Re:Believe it or not, Apple's DRM doesn't bother m
on
iTunes DRM Hole Closed
·
· Score: 1
With bandwidth costs going down, this may not be an issue forever, but I think the major reason for compressing music is so it downloads quickly.
If the aim is just to get lossless (well, there's no such thing, but lossless in terms of being encoded at a reasonable PCM bitrate) content, and convenience is not an issue (and loading 300Mb at a time, which is the best you'd get for a FLAC-style compressed CD's worth of music), then CDs remain the best way of getting the content. After fiddling with iTMS, downloading a few tunes, finding the DRM is enough of a PITA to make the whole "convenience" selling point bunk - to me - I have to admit that's pretty much the route I ended up taking. Yes, I could have used DVD Jon's hacks, but I tend to go for the "If it's breaking the law (even an insane law), and there's a legal alternative to do the same thing, go for the latter."
Camino's better integrated with OS X and probably fractionally lighter on the resource requirements. Integration can mean more than you think - until relatively recent nightly builds, the OS X version of Firefox didn't support the middle mouse button, for example.
OS X manages things like proxies and other network settings as part of the OS, so you can relatively easily (but not seamlessly, alas) switch from Ethernet to WLAN to modem connections just by selecting the configuration you're using from the Apple menu. Firefox doesn't pick up the proxy settings itself if you do this. Camino would.
Now, that said, Camino isn't compatable with Firefox's plug-ins, and I don't know about you but I've found it's gotten hard to browse without Adblock and Flashblock having tried them. (Hacks to add crippled functionality similar, but not really, to the two are usually quoted when I mention this, but the full blown "I want to be able to get a flash animation to start only when I click it" and "I want to be able to easily add wildcards to my ad blocking script" functionality just isn't present with these "answers.")
I don't think anyone's arguing there's never a reason to use XML. The issue is more that there's a mentality at the moment that XML is the solution to everything.
Sure, you can store extremely heirarchical data in XML. Sure you can store Unicode in XML too. But what if, say, you're storing relatively simple information that doesn't really have either of those issues?
In response to an AC below, I suggested the fstab as an example of a file that really shouldn't be rewritten in XML. The advantages and disadvantages of both approaches would be:
fstab:
Easily parsed by any language, OOP or otherwise. C and shell scripts can easily parse it.
Catting it generally results in less than a screenful of information (any verbosity is usually down to copyright notices), meaning that all the information a sysadmin needs is available at a glance.
Easy to manually scan - finding/mnt/cdrom is a matter of glancing down the first column. All related information to it is thus on the first line. You can answer the question "Is/mnt/cdrom mounted upon boot?" in a second or two.
"fsxml"
Can be parsed by a standards compliant library, but such libraries are not available in a form easier to use than manual scanning of fstab for lower level HLLs like C.
Can be validated automatically, though testing validity is easy for fstab as mount can be run manually and will complain about problems.
Is potentially self documenting, assuming obvious tags are chosen, though comments above column headings in fstab makes this redundant.
The key thing with fstab, of course, is that mount, a simple C program, and some start up scripts, written (of course) in Bourne Shell, need to read it. That's pretty much the clincher - and it's true, actually of a lot of files in/etc. It needs to be simple, and XML isn't just overengineered for the purpose, it's positively unsuitable.
If XML is as great as its more enthusiastic supporters suggest, then Oracle needs to be rewritten so its tables are stored in XML. Its indexes too. ELF needs to be discarded in favour of XML. File systems like ext2fs and ReiserFS need to be replaced with XML too.
Clearly that's over the top, and no supporter, no matter how enthusiastic, really would suggest such a thing (though I wouldn't put it past someone to suggest replacing ELF...), but the point is that it really isn't suitable for every application. More to the point, it's not an optimal solution (like all portable, extendable, generic formats) for the vast majority of applications. It's a useful solution, but it's not even the only portable, extendable, format, and those who propose it knee-jerk for everything are doing a grave disservice to the format itself.
I hear "difficult to parse" a lot. Why aren't you using an off-the-shelf parser? Not having to write a parser is one of the benefits of XML!
Because we're talking about lightwieght programs. Programs like "mount". "Off the shelf parsers" are generally not available in any usable form for non-OOP languages.
As for difficult to read, well, sorry, that's a matter of taste, and there are more than enough people that disagree with you. You can't use that as an objective reason to dismiss XML.
No, I can. XML is difficult to read because the data is mixed with verbose, redundant, information. Now, I actually gave you an example. I suggested you write an XML equivalent of the fstab file. Don't pretend I wasn't being objective here!
