Because there arn't any solid implementations around? Well, maybe there are somewhere, who knows, but nothing that I can just 'apt-get'. Guile tried to provide different language front-ends, but completle failed with that, maybe because Guile itself wasn't much of a good Scheme implementation who knows.
### it's easiest just to use S-expressions.
S-Expressions really have little to no advantage for most problems. Sure its cool that I can write my own language inside Lisp, but truth to be told, this is completly useless for 99% of the cases. Back then when I used Guile a bit more I ended up using macros to fix language defects and missing things (lack of for-loop construct and the like), it was cool that the language provided the power todo so, but if the language wouldn't miss so many day-to-day things, I wouldn't have needed to fix it in the first place. Rubys blocks for example solve basically the same problems which I solved with macros in Scheme or Lisp, sure they are not as powerfull, for most problems they are more then enough and far more elegant to use. S-Expressions are really the biggest problem of Lisp, sure they are also what makes Lisp so powerfull, but being forced to even write simple stuff in S-Expressions really makes the language as a whole rather hard to read and to write.
### The goal itself is flawed.
Don't see much flaw in this, for most part APIs are all the same, no matter what language you use, differences are really small. So having all of CPAN available in Python might be quite a cool thing.
### You mean the choice of editing Perl in Notepad and editing Perl in vi? Where is the "choice" in that?
I can use the languages like I want and am not forced to first learn how to start the VM, how to launch the editor and what else. With Perl, Python, Tcl/Tk and all other popular languages its easy hand over 10 lines of script to demonstrage their strength. Its also easy to add them as scripting languages to a programm. With Smalltalk, Lisp and Co. this is a whole lot more difficult. If there are any 10 line programms that show the true power of the languages, how to embedd them, how to script with them, etc and are only an 'apt-get' away, I would like to know.
### And what "freedom" does Perl give me?
The freedom to get the job done. Smalltalk and Lisp are good for writing Smalltalk and Lisp programms, they havn't been all that good at solving the common mans problems, either due to lack of APIs/libraries, due to being to difficult to install, VM overhead or whatever.
Don't get me wrong, I like Lisp and often whish I had the power of Lisp at my hands when writing code in other less powerfull languages, however that power has such a high price in reduced readabilty that for most tasks I prefer to not use Lisp.
ACK, fully agree with you on that. I really enjoyed the pilot and the beginning of the series, but it quickly turned into some Cylon psycho-game with mixed in religious yadayada talk, just didn't make any sense. And now that they want to go even more into religious stuff in the next seasons I am really not sure if I am even watching it any longer, the last few episodes of the first season already stopped being fun to watch.
The show really started out great and had potential, but as it stands now they will have a really hard time explaining only half the Cylon stuff they did. And well, if they don't end up explaining it, the show really lost it, just doesn't make any fun if the show can turn at completly random point into whatever direction the writer wants, because Cylons can do whatever they want.
And heck, why did the Cylons first blow up the collonies and now end up playing cat & mouse with Galactica instead of simply blowing it up (Baltar after all has access to a nuke).
I kind of doubt it, even after the first season they are already running extremly short on fighter pilots and Vipers (can't remember the exact number, but it was some low two digit one). So unless some other Battlestar pops out of nowwhere to lend them some, they will have huge throuble standing against the Cylons much longer.
They can of course let the rest of the show play inside with internal conflicts and endless talk or have some miraculus event save them all, but if they continue as they did, they will simply run out of useable excuses how they escaped the Cylons yet again.
### Starbuck injures her knee in one episode and spends most of the rest of the season hobbling around in pain.
Which I found kind of annoying. I mean its ok for two or three episodes at most, but after she didn't seem to have much knee problems when outperforming Apollo in the Cylon fighter, I found it extremly weird that the wasn't able to pilot a Viper a whole lot episodes later. Just didn't seem natural.
I for one love that style ever since I have seen it for the first time in Firefly. I can understand if people might find it annoying because its different, but I for one enjoy that the show doesn't look like every other show before and instead tries something different.
