Mozilla is the better browser?! Ugh. I tried to use Mozilla the other day, and I was stunned at how hard it was to use. I got away from Mozilla to Galeon long ago because it was so slow and klunky, and only switched back to Firefox because it absorbed a lot of the developer mindset of Galeon (separating searching from URLs; allowing for quick search selection between dictionary, shopping, Web, etc.; vastly simplified, and yet still powerful bookmark management (though I want Galeon's make bookmark here without having to install an extension); and most of all, it's FAST and relatively small).
If you LIKE Mozilla better, that's fine... preference is good, but calling it the "better browser is, I think, a bit silly".
Right, and in the 90s that 5% of people motivated by the sheer joy of what they do were, largely being swept up into the growth of the Internet. It was a big shiny thing. Now, it turns out that in technical fields, most of those people are men. In liberal arts fields, it's more of an even mix and in health-care it's more weighted toward women.
If, let's just say for example's sake, 20% of men in Internet-related businesses were in it for non-monetatry reasons and 2% of women were. If half of the money-motivated people left, that would mean a 48% exodus of women while only 40% of men would leave.
I happen to think that the numbers are EVEN MORE skewed than that, and I think that there are also shades of gray. There are many men who are in it for the money, but also have some "this is cool work" motivation, keeping them more tightly tied to the industry.
Sadly BitTorrent is only half of what I look for in a P2P client. To get the full picture, I use gtk-gnutella to talk to the Gnutella network. Easy searching, good bandwith-distributed downloads, excellent resiliancy and just all around nice. The last time I fired up BT was for a FC1 download, but these days (thanks to people like me) even full OS distributions show up within minutes on Gnutella.
What BitTorrent DOES give you is a single point of control. This can be useful, and is why I keep a BT client installed too.
"and not having any physical media on which to store our data sounds good, right up until the first datacenter fire that loses me last week's data storage."
Not if its RAID-ed across the 'net.
Oh good, now we don't even need a datacenter fire, since you've introduced a single point of failure!
We at TrustUsWithYourMedia would like to appologize for the interruption of service. Our technitions tell us that, inside our new keen software GlobalRAID++, the parity calculations between highly redundant datacenters lost track of their table of contents or something like that... we don't understand what they said, but the net result is that we've lost 3 days of data, and nothing will be available for 3 days while we recover. However, we do want to point out that the corrupt data was written to both datacenters, and thus replicated perfectly!
We understand that you had a lot of options, and we're very happy you chose our servi... uh... hello, is anyone out there? Hello...?
"While he may make optomistic comments about the lawsuit filed by SCO, from speaking with hundreds of technical decision makers, including CIO's, the lawsuits have actually been a stumbling block in using a fullblown linux back end for alot of companies."
I think you're looking at this upside down. How many of those people who are using this as an excuse didn't need this event to make an execuse? Would the fact that Red Hat is a new company or IBM might go back to the Windows camp or something else have been the issue otherwise? I suspect so.
There are many people I deal with who look at the giant farms of Linux servers that I interact with and say, "hey, that's scary stuff!" Invariably, they are the dinosaurs who are busy being rendered obsolete. IBM mainframes were scary stuff at one point too, and no one could understand why you would want to stake your business on a MACHINE that could make MISTAKES... until a few dozen companies made it clear that NOT going that way was a ticket to extinction.
Windows desktops were the same way.
Unix servers: same story.
Web-based business transactions: same thing.
Every new technology requires a period of early-adoptors, and we're exiting (or just starting to exit) that phase with Linux. That's a scary time. These guys see the writing on the wall, and they're trying to make any rationalization they can to avoid the descision that they know they have to make. Tough nuts, industry doesn't care about their rationalization, only results.
First, the idea that we will throw away out current media has been floated since the days of the floppy. It's always a correct prediction, but only because a better physical medium comes along.
Second, the idea that we're going to be OK with just using storage on the Net and not having any physical media on which to store our data sounds good, right up until the first datacenter fire that loses me last week's data storage. It's also a terrible idea to keep your wares and copyrighted porn on someone else's servers;-) and that bring us to:
Third, PRIVACY. There's no single reason why networked media will never win over good-old local storage that beats the desire for privacy.
"A complex list comprehension can lead to some powerful code that does an amazing amount of stuff in a neat little package. To me this is different than the arbitrary line-noise you get with Perl."
There you just lost me. The term "line noise" implies a low signal-to-noise ratio, when in fact Perl presents exactly the opposite. The SIGNAL is in fact, so high that many programmers find it difficult to cope with. That's fine, but let's not confuse that with actuall NOISE.