I can't figure out what you mean by this. Getting the filesystem to appear on part of a screen? What?
cat fstab
Now cat your XML equivalent.
Sure. You do know XML is human-readable, right? Some morons on Slashdot like to post things like "you need special software to edit XML, I want vi!", so it's obvious not everybody who likes to criticise XML has a clue about it.
Did I say it wasn't? I was comparing the two, and pointing out that the simpler design is the more readable.
cat fstab, and at a glance you can see everything about/mnt/cdrom. It takes a second to find the line refering to it and understand how/mnt/cdrom is set up.
cat "fsxml", on the other hand, and you'll have to plough through a large amount of what's essentially garbage to find the text you're looking for. Now, before you reply, please, do what I'm suggesting. Take your fstab file and convert it to the most obvious form of XML. It's not difficult. It'll take you five to ten minutes. Once you've done this, we'll be on the same page. You'll see instantly what I'm refering to. You'll see you can no longer easily see, at a glance, the information you're trying to find. You'll no longer be able to fit it on one screen.
The format of fstab is hardly wonderful, but compared to XML it is instantly, easily, readable.
For the specific application of listing available file systems in a Unix environment together with their security and mounting attributes, fstab is a more optimal solution than XML.
XML doesn't require C++, so that question isn't even on the table.
Technically it's possible to write an XML parser in regular C. It's hard. It's complicated. It's difficult to do without producing a complex, difficult to maintain, large program.
Realistically, you wouldn't do it.
And if you keep it in C, it's going to be a PITA to actually parse that fstab file.
Why? One basic model for pulling values out of an XML config file is this:
1. Instantiate parser.
C, not C++. C. You can't "instantiate" instances in C. It has no concept. That's the point.
No, you are wrong, it still looks like needless wheel reinventing. And you forgot to multiply that "classic example" by... ~200 config files in my/etc (with virtually no shared syntaxes).
I'm surprised anyone would describe it as "reinventing the wheel". XML came after the CSV or white-space delimited file. XML is infinitely more complex, is objectively harder to read, and objectively more difficult to implement. While off-the-shelf libraries provide a semi-solution to the latter, XML's complexity and required Unicode support means that it's a bugger to use for simpler languages like C.
Asserting it isn't, and then quoting what you'd do in C++ and pretending it's the same in C as a justification, ignoring (deliberately?) an objective, relevent, example, and - irony of ironies - complaining that white-space delimited files are "reinventing the wheel" as if to pretend XML isn't, strikes me as not very convincing.
Implement my example. For shits and giggles, feel free to modify the mount command to support it (you'll have to learn C first though) before responding. I think you'll see my point.
Oh for heaven's sake, what's with the anti-XML bigots on Slashdot?
You need to see a doctor about that jerking knee of yours.
XML is a good solution for a small set of particular formats. For everything else it's either overkill or too limited.
XML is being treated right now by a lot of programmers as the be-all and end-all solution to everything. It isn't. It's difficult to parse for otherwise lightwieght programs, it's difficult to read, and it's verbose. If you doubt this, try converting, as an example, fstab to XML. It's deceptively easy and the solution is apparently obvious.
Then, once you've done that, play around with it. Consider the consequences of what you've just done. Can you get the entire file system to appear on part of one screen any more, like you could the old fstab? Can you easy, at a glance, see the settings related to/mnt/cdrom?
Now consider all the programs that use fstab. Do you really want to rewrite "mount" in C++? You're not going to be able to rewrite it in Python, it's too low level. And if you keep it in C, it's going to be a PITA to actually parse that fstab file.
Suddenly this classic example of a "horrible" old-style configuration file doesn't look so bad, does it?
I'm not an XML bigot. I'm a realist. I appreciate it annoys the thought-free fan-boys who jump on a new technology, see some tiny part of it that looks good, and suddenly enthuse about how it can revolutionise everything, but just as OOP is a great way of writing interactive GUI-based apps, but pretty ugly the further you get away from that, XML is a great way of expressing certain types of complex hierarchical textual information, but pretty ugly the further you get from that. If it's simple, you don't want to express it in XML.
Remember I made the criticism of programmers who think XML is better than simplicity? You kind of demonstrated the issue with your example. Not only is what you wrote harder to read than the equivalent ".INI", but it's actually only a small portion - there's the XML headers to take care of too.
XML is cluttered and complicated. It is possible to "cleanly" express some content types better in XML, but for the vast majority of cases, it's overkill and unnecessary added complexity. What's worse, it's not even a real "solution" to complex data - complex data is rarely limited to information best expressed textually.
