But on the other hand, neither form is something that a beginner would know without being told, really you're going to need to be handed an example to learn it, so the actual practical difference between the two isn't that huge.
A beginner wouldn't know that a "read" method exists in a "file" object. But when confronted with a "file.read()" invocation, the meaning seems pretty straightforward to intuit.
Now, "local $/;" gives you absolutely nothing to reason over it. Only after your explanation I have a grasp on what's happening (which in Python could be represented with "".join(file)) , though I still don't understand how the $/ syntax means $INPUT_RECORD_SEPARATOR = undef.
It's mostly because Perl is line-by-line by default. Most of the time you don't want to read a whole file into memory, do you?
[...]
You could also do my @data = <$file>; which slurps the whole file, line-by-line, into an array, if that's what you need.
I can't say "most of the time", my guess is I use equally reading the whole file, reading chunks of binary files, and reading lines.
local $/; switches to slurp mode (temporarily).
Ahhh, of course!!! That's obvious! </sarcasm>
Not only the syntax is cryptic, the terminology is terrible! "Slurp" is a funny way to describe the process, but introduces one more step (two more for non-native speakers) to understanding it. Compare that to Python's `file.read()` and `file.readlines()` and you'll understand why Python is better for beginners and for people that want maintainable code.
Shakespeare's style might be great for poetry, but if you use it to write guides, don't be surprised if people get lost.
a) Scientists are terrible at Perl. "You can write FORTRAN in any language" applies to them; they often write Perl as if it were C. Or that's my experience based on Stack Overflow and Perlmonks questions. They aren't programmers, they're scientists, and it shows.
Yep, that's usually the case.
b) A rewrite is always clearer, no matter the language.
You are right, of course, and as I used the word "translated", that could imply a rewrite. I should have used "transliterated", as I kept the original structure. I tried to find these scripts so I could give a concrete example, but I could only find my Python versions. But it was stuff in Perl like open($file, "file.txt"); my $data = do { local $/; <$file> };
that in Python becomes data = open("file.txt").read()
that makes you think, wtf Perl??? Really, why do you need all that crap just to read the whole file?
Thanks for your comment, it expresses my view pretty well. I want to add a note about this:
Also, Python has done pretty well as a first programming language, even if the design runs out of steam at certain points. In contrast, we tend to think of Perl (especially Perl 6) more as a last programming language, the language of choice for people who need a language that won't give up when the going gets tough.
I started coding with BASIC and Z80 Assembly in the late 80's, then coded mostly C until 97 when I switched to C++, then finally in 2006 I switched to Python as main programming language. In these ~30 years I obviously had to program in several other languages, and Perl is by far the one I most hated. After my bad experience with Perl I even translated some Perl scripts coded by a biologist friend (Perl seems to be strong in the field) to show her how much more clear and concise they can be made in Python, and the difference is undeniable.
So it seems to me that Python (with occasional C extensions for speed) is a suitable "last language", though what the future holds no one knows. Except that it will be neither Perl 5 nor 6.
P.S.: JavaScript as it was in the beginning was pretty terrible too, even if it had a saner syntax. But it redeemed itself in the latest ES standards, and with the recent frameworks it is actually quite enjoyable.
It's even more interesting when you contrast it with the way the media conflate copyright infringement with stealing. When someone makes a bootleg copy of a movie, the original is still accessible. When someone steals something from you, you lose access to what has been stolen.
So, when will we read the news "Microsoft will steal workout videos from consumers"?
The FBI is only one of the cogs. Isn't it troubling when you can read either as...
The secret *government requests* for customer information
or
The *secret government* requests for customer information
because the FISA Court allows for exactly that, a secret and unaccountable government. Some day they'll swap "Foreign" for "Federal" in that acronym and nobody will notice any difference.
It would be so much better if Musk took up a conventional religion with decent moral imperatives
The intersection you want is empty, unfortunately. Unless you consider Jainism conventional. Religions in general have a tendency to consider moral the killing of people with certain (arbitrary) characteristics. IMHO, if you believe that you're as immoral as you could be.
Yes, MS loves UNIX standards. Like when they decided to use \ as folder delimiter, just to be different. And instead of using about any-fucking-thing else, they had to pick exactly the escape character. The Devil himself was envious of their ingenuity, which was eclipsed only many years later, with the launch of IE6.
Yes, just like Windows and Mac OS X, users are enjoying their 0-day releases of Mozilla Firefox and don't have to wait for package maintainers of a particular GNU/Linux distribution to update the software in the main repositories.
