Ah, that looks much nicer. I got the colour by using the colour picker with a large (15) averaged sample on the guys' white shirt. It's not actually a red tint, it's orange.
I think you meant to say big-endian machines, and the answer is still yes - the Dec/Compaq/HP Alpha was FreeBSD's second platform (if you don't count PC98 since that's still x86). 5.0 will add Sparc64 and PPC.
Oops, yes. I meant big-endian. That'll teach me for posting in the wee hours of the morning when my brain needs sleep.
Has it not been ported to 68k? That's often the other processor popular and accessable to hackers (mac, amiga, atari, and many old unix boxen were 68k).
...any PCI devices supported on FreeBSD x86 should work just fine on any other platform.
From skimming the LKML and Kernel Traffic for years, I gather the only big problem will perhaps be the endian difference. Has FreeBSD been ported to little-endian machines before?
Java runs nearly as fast as compiled code for most things.
Yeah, I've heard that a lot before. But I still find Java apps as slow as crap, so I avoid them like the plague. Is there some secret to get them running them "nearly as fast as compiled code"?
In these days of 2Ghz cpus on $700 computers, bitching about performance on your 6-year old PC is really not realistic.
<rant>
I find that comment highly arrogant. Let me guess, you're an American? Here in Oz, expect to pay closer to $1500-$2000 for a "2GHz" computer. My 750MHz Athlon is 2.5 years old and is fine for "most things", but Java apps still crawl. I've been unemployed for a while and even a couple hundred dollars is too much for me. I make my old (but recent) hardware do what I can, thankyou very much. I don't need some pretentious Java programmer telling me to upgrade just so that he can program in a trendy language and compile to an inefficient virtual machine code.
</rant>
1) Used by only one vendor and is going to remain that way
That one vendor is the UNIX volume leader.
2) It is tied to a particular hardware model and thus is likely to be outdated within a decade No, Quartz is quite hardware-independent.
So, until that "volume leader" ports to anything else than its own hardware, that point is kinda moot, isn't it?
# No matter how hard I try, I cannot grok the date-time functions, which I find to be extremely cryptic. For example, the simplest way I've found to calculate the number of days elapsed between today and a given date is...
Er, so you found the date_part() function, but you couldn't just use it to get the 'days' from your calculation? Try this simpler version:
date_part('days',now()-mydate)
BTW, I just checked that snippet on a (slowly growing) table of temperature readings. Not only does it work, but it doesn't round up like the more complicated version.
Genius, like it to you buying a VCR, somebody stealing it from you and then offering you to buy it from them, effectivly paying for it again..
The day after the hugely successful initial sale of shares (25%? 33%?), a female comedian on the radio made a similar analogy. I think it was Judith Lucy on Martin/Molloy, and she likened it to going down to the pawn shop the day after your house was burgled to buy all your stuff back.
Of course the shares sold like crazy. It's always been ours, and we didn't want to see Telstra added to the already-too-long list of Australian companies owned/controlled by overseas interests. Nothing racist or xenophobic, we just want the profits to stay within the country.
erm, not likely. Flash memory is pretty slow the last time I checked. Certainly not faster than a modern hard disk, even a 5400rpm one. And Flash memory only has a limited life time. It can only be rewritten by clearing a whole page (32k or so), and that can only be done a certain number of times (10k? 100k?) before it "wears out". Don't ask me why. It's weird.
Just get a nice 15K SCSI disk for root and swap. Now if only I had the money...:(
I'd love to have xawtv running on my screen, or to watch mplayer on the TV.
And what would you watching with MPlayer? A DVD perhaps? The DVD-CCA (or whoever) would be on ATi's back to make sure that Macrovision is always turned on and not easily disabled. Otherwise, you would be able to record a DVD onto VHS tape. Oh the horror!
I believe a number of years ago Matrox refused to release docs about some of their tv-out hardware to some Linux developers for the same reason. Well, guess what? The Linux hackers reverse-engineered the hardware and found out enough info on how to drive the hardware without Matrox's help. And they found (I think) the Macrovision enable/disable bit in one of the control registers.
Re:Open Source Pioneers? Or $$$ Saving?
on
Film Gimp
·
· Score: 1
Strange that you equate loving open source with hating Microsoft.