I make no secret of not liking XML. I'm not saying it's never appropriate, but most of the time, when I see it, you've got to wonder what crack the user was smoking when he or she chose it. It's a modern day programmers fetish that's inappropriate for most of the applications it's used for.
Whatever solution you come up with, it's unlikely to remain clean for long. I'm sure the inventors of the Windows Registry thought they'd come up with a clean solution, and I know the GConf people did too. Are they clean? Do you like them?
I don't.
Leave/etc as it is. It works. It's done the right job now for coming on 30 years. It's had one major clean up (executables moved to/sbin), and otherwise it's nice. You can grep it. Help is a manpage away. With the exception of a handful of recent apps from programmers who don't "get it" and think XML is better than simplicity, the config files are, by and large, consistant, within a certain mindset.
There's lots of things that could be improved about Unix. Destroying/etc doesn't really improve it.
Yea, but it's kinda hard to sell it when you already gave it away for free.
He was telling the AC that the reason he was chosing the GPL over the BSD license was that if he gave it away for free, ie licensed under the BSD license, nobody would "buy" copies from him.
You then responded to him with this:
Well licensing under GPL is not giving it away for free. It is licensing the use of the code.
If you are wanting to use the code in a traditional commercial closed source sense then the GPL is no good so you may very well be prepared to pay for use.
I assumed you'd misread (and still believe you're misreading) and thought Greger47's point was that giving the software under the GPL was "giving it away for free". Otherwise there'd have been no point in making the comment.
Gut feeling - most Amigans not hostile to the concept of Free Software and Open Source moved to Free Software and Open Source. Matt Dillon would be an example, a pioneer in the Amiga world, responsible for closed and open software alike, he's now very big in the BSD community.
Amiga was a big Free Software platform in its time. Most of the networking code was Free Software, and there was fury in the community when the AmiTCP authors took AmiTCP 4 proprietary.
It's been increasingly difficult to be an Amiga user in the last ten years. The platform has practically zero support. Anyone with choices has moved on. Those who loved the platform moved to things like GNU/Linux and got working on trying to make their new platform have the features they missed from the old one.
It's not altogether surprising that a significant minority amongst those left would be FOSS hostile. If they weren't, they'd probably not still be using the Amiga.
Shame about MorphOS though. It'd be nice to see a Free Software attempt at a next-generation AmigaOS (I don't mean AROS, I mean to AmigaOS what Mac OS X is to classic Mac OS, or Windows NT was to DOS/Windows.)
Re:Good reasons for chosing GPL over BSD
on
Tracking GPL Violators
·
· Score: 2, Insightful
I think you're still misunderstanding the point. He was saying that giving it under the BSD license would be "giving it away for free". If, as he does, he licenses it under the GPL, he's able to sell to companies that want to make a proprietary version. If he licenses it under the BSD, there's no point, because same companies don't have to pay him to produce proprietary versions, they can just get the BSD licensed version.
(I posted this earlier but can't see it. Slashdot eating posts? Mod on crack?)
You're modded as funny, but it really isn't GPL compatable and arguably isn't Open Source or Free Software either. It's incompatable with clause 6 of the GPL:
6. Each time you redistribute the Program (or any work based on the Program), the recipient automatically receives a license from the original licensor to copy, distribute or modify the Program subject to these terms and conditions.
You may not impose any further restrictions on the recipients' exercise of the rights granted herein. You are not responsible for enforcing compliance by third parties to this License.
(My emphasis.)
I think the "no whining" thing is dumb comment to put in a license and shows a pointlessly contentious attitude by the Bitkeeper people, something they've shown ever since the question of "Why is a major Free Software project being developed using and depending on unfree software?" issue was raised. They'd have been better off saying "This is our business model. We appreciate it will not be considered appropriate for Free Software, and suggest Free Software developers look elsewhere", and leaving it at that. The criticisms, for the most part, were of Linus and his decision to use Bitkeeper until the Bitkeeper people wieghed in.
Not that it was necessarily bad for Open Source and Free Software that they did. There's nothing more helpful to your cause for the proponents of what you oppose to act like an ass.
You're modded as funny, but it really isn't GPL compatable and arguably isn't Open Source or Free Software either. It's incompatable with clause 6 of the GPL:
6. Each time you redistribute the Program (or any work based on the Program), the recipient automatically receives a license from the original licensor to copy, distribute or modify the Program subject to these terms and conditions.
You may not impose any further restrictions on the recipients' exercise of the rights granted herein. You are not responsible for enforcing compliance by third parties to this License.
(My emphasis.)