Am I the only one who found the ending rather boring and pointless? I mean dropping a nuke into a Cylon ship sounds intersting, but they end up turning it into a 'meet some naked Boomer' kind of thing and well, then Boomer getting back and shooting Adama, well, was a suprise for sure, but still knowing nothing about Boomers background it was just yet another completly random thing. Since Boomer has gone crazy before it also wasn't really much news, more a 'Ok, Boomer crazy again, how interesting... well not'. At leasts the people should now finally have a clue that Boomer is a Cylon, after the first two moments to find it out people prefered to keep their moth shut for no good reason.
I found the religios stuff in BG extremly annoying. Its nothing wrong with the Cylons having some motivation and stuff, but this constant talk about god didn't sound any interesting at all, half the time they could have inserted random yadada and it would have made the same sense. Worst of all the religion isn't only the background motivation, but it seems to have actually impact, ie. there is a god. We won't find out until next season, but Starbuck jumping back to earth to search for some historic error, prophecies turning out true and Boomer getting pragnant just was to much, especially in a show that should focus on the people. I don't mind a little of such yadada, but it seemsed to be the driving factor in the last few episodes, worse then worst techno babble ever on StarTrek.
Sure, being able to wrap everything is nice and cool, however that was not the question. The question was why I have to define everything seperatly and can't use the slot name for everything in the first place, still don't see any use for that.
The throuble with Lisp is that its 'ugly at default', defclass is just one of the many areas where it is simply quite a bit more ugly then necesarry.
No, in your case the man with the book might only now 4 options, but the man with the man page wouldn't even now that the command exist in the first place. Man pages are good to lookup the options when you already know what you want, for anything else they are quite horrible.
(defclass new-class-name (parent-class)
one
two
three))
I really don't understand why you need to have a seperate name for the slot, for the keyword, for the getter and for the setter, while in the end you end up naming all of them the same anyway.
### By comparison with systems like Lisp and Dylan, for example, the Parrot system is still enormously complex, limited, and cumbersome from the programmer's point of view.
Well, Parrot is a virtual machine for interpreted languages, and as such rather language neutral, ie. we most likly see Python, Perl and Ruby programms running ontop of Parrot, good thing IMHO. While Lisp is nice and interesting, its also rather limited in terms of choice, you either have to do it the Lisp-way or don't do it at all and well, most people don't seem to like the Lisp-way all that much. So I don't really see how Lisp is in any way superiour to the Parrot way. And Guiles concept of 'translating all languages to Scheme' hasn't been all that successfull either, I think there is some Basic translator now, but thats it, nobody really seemed to be interested in that concept. If Parrot turns out well it might end up uniting all the scripting languages out there, while from an academic point of view maybe not the most interesting thing ever, practically it might make a lot of peoples lives much easier, something that cannot be said about Lisp and Smalltalk.
### And compared to Smalltalk, Perl/Parrot isn't even in the same league when it comes to programming environments, browsers, and other tools (in fact, very little really is).
I kind of agree on that, but again Perl and friends offer you choice about the environment, which Smalltalk doesn't all that much, which in turn is IMHO one of the major points why Perl is successfull while Smalltalk is not so much. People like freedom, people like doing it there way. Smalltalk and Lisp simply force far to much into doing things their way.
PS: Dylan is just Lisp+CLOS with the () removed, but still hardly readable, since the CLOS syntax for defining classes is really rather unreadable and all but elegant.
So what are the lessions to be learned from languages written in the past?
- API/Libraries are important, more important than the language itself, no matter how good your language is, if you don't have a bunch of libaries ready to use the common man will solve his problems faster and better in another language. (Perl/CPAN)
- good syntax is important, do/end are no fun, {}'s are easier to read to the common man (C)
- interoperability with other languages is important (C-libraries exported to scripting languages)
At least for me that seems to be the points that make a language successfull, while not necesarrily beatifull. Most of the powerfull, but mostly failed languages, of the past (Smalltalk, Lisp) seem to either ignore most or all of these points, worse they come with their own VM, their own development environment and such, so unless you do it their way you are mostly (hard to write or ship a few ten-line long script, hard/impossible into a native-binary, etc.).
Can't say for the original LISP, but todays Lisps (Common Lisp, Scheme, etc.) are all but 'pure functional' languages. Sure they support functional programming style too, but its neither the only or even the most common style, Common Lisp doesn't even gurantee tail-recursion if I remember correctly:
(defvar a 2) (setq a (+ a 5)) (princ a) => 7
If you add CLOS or Goops into the mix there there is even less functional programming. There is of course still a lot of functional programming in Lisp, due to it being based on lists, but an average Lisp programm doesn't look all that much more functional than a C++ with STL, Python or Ruby programm, especially when compared to Haskell and the like.