Really? I mean, are you guys so fucking retarded that you're not able to setup your own.deb repositories? Fuck, you can even use Launchpad's PPAs!
And it seems it's inconsequential to them as well. And from that we can extrapolate and say that those users, that were willing to pay but now are probably going to get the content by illegal means, doesn't really matter. It's a win-win-win scenario!
It doesn't justify the cost... you only need to test the cable once, when you're buying it. If your work involves dealing with lots of cables, then buy a tester.
My thoughts exactly. What we actually need is cheap hardware testers, available at any store so the consumer can test the cables before bringing them home. Online stores could state more confidently their cables were actually tested too.
Me too. And I guess any programmer feels the same, as we're trained to express what we're thinking without errors (as doing otherwise makes the compiler/tests scream at us).
A hyperbolic analogy may help you understand the reaction: imagine you know a guy that for three decades has been raping children with impunity, and now he starts giving away lots of candy... would you trust him?
Dude, you're mixing two things up: the "stereotypification" and the prejudice.
Inferring one's ethnicity by their name is not racism, it's just pattern recognition. Assuming something is bad because of the specific pattern, that might be racist. Just to be clear: stereotype("Palihapitiya") -> "Indian" # Wrong (the guy is not Indian), but OK! groupthink("Indian") -> "Everything must seem like crap!" # plain stupid, and racist
It's not a multiuser system, and it's highly unlikely someone could damage it remotely. Even if it happened, there's nothing of much importance in it (it's mostly a secondary dns, mx, and xmpp server). Userland libs and services are kept up-to-date and restarted as needed, though.
Slashdot Mirror
Franceâ(TM)s
And the walk of shame continues... bling bling bling! (https://slashdot.org/comments.pl?sid=9511599&cid=52681371)
I hate Perl, but even I know that you could solve (palliatively) this disgrace with a simple:
$post =~ s/â\(TM\)/'/g;
Or just use SoylentNewsâ(TM) [yeah, it was on purpose] version of Rehash, as they fixed this ages ago.
It's 2016, yet...
[...] Jet.comâ(TM)s [...] Samâ(TM)s Club. [...] companyâ(TM)s website[...]
...still happens.
- shame... shame... shame... bling bling bling
Very well said. I have no mod points, so here's my virtual +1.
But on the other hand, neither form is something that a beginner would know without being told, really you're going to need to be handed an example to learn it, so the actual practical difference between the two isn't that huge.
A beginner wouldn't know that a "read" method exists in a "file" object. But when confronted with a "file.read()" invocation, the meaning seems pretty straightforward to intuit.
Now, "local $/;" gives you absolutely nothing to reason over it. Only after your explanation I have a grasp on what's happening (which in Python could be represented with "".join(file)) , though I still don't understand how the $/ syntax means $INPUT_RECORD_SEPARATOR = undef.
It's mostly because Perl is line-by-line by default. Most of the time you don't want to read a whole file into memory, do you?
[...]
You could also do my @data = <$file>; which slurps the whole file, line-by-line, into an array, if that's what you need.
I can't say "most of the time", my guess is I use equally reading the whole file, reading chunks of binary files, and reading lines.
local $/; switches to slurp mode (temporarily).
Ahhh, of course!!! That's obvious! </sarcasm>
Not only the syntax is cryptic, the terminology is terrible! "Slurp" is a funny way to describe the process, but introduces one more step (two more for non-native speakers) to understanding it. Compare that to Python's `file.read()` and `file.readlines()` and you'll understand why Python is better for beginners and for people that want maintainable code.
Shakespeare's style might be great for poetry, but if you use it to write guides, don't be surprised if people get lost.
a) Scientists are terrible at Perl. "You can write FORTRAN in any language" applies to them; they often write Perl as if it were C. Or that's my experience based on Stack Overflow and Perlmonks questions. They aren't programmers, they're scientists, and it shows.
Yep, that's usually the case.
b) A rewrite is always clearer, no matter the language.
You are right, of course, and as I used the word "translated", that could imply a rewrite. I should have used "transliterated", as I kept the original structure.
I tried to find these scripts so I could give a concrete example, but I could only find my Python versions. But it was stuff in Perl like
open($file, "file.txt");
my $data = do { local $/; <$file> };
that in Python becomes
data = open("file.txt").read()
that makes you think, wtf Perl??? Really, why do you need all that crap just to read the whole file?