I think the "no whining" thing is dumb comment to put in a license and shows a pointlessly contentious attitude by the Bitkeeper people, something they've shown ever since the question of "Why is a major Free Software project being developed using and depending on unfree software?" issue was raised. They'd have been better off saying "This is our business model. We appreciate it will not be considered appropriate for Free Software, and suggest Free Software developers look elsewhere", and leaving it at that. The criticisms, for the most part, were of Linus and his decision to use Bitkeeper until the Bitkeeper people wieghed in.
Not that it was necessarily bad for Open Source and Free Software that they did. There's nothing more helpful to your cause for the proponents of what you oppose to act like an ass.
The GPL doesn't have to be agreed to in order to obtain content licensed under the GPL. If you don't, you don't have the right to redistribute copies of the software you received. But you still have fair use rights.
The iTMS system, on the other hand, requires you agree to the ToS just to obtain the music, and that ToS does remove certain rights you'd otherwise have.
So, to me, the GPL is a fair legal document to agree to. The iTMS isn't. If the entire world goes GPL or uses the same principles, no harm is done. If the entire world goes iTMS and uses the same principles, the world will be a worse place and people, in the end, will end up with less rights.
I don't think you should be obliged to agree to a contract or license simply to listen to music you choose and have copies yourself. This is not to suggest that (a) I think that if you've agreed then you should violate that contract - at least, not for music or (b) there shouldn't be some framework that requires people who listen to music produced on the assumption that the legal framework of copyright that exists today can ensure the artists can compell listeners to contribute to the costs in some way shouldn't contribute to the costs in some way (geez, is that a fucking awful sentence or what?) I have no time for pirates (first entry, sub-definition 3).
Apple's Fairplay may not be the worst DRM in the world, but it exists, and it does hamper legitimate usage of music. Rather than employ the hacks DVD Jon is proposing in this article though, I'd rather people simply avoid the store.
Google is widely used but it doesn't control any markets that I'm aware of. And that's the rub - monopoly doesn't mean "is only supplier" or "is by far the biggest supplier taking up almost all of the market", it has to do with control over it.
The truth is if I wanted to create a search engine, I can't see anything Google could do that would damage it, beyond normal, healthy, competition (ie being better.)
I'm looking at them again and can't really see how anyone would think they're not highly Luna like. I agree that the lack of colouring on the window control buttons is a vast improvement on Microsoft's original, but it, and the slightly less rounded corners, appear to be the only serious changes. It's a shame the whole blue frames and seriously ugly "shadowed" text was copied wholesale. It's not as if the buttons are the only aspects of Luna that are hideous and unprofessional.
Looks ugly to me, but I must admit I find most modern GUIs butt-ugly. That said, I thought most people agreed about Luna. It's one thing for geeks to immediately retheme an OS the moment they install it, but when every Joe User is doing it too, then there's a problem.
It's just KDE's commitment to being ahead of everyone else. KDE's developers have been running some complex global warming simulators, and it turns out that in 2007, sea levels are going to rise drastically. This, and that giant Earthquake in 2008, will cause Britain to disappear from the map.
You'll notice Florida's been reduced to a small stump as well. Same reason. Well, no earthquake this time, just the rising sea levels.
I am so glad they've copied the Luna theme from Windows XP. This is highly popular amongst Windows XP users, I've never seen anyone turn it off and switch to classic mode yelling "Damn this ugly blue #$%@! What was Microsoft thinking?"
No, not ever. Definitely not. Not geeks and ordinary users alike. Oh no. Never happens. It isn't the case that almost everyone in the office with XP - from the programmers to our product support people, etc - has switched to the classic theme and switches to it the moment XP finishes installing.
Raskin was originally a one button mouse proponent, though he changed his mind after the Mac shipped. Actually, to be absolutely clear, he wasn't a proponent of mice at all, but he was prepared to tolerate the device as long as it was one button. Some of his reasoning can be found in this interview:
As for the one-button mouse, I'd observed at Xerox Parc which had a 3-button mouse, that people were very confused as to its use and when I was designing the software for the Macintosh, in designing the interface, I figured that if there was only one button, there would never be any question on what you have to press the number of ways of using a one-button mouse. I think this was probably a mistake, in fact there is an appendix in my book which discusses why I think this was a mistake and what I think I should have done. One of the reasons I made the mistake is that there is a certain school of industrial design dating back to the Bauhaus which says that designs have to be simple, uncluttered, and clean. In particular, don't put writing on it except for brand names or logos. If we had had a multiple-button mouse with two keys, labeled something like "select" and "activate," it would have been much easier to use, but the idea of putting writing on keys did not occur to anybody, including me. So if I was designing one today, it would have two buttons and they would be labeled.