I have seen it through and I have to say that it gets worse and worse with every episode. The pilot/mini-series started of really well, but the series itself quickly turned into some pseudo religious mambo-jumbo non-sense mixed with some rather weird technical issues, especially in the end (Starbuck beating Apollo in a Cylon fighter which she is flying for the first time ever, start to search for fuel only short before its completly out, human-looking Cylons being near impossible to detect, Dr. Baltar's fantasies and weird behaviour, yet nobody seems to care, Cylons constantly talking about god and fate, no humans in the cities left, Boomer being cylon, Boomer being pregnant, etc.).
The series has its good moments and the beginning was quite impressive, but it has way to much Cylon mindgames going on, with little or no explanation what the Cylons are all about in the first place. I mean first they nearly wipe out human race, then they have multiple Cylon infiltrators on Galactica, yet instead of blowing away the rest of man-kind they start playing weird games with them instead, just doesn't make all that much sense.
Yet NeXTSTEP, written in Objective-C, outperforms Gnome, written in C, quite well and is also more flexible. In theory Objective-C should be quite a bit slower due to the late binding nature and such, but seems like its actually better when done right. Its really more about the algorithm then about low-level coding vs high-level one.
PS: Wasn't there something that KDE applications loading extremly slow was in most part caused by the dynamic loader not handling C++ symbol tables very well?
### Applications back then fly today, but seem like a small insect when it comes to functionality.
Well, a 10 year old NeXTSTEP computer can do a lot of stuff that I still can't do with Gnome and KDE, and yet it is still faster doing some things than todays computer. Same with Inkscape, I tried to play around with some of the stuff I did on a P90 with 24mb RAM in CorelDraw years ago, Inkscape turned out to have huge problems rendering the stuff on a 1Ghz Athlon with 768mb RAM, was almost unusable. Its true that there are some things that I can do today that would have been impossible some years ago (fullscreen video, todays games, etc.), yet there are many many things that are basically exactly the same as years ago, only that they havn't speeded up at all or even got slower, even if the CPU speed itself has massivly increased.
And yet a 10 year old NeXT computer provides a user experience at least as smooth, if not smoother then todays computers.
A bit CPU and RAM certainly helps in smoothig the user experience (ie. having a dozens webpages open at once is an improvement). However there is little excuse for having some trivial app like GEdit take five seconds to load when most of Gnome is already loaded into memory, something just went quite wrong in the evolution of software. People tend to add more and more feature, rewrite pieces, redesign GUIs, etc. yet seldomly they take a step back to look at the broader picture on what they are doing and how exactly it benefits the user or how it duplicates work already done by other parts of the OS. There also seems to be little interest in optimizing the user experience itself, old programms sometimes loaded up a screenshot of the programm and displayed that to the user before the real application was started, since that screenshot was a lot faster to load than the real application, the result is that the user can already adopt to the interface, while the computer is actually loading the real application in the background. If done right the moment the user has adopted to the interface, which normally takes a few seconds, the real programm will be ready to work and seamlessly replace the screenshot. Today however such stuff is not done, so the user has first to wait till the app is loaded and only then can start 'adopting' to the interface, thus wasting time.
### But having the source does make it possible for you to modify the software to do these things, if that is something important enough to your business.
Sure it doesn't hurt having access to the source, but what you need is a specified interface to let programms interact with each other and some standard way of interchanging data. Source code alone doesn't provide that.
If I want to have a drag&drop mechanism that works across application boundaries I need a way to implement it. But not just a way that works for two or three applications, but for basically everything, so that not only I can start implementing it, but everybody else too. A good mechanism for interoperability doesn't need access to the source code to work, it works out-of-the-box without both application developers ever having worked together, it doesn't even require that the other party know that they exist.
The throuble is that interoperability isn't something you can quickly patch into a single programm, its something that almost all programms have to agree on how to do it and source code doesn't help with that.
### these formats are not really intended to be opened by other programs!
We are in the OSS world and there really shouldn't be a need for something secret obscure format or even if there is something it shouldn't be to much additional work to provide a libgimp or the like that plugs into Imagemagick or whatever to offer all image editing programms the possibility to interact with.xcf files. People just havn't paid all that much attention to things like this in the past. PS: Gimp people want to change there format to something like.zip of png files + xml, so it might be much easier to extract data from them in the future.