Not at all. Microsoft has all but declared war on Open Source software. So why shouldn't Open Source/Free Software users and enthusiasts return the sentiment?
Is that what your kindergarten teacher said to you today?
Ooh, you're really trolling aren't you? Luckily you included something more substantial, so I'll respond to that.
Apple took Unix [slashdot.org] and made it easy to use within what, a year?
Correct me if I'm wrong, but isn't a lot of the Aqua stuff based on NeXTstep/OpenStep that Steve Jobs brought with him from NeXT? That stuff went through a number of versions (version 4 was the last I think) over perhaps a decade. And my recollection might be off, but I thought it was longer than a year. Use your head, no one can come up with something like OS/X in only a year.
3. Reader is free, and the PDF spec API is open - there are some freeware products [miktex.org] that create PDF already
Most of which are crappy, and produce essentially "jpeg" PDF's - where the document is rendered as an image and stored in an PDF. Lame.
I use PDFlatex, you obviously haven't. For your information, PDFlatex does not simply render to an image and embed it into a PDF. I use it from LyX to produce perfectly fine PDF files. I just have to set the document font to "times" and it uses the default Times Roman font of the reader instead of embedding its own chunky bitmap rendition of the CM font. With that small change it produces nice, small files.
Having an opinion is fine, but go spread your misinformation and assumptions somewhere else, dan.
Every time there's an attempt to make it "easy to use", failure is the result. What was the name of the last company trying to make it easy to use? Easel, wasn't it? It was started up by the guys that started up Apple. They, like so many before them, failed. And you know where they ended up working? Apple.
That's right. Just ignore the fact that Easel had a hopeless business model. (Selling a file manager? get real). No, they were making Linux easy, and they failed. So everyone trying to make Linux easy is doomed to failure.
Lets all say it unison so that we learn it right:
Lots of companies sprung up in the dot-com era. Some of them were "Linux companies". Most of them had business models with un-sustainable and unrealistic revenue sources. Most of them crashed when the dot-com "bubble" burst. Don't be so quick to attribute their failures to Linux.
I agree that interopability is important for Linux. But I don't think you give good examples. Those problems were limited to specific platforms. Is it the kernel developers' job to work around other developers' errors? Why should they muddy and complicate the linux kernel (or whatever else) to work around someone elses errors?
In both cases, I think David's stand on the issues caused the respective companies and developers to move faster to fix their problems with their products, and the people using them.
Yes, Debian is volunteer driven, but that doesn't mean it couldn't be managed much better.
True, managing people and resources is certainly becoming a problem as Open Source projects grow larger and become more important. They could be managed better in some ways. But you can't go too far or you will stifle the openness, creativity and (gasp!) innovativeness of Open Source/Free software.
First is it has way too many packages it tries to support...
Second is that Debain tries to support way too many platforms...
...and where is your reasoning for your little piece of insight? You ramble a little about numbers, but nothing that you say (or that I've seen) backs up your assertions that Debian has too many packages or too many platforms. Is there too many bugs? Is it unstable? Is the release cycle too slow? What man, what?
I might point out that you seem to be forgetting that Debian is volunteer-driven. There is no central office that says "we will do this...". If people want to do it, they will do it.
(Yes, there is a central release manager that has final say over what packages and architectures go into the official release. But the decision is made based on stability and other qualities rather than some corporate goal or some such).
The difference between this and what browser you use is that your skin color or physical ability usually isn't your choice. Most people don't wake up one day and say "I think I will be disabled today" or "I think I will be a minority race today". You pretty much are or you aren't, and you can't usually change it easily.
Ok then, lets combine these two topics. What about blind users? Blind and visually impaired users typically (from what I understand) use a text browser with a speech synth. As you say, these people didn't wake up one day and decide to be unable to see properly. And because of their disability they must use an "alternative" web browser platform to view the web.
Oh, but visually impaired web people probably make up a few percent of all web users. So by your (and all the other troll's) reasoning it's not worth it to the banks to support these people.
What about other platforms? We're seeing an increasing number of non-windows, non-mac platforms appear with internet connectivity and web browsing capability. Game consoles? Mobile (cell) phones? Set-top boxes? PDAs? Fridges? Thankfully, MS can't possibly weasle wince onto all of these platforms, so IE won't always be "available". Face it, the future of the internet and the web will not be MS-only. Only by following real standards will all of this work together.