"So Perl 6 lets you define your own syntax so that someone reading your code neads to figure out what your ideas of the Right Language is?"
No. You wholy misunderstood the concept.
In Perl 6, you will have full access to the grammar, so you could enforce your local stylistic conventions. You would obviously not want to make INCOMPATIBLE changes so that your code is still valid Perl, but you could write your own "strict".
Think of it this way. Imagine a C++ header that caused all uses of operator overloading outside of a limited few "neccessary" to be illegal, or that issued a compiler warning on every use of an iterator initialization outside of a for loop. These are just simple (and not very useful) examples, but they serve to illustrate the point: you can instrument the compiler just as fully as you can instrument your code. Don't like the type checking in Perl 6? Make it stricter.
I'm sure that there will be someone who will publish the "python-like bondage" module 15 minutes after Perl 6 is released. If you're into that sort of thing, then your company can take full advantage of it, while still getting all the value of Perl 6 like LISP-style currying and macros, Ruby-style mixins, cross-langauge bindings through Parrot, boxed and unboxed type constraints on standard Perl scalars, full multi-method dispatch, etc., etc.
"Common LISP and Scheme's macro facilities can be used to define your own language constructs"
Yes, but we're not talking about defining language constructs. We're talking about changing the behavior of the compiler in structured, standardized ways that aren't just implementation hacks. Don't get me wrong. Common LISP is on my list of cool languages to learn more about right below Python and Ruby. I'm just saying that these particular Perl 6 features bear a bit more looking at.
"Or hell, TeX lets you redefine the world if you are so twisted, though to me TeX is more unreadable than Perl."
TeX is actually a good example of what I'm talking about. TeX is very readable for a full typesetting system, but most of us could not care less about typesetting. When you need to do specific tasks that INVOLVE typesetting, but you don't really need all of that power and flexibility, you step up a layer of abstraction and turn TeX into LaTeX. LaTeX is valid TeX, so it's not quite the same, but the idea of limiting a powerful system in order to step back a level of abstraction holds.
Perl 6 will provide all of the power that you need from a modern high-level programming langauge, but let you manage that complexity. You might decide, for example, to restrict Perl 6 in your programs to just the facilities that make sense for scientific calculation. You might even introduce a special syntax/grammar for putting differential equations directly into your program without having to quote around them and hand them off to a seperate processing tool (object, module, what-have-you).
None of this is useful for your average 1000-line CGI program, but for the company that produces tens to millions of lines of structured libraries upon which new software is built and re-factored over time, this will all be a godsend.
Much of what Perl 6 brings to the table, Common LISP has done for years, but some of it is either gathered from other, more recent langauges (e.g. Python, Ruby, Scheme, Java, etc.) or is, as far as I can tell, unique. I hope you give it a try and throw away your naive ideas of "line noise" in favor of considering the value to your productivity and the maintainability of your code base.
LISP has always had the same exact problem. In fact, the creator of Perl, Larry Wall, once said that LISP had all of the visual appeal of oatmeal with fingernail clippings mixed in (an obvious jab at the use of parens which is no more valid that the jabs against Perl for the use of $). And now, Perl 6 is borrowing HEAPING amounts of old wisdom from Common LISP in order to improve the power and flexibility of the langauge (from the way it does macros to currying to the way blocks are built as functional units instead of as C-style lexical scopes).
I'm not putting Larry down, here, just pointing out that languages which make it easy to do a large job in fairly little real eastate are often beset with this kind of short-sighted thinking from even otherwise level-headed folk. It's a good thing that we grow up and get past it.
"Of course you can write some incredibly terse code in Python too, especially with recent language additions."
Yep. EVERY language can be bent to horrid non-readability.
If you want a language that enforces readability, I would suggest using Perl 6 (the first MAJOR step in the actual implementation of which was just made with the creation of a Haskell-hosted compiler called pugs) and writing a module which enforces your idea of readability. Perl 6 will be the first language to give you enough control that you can write this kind of requirement in the language itself. Common LISP comes in a close second, so use that in the mean-time if you want such levels of dictatorial control. Of course, you could write a module that allowed the rest of your program to be written in CL in P6 quite easily;-)
PS: y=x+++1 Guess what language that is. Answer: C, C++, Python and Java. Perl too if you just add $ before y and x. Programming languages are ugly.
As they do with C, C++, Java, Snobol, Forth, APL, C#, etc, etc.