### but why would i want to paste the color, font and background image together with the text??
To preserve table-layout, images and links, extremly usefull if one wants to print a webpage and cut out the advertisment banners before doing so. As I just found out OpenOffice actually supports that to some degree.
That man is right or at least not totally wrong. Just because you have the source it doesn't automatically make your software work together. Simple examples:
- open a OpenOffice document in AbiWord - copy&paste between different applications - embbed an Gnumeric chart into some OpenOffice document - try to edit a LaTeX document with Abiword or OpenOffice - try to open a Gimp xcf in anything beside Gimp - try to copy&paste some webpage in a Office application and get something more then plain-text -...
None of this works or only in a much less smooth way then it does under Windows or MacOSX with similar software. Free Software has improved a lot in these regions in the last years, but there is still lots and lots of software floating around that doesn't operate much with other software at all. Sure, you can always export to.png or plain-text and somehow get the job done, but smooth interoperability is something else.
The reason to go for vector graphics is not to get more/less speed or disc usage for graphics, but graphics that are independend of the displays resolution. Today if you switch from a 17" LCD with 1024x768 to a 17" LCD with 1600x1200 you will get tiny fonts, tiny GUI elements and a whole lot of nasty side-effects. Todays GUIs provide a bit of help to get the fontsize up again, but thats mostly it, icons often stay tiny, stuff like some themed music players often even get unusable on to large resolutions.
With a fully vectorized GUI if you make the same switch you would have exactly no sideeffects from a higher resolution, it simply would look sharper, no disortion, shrinking, etc. of GUI elements or fonts. The OS would take the DPI of the monitor and just adopt the rendering to it.
The issue gets even more dramatic if you have a 1600:1200 display with a ration of 16:9 or the like (aka non-quadratic pixels). With todays pixel based interfaces it simply would look awfully streched, with vector based GUI on the other side it would simply work, no streching since the GUI can adopt automatically to whatever you throw at it.
### If the Linux community are so short on games, why don't they do their usual collaborative thing and make game engines.
They are doing it, we have Stratagus, Flightgear, Cube, CrystalSpace, Ogre and a whole bunch of other games or engines that are more or less ready to use. However none of them really is up to commercial standards and most of them seriously lack content creation tools. Having a good engine is only half the work, you still need to have a good level editor.
The free software world simply lacks the man-power to do all that and do it right. We don't see high quality free games or game engines for the same reason we don't see drop-in Photoshop replacements or Microsoft Office killers. Sure we have Gimp and OpenOffice, but beside being free they don't have all that much to offer. Beside from that in the games world stuff gets quickly out of date, an Office can be maintained for quite some years without huge rewrites, game engines on the other side need to be mostly rewritten from scratch every now an then cause the underlying hardware changes quite frequently.
### I don't think the resolution would be even tolerable if you take an image designed for a 6 inch screen and threw it on something 5 times or more larger than that.
The resolution isn't much of a problem. Sure you might never get full Playstation2 quality out of that thing, but you don't have to. Sometimes its more then enough to simply have a bigger screen instead of being forced to look on the small one, a few pixels don't matter much. I for one have enjoyed playing Gameboy and GBA games on TV via the SuperGameboy or now the GameboyPlayer.
The LGPL isn't a problem since it is compatible with the GPL. The throuble comes with incompatible licenses and a GPLv2 without the 'any later version' would be incompatible with a GPLv3, ie. neither the GPLv3 programm could add GPLv2 code nor could the GPLv2 programm add GPLv3 code.
So far there are quite a few programms around who have removed the 'any later version' clause and thus might get quite some throuble on an upgrade (Linux being probally the most famous example).
While Debian provides a lot, it by far does not provide everything, especially the stable tree is completly outdated. Unstable on the other side provides much more, but still far from everything and isn't really suited for newbies (well, stable isn't either for numerous reasons). Debian is a good starting point, but nothing more.
Slashdot Mirror
### Why aren't people using those?
Because there arn't any solid implementations around? Well, maybe there are somewhere, who knows, but nothing that I can just 'apt-get'. Guile tried to provide different language front-ends, but completle failed with that, maybe because Guile itself wasn't much of a good Scheme implementation who knows.
