libraries problem).
I wasn't quite clear that this wasn't the only reason he prefers film, but the point is that digital can be a crutch that allows you to only get 15 good pictures out of 500 rather than a much higher percentage.
Actually digital would make more sense for your friend. With a large memory card, just keep shooting. You can't plan for something like groom's face unless you are shooting for it, trying not to miss it.
Anyway, coming to this one, digital or film, type of camera will not magically improve your picture quality unless you teach yourself how to shoot. The best way, as far as I can tell, is shooting often, shooting many shots of a single subject, noting down what's changed, camera settings etc. Digital (with EXIF) makes one stage of this training easier. The second stage is, again, the cost. I couldn't afford more than a roll when I was a student. If I was a student now, armed with a cheap 3MPixel camera with full manual override I would manage to train myself much quicker.
No photographer, with or without an LCD, will stop and look at the outcome if he is planing to take a number of pictures very rapidly. Again back to your friend, he will never stop to see if groom was looking good in the picture he just shot. He would just keep on shooting, doing the selection at his studio, on a large screen. LCD is not a factor in his case. LCD is a factor for a typical point-and-shoot user and for the ones who are training themselves. It is priceless.
Real photographers don't worry about the shots. They take a lot of shots and throw not-that-good ones away. Digital makes it easier.
What you describe is the budget photographer, people like you and me, who really can't afford to take 10 rolls a day and get them printed to select one or two out of the lot. I owned a film camera for years and used to finish a roll or two a month, thinking about most of the shots very carefully, if I had the time. I used to play with speed and aperture of the camera (because I have a full-manual camera, not these fancy auto-focus/auto-everything toys Canon and Nikon usually sells these days). I used to get about 15 good pictures out of my 36 picture roll.
Now I shoot around 500 pictures a month using my digital camera (not so high end either) but I still end up with around 15 good pictures a month. I still (almost all the time) take pictures with full-manual mode, apart from auto-focus. I am satisfied.
You are twisting the truth a little bit. Usually the resolution of a film is limited with the granularity, higher ISO rating, worse is the resolution. I have to agree that below 5 megapixels the printouts don't look that good on an A4, I can't remember when I managed to print a single picture of mine (at an affordable cost) on an A4 size photographic paper before the advent of photo-quality colour printers so I can't really compare the quality. What I can compare is if I scan a medium size (postcard size) picture printed at a shop and compare with a similar picture taken with my 3 megapixel camera, most of the times my camera has worse resolution. On the other hand I can clearly see the granules in my scanned picture when I zoom to 100% (200ASA film, 1200dpi Canon Lide 30 scanner)
I have a Fuji with a SCCD and I am very unsatisfied with it. Its RAW tools out of the box are crap and I couldn't find a linux utility to conver RAWs last time I looked for. As a result I am using JPG compression and it has no way of saving the files as "finer". They tend to come out pretty bad if taken under low-light conditions, with lots of artifacts. Canon on the hand, even with their lower-end A95 cameras deliver what you pay for. This is probably first and the last time I got myself a Fuji.
Given enough memory it would circle around Win3.1 and Win'95 and Win NT when finally it come out and wouldn't perform where Warp would.
OS/2 always got hammered because it needed 16MB to be comfortable and those days a server usually had 8MB. I had 8 and I was running a BBS on my PC. It was significantly smoother, never dropped a single package over the modem while I was working on my CAD software (which alone used over 8MB of RAM), constantly swapping in and out. Win3.1 even couldn't handle me moving the mouse with a user downloading. Win95 wasn't an improvement.
Most of the Win95 and OS/2 users were single-task users. It really showed its power when you used it as a server or a real multi-task environment. Later on I ran MUDs and httpd daemons on it and it always performed faster than anything Microsoft could supply. The lack of graphics card driver support really doesn't matter if you are content with a VGA screen, who needs graphics on servers in any case?
Where it failed is the developers. Steve Balmer wasn't shouting "Developers! Developers! Developers!" for no reason. IBM's expensive compilers and other suppliers' (i.e., Borland) lack of commitment effectively what killed OS/2. There was a limit on what you really wanted to do with gcc.
XFree86 worked fine under OS/2 and for some time I used it as a X server pretty happily. Once you have XFree, porting GTK or QT shouldn't be a big issue, probably already done (I haven't checked recently, I turned the power off of the latest OS/2 box I was administering around year 2000).
Don't forget that MS's most blatant violations of anti-competitive laws are actually from this period in history. Once it crushed IBM and other ISVs, now it is the established monopoly. Linux is not a company so you can't be anti-competitive against it, can you?:)
IBM's lack of drive really drove me nuts. Not having a decent compiler (which didn't cost an arm and a leg) for OS/2 drove me to gcc, once I got convinced that it would actually make more sense to run Linux full time because half of my time was spent debugging and modifying code so that they would work with OS/2 I did the switch. Also Yggdrassil and Slackware's totally-usable distibutions were a pleasure to use and my 486 performed many times better than with OS/2, for a while at least. Also if I remember correctly I couldn't (easily) compile my pascal code to OS/2 binaries either. I finally had all of the tools but by that time I was pretty fed up with it.