And if we want to make comparisons, I also code my small web site in XHTML and CSS2. The only references I use are the W3C docs on HTML 4.0, XHTML 1.0, and CSS 2. The only browsers I can test on are Mozilla and occasionally Konqueror since I only run Linux. Guess what? My site renders fine in IE and I didn't even try.
If the TCP-IP protocol were released under GPL, then anyone writing an application that used TCPIP in any fashion would have to be GPL as well. That means : No commercial web browsers. No commercial email programs.
Bzzt. Wrong, but thanks for playing.
You seem to be under the impression that programs using TCP/IP link in the TCP/IP stack. They don't. They use the interfaces provided by the operating system or by a seperate TCP/IP module (winsock?) of some sort. So commercial applications would not have been affected by a GPL'd TCP/IP stack. Operating systems are another matter however. As much as I prefer the GPL, I believe that the BSD license is more appropriate for material such as the reference TCP/IP stack.
As for you question at the end: I understand that the GPL only applies to source code.
Re:The secret of ./'s success....
on
Slashdot Turns 5
·
· Score: 1
Holy crap! A two digit ID number. And I thought my three-digit number was old. You don't see too many of us 'old timers' around here now.
...isn't it the case that you memcopy the entire address space of the forking process for each fork()...
No. There's this little trick called Copy On Write (COW). It uses the MMU and works like this.
Process fork()'s
Kernel sets up the child process.
Instead of copying the whole memory space over, it simply points the childs pagetable entries at the parents pages. Both the parents and childs pages are also marked as read-only (this is the important part).
Now, when either process goes to write to data in their memory spaces:
An exception is raised because the page is marked as read-only and the CPU jumps into the MM code of the kernel.
The kernel now does the copy for real, but only of that page.
The parent and child now have there own seperate copy of this page of data that they can work with.
Tada! Data is only copied when it's needed. Depending on the program involved, this may be a lot or a little. For programs that will immedediately perform an exec() after the fork, vfork() gives even better performance. Under Linux (according to the vfork(2) man page), vfork doesn't even copy the pagetables and it suspends the parent until an exec() is performed. It's kind of a hack, but it saves some work when all you want to do is spawn a new program. Which isn't all that uncommon in the Unix/Linux environment.
Why would a company devote time and resources for only a 10% return where they could spend 100% effort into marketing to a 90% MS desktop market.
I'll tell you why.
Because MS could decide that your neat little app is something that they'd like. They then either buy your company and fire the redundant employees, or they bundle (read "integrate") a wannabe clone with the next version of Windows/Office/Media Player/whatever.
Because Linux is a stable platform. How many changes has windows been through in the last 5 years? Sure, Linux advances as well, but a lot of MS's changes are to disadvantage competitors or would-be competitors. See previous point about becoming a would-be competitor.
Linux is an even playing field. Want to know how something works? It's all available for you to look at. Need more than that? Get on mailing lists or IRC channels. Hell, even hire one of the hackers yourself to guarantee a connection to the developers. You get "preview" access to betas and pre-release versions. Be ready when the next version of GNOME/KDE/X/the kernel/etc is released.
My point being that MS and its monopoly is a formidable force to go up against, so you don't want to gain their attention.
I don't know the answer either. I think that throwing Open Source software into the commercial scene (as is starting to happen) will shake it all up. Who knows what the result will be. We will probably find that certain types of software (niche apps, military) should be commercially developed, while other parts (OS, common libs, desktops, frameworks) should be community developed and Open Source.
You yourself have disassembled and audited the entire system, including CPU microcode.
You yourself have personaly programed, using only hardware (no software) that you yourself have audited, the entire system, including CPU microcode.
I think you're going just a little too far going down to microcode. To what level would some part of a computer system be able to recognise and/or modify a vote?
GUI toolkit? Probably not, but I'm no GUI programmer.
RDBMS? Probably, look at queries before they're parsed. Do a regexp-based scan of the SQL and change candidate A with candidate B a certain percentage of the time. That is, if the name is actually used instead of just an ID #.
OS? Probably not, too low level.
Compiler? Yes I know about the famous Ken Thomson hack with login and cc. But you have to know pretty much the exact code to look for.
Processor core/microcode? That's just getting silly.
Stick to paper. Maybe scan/count it electronicaly, but keep an audit trail that can't be modified electronicaly.