Half of the people programming are below-average programmers. Bad programmers can make life HELL in C, and by the same token good programmers can make life quite easy in Perl.
That said, Perl gets a bad rep, not because good code is hard to read, but because a) bad code is more common in any language which is easy to learn and b) Perl has several features which people mistake for non-readability (that is, non- or inexperienced Perl programmers assume that code is hard to read because they don't know Perl and see these things which scare them):
Regular expressions - This is a common feature in almost all languages these days, but the ease with which they are integrated into Perl syntax makes them more common in Perl programs. Of course, you have to be careful about the amount to which you let such constructs take over your code, but Perl was the one to introduce whitespace formatting and comments into regular expressions for just this reason.
Typing glyphs - Many programmers from the C/Pascal/Fortran - derived world take exception to the prefix-characters in Perl. Some Perl programmers find it hard to read C code that uses subroutines, complex data structures and simple types without any indication of what's being accesed or how. It's a matter of time and exposure on BOTH sides, and being a programmer in both worlds, I can tell you that both are equally readable with sufficient exposure.
Context sensitivity - Perl's context sensitivity spans every level from the way the tokenizer works all the way up to the handling of large-scale data structures. This is the nature of a language designed by a linguist. It "reads well", but only when you start trying to read it like a spoken language, and not a mathematical code. Software source code has always been somewhere in the middle-ground between those two extremes, and it's jarring at first that Perl moved the line so much, but give it a while and you find that it's much easier to think and communicate complex ideas (by complex, I mean cognitive complexity, not code complexity) in Perl than in any other programming language.
Weak typing - This is a matter of religion, but I'll make a footnote of it anyway: Perl is weakly typed, and that frustrates many who are used to strong typing. Neither is better or worse, though the fact that Perl 6 will (as Common LISP has done for quite some time) give you both is a boon, I think.
"The new generation of kids don't seem to like anything that you don't pop a CD into and have a game controller attached to."
You're generalizing. Video games are very popular, but so are collectable card games, comics, tabletop roleplaying, miniatures gaming (D&D Minis are HOT on eBay right now, and kids are buying most of them for "warbands"), etc.
I suggest you think of kids as very small adults, and then imagine the generalization, "these adults these days aren't interested in any OS that doesn't have a "Start" button."
This however, goes after a document that I think is invaluable, the methods FAQ. This was a wonderful document that gave the attention-getters a list of the things that would hurt and/or disfigure and/or kill (and thus to avoid) while also giving those serious about suicide a list of the things that would work quickly and painlessly.
Personally, I'm of the opinion that if someone wants to end their life, we should try to help them through the trouble that lead to that decision, but at the same time, provide them with accurate and helpful information that prevents some of the life-shatteringly bad decisions I've seen (like people who blow their face off with a shotgun and end up in a hospital for 20 years).
My point was that, here in the states, we don't care if the title is Boogie Nights or T2 or any other awful thing... we go to a movie based on word of mouth and/or advertising. If someone says "hey, that Japanese film, School Girl Mud Bath, is really good," we'll go see it... ok, there are a lot of guys that would see it anyway, but that's not my point;-)
[Examples of movies that translate poorly to English, as a reason that they'll "have a hard time" in the US]
Porco Rosso [...] Princess Mononoke [...] My neighbor Totoro / Totoro [...] Naussica of the Valley of the Winds [...] Ghost in the Shell [...] Perfect Blue [...] Wings of Honneamise
Ok, so let's look at the movies that we Americans DO like for examples of those excellent title ideas, shall we?
Harry Potter and the Sorcerer's Stone -- Top grossing movie in the US, 2001. WAY too long, name sounds like someone who should be living on the street. Harkens back to all those bad 70s movies that tried to bring european legends to the big screen. Yuck!
The Lord of the Rings: The Fellowship of the Ring -- Second highest grossing movie of 2001 in the US. Again, WAY too long. Sounds religious. "Fellowship of the Ring" sounds like it should be the title of a gay porn movie.
Shrek -- Third highest grossing movie in the US of 2001. I make noises like this when I sneeze. WHAT THE UNDEAD MOTHER OF STARBURST FRUIT-CHEWS WERE THEY THINKING?!
Monsters, Inc. -- Fourth highest grossing movie in the US of 2001. Here we combine the child-unsafe hint ("Monsters", thus excluding many of your target audience by parent decree) with the mundane horror of American corporate life ("Inc."). Clearly either a) someone was on crack or b) this was translated from the Hindi.