First versions of WinNT had the capability to run OS/2 1.x code, mainly because of their shared codebase and standards. IBM also got Win16 applications run on OS/2 but this was mainly done by installing a normal (sometimes special) version of Windows 3.1, which made MS incredibly happy because every OS/2 sold meant an Win3.1 sold as well.
I would love to see something like KDE on OS/2's kernel. It would rock. WPS was much better than Win3.1 AND Win95 but you have to get used to it. Rexx was built-in to the system and it was beautiful.
I had a BBS running on OS/2 and Maximus/2 with terminal servers coming in as telnet connections + 8 modem lines. It worked brilliantly. We shut the whole thing down after Y2K started to kill other BBS software around the country and there were almost none to communicate with. Ocassionally I'm still blamed with killing the community. They should have moved on, I already had.
OS/2 is out of favour for some time now. It's been years since I used it as my primary OS (well, since early Linux kernels actually). As a result most of the/. readers do not even know what a nice system OS/2 Warp was and don't care. Only some people will remember OS/2 with fond memories. Simon doesn't.
My HF radioes don't have DSB with no carrier but one has DSP, all have DSB with carrier not supressed (AM) but I never used it that way, it also reduces the PEP to 25W on my solid-state rig.:-)
Not really. Most of their EMEA people were a) Linux people came to work with them because of a steady salary (I was in this group and I needed the money b) OpenUnix people who knew almost nothing about any modern Unix system. They were constantly amazed with the features of Unixware (which was bought from Novell).
Their sales pitch were pathetic and especially my boss was a complete dickhead and was joking about "Linux being a toy OS". I tried to explain what was going to happen for months and he would just laugh. I'm pretty glad that he laughed all his way to bankrupcy.
What (original) SCO had was the marketing channel all the way to the customers which were desperate to move away from OpenUnix. Also assuming that you would be able to train group B of their personnel, they would also have a lot of trained people on the ground when needed.
Slashdot Mirror
The annoying thing is, that's what federal funds come from: Average people's tax. No government should be able to limit research just based on some arbitrary pseudo-religious/ethical argument.
Noooooooooooooooooooooooooooooooooooooooooooo!
OpenServer IS evil.
If you have ever used it, you would understand what I mean.
Slowly and surely, I'm teaching my colleauge sitting next to me with the arcane art of shells. He has improved quite a lot. These days he can ssh into a Linux box, do what he wants to do and exit without damaging anything, even if he logins as root.
Readme file? Come on. /etc/motd...
I was just about to mention TORA as well. On the other hand Tora in sourceforge is GPL'ed code. If you can get it compiled on Windows I can'see where the problem would be (apart from QT libraries problem).
Actually digital would make more sense for your friend. With a large memory card, just keep shooting. You can't plan for something like groom's face unless you are shooting for it, trying not to miss it.
Anyway, coming to this one, digital or film, type of camera will not magically improve your picture quality unless you teach yourself how to shoot. The best way, as far as I can tell, is shooting often, shooting many shots of a single subject, noting down what's changed, camera settings etc. Digital (with EXIF) makes one stage of this training easier. The second stage is, again, the cost. I couldn't afford more than a roll when I was a student. If I was a student now, armed with a cheap 3MPixel camera with full manual override I would manage to train myself much quicker.
No photographer, with or without an LCD, will stop and look at the outcome if he is planing to take a number of pictures very rapidly. Again back to your friend, he will never stop to see if groom was looking good in the picture he just shot. He would just keep on shooting, doing the selection at his studio, on a large screen. LCD is not a factor in his case. LCD is a factor for a typical point-and-shoot user and for the ones who are training themselves. It is priceless.
What you describe is the budget photographer, people like you and me, who really can't afford to take 10 rolls a day and get them printed to select one or two out of the lot. I owned a film camera for years and used to finish a roll or two a month, thinking about most of the shots very carefully, if I had the time. I used to play with speed and aperture of the camera (because I have a full-manual camera, not these fancy auto-focus/auto-everything toys Canon and Nikon usually sells these days). I used to get about 15 good pictures out of my 36 picture roll.
Now I shoot around 500 pictures a month using my digital camera (not so high end either) but I still end up with around 15 good pictures a month. I still (almost all the time) take pictures with full-manual mode, apart from auto-focus. I am satisfied.