I agree. Several other people have made good suggestions of a system that uses punched/stamped cards that are taken to a machine that scans and verifies the vote. Personally, I'd prefer a system that uses my PGP/GPG key to verify my vote. But I guess it will be a long time before a significant portion of the population have their own crypto keys.
Slashdot Mirror
Also in the gimp:
Ah, that looks much nicer. I got the colour by using the colour picker with a large (15) averaged sample on the guys' white shirt. It's not actually a red tint, it's orange.
Oops, yes. I meant big-endian. That'll teach me for posting in the wee hours of the morning when my brain needs sleep.
Has it not been ported to 68k? That's often the other processor popular and accessable to hackers (mac, amiga, atari, and many old unix boxen were 68k).
From skimming the LKML and Kernel Traffic for years, I gather the only big problem will perhaps be the endian difference. Has FreeBSD been ported to little-endian machines before?
Yeah, I've heard that a lot before. But I still find Java apps as slow as crap, so I avoid them like the plague. Is there some secret to get them running them "nearly as fast as compiled code"?
<rant>
I find that comment highly arrogant. Let me guess, you're an American? Here in Oz, expect to pay closer to $1500-$2000 for a "2GHz" computer. My 750MHz Athlon is 2.5 years old and is fine for "most things", but Java apps still crawl. I've been unemployed for a while and even a couple hundred dollars is too much for me. I make my old (but recent) hardware do what I can, thankyou very much. I don't need some pretentious Java programmer telling me to upgrade just so that he can program in a trendy language and compile to an inefficient virtual machine code.
</rant>
So, until that "volume leader" ports to anything else than its own hardware, that point is kinda moot, isn't it?
Er, so you found the date_part() function, but you couldn't just use it to get the 'days' from your calculation? Try this simpler version:
BTW, I just checked that snippet on a (slowly growing) table of temperature readings. Not only does it work, but it doesn't round up like the more complicated version.
Welcome to the world of Slashdot, where trolls drag almost anything into a Linux/MS issue.
Zachary Beane of GIMP fame, has a MySQL to PostgreSQL migration page with a Perl script and some advice.
The day after the hugely successful initial sale of shares (25%? 33%?), a female comedian on the radio made a similar analogy. I think it was Judith Lucy on Martin/Molloy, and she likened it to going down to the pawn shop the day after your house was burgled to buy all your stuff back.
Of course the shares sold like crazy. It's always been ours, and we didn't want to see Telstra added to the already-too-long list of Australian companies owned/controlled by overseas interests. Nothing racist or xenophobic, we just want the profits to stay within the country.
erm, not likely. Flash memory is pretty slow the last time I checked. Certainly not faster than a modern hard disk, even a 5400rpm one. And Flash memory only has a limited life time. It can only be rewritten by clearing a whole page (32k or so), and that can only be done a certain number of times (10k? 100k?) before it "wears out". Don't ask me why. It's weird. :(
Just get a nice 15K SCSI disk for root and swap. Now if only I had the money...
And what would you watching with MPlayer? A DVD perhaps? The DVD-CCA (or whoever) would be on ATi's back to make sure that Macrovision is always turned on and not easily disabled. Otherwise, you would be able to record a DVD onto VHS tape. Oh the horror!
I believe a number of years ago Matrox refused to release docs about some of their tv-out hardware to some Linux developers for the same reason. Well, guess what? The Linux hackers reverse-engineered the hardware and found out enough info on how to drive the hardware without Matrox's help. And they found (I think) the Macrovision enable/disable bit in one of the control registers.
Not at all. Microsoft has all but declared war on Open Source software. So why shouldn't Open Source/Free Software users and enthusiasts return the sentiment?
Ooh, you're really trolling aren't you? Luckily you included something more substantial, so I'll respond to that.
Correct me if I'm wrong, but isn't a lot of the Aqua stuff based on NeXTstep/OpenStep that Steve Jobs brought with him from NeXT? That stuff went through a number of versions (version 4 was the last I think) over perhaps a decade. And my recollection might be off, but I thought it was longer than a year. Use your head, no one can come up with something like OS/X in only a year.