I'm not sure it's fair to say Jef and Steve had "furious disagreements" as you suggest on this specific issue, at least, not with Jef being in favour of multiple buttons and Steve not. Both Jef and Steve, after the Mac went into production, changed their minds. Steve Jobs went on to found NeXT, producing a computer with a multi-button mouse shipped by default. On retaking-over Apple, he didn't switch them from single buttons yet, but that doesn't mean he wasn't a multiple button supporter - and the fact Mac OS X is so much more usable with multiple buttons makes me think he's been planning this for a while.
It's probably the same reason that everything about applications for generic Unix systems get filed under the "Linux" section. The current list of sections is a little poorly chosen, to put it mildly. There is no IBM or Power section, or anything that really would make a good home for Power related topics.
Re:I think he came off as having OCD
on
Donald Knuth On NPR
·
· Score: 3, Insightful
If he's some schmoe working at a deli, and he's obsessing about the pickles to the point that he's not able to make sandwiches, then he's in the wrong job. There's nothing to defend.
It only becomes an issue if you consider the perfectionism to be a mental illness. Which you do, and I don't. Someone not being the right person for the job is not a mental illness.
Re:Abandoning Email is Stupid
on
Donald Knuth On NPR
·
· Score: 4, Informative
He actually abandoned email in 1990. The complaint was that email is for people who want to get on top of things, and he's the type of person who wants to get to the bottom of things.
In other words, he was getting legitimate email, and it was a distraction for that reason.
I'm pretty sure that if the problem was spam, Knuth is one of the few people who'd actually create a system that can, actually, filter spam and spam only.
Re:I think he came off as having OCD
on
Donald Knuth On NPR
·
· Score: 5, Insightful
The other way of looking at this is that being good at mathematics is a mental illness. I'm having difficulty understanding why you and the grandparent are considering Knuth's perfectionalism a bad thing. Is he anti-social as a result? Does his perfectionism prevent him from leading a safe, furfilling, life?
I see no evidence that it's doing any such thing. He's a brilliant mathematician and computer scientist, and that's all. The world is full of different people. It's also full of arrogant, scared, jerks who do not like differences.
which means that market is regulated in such a way as to ensure that no real competitor to the "national operator" would be allowed to grow.
This is illegal under current EU competition law, so I doubt the situation is exactly as described.
From my point of view heavy regulation is for example the fact that in some cases (e.g. Poland, as far as I know also Czech Rep.) licenses issued (effectively agreements between the state and the operator) included a promise from the state that for a given period of time no new cellular operators would be allowed to enter the market.
Poland and the CR joined in 2004. Unless you're suggesting these laws were passed in the last few months, I don't think they're indicative of the way the EU works. They certainly would violate EU competition law (again!)
Also, from the point of view of marketing departments of a GSM operator playing in such a market its competitive options are very limited. All other (two or three) operators use the same technology, same phones with same capabilities over the same bands with very similar coverage. You can cut prices only a bit, because doing so dramatically is out of question - it would create a price war on which everybody would loose in the end - and the margins are huuuuuge, believe me. So the only way you can try do differentiate from your competitors is by creating various add-ons - hence the premium SMS-es, which serve as micro-payment medium for many services, ringtones and images etc.
There's nothing to stop GSM operators from adding on pretty much any services they want. There's a requirement for interoperability which means the operators must make their networks compatable with GSM, but that's not the same thing as being limited to offering only those services that have an official ITU standard associated with them.
It's also a little ironic. If we're comparing it to the US, what system in common use in the USA actually features more services than GSM? The only one I can think of, off the top of my head, is iDEN, a close relative. The choice for most users in the US is GSM or IS-95 ("CDMA"), the latter of which is provided in an operator hobbled form (with the full blessing of Qualcomm) with restrictions imposed on end users and their ability to choose their own equipment.
Whatever the case, the reason for higher per-minute charges in the EU is the one I gave - the relative unwillingness of most Europeans to commit to a larger monthly spend. It's a real cultural difference, and not one I expect people who haven't lived in both countries to understand. For much the same reason, DSL is considerably cheaper there than it is in America, so it works both ways.
Slashdot Mirror
If the aim is just to get lossless (well, there's no such thing, but lossless in terms of being encoded at a reasonable PCM bitrate) content, and convenience is not an issue (and loading 300Mb at a time, which is the best you'd get for a FLAC-style compressed CD's worth of music), then CDs remain the best way of getting the content. After fiddling with iTMS, downloading a few tunes, finding the DRM is enough of a PITA to make the whole "convenience" selling point bunk - to me - I have to admit that's pretty much the route I ended up taking. Yes, I could have used DVD Jon's hacks, but I tend to go for the "If it's breaking the law (even an insane law), and there's a legal alternative to do the same thing, go for the latter."