Thanks for your comment, it expresses my view pretty well. I want to add a note about this:
Also, Python has done pretty well as a first programming language, even if the design runs out of steam at certain points. In contrast, we tend to think of Perl (especially Perl 6) more as a last programming language, the language of choice for people who need a language that won't give up when the going gets tough.
I started coding with BASIC and Z80 Assembly in the late 80's, then coded mostly C until 97 when I switched to C++, then finally in 2006 I switched to Python as main programming language. In these ~30 years I obviously had to program in several other languages, and Perl is by far the one I most hated. After my bad experience with Perl I even translated some Perl scripts coded by a biologist friend (Perl seems to be strong in the field) to show her how much more clear and concise they can be made in Python, and the difference is undeniable.
So it seems to me that Python (with occasional C extensions for speed) is a suitable "last language", though what the future holds no one knows. Except that it will be neither Perl 5 nor 6.
P.S.: JavaScript as it was in the beginning was pretty terrible too, even if it had a saner syntax. But it redeemed itself in the latest ES standards, and with the recent frameworks it is actually quite enjoyable.
It's even more interesting when you contrast it with the way the media conflate copyright infringement with stealing. When someone makes a bootleg copy of a movie, the original is still accessible. When someone steals something from you, you lose access to what has been stolen.
So, when will we read the news "Microsoft will steal workout videos from consumers"?
Indeed, Microsoft has been sloppy, as always. But don't worry, next time there will be no evidence of spying being done.
The FBI is only one of the cogs. Isn't it troubling when you can read either as ...
The secret *government requests* for customer information
or
The *secret government* requests for customer information
because the FISA Court allows for exactly that, a secret and unaccountable government. Some day they'll swap "Foreign" for "Federal" in that acronym and nobody will notice any difference.
It would be so much better if Musk took up a conventional religion with decent moral imperatives
The intersection you want is empty, unfortunately. Unless you consider Jainism conventional. Religions in general have a tendency to consider moral the killing of people with certain (arbitrary) characteristics. IMHO, if you believe that you're as immoral as you could be.
That would actually be the AGPLv3.
Yes, MS loves UNIX standards. Like when they decided to use \ as folder delimiter, just to be different. And instead of using about any-fucking-thing else, they had to pick exactly the escape character. The Devil himself was envious of their ingenuity, which was eclipsed only many years later, with the launch of IE6.
The article's author lives in the Sahara, 30*C is already too cold!
(it should be a degree glyph instead of an asterisk, but /. thinks unicode is bad)
Yes, just like Windows and Mac OS X, users are enjoying their 0-day releases of Mozilla Firefox and don't have to wait for package maintainers of a particular GNU/Linux distribution to update the software in the main repositories.
Really? I mean, are you guys so fucking retarded that you're not able to setup your own .deb repositories? Fuck, you can even use Launchpad's PPAs!
Because the rights holders demand it.
And it seems it's inconsequential to them as well. And from that we can extrapolate and say that those users, that were willing to pay but now are probably going to get the content by illegal means, doesn't really matter. It's a win-win-win scenario!
It doesn't justify the cost... you only need to test the cable once, when you're buying it. If your work involves dealing with lots of cables, then buy a tester.
My thoughts exactly. What we actually need is cheap hardware testers, available at any store so the consumer can test the cables before bringing them home. Online stores could state more confidently their cables were actually tested too.
Me too. And I guess any programmer feels the same, as we're trained to express what we're thinking without errors (as doing otherwise makes the compiler/tests scream at us).
A hyperbolic analogy may help you understand the reaction: imagine you know a guy that for three decades has been raping children with impunity, and now he starts giving away lots of candy... would you trust him?
Dude, you're mixing two things up: the "stereotypification" and the prejudice.
Inferring one's ethnicity by their name is not racism, it's just pattern recognition. Assuming something is bad because of the specific pattern, that might be racist.
Just to be clear:
stereotype("Palihapitiya") -> "Indian" # Wrong (the guy is not Indian), but OK!
groupthink("Indian") -> "Everything must seem like crap!" # plain stupid, and racist
It's not a multiuser system, and it's highly unlikely someone could damage it remotely. Even if it happened, there's nothing of much importance in it (it's mostly a secondary dns, mx, and xmpp server). Userland libs and services are kept up-to-date and restarted as needed, though.
dmbasso@raspi ~ $ uptime
20:03:06 up 399 days, 20:50, 1 user, load average: 0.03, 0.07, 0.06
Booting speed is not an issue to me. :)
Now I can finally remove the [FUCK BETA] tag from my sig. Took only two years, but better late than never! :)
Wow, now I see. It is indeed revolutionary!