### it's easiest just to use S-expressions.
S-Expressions really have little to no advantage for most problems. Sure its cool that I can write my own language inside Lisp, but truth to be told, this is completly useless for 99% of the cases. Back then when I used Guile a bit more I ended up using macros to fix language defects and missing things (lack of for-loop construct and the like), it was cool that the language provided the power todo so, but if the language wouldn't miss so many day-to-day things, I wouldn't have needed to fix it in the first place. Rubys blocks for example solve basically the same problems which I solved with macros in Scheme or Lisp, sure they are not as powerfull, for most problems they are more then enough and far more elegant to use. S-Expressions are really the biggest problem of Lisp, sure they are also what makes Lisp so powerfull, but being forced to even write simple stuff in S-Expressions really makes the language as a whole rather hard to read and to write.
### The goal itself is flawed.
Don't see much flaw in this, for most part APIs are all the same, no matter what language you use, differences are really small. So having all of CPAN available in Python might be quite a cool thing.
### You mean the choice of editing Perl in Notepad and editing Perl in vi? Where is the "choice" in that?
I can use the languages like I want and am not forced to first learn how to start the VM, how to launch the editor and what else. With Perl, Python, Tcl/Tk and all other popular languages its easy hand over 10 lines of script to demonstrage their strength. Its also easy to add them as scripting languages to a programm. With Smalltalk, Lisp and Co. this is a whole lot more difficult. If there are any 10 line programms that show the true power of the languages, how to embedd them, how to script with them, etc and are only an 'apt-get' away, I would like to know.
### And what "freedom" does Perl give me?
The freedom to get the job done. Smalltalk and Lisp are good for writing Smalltalk and Lisp programms, they havn't been all that good at solving the common mans problems, either due to lack of APIs/libraries, due to being to difficult to install, VM overhead or whatever.
Don't get me wrong, I like Lisp and often whish I had the power of Lisp at my hands when writing code in other less powerfull languages, however that power has such a high price in reduced readabilty that for most tasks I prefer to not use Lisp.
ACK, fully agree with you on that. I really enjoyed the pilot and the beginning of the series, but it quickly turned into some Cylon psycho-game with mixed in religious yadayada talk, just didn't make any sense. And now that they want to go even more into religious stuff in the next seasons I am really not sure if I am even watching it any longer, the last few episodes of the first season already stopped being fun to watch.
The show really started out great and had potential, but as it stands now they will have a really hard time explaining only half the Cylon stuff they did. And well, if they don't end up explaining it, the show really lost it, just doesn't make any fun if the show can turn at completly random point into whatever direction the writer wants, because Cylons can do whatever they want.
And heck, why did the Cylons first blow up the collonies and now end up playing cat & mouse with Galactica instead of simply blowing it up (Baltar after all has access to a nuke).
I kind of doubt it, even after the first season they are already running extremly short on fighter pilots and Vipers (can't remember the exact number, but it was some low two digit one). So unless some other Battlestar pops out of nowwhere to lend them some, they will have huge throuble standing against the Cylons much longer.
They can of course let the rest of the show play inside with internal conflicts and endless talk or have some miraculus event save them all, but if they continue as they did, they will simply run out of useable excuses how they escaped the Cylons yet again.
### Starbuck injures her knee in one episode and spends most of the rest of the season hobbling around in pain.
Which I found kind of annoying. I mean its ok for two or three episodes at most, but after she didn't seem to have much knee problems when outperforming Apollo in the Cylon fighter, I found it extremly weird that the wasn't able to pilot a Viper a whole lot episodes later. Just didn't seem natural.
I for one love that style ever since I have seen it for the first time in Firefly. I can understand if people might find it annoying because its different, but I for one enjoy that the show doesn't look like every other show before and instead tries something different.
### SPOILER AHEAD ###
Am I the only one who found the ending rather boring and pointless? I mean dropping a nuke into a Cylon ship sounds intersting, but they end up turning it into a 'meet some naked Boomer' kind of thing and well, then Boomer getting back and shooting Adama, well, was a suprise for sure, but still knowing nothing about Boomers background it was just yet another completly random thing. Since Boomer has gone crazy before it also wasn't really much news, more a 'Ok, Boomer crazy again, how interesting... well not'. At leasts the people should now finally have a clue that Boomer is a Cylon, after the first two moments to find it out people prefered to keep their moth shut for no good reason.