(Running "Mozilla/5.0 (Macintosh; U; PPC Mac OS X Mach-O; en-US; rv:1.8a6) Gecko/20041222 Firefox/1.0+" here)
The next point release of Firefox should support it "officially."
OS X manages things like proxies and other network settings as part of the OS, so you can relatively easily (but not seamlessly, alas) switch from Ethernet to WLAN to modem connections just by selecting the configuration you're using from the Apple menu. Firefox doesn't pick up the proxy settings itself if you do this. Camino would.
Now, that said, Camino isn't compatable with Firefox's plug-ins, and I don't know about you but I've found it's gotten hard to browse without Adblock and Flashblock having tried them. (Hacks to add crippled functionality similar, but not really, to the two are usually quoted when I mention this, but the full blown "I want to be able to get a flash animation to start only when I click it" and "I want to be able to easily add wildcards to my ad blocking script" functionality just isn't present with these "answers.")
Sure, you can store extremely heirarchical data in XML. Sure you can store Unicode in XML too. But what if, say, you're storing relatively simple information that doesn't really have either of those issues?
In response to an AC below, I suggested the fstab as an example of a file that really shouldn't be rewritten in XML. The advantages and disadvantages of both approaches would be:
fstab:
Easily parsed by any language, OOP or otherwise. C and shell scripts can easily parse it.
Catting it generally results in less than a screenful of information (any verbosity is usually down to copyright notices), meaning that all the information a sysadmin needs is available at a glance.
Easy to manually scan - finding /mnt/cdrom is a matter of glancing down the first column. All related information to it is thus on the first line. You can answer the question "Is /mnt/cdrom mounted upon boot?" in a second or two.
"fsxml"
Can be parsed by a standards compliant library, but such libraries are not available in a form easier to use than manual scanning of fstab for lower level HLLs like C.
Can be validated automatically, though testing validity is easy for fstab as mount can be run manually and will complain about problems.
Is potentially self documenting, assuming obvious tags are chosen, though comments above column headings in fstab makes this redundant.
The key thing with fstab, of course, is that mount, a simple C program, and some start up scripts, written (of course) in Bourne Shell, need to read it. That's pretty much the clincher - and it's true, actually of a lot of files in /etc. It needs to be simple, and XML isn't just overengineered for the purpose, it's positively unsuitable.
If XML is as great as its more enthusiastic supporters suggest, then Oracle needs to be rewritten so its tables are stored in XML. Its indexes too. ELF needs to be discarded in favour of XML. File systems like ext2fs and ReiserFS need to be replaced with XML too.
Clearly that's over the top, and no supporter, no matter how enthusiastic, really would suggest such a thing (though I wouldn't put it past someone to suggest replacing ELF...), but the point is that it really isn't suitable for every application. More to the point, it's not an optimal solution (like all portable, extendable, generic formats) for the vast majority of applications. It's a useful solution, but it's not even the only portable, extendable, format, and those who propose it knee-jerk for everything are doing a grave disservice to the format itself.
Now cat your XML equivalent.
Did I say it wasn't? I was comparing the two, and pointing out that the simpler design is the more readable.cat fstab, and at a glance you can see everything about /mnt/cdrom. It takes a second to find the line refering to it and understand how /mnt/cdrom is set up.
cat "fsxml", on the other hand, and you'll have to plough through a large amount of what's essentially garbage to find the text you're looking for. Now, before you reply, please, do what I'm suggesting. Take your fstab file and convert it to the most obvious form of XML. It's not difficult. It'll take you five to ten minutes. Once you've done this, we'll be on the same page. You'll see instantly what I'm refering to. You'll see you can no longer easily see, at a glance, the information you're trying to find. You'll no longer be able to fit it on one screen.
The format of fstab is hardly wonderful, but compared to XML it is instantly, easily, readable.
For the specific application of listing available file systems in a Unix environment together with their security and mounting attributes, fstab is a more optimal solution than XML.
Technically it's possible to write an XML parser in regular C. It's hard. It's complicated. It's difficult to do without producing a complex, difficult to maintain, large program.Realistically, you wouldn't do it.
C, not C++. C. You can't "instantiate" instances in C. It has no concept. That's the point. I'm surprised anyone would describe it as "reinventing the wheel". XML came after the CSV or white-space delimited file. XML is infinitely more complex, is objectively harder to read, and objectively more difficult to implement. While off-the-shelf libraries provide a semi-solution to the latter, XML's complexity and required Unicode support means that it's a bugger to use for simpler languages like C.Asserting it isn't, and then quoting what you'd do in C++ and pretending it's the same in C as a justification, ignoring (deliberately?) an objective, relevent, example, and - irony of ironies - complaining that white-space delimited files are "reinventing the wheel" as if to pretend XML isn't, strikes me as not very convincing.