I use PDFlatex, you obviously haven't. For your information, PDFlatex does not simply render to an image and embed it into a PDF. I use it from LyX to produce perfectly fine PDF files. I just have to set the document font to "times" and it uses the default Times Roman font of the reader instead of embedding its own chunky bitmap rendition of the CM font. With that small change it produces nice, small files.
Having an opinion is fine, but go spread your misinformation and assumptions somewhere else, dan.
That's right. Just ignore the fact that Easel had a hopeless business model. (Selling a file manager? get real). No, they were making Linux easy, and they failed. So everyone trying to make Linux easy is doomed to failure.
Lets all say it unison so that we learn it right:
Silly me, I thought I read that they're using diskless Linux Terminals. I must have read a different article to you.
I agree that interopability is important for Linux. But I don't think you give good examples. Those problems were limited to specific platforms. Is it the kernel developers' job to work around other developers' errors? Why should they muddy and complicate the linux kernel (or whatever else) to work around someone elses errors?
In both cases, I think David's stand on the issues caused the respective companies and developers to move faster to fix their problems with their products, and the people using them.
I might point out that you seem to be forgetting that Debian is volunteer-driven. There is no central office that says "we will do this...". If people want to do it, they will do it.
(Yes, there is a central release manager that has final say over what packages and architectures go into the official release. But the decision is made based on stability and other qualities rather than some corporate goal or some such).
Ok then, lets combine these two topics. What about blind users? Blind and visually impaired users typically (from what I understand) use a text browser with a speech synth. As you say, these people didn't wake up one day and decide to be unable to see properly. And because of their disability they must use an "alternative" web browser platform to view the web.
Oh, but visually impaired web people probably make up a few percent of all web users. So by your (and all the other troll's) reasoning it's not worth it to the banks to support these people.
What about other platforms? We're seeing an increasing number of non-windows, non-mac platforms appear with internet connectivity and web browsing capability. Game consoles? Mobile (cell) phones? Set-top boxes? PDAs? Fridges? Thankfully, MS can't possibly weasle wince onto all of these platforms, so IE won't always be "available". Face it, the future of the internet and the web will not be MS-only. Only by following real standards will all of this work together.
And if we want to make comparisons, I also code my small web site in XHTML and CSS2. The only references I use are the W3C docs on HTML 4.0, XHTML 1.0, and CSS 2. The only browsers I can test on are Mozilla and occasionally Konqueror since I only run Linux. Guess what? My site renders fine in IE and I didn't even try.
Bzzt. Wrong, but thanks for playing.
You seem to be under the impression that programs using TCP/IP link in the TCP/IP stack. They don't. They use the interfaces provided by the operating system or by a seperate TCP/IP module (winsock?) of some sort. So commercial applications would not have been affected by a GPL'd TCP/IP stack. Operating systems are another matter however. As much as I prefer the GPL, I believe that the BSD license is more appropriate for material such as the reference TCP/IP stack.
As for you question at the end: I understand that the GPL only applies to source code.
Holy crap! A two digit ID number. And I thought my three-digit number was old. You don't see too many of us 'old timers' around here now.
No. There's this little trick called Copy On Write (COW). It uses the MMU and works like this.
Now, when either process goes to write to data in their memory spaces:
Tada! Data is only copied when it's needed. Depending on the program involved, this may be a lot or a little. For programs that will immedediately perform an exec() after the fork, vfork() gives even better performance. Under Linux (according to the vfork(2) man page), vfork doesn't even copy the pagetables and it suspends the parent until an exec() is performed. It's kind of a hack, but it saves some work when all you want to do is spawn a new program. Which isn't all that uncommon in the Unix/Linux environment.
I'll tell you why.
My point being that MS and its monopoly is a formidable force to go up against, so you don't want to gain their attention.
I don't know the answer either. I think that throwing Open Source software into the commercial scene (as is starting to happen) will shake it all up. Who knows what the result will be. We will probably find that certain types of software (niche apps, military) should be commercially developed, while other parts (OS, common libs, desktops, frameworks) should be community developed and Open Source.
I think you're going just a little too far going down to microcode. To what level would some part of a computer system be able to recognise and/or modify a vote?
I agree. Several other people have made good suggestions of a system that uses punched/stamped cards that are taken to a machine that scans and verifies the vote. Personally, I'd prefer a system that uses my PGP/GPG key to verify my vote. But I guess it will be a long time before a significant portion of the population have their own crypto keys.