### SPOILER AHEAD... ###
I found the religios stuff in BG extremly annoying. Its nothing wrong with the Cylons having some motivation and stuff, but this constant talk about god didn't sound any interesting at all, half the time they could have inserted random yadada and it would have made the same sense. Worst of all the religion isn't only the background motivation, but it seems to have actually impact, ie. there is a god. We won't find out until next season, but Starbuck jumping back to earth to search for some historic error, prophecies turning out true and Boomer getting pragnant just was to much, especially in a show that should focus on the people. I don't mind a little of such yadada, but it seemsed to be the driving factor in the last few episodes, worse then worst techno babble ever on StarTrek.
### then you could trivially wrap defclass
Sure, being able to wrap everything is nice and cool, however that was not the question. The question was why I have to define everything seperatly and can't use the slot name for everything in the first place, still don't see any use for that.
The throuble with Lisp is that its 'ugly at default', defclass is just one of the many areas where it is simply quite a bit more ugly then necesarry.
### actually, man pages are great for learning.
No, in your case the man with the book might only now 4 options, but the man with the man page wouldn't even now that the command exist in the first place. Man pages are good to lookup the options when you already know what you want, for anything else they are quite horrible.
### Unreadable? How so?
Far to much redundance in there, why not just:
(defclass new-class-name (parent-class)
one
two
three))
I really don't understand why you need to have a seperate name for the slot, for the keyword, for the getter and for the setter, while in the end you end up naming all of them the same anyway.
### By comparison with systems like Lisp and Dylan, for example, the Parrot system is still enormously complex, limited, and cumbersome from the programmer's point of view.
Well, Parrot is a virtual machine for interpreted languages, and as such rather language neutral, ie. we most likly see Python, Perl and Ruby programms running ontop of Parrot, good thing IMHO. While Lisp is nice and interesting, its also rather limited in terms of choice, you either have to do it the Lisp-way or don't do it at all and well, most people don't seem to like the Lisp-way all that much. So I don't really see how Lisp is in any way superiour to the Parrot way. And Guiles concept of 'translating all languages to Scheme' hasn't been all that successfull either, I think there is some Basic translator now, but thats it, nobody really seemed to be interested in that concept. If Parrot turns out well it might end up uniting all the scripting languages out there, while from an academic point of view maybe not the most interesting thing ever, practically it might make a lot of peoples lives much easier, something that cannot be said about Lisp and Smalltalk.
### And compared to Smalltalk, Perl/Parrot isn't even in the same league when it comes to programming environments, browsers, and other tools (in fact, very little really is).
I kind of agree on that, but again Perl and friends offer you choice about the environment, which Smalltalk doesn't all that much, which in turn is IMHO one of the major points why Perl is successfull while Smalltalk is not so much. People like freedom, people like doing it there way. Smalltalk and Lisp simply force far to much into doing things their way.
PS: Dylan is just Lisp+CLOS with the () removed, but still hardly readable, since the CLOS syntax for defining classes is really rather unreadable and all but elegant.
So what are the lessions to be learned from languages written in the past?
- API/Libraries are important, more important than the language itself, no matter how good your language is, if you don't have a bunch of libaries ready to use the common man will solve his problems faster and better in another language. (Perl/CPAN)
- good syntax is important, do/end are no fun, {}'s are easier to read to the common man (C)
- interoperability with other languages is important (C-libraries exported to scripting languages)
At least for me that seems to be the points that make a language successfull, while not necesarrily beatifull. Most of the powerfull, but mostly failed languages, of the past (Smalltalk, Lisp) seem to either ignore most or all of these points, worse they come with their own VM, their own development environment and such, so unless you do it their way you are mostly (hard to write or ship a few ten-line long script, hard/impossible into a native-binary, etc.).
Can't say for the original LISP, but todays Lisps (Common Lisp, Scheme, etc.) are all but 'pure functional' languages. Sure they support functional programming style too, but its neither the only or even the most common style, Common Lisp doesn't even gurantee tail-recursion if I remember correctly:
(defvar a 2)
(setq a (+ a 5))
(princ a)
=> 7
If you add CLOS or Goops into the mix there there is even less functional programming. There is of course still a lot of functional programming in Lisp, due to it being based on lists, but an average Lisp programm doesn't look all that much more functional than a C++ with STL, Python or Ruby programm, especially when compared to Haskell and the like.