Implement my example. For shits and giggles, feel free to modify the mount command to support it (you'll have to learn C first though) before responding. I think you'll see my point.
XML is a good solution for a small set of particular formats. For everything else it's either overkill or too limited.
XML is being treated right now by a lot of programmers as the be-all and end-all solution to everything. It isn't. It's difficult to parse for otherwise lightwieght programs, it's difficult to read, and it's verbose. If you doubt this, try converting, as an example, fstab to XML. It's deceptively easy and the solution is apparently obvious.
Then, once you've done that, play around with it. Consider the consequences of what you've just done. Can you get the entire file system to appear on part of one screen any more, like you could the old fstab? Can you easy, at a glance, see the settings related to /mnt/cdrom?
Now consider all the programs that use fstab. Do you really want to rewrite "mount" in C++? You're not going to be able to rewrite it in Python, it's too low level. And if you keep it in C, it's going to be a PITA to actually parse that fstab file.
Suddenly this classic example of a "horrible" old-style configuration file doesn't look so bad, does it?
I'm not an XML bigot. I'm a realist. I appreciate it annoys the thought-free fan-boys who jump on a new technology, see some tiny part of it that looks good, and suddenly enthuse about how it can revolutionise everything, but just as OOP is a great way of writing interactive GUI-based apps, but pretty ugly the further you get away from that, XML is a great way of expressing certain types of complex hierarchical textual information, but pretty ugly the further you get from that. If it's simple, you don't want to express it in XML.
XML is cluttered and complicated. It is possible to "cleanly" express some content types better in XML, but for the vast majority of cases, it's overkill and unnecessary added complexity. What's worse, it's not even a real "solution" to complex data - complex data is rarely limited to information best expressed textually.
I make no secret of not liking XML. I'm not saying it's never appropriate, but most of the time, when I see it, you've got to wonder what crack the user was smoking when he or she chose it. It's a modern day programmers fetish that's inappropriate for most of the applications it's used for.
I don't.
Leave /etc as it is. It works. It's done the right job now for coming on 30 years. It's had one major clean up (executables moved to /sbin), and otherwise it's nice. You can grep it. Help is a manpage away. With the exception of a handful of recent apps from programmers who don't "get it" and think XML is better than simplicity, the config files are, by and large, consistant, within a certain mindset.
There's lots of things that could be improved about Unix. Destroying /etc doesn't really improve it.
Greger47 said:
He was telling the AC that the reason he was chosing the GPL over the BSD license was that if he gave it away for free, ie licensed under the BSD license, nobody would "buy" copies from him.You then responded to him with this:
I assumed you'd misread (and still believe you're misreading) and thought Greger47's point was that giving the software under the GPL was "giving it away for free". Otherwise there'd have been no point in making the comment.I think you've misunderstood Greger47's point.
Amiga was a big Free Software platform in its time. Most of the networking code was Free Software, and there was fury in the community when the AmiTCP authors took AmiTCP 4 proprietary.
It's been increasingly difficult to be an Amiga user in the last ten years. The platform has practically zero support. Anyone with choices has moved on. Those who loved the platform moved to things like GNU/Linux and got working on trying to make their new platform have the features they missed from the old one.
It's not altogether surprising that a significant minority amongst those left would be FOSS hostile. If they weren't, they'd probably not still be using the Amiga.
Shame about MorphOS though. It'd be nice to see a Free Software attempt at a next-generation AmigaOS (I don't mean AROS, I mean to AmigaOS what Mac OS X is to classic Mac OS, or Windows NT was to DOS/Windows.)
I think you're still misunderstanding the point. He was saying that giving it under the BSD license would be "giving it away for free". If, as he does, he licenses it under the GPL, he's able to sell to companies that want to make a proprietary version. If he licenses it under the BSD, there's no point, because same companies don't have to pay him to produce proprietary versions, they can just get the BSD licensed version.
You're modded as funny, but it really isn't GPL compatable and arguably isn't Open Source or Free Software either. It's incompatable with clause 6 of the GPL:
(My emphasis.)I think the "no whining" thing is dumb comment to put in a license and shows a pointlessly contentious attitude by the Bitkeeper people, something they've shown ever since the question of "Why is a major Free Software project being developed using and depending on unfree software?" issue was raised. They'd have been better off saying "This is our business model. We appreciate it will not be considered appropriate for Free Software, and suggest Free Software developers look elsewhere", and leaving it at that. The criticisms, for the most part, were of Linus and his decision to use Bitkeeper until the Bitkeeper people wieghed in.
