We have chips, in use, that were of Motorola design when they were in cahoots with IBM and Apple. Then Freescale. Then NXP. Now Qualcomm. They just now updated their default install location from C:\Freescale to C:\NXP.
Maybe moving to Renesas won't be too terrible... then again they were NEC.
Lets bring back gasification cars that got us through gas shortages. Go down and buy myself some Grade A West Virginia coal and put all those hard working ditch diggers back to work.
In the olden days we called this a 'ledger'. Bitcoin itself, I bought some at $500 for fun, it could crash, it could go to the moon. I've cashed out break even.
Now the blockchain is where I'm excited for my line of work (embedded automotive/industrial/aerospace). Accountable, recorded, distributed tracking of who signed off on what calibration and when.
The current crop of tools AVL CRETA and Vector vCDM are traceability abominations. Digging into the underlying system it's just a terrible wrapper on a SQL database. A DB admin could go in and flip the "Violate Diesel Emissions" bit without having to go through the front end.
When Grandma's self driving car goes through the Farmers Market the NTSB is going to start tracking down exactly when and who made the brake force calibration. I absolutely hope there's a block chain that points out it wasn't my decision to change it but Bob in accounting.
You somehow think this is better, the correction under it?
sub eye {
push @_, $_[0] if @_
my $pdl = zeroes(@_);
$pdl->diagonal(0,1).= 1;
$pdl; }
The point stands that PDL is nowhere near as complete of a toolbox as numpy or MATLAB. You may be able to do some light number crunching with it but academics aren't going to be using Perl for any numerics any time soon.
In my last 15+ years on Slashdot I've found that commenter fall into one of two categories.
I think it's one of those theoretical arguments by people who don't like the language
and
No wireless. Less space than a nomad. Lame.
Python is here to stay. These Perl luddites are more than happy to go into their corners of industry and die like the COBOL and FORTRAN people have. No one is ever going to touch their code because 'it works' and some unlucky schmuck will be stuck maintaining it. But for the rest of us Python is trivially easy to read and write.
No one is teaching middle schoolers Perl. There aren't re-education and retraining syllabuses designed around Perl. There aren't dozens if not hundreds of tutorials and training around how to use Perl. The best responses most of them have in this thread are "It's in CPAN, RTFM".
Colleges aren't moving their freshmen engineering programming courses from MATLAB/FORTRAN to Perl, they're moving it to Python.
High school and middle school students doing code combat in Perl.
It's nice that you've had your head buried in what ever you've been doing for the last 5 years but Python is here and it's here to stay big. It's the lingua franca of programming for the next decade or so. They have a higher chance of finding a tutorial on on what they're trying to do with 'How to ____ in Python".
The only people picking up Perl like those picking up FORTRAN & COBOL, they're doing it because they want a niche in industry, not because it's actually going to become more popular. Perl was a nice stepping stone but most people don't want to deal with it.
have yet to find a data processing library/module/implementation
numpy, matplotlib, any of the neural network libraries.
whatever fad language is popular at the moment.
Yeah, Python that fad language from 1991.
2000 if you're counting Python 2. 2008 for Python 3.
Python is hitting critical mass and being used everywhere. Perl is going the way of COBOL. There will always be profit in using it and always be someone doing it. But they're not teaching Perl to middle schoolers, for good reason.
CPAN is twice the size of PyPI measured by discrete modules. Determination of usefulness is left as an exercise for the reader.
Where does npm fit if you want to do shear numbers? How many of those packages are still maintained?
For the most part Perl is being used as a wrapper for something else. It's a thin script to automate some bulk of work done *elsewhere* in the system.
I want to know where Perl beats Python hands down in performance that doesn't include extra libraries.
When your peers and successors only need to pick up C# to get their work done then C# is all that matters. With Perl being 'write only' that is a near impossible task. Given the choice of fixing existing Perl or re-writing it in Python (or even C# in your case) I'll always chose the latter.
I write wrappers like that all day long. Python wrapping C# for dSpace, ETAS and CRETA. Python wrapping C for Vector CANape. Python wrapping C for Matlab/Simulink. When my peers need to use any of those tools all they need to know is Python. There's even a chance they can fix or improve it. With Perl that's mostly not an option.