Rush Hour 2 -- Fifth highest grossing movie in the US of 2001. This is a sequel, so you can't account for naming, but still... a movie about the worst part of everyone's day?! And they TOLD YOU THAT in the title?! Again. Gotta be the crack.
Go ahead. Pick a year. You can do this with ANY set of movies.
Japanese movies will continue to do well in the US, with the only barrier being how well they are promoted (of course, the bad Japanese movies won't do any better in the US than they do in Japan... for the most part).
Cloaking is when you pretend to the search engine that you're a different kind of site so that you get ranked in with that kind of site.
No, cloaking is when you present different content to clients depending upon whether they are a search engine bot or a normal visitor. This is definitely cloaking.
Ok, then perhaps someone needs to come up with a reason why that's a bad thing, rather than just throwing the name around. I was introduced to the concept of cloaking by example. Company A wants company B to index them so that they show up when people are looking for radishes, so when a search engine comes along, they "cloak" themselves so that they look (in terms of what search engines care about) like a radish-lovers site. Then when a user comes along, they pop up their pr0n ads.
Google is just setting some titles ON THEIR OWN SITES FOR THEIR OWN INDEXER... this goes way beyond harmless, and I simply don't see why the two actions are being compared.
This is OBVIOUSLY a criteria Google can never meet. For them to recommend it to others is one thing, but how are they supposed to write ANYTHING with that criteria in mind? Hmmm... how would I do this if I didn't exist? Well, I guess I wouldn't... Wow, that was easy, I'm going on vacation!
Seriously, this is Google talking to Google. Take a deep breath and relax. They have not taken Poland yet.
why do the GNOME/KDE/whomever developers have to design their own clipboards rather than use what X provides?[...] Is it really that difficult?
No, it's not difficult at all. It's also non-functional. The X-provided selection mechanism is primative at best, and lacks almost all of the features that a high-level desktop environment exists to provide. This is why Gnome and KDE must roll their own. They have also made great strides in interoperability between the two, and standards now exist which ANY desktop can conform to in order to interoperate with all of the above.
What applications are you refering to when you say that interoperability isn't available?
I routinely cut-and-paste and I have no problems at all between Gaim, Firefox, XChat, Evolution, Gnumeric, gnome-terminal, gvim, etc.
Here's the problem with Linux users: they get a distribution with "everything" because they think that's a good thing, and then they complain because "everything" doesn't work. Try not installing everything, but just the bare-bones tools that the distribution recommeneds that you need to get work done. Your life will be easier.
More to the point: ignorning the fact that this is google sending keywords to google; ignoring the fact that this their site and they can do whatever they want; ignoring the fact that people are using words like "evil" to describe something that affects exactly no one;...
The key point here is relevancy. The keywords are relevant and accurate. You might say that this breaks Google's style guidelines, and that's a good reason for them to bug-fix it. But, I fail to see how this is some great transgression on Google's part. This is USEFUL INFORMATION that they are putting in the title. Ugly, sure. I hate when eBay does the same thing. It's still not keyword spam, and it's still not cloaking. Cloaking is when you pretend to the search engine that you're a different kind of site so that you get ranked in with that kind of site. It's not putting keywords in ugly user-visible places when they are relevant.
Please return to your useless ranting about Microsoft or something.
Keyword spamming is when you put UNRELATED keywords in the title or "keywords" headers of a page.
For example, if your page is a pile of ads for random stuff and your keywords are "tequila, mp3, oscars", then that's keyword spam. Putting the keywords in the title was a way to get around anti-keyword spamming techniques for a while. Many have said that putting keywords in the title is a bad thing because it results in unreadable titles, which is true.
Google has no circumvented that by putting readable, usable titles in the pages served to users and relevant, but verbose titles in pages served to crawlers... and this is related to keyword spamming how?!
Warren Ellis reviewed just this (I'm not going to link to his site, it's already slow), and said that Americans would almost certainly not enjoy it. It's (as far as he's concerned) a return to many things that're right and wrong with Brittish television, but far too Brittish for most of the rest of the world to enjoy, especially us Yanks.
"Word is that Sci-Fi Channel declined to acquire this new DOCTOR WHO series. And I can see why. It's too damned English." -Ellis
C-front had no bearing on GCC. There were, in fact, other true compilers for C++ before GCC, and GCC was never a C++-to-C translator like C-front. Did C-front back-end to gcc by default? I don't recall, but possibly. Still, it was never part of the GCC suite.
I never heard anyone calling M.T.'s code crap at the time, but that's almost irrelevant. It's clear that, even if he's the worst programmer ever, he is so prolific (GDB works on C++ because of him) and so able to build community (he was the reason that GCC 3.0 happened) that it would be worth it.