### WARNING: POSSIBLE SPOILERS AHEAD ###
I have seen it through and I have to say that it gets worse and worse with every episode. The pilot/mini-series started of really well, but the series itself quickly turned into some pseudo religious mambo-jumbo non-sense mixed with some rather weird technical issues, especially in the end (Starbuck beating Apollo in a Cylon fighter which she is flying for the first time ever, start to search for fuel only short before its completly out, human-looking Cylons being near impossible to detect, Dr. Baltar's fantasies and weird behaviour, yet nobody seems to care, Cylons constantly talking about god and fate, no humans in the cities left, Boomer being cylon, Boomer being pregnant, etc.).
The series has its good moments and the beginning was quite impressive, but it has way to much Cylon mindgames going on, with little or no explanation what the Cylons are all about in the first place. I mean first they nearly wipe out human race, then they have multiple Cylon infiltrators on Galactica, yet instead of blowing away the rest of man-kind they start playing weird games with them instead, just doesn't make all that much sense.
Yet NeXTSTEP, written in Objective-C, outperforms Gnome, written in C, quite well and is also more flexible. In theory Objective-C should be quite a bit slower due to the late binding nature and such, but seems like its actually better when done right. Its really more about the algorithm then about low-level coding vs high-level one.
PS: Wasn't there something that KDE applications loading extremly slow was in most part caused by the dynamic loader not handling C++ symbol tables very well?
### Applications back then fly today, but seem like a small insect when it comes to functionality.
Well, a 10 year old NeXTSTEP computer can do a lot of stuff that I still can't do with Gnome and KDE, and yet it is still faster doing some things than todays computer. Same with Inkscape, I tried to play around with some of the stuff I did on a P90 with 24mb RAM in CorelDraw years ago, Inkscape turned out to have huge problems rendering the stuff on a 1Ghz Athlon with 768mb RAM, was almost unusable. Its true that there are some things that I can do today that would have been impossible some years ago (fullscreen video, todays games, etc.), yet there are many many things that are basically exactly the same as years ago, only that they havn't speeded up at all or even got slower, even if the CPU speed itself has massivly increased.
And yet a 10 year old NeXT computer provides a user experience at least as smooth, if not smoother then todays computers.
A bit CPU and RAM certainly helps in smoothig the user experience (ie. having a dozens webpages open at once is an improvement). However there is little excuse for having some trivial app like GEdit take five seconds to load when most of Gnome is already loaded into memory, something just went quite wrong in the evolution of software. People tend to add more and more feature, rewrite pieces, redesign GUIs, etc. yet seldomly they take a step back to look at the broader picture on what they are doing and how exactly it benefits the user or how it duplicates work already done by other parts of the OS. There also seems to be little interest in optimizing the user experience itself, old programms sometimes loaded up a screenshot of the programm and displayed that to the user before the real application was started, since that screenshot was a lot faster to load than the real application, the result is that the user can already adopt to the interface, while the computer is actually loading the real application in the background. If done right the moment the user has adopted to the interface, which normally takes a few seconds, the real programm will be ready to work and seamlessly replace the screenshot. Today however such stuff is not done, so the user has first to wait till the app is loaded and only then can start 'adopting' to the interface, thus wasting time.
### But having the source does make it possible for you to modify the software to do these things, if that is something important enough to your business.
Sure it doesn't hurt having access to the source, but what you need is a specified interface to let programms interact with each other and some standard way of interchanging data. Source code alone doesn't provide that.
If I want to have a drag&drop mechanism that works across application boundaries I need a way to implement it. But not just a way that works for two or three applications, but for basically everything, so that not only I can start implementing it, but everybody else too. A good mechanism for interoperability doesn't need access to the source code to work, it works out-of-the-box without both application developers ever having worked together, it doesn't even require that the other party know that they exist.
The throuble is that interoperability isn't something you can quickly patch into a single programm, its something that almost all programms have to agree on how to do it and source code doesn't help with that.