Not that it was necessarily bad for Open Source and Free Software that they did. There's nothing more helpful to your cause for the proponents of what you oppose to act like an ass.
I think the "no whining" thing is dumb comment to put in a license and shows a pointlessly contentious attitude by the Bitkeeper people, something they've shown ever since the question of "Why is a major Free Software project being developed using and depending on unfree software?" issue was raised. They'd have been better off saying "This is our business model. We appreciate it will not be considered appropriate for Free Software, and suggest Free Software developers look elsewhere", and leaving it at that. The criticisms, for the most part, were of Linus and his decision to use Bitkeeper until the Bitkeeper people wieghed in.
Not that it was necessarily bad for Open Source and Free Software that they did. There's nothing more helpful to your cause for the proponents of what you oppose to act like an ass.
The GPL doesn't have to be agreed to in order to obtain content licensed under the GPL. If you don't, you don't have the right to redistribute copies of the software you received. But you still have fair use rights.
The iTMS system, on the other hand, requires you agree to the ToS just to obtain the music, and that ToS does remove certain rights you'd otherwise have.
So, to me, the GPL is a fair legal document to agree to. The iTMS isn't. If the entire world goes GPL or uses the same principles, no harm is done. If the entire world goes iTMS and uses the same principles, the world will be a worse place and people, in the end, will end up with less rights.
I don't think you should be obliged to agree to a contract or license simply to listen to music you choose and have copies yourself. This is not to suggest that (a) I think that if you've agreed then you should violate that contract - at least, not for music or (b) there shouldn't be some framework that requires people who listen to music produced on the assumption that the legal framework of copyright that exists today can ensure the artists can compell listeners to contribute to the costs in some way shouldn't contribute to the costs in some way (geez, is that a fucking awful sentence or what?) I have no time for pirates (first entry, sub-definition 3).
Apple's Fairplay may not be the worst DRM in the world, but it exists, and it does hamper legitimate usage of music. Rather than employ the hacks DVD Jon is proposing in this article though, I'd rather people simply avoid the store.
The truth is if I wanted to create a search engine, I can't see anything Google could do that would damage it, beyond normal, healthy, competition (ie being better.)
Looks ugly to me, but I must admit I find most modern GUIs butt-ugly. That said, I thought most people agreed about Luna. It's one thing for geeks to immediately retheme an OS the moment they install it, but when every Joe User is doing it too, then there's a problem.
You'll notice Florida's been reduced to a small stump as well. Same reason. Well, no earthquake this time, just the rising sea levels.
No, not ever. Definitely not. Not geeks and ordinary users alike. Oh no. Never happens. It isn't the case that almost everyone in the office with XP - from the programmers to our product support people, etc - has switched to the classic theme and switches to it the moment XP finishes installing.
Truly, they've copied a winner!
It's probably the same reason that everything about applications for generic Unix systems get filed under the "Linux" section. The current list of sections is a little poorly chosen, to put it mildly. There is no IBM or Power section, or anything that really would make a good home for Power related topics.
It only becomes an issue if you consider the perfectionism to be a mental illness. Which you do, and I don't. Someone not being the right person for the job is not a mental illness.
In other words, he was getting legitimate email, and it was a distraction for that reason.
I'm pretty sure that if the problem was spam, Knuth is one of the few people who'd actually create a system that can, actually, filter spam and spam only.
I see no evidence that it's doing any such thing. He's a brilliant mathematician and computer scientist, and that's all. The world is full of different people. It's also full of arrogant, scared, jerks who do not like differences.
It's also a little ironic. If we're comparing it to the US, what system in common use in the USA actually features more services than GSM? The only one I can think of, off the top of my head, is iDEN, a close relative. The choice for most users in the US is GSM or IS-95 ("CDMA"), the latter of which is provided in an operator hobbled form (with the full blessing of Qualcomm) with restrictions imposed on end users and their ability to choose their own equipment.
Whatever the case, the reason for higher per-minute charges in the EU is the one I gave - the relative unwillingness of most Europeans to commit to a larger monthly spend. It's a real cultural difference, and not one I expect people who haven't lived in both countries to understand. For much the same reason, DSL is considerably cheaper there than it is in America, so it works both ways.
Honeywell made the computers that featured in Billion Dollar Brain, the last Harry Palmer film. FWIW. I'm sure a few Slashdotters have watched that.