Was most of his code in GCC re-written? That would surprise me, but it's possible. 3.4 is a very different beast from 3.0, but 3.4 exists because of 3.0, and the state of stagnation that 2.95 was in when Tiemann (and his team at Cygnus) gave up and forked egcs to push the community forward was not going to produce 3.0 any time soon.
Just following up because the mod of "troll" seemed kind of shocking to me. I want to make it clear here that I firmly believe in what I said. Fedora has a much larger audience than RHL ever did, and many of its users are very loyal. Red Hat may have pissed some people off by EOLing RHL, but that is no reason to ignore the fact that they also PLEASED a lot of people by releasing Fedora.
Oh please, just stop. If you have no appreciation for the history involved, then you're not going to be able to contribute to this in an informed way.
Michael Tiemann is the founder of Cygnus Software (which was bought by Red Hat). If you want his OSS credentials, go to any copy of the GCC source and use grep. He's not heading this group because he's a corporate drone for Red Hat, he's heading this group because he's a better choice than ANY OF US!
It's an interesting article with a fairly cogent -- if subjective -- thought about socio-economic origins of prejudice... and yet it's said with all the tact of a true geek. Heh.
Well, at least he understood that people were not taking it as intended, and took it down. Quite a few people around here would have left it up, saying, "what's the big deal?"
Slashdot Mirror
Mozilla is the better browser?! Ugh. I tried to use Mozilla the other day, and I was stunned at how hard it was to use. I got away from Mozilla to Galeon long ago because it was so slow and klunky, and only switched back to Firefox because it absorbed a lot of the developer mindset of Galeon (separating searching from URLs; allowing for quick search selection between dictionary, shopping, Web, etc.; vastly simplified, and yet still powerful bookmark management (though I want Galeon's make bookmark here without having to install an extension); and most of all, it's FAST and relatively small).
If you LIKE Mozilla better, that's fine... preference is good, but calling it the "better browser is, I think, a bit silly".
Right, and in the 90s that 5% of people motivated by the sheer joy of what they do were, largely being swept up into the growth of the Internet. It was a big shiny thing. Now, it turns out that in technical fields, most of those people are men. In liberal arts fields, it's more of an even mix and in health-care it's more weighted toward women.
If, let's just say for example's sake, 20% of men in Internet-related businesses were in it for non-monetatry reasons and 2% of women were. If half of the money-motivated people left, that would mean a 48% exodus of women while only 40% of men would leave.
I happen to think that the numbers are EVEN MORE skewed than that, and I think that there are also shades of gray. There are many men who are in it for the money, but also have some "this is cool work" motivation, keeping them more tightly tied to the industry.
Sadly BitTorrent is only half of what I look for in a P2P client. To get the full picture, I use gtk-gnutella to talk to the Gnutella network. Easy searching, good bandwith-distributed downloads, excellent resiliancy and just all around nice. The last time I fired up BT was for a FC1 download, but these days (thanks to people like me) even full OS distributions show up within minutes on Gnutella.
What BitTorrent DOES give you is a single point of control. This can be useful, and is why I keep a BT client installed too.
Not if its RAID-ed across the 'net.
Oh good, now we don't even need a datacenter fire, since you've introduced a single point of failure!Yay technology!
"While he may make optomistic comments about the lawsuit filed by SCO, from speaking with hundreds of technical decision makers, including CIO's, the lawsuits have actually been a stumbling block in using a fullblown linux back end for alot of companies."
I think you're looking at this upside down. How many of those people who are using this as an excuse didn't need this event to make an execuse? Would the fact that Red Hat is a new company or IBM might go back to the Windows camp or something else have been the issue otherwise? I suspect so.
There are many people I deal with who look at the giant farms of Linux servers that I interact with and say, "hey, that's scary stuff!" Invariably, they are the dinosaurs who are busy being rendered obsolete. IBM mainframes were scary stuff at one point too, and no one could understand why you would want to stake your business on a MACHINE that could make MISTAKES... until a few dozen companies made it clear that NOT going that way was a ticket to extinction.
Windows desktops were the same way.
Unix servers: same story.
Web-based business transactions: same thing.
Every new technology requires a period of early-adoptors, and we're exiting (or just starting to exit) that phase with Linux. That's a scary time. These guys see the writing on the wall, and they're trying to make any rationalization they can to avoid the descision that they know they have to make. Tough nuts, industry doesn't care about their rationalization, only results.