You are twisting the truth a little bit. Usually the resolution of a film is limited with the granularity, higher ISO rating, worse is the resolution. I have to agree that below 5 megapixels the printouts don't look that good on an A4, I can't remember when I managed to print a single picture of mine (at an affordable cost) on an A4 size photographic paper before the advent of photo-quality colour printers so I can't really compare the quality. What I can compare is if I scan a medium size (postcard size) picture printed at a shop and compare with a similar picture taken with my 3 megapixel camera, most of the times my camera has worse resolution. On the other hand I can clearly see the granules in my scanned picture when I zoom to 100% (200ASA film, 1200dpi Canon Lide 30 scanner)
I have a Fuji with a SCCD and I am very unsatisfied with it. Its RAW tools out of the box are crap and I couldn't find a linux utility to conver RAWs last time I looked for. As a result I am using JPG compression and it has no way of saving the files as "finer". They tend to come out pretty bad if taken under low-light conditions, with lots of artifacts. Canon on the hand, even with their lower-end A95 cameras deliver what you pay for. This is probably first and the last time I got myself a Fuji.
OS/2 always got hammered because it needed 16MB to be comfortable and those days a server usually had 8MB. I had 8 and I was running a BBS on my PC. It was significantly smoother, never dropped a single package over the modem while I was working on my CAD software (which alone used over 8MB of RAM), constantly swapping in and out. Win3.1 even couldn't handle me moving the mouse with a user downloading. Win95 wasn't an improvement.
Most of the Win95 and OS/2 users were single-task users. It really showed its power when you used it as a server or a real multi-task environment. Later on I ran MUDs and httpd daemons on it and it always performed faster than anything Microsoft could supply. The lack of graphics card driver support really doesn't matter if you are content with a VGA screen, who needs graphics on servers in any case?
Where it failed is the developers. Steve Balmer wasn't shouting "Developers! Developers! Developers!" for no reason. IBM's expensive compilers and other suppliers' (i.e., Borland) lack of commitment effectively what killed OS/2. There was a limit on what you really wanted to do with gcc.
Since IBM couldn't afford pushing OS/2 instead of Windows because Microsoft would cut its Win supply.
Who cares... I don't.
You mean african-american, obviously. Or nigger.
You mean they started to steal our jobs much much much earlier???
:-)
XFree86 worked fine under OS/2 and for some time I used it as a X server pretty happily. Once you have XFree, porting GTK or QT shouldn't be a big issue, probably already done (I haven't checked recently, I turned the power off of the latest OS/2 box I was administering around year 2000).
I sometimes think Simon never used OS/2.
IBM's lack of drive really drove me nuts. Not having a decent compiler (which didn't cost an arm and a leg) for OS/2 drove me to gcc, once I got convinced that it would actually make more sense to run Linux full time because half of my time was spent debugging and modifying code so that they would work with OS/2 I did the switch. Also Yggdrassil and Slackware's totally-usable distibutions were a pleasure to use and my 486 performed many times better than with OS/2, for a while at least. Also if I remember correctly I couldn't (easily) compile my pascal code to OS/2 binaries either. I finally had all of the tools but by that time I was pretty fed up with it.
First versions of WinNT had the capability to run OS/2 1.x code, mainly because of their shared codebase and standards. IBM also got Win16 applications run on OS/2 but this was mainly done by installing a normal (sometimes special) version of Windows 3.1, which made MS incredibly happy because every OS/2 sold meant an Win3.1 sold as well.
I would love to see something like KDE on OS/2's kernel. It would rock. WPS was much better than Win3.1 AND Win95 but you have to get used to it. Rexx was built-in to the system and it was beautiful.
I had a BBS running on OS/2 and Maximus/2 with terminal servers coming in as telnet connections + 8 modem lines. It worked brilliantly. We shut the whole thing down after Y2K started to kill other BBS software around the country and there were almost none to communicate with. Ocassionally I'm still blamed with killing the community. They should have moved on, I already had.
It was Yggdrasil what convinced me to switch completely. Ah.. Good old days. :)
OS/2 is out of favour for some time now. It's been years since I used it as my primary OS (well, since early Linux kernels actually). As a result most of the /. readers do not even know what a nice system OS/2 Warp was and don't care. Only some people will remember OS/2 with fond memories. Simon doesn't.
My HF radioes don't have DSB with no carrier but one has DSP, all have DSB with carrier not supressed (AM) but I never used it that way, it also reduces the PEP to 25W on my solid-state rig. :-)
Their sales pitch were pathetic and especially my boss was a complete dickhead and was joking about "Linux being a toy OS". I tried to explain what was going to happen for months and he would just laugh. I'm pretty glad that he laughed all his way to bankrupcy.
What (original) SCO had was the marketing channel all the way to the customers which were desperate to move away from OpenUnix. Also assuming that you would be able to train group B of their personnel, they would also have a lot of trained people on the ground when needed.