### these formats are not really intended to be opened by other programs!
.xcf files. People just havn't paid all that much attention to things like this in the past. PS: Gimp people want to change there format to something like .zip of png files + xml, so it might be much easier to extract data from them in the future.
We are in the OSS world and there really shouldn't be a need for something secret obscure format or even if there is something it shouldn't be to much additional work to provide a libgimp or the like that plugs into Imagemagick or whatever to offer all image editing programms the possibility to interact with
### but why would i want to paste the color, font and background image together with the text??
To preserve table-layout, images and links, extremly usefull if one wants to print a webpage and cut out the advertisment banners before doing so. As I just found out OpenOffice actually supports that to some degree.
That man is right or at least not totally wrong. Just because you have the source it doesn't automatically make your software work together. Simple examples:
...
.png or plain-text and somehow get the job done, but smooth interoperability is something else.
- open a OpenOffice document in AbiWord
- copy&paste between different applications
- embbed an Gnumeric chart into some OpenOffice document
- try to edit a LaTeX document with Abiword or OpenOffice
- try to open a Gimp xcf in anything beside Gimp
- try to copy&paste some webpage in a Office application and get something more then plain-text
-
None of this works or only in a much less smooth way then it does under Windows or MacOSX with similar software. Free Software has improved a lot in these regions in the last years, but there is still lots and lots of software floating around that doesn't operate much with other software at all. Sure, you can always export to
The reason to go for vector graphics is not to get more/less speed or disc usage for graphics, but graphics that are independend of the displays resolution. Today if you switch from a 17" LCD with 1024x768 to a 17" LCD with 1600x1200 you will get tiny fonts, tiny GUI elements and a whole lot of nasty side-effects. Todays GUIs provide a bit of help to get the fontsize up again, but thats mostly it, icons often stay tiny, stuff like some themed music players often even get unusable on to large resolutions.
With a fully vectorized GUI if you make the same switch you would have exactly no sideeffects from a higher resolution, it simply would look sharper, no disortion, shrinking, etc. of GUI elements or fonts. The OS would take the DPI of the monitor and just adopt the rendering to it.
The issue gets even more dramatic if you have a 1600:1200 display with a ration of 16:9 or the like (aka non-quadratic pixels). With todays pixel based interfaces it simply would look awfully streched, with vector based GUI on the other side it would simply work, no streching since the GUI can adopt automatically to whatever you throw at it.
### If the Linux community are so short on games, why don't they do their usual collaborative thing and make game engines.
They are doing it, we have Stratagus, Flightgear, Cube, CrystalSpace, Ogre and a whole bunch of other games or engines that are more or less ready to use. However none of them really is up to commercial standards and most of them seriously lack content creation tools. Having a good engine is only half the work, you still need to have a good level editor.
The free software world simply lacks the man-power to do all that and do it right. We don't see high quality free games or game engines for the same reason we don't see drop-in Photoshop replacements or Microsoft Office killers. Sure we have Gimp and OpenOffice, but beside being free they don't have all that much to offer. Beside from that in the games world stuff gets quickly out of date, an Office can be maintained for quite some years without huge rewrites, game engines on the other side need to be mostly rewritten from scratch every now an then cause the underlying hardware changes quite frequently.
### I don't think the resolution would be even tolerable if you take an image designed for a 6 inch screen and threw it on something 5 times or more larger than that.
The resolution isn't much of a problem. Sure you might never get full Playstation2 quality out of that thing, but you don't have to. Sometimes its more then enough to simply have a bigger screen instead of being forced to look on the small one, a few pixels don't matter much. I for one have enjoyed playing Gameboy and GBA games on TV via the SuperGameboy or now the GameboyPlayer.
The LGPL isn't a problem since it is compatible with the GPL. The throuble comes with incompatible licenses and a GPLv2 without the 'any later version' would be incompatible with a GPLv3, ie. neither the GPLv3 programm could add GPLv2 code nor could the GPLv2 programm add GPLv3 code.
So far there are quite a few programms around who have removed the 'any later version' clause and thus might get quite some throuble on an upgrade (Linux being probally the most famous example).
### EVERYTHING is included.
While Debian provides a lot, it by far does not provide everything, especially the stable tree is completly outdated. Unstable on the other side provides much more, but still far from everything and isn't really suited for newbies (well, stable isn't either for numerous reasons). Debian is a good starting point, but nothing more.