There are three reasons this is bunk:
;-) and that bring us to:
First, the idea that we will throw away out current media has been floated since the days of the floppy. It's always a correct prediction, but only because a better physical medium comes along.
Second, the idea that we're going to be OK with just using storage on the Net and not having any physical media on which to store our data sounds good, right up until the first datacenter fire that loses me last week's data storage. It's also a terrible idea to keep your wares and copyrighted porn on someone else's servers
Third, PRIVACY. There's no single reason why networked media will never win over good-old local storage that beats the desire for privacy.
"A complex list comprehension can lead to some powerful code that does an amazing amount of stuff in a neat little package. To me this is different than the arbitrary line-noise you get with Perl."
There you just lost me. The term "line noise" implies a low signal-to-noise ratio, when in fact Perl presents exactly the opposite. The SIGNAL is in fact, so high that many programmers find it difficult to cope with. That's fine, but let's not confuse that with actuall NOISE.
"So Perl 6 lets you define your own syntax so that someone reading your code neads to figure out what your ideas of the Right Language is?"
No. You wholy misunderstood the concept.
In Perl 6, you will have full access to the grammar, so you could enforce your local stylistic conventions. You would obviously not want to make INCOMPATIBLE changes so that your code is still valid Perl, but you could write your own "strict".
Think of it this way. Imagine a C++ header that caused all uses of operator overloading outside of a limited few "neccessary" to be illegal, or that issued a compiler warning on every use of an iterator initialization outside of a for loop. These are just simple (and not very useful) examples, but they serve to illustrate the point: you can instrument the compiler just as fully as you can instrument your code. Don't like the type checking in Perl 6? Make it stricter.
I'm sure that there will be someone who will publish the "python-like bondage" module 15 minutes after Perl 6 is released. If you're into that sort of thing, then your company can take full advantage of it, while still getting all the value of Perl 6 like LISP-style currying and macros, Ruby-style mixins, cross-langauge bindings through Parrot, boxed and unboxed type constraints on standard Perl scalars, full multi-method dispatch, etc., etc.
"Common LISP and Scheme's macro facilities can be used to define your own language constructs"
Yes, but we're not talking about defining language constructs. We're talking about changing the behavior of the compiler in structured, standardized ways that aren't just implementation hacks. Don't get me wrong. Common LISP is on my list of cool languages to learn more about right below Python and Ruby. I'm just saying that these particular Perl 6 features bear a bit more looking at.
"Or hell, TeX lets you redefine the world if you are so twisted, though to me TeX is more unreadable than Perl."
TeX is actually a good example of what I'm talking about. TeX is very readable for a full typesetting system, but most of us could not care less about typesetting. When you need to do specific tasks that INVOLVE typesetting, but you don't really need all of that power and flexibility, you step up a layer of abstraction and turn TeX into LaTeX. LaTeX is valid TeX, so it's not quite the same, but the idea of limiting a powerful system in order to step back a level of abstraction holds.
Perl 6 will provide all of the power that you need from a modern high-level programming langauge, but let you manage that complexity. You might decide, for example, to restrict Perl 6 in your programs to just the facilities that make sense for scientific calculation. You might even introduce a special syntax/grammar for putting differential equations directly into your program without having to quote around them and hand them off to a seperate processing tool (object, module, what-have-you).
None of this is useful for your average 1000-line CGI program, but for the company that produces tens to millions of lines of structured libraries upon which new software is built and re-factored over time, this will all be a godsend.
Much of what Perl 6 brings to the table, Common LISP has done for years, but some of it is either gathered from other, more recent langauges (e.g. Python, Ruby, Scheme, Java, etc.) or is, as far as I can tell, unique. I hope you give it a try and throw away your naive ideas of "line noise" in favor of considering the value to your productivity and the maintainability of your code base.
LISP has always had the same exact problem. In fact, the creator of Perl, Larry Wall, once said that LISP had all of the visual appeal of oatmeal with fingernail clippings mixed in (an obvious jab at the use of parens which is no more valid that the jabs against Perl for the use of $). And now, Perl 6 is borrowing HEAPING amounts of old wisdom from Common LISP in order to improve the power and flexibility of the langauge (from the way it does macros to currying to the way blocks are built as functional units instead of as C-style lexical scopes).
I'm not putting Larry down, here, just pointing out that languages which make it easy to do a large job in fairly little real eastate are often beset with this kind of short-sighted thinking from even otherwise level-headed folk. It's a good thing that we grow up and get past it.
"Of course you can write some incredibly terse code in Python too, especially with recent language additions."
;-)
Yep. EVERY language can be bent to horrid non-readability.
If you want a language that enforces readability, I would suggest using Perl 6 (the first MAJOR step in the actual implementation of which was just made with the creation of a Haskell-hosted compiler called pugs) and writing a module which enforces your idea of readability. Perl 6 will be the first language to give you enough control that you can write this kind of requirement in the language itself. Common LISP comes in a close second, so use that in the mean-time if you want such levels of dictatorial control. Of course, you could write a module that allowed the rest of your program to be written in CL in P6 quite easily
PS: y=x+++1 Guess what language that is. Answer: C, C++, Python and Java. Perl too if you just add $ before y and x. Programming languages are ugly.
You are correct. It's also not dependent on who you ask. It means BOTH things, as stated by the creator of Perl, Larry Wall.
As they do with C, C++, Java, Snobol, Forth, APL, C#, etc, etc.
Half of the people programming are below-average programmers. Bad programmers can make life HELL in C, and by the same token good programmers can make life quite easy in Perl.
That said, Perl gets a bad rep, not because good code is hard to read, but because a) bad code is more common in any language which is easy to learn and b) Perl has several features which people mistake for non-readability (that is, non- or inexperienced Perl programmers assume that code is hard to read because they don't know Perl and see these things which scare them):
"The new generation of kids don't seem to like anything that you don't pop a CD into and have a game controller attached to."
You're generalizing. Video games are very popular, but so are collectable card games, comics, tabletop roleplaying, miniatures gaming (D&D Minis are HOT on eBay right now, and kids are buying most of them for "warbands"), etc.
I suggest you think of kids as very small adults, and then imagine the generalization, "these adults these days aren't interested in any OS that doesn't have a "Start" button."
This however, goes after a document that I think is invaluable, the methods FAQ. This was a wonderful document that gave the attention-getters a list of the things that would hurt and/or disfigure and/or kill (and thus to avoid) while also giving those serious about suicide a list of the things that would work quickly and painlessly.
Personally, I'm of the opinion that if someone wants to end their life, we should try to help them through the trouble that lead to that decision, but at the same time, provide them with accurate and helpful information that prevents some of the life-shatteringly bad decisions I've seen (like people who blow their face off with a shotgun and end up in a hospital for 20 years).
My point was that, here in the states, we don't care if the title is Boogie Nights or T2 or any other awful thing... we go to a movie based on word of mouth and/or advertising. If someone says "hey, that Japanese film, School Girl Mud Bath, is really good," we'll go see it... ok, there are a lot of guys that would see it anyway, but that's not my point ;-)
[Examples of movies that translate poorly to English, as a reason that they'll "have a hard time" in the US]
Porco Rosso [...] Princess Mononoke [...] My neighbor Totoro / Totoro [...] Naussica of the Valley of the Winds [...] Ghost in the Shell [...] Perfect Blue [...] Wings of Honneamise
Ok, so let's look at the movies that we Americans DO like for examples of those excellent title ideas, shall we?
Harry Potter and the Sorcerer's Stone -- Top grossing movie in the US, 2001. WAY too long, name sounds like someone who should be living on the street. Harkens back to all those bad 70s movies that tried to bring european legends to the big screen. Yuck!
The Lord of the Rings: The Fellowship of the Ring -- Second highest grossing movie of 2001 in the US. Again, WAY too long. Sounds religious. "Fellowship of the Ring" sounds like it should be the title of a gay porn movie.
Shrek -- Third highest grossing movie in the US of 2001. I make noises like this when I sneeze. WHAT THE UNDEAD MOTHER OF STARBURST FRUIT-CHEWS WERE THEY THINKING?!
Monsters, Inc. -- Fourth highest grossing movie in the US of 2001. Here we combine the child-unsafe hint ("Monsters", thus excluding many of your target audience by parent decree) with the mundane horror of American corporate life ("Inc."). Clearly either a) someone was on crack or b) this was translated from the Hindi.
Rush Hour 2 -- Fifth highest grossing movie in the US of 2001. This is a sequel, so you can't account for naming, but still... a movie about the worst part of everyone's day?! And they TOLD YOU THAT in the title?! Again. Gotta be the crack.
Go ahead. Pick a year. You can do this with ANY set of movies.
Japanese movies will continue to do well in the US, with the only barrier being how well they are promoted (of course, the bad Japanese movies won't do any better in the US than they do in Japan... for the most part).
Google is just setting some titles ON THEIR OWN SITES FOR THEIR OWN INDEXER... this goes way beyond harmless, and I simply don't see why the two actions are being compared.
Seriously, this is Google talking to Google. Take a deep breath and relax. They have not taken Poland yet.
No, it's not difficult at all. It's also non-functional. The X-provided selection mechanism is primative at best, and lacks almost all of the features that a high-level desktop environment exists to provide. This is why Gnome and KDE must roll their own. They have also made great strides in interoperability between the two, and standards now exist which ANY desktop can conform to in order to interoperate with all of the above. What applications are you refering to when you say that interoperability isn't available? I routinely cut-and-paste and I have no problems at all between Gaim, Firefox, XChat, Evolution, Gnumeric, gnome-terminal, gvim, etc. Here's the problem with Linux users: they get a distribution with "everything" because they think that's a good thing, and then they complain because "everything" doesn't work. Try not installing everything, but just the bare-bones tools that the distribution recommeneds that you need to get work done. Your life will be easier.
More to the point: ignorning the fact that this is google sending keywords to google; ignoring the fact that this their site and they can do whatever they want; ignoring the fact that people are using words like "evil" to describe something that affects exactly no one; ...
The key point here is relevancy. The keywords are relevant and accurate. You might say that this breaks Google's style guidelines, and that's a good reason for them to bug-fix it. But, I fail to see how this is some great transgression on Google's part. This is USEFUL INFORMATION that they are putting in the title. Ugly, sure. I hate when eBay does the same thing. It's still not keyword spam, and it's still not cloaking. Cloaking is when you pretend to the search engine that you're a different kind of site so that you get ranked in with that kind of site. It's not putting keywords in ugly user-visible places when they are relevant.
Please return to your useless ranting about Microsoft or something.
This is NOT keyword spamming.
Keyword spamming is when you put UNRELATED keywords in the title or "keywords" headers of a page.
For example, if your page is a pile of ads for random stuff and your keywords are "tequila, mp3, oscars", then that's keyword spam. Putting the keywords in the title was a way to get around anti-keyword spamming techniques for a while. Many have said that putting keywords in the title is a bad thing because it results in unreadable titles, which is true.
Google has no circumvented that by putting readable, usable titles in the pages served to users and relevant, but verbose titles in pages served to crawlers... and this is related to keyword spamming how?!
Warren Ellis reviewed just this (I'm not going to link to his site, it's already slow), and said that Americans would almost certainly not enjoy it. It's (as far as he's concerned) a return to many things that're right and wrong with Brittish television, but far too Brittish for most of the rest of the world to enjoy, especially us Yanks.
"Word is that Sci-Fi Channel declined to acquire this new DOCTOR WHO series. And I can see why. It's too damned English." -Ellis
Hmmm... not really.
C-front had no bearing on GCC. There were, in fact, other true compilers for C++ before GCC, and GCC was never a C++-to-C translator like C-front. Did C-front back-end to gcc by default? I don't recall, but possibly. Still, it was never part of the GCC suite.
I never heard anyone calling M.T.'s code crap at the time, but that's almost irrelevant. It's clear that, even if he's the worst programmer ever, he is so prolific (GDB works on C++ because of him) and so able to build community (he was the reason that GCC 3.0 happened) that it would be worth it.
Was most of his code in GCC re-written? That would surprise me, but it's possible. 3.4 is a very different beast from 3.0, but 3.4 exists because of 3.0, and the state of stagnation that 2.95 was in when Tiemann (and his team at Cygnus) gave up and forked egcs to push the community forward was not going to produce 3.0 any time soon.
Just following up because the mod of "troll" seemed kind of shocking to me. I want to make it clear here that I firmly believe in what I said. Fedora has a much larger audience than RHL ever did, and many of its users are very loyal. Red Hat may have pissed some people off by EOLing RHL, but that is no reason to ignore the fact that they also PLEASED a lot of people by releasing Fedora.
Oh please, just stop. If you have no appreciation for the history involved, then you're not going to be able to contribute to this in an informed way.
Michael Tiemann is the founder of Cygnus Software (which was bought by Red Hat). If you want his OSS credentials, go to any copy of the GCC source and use grep. He's not heading this group because he's a corporate drone for Red Hat, he's heading this group because he's a better choice than ANY OF US!
It's an interesting article with a fairly cogent -- if subjective -- thought about socio-economic origins of prejudice... and yet it's said with all the tact of a true geek. Heh.
Well, at least he understood that people were not taking it as intended, and took it down. Quite a few people around here would have left it up, saying, "what's the big deal?"