This is the PhD gamble. You hope that you learn enough and live long enough for your cutting edge research to find a practical purpose.
Back in 2004 DARPA sponsored a 'small' project to drive cars autonomously. Lots of companies and schools threw warm bodies at the problem and for a few years it some of it was purely theoretical research.
Then it reached a tipping point that a profitable end was in sight.
Slashdot Mirror
Sigh.
We have chips, in use, that were of Motorola design when they were in cahoots with IBM and Apple. Then Freescale. Then NXP. Now Qualcomm. They just now updated their default install location from C:\Freescale to C:\NXP.
Maybe moving to Renesas won't be too terrible... then again they were NEC.
Lets bring back gasification cars that got us through gas shortages. Go down and buy myself some Grade A West Virginia coal and put all those hard working ditch diggers back to work.
How many of those 'mistakes' make national news? Mistakes are made all the time. Mistakes don't get the attention of regulatory bodies and nations.
it is by far the rare exception.
Where have you been recently?
So VW just accidentally fat fingered a calibration into their diesel software?
Kobe Steel accidentally found some inaccurate data in their metallurgy?
GM's ignition switch was accidentally ignored?
In the olden days we called this a 'ledger'. Bitcoin itself, I bought some at $500 for fun, it could crash, it could go to the moon. I've cashed out break even.
Now the blockchain is where I'm excited for my line of work (embedded automotive/industrial/aerospace). Accountable, recorded, distributed tracking of who signed off on what calibration and when.
The current crop of tools AVL CRETA and Vector vCDM are traceability abominations. Digging into the underlying system it's just a terrible wrapper on a SQL database. A DB admin could go in and flip the "Violate Diesel Emissions" bit without having to go through the front end.
When Grandma's self driving car goes through the Farmers Market the NTSB is going to start tracking down exactly when and who made the brake force calibration. I absolutely hope there's a block chain that points out it wasn't my decision to change it but Bob in accounting.
You somehow think this is better, the correction under it?
sub eye
{
push @_, $_[0] if @_
my $pdl = zeroes(@_);
$pdl->diagonal(0,1) .= 1;
$pdl;
}
The point stands that PDL is nowhere near as complete of a toolbox as numpy or MATLAB. You may be able to do some light number crunching with it but academics aren't going to be using Perl for any numerics any time soon.
That's because I had to switch it up because in your world PDL is close to Numpy.
Based on the PDL 2.007 release comments this is somehow the appropriate way to create an identity matrix with PDL:
sub eye .= 1;
{
push @_, $_[0] if @_ diagonal(0,1)
$pdl;
}
I can't imagine what other horrors lay beneath trying to do any sort of controls work.
In my last 15+ years on Slashdot I've found that commenter fall into one of two categories.
I think it's one of those theoretical arguments by people who don't like the language
and
No wireless. Less space than a nomad. Lame.
Python is here to stay. These Perl luddites are more than happy to go into their corners of industry and die like the COBOL and FORTRAN people have. No one is ever going to touch their code because 'it works' and some unlucky schmuck will be stuck maintaining it. But for the rest of us Python is trivially easy to read and write.
No one is teaching middle schoolers Perl. There aren't re-education and retraining syllabuses designed around Perl. There aren't dozens if not hundreds of tutorials and training around how to use Perl. The best responses most of them have in this thread are "It's in CPAN, RTFM".
How about this, it's being taught.
Colleges aren't moving their freshmen engineering programming courses from MATLAB/FORTRAN to Perl, they're moving it to Python.
High school and middle school students doing code combat in Perl.
It's nice that you've had your head buried in what ever you've been doing for the last 5 years but Python is here and it's here to stay big. It's the lingua franca of programming for the next decade or so. They have a higher chance of finding a tutorial on on what they're trying to do with 'How to ____ in Python".
The only people picking up Perl like those picking up FORTRAN & COBOL, they're doing it because they want a niche in industry, not because it's actually going to become more popular. Perl was a nice stepping stone but most people don't want to deal with it.
still work. ;)
If you call what clear ClearCase does 'working'.
Also, how in god's name do you consider that easier to read and simpler than https://numba.pydata.org/?
Ctrl-F CUDA.
I didn't see CUDA support. Were those instructions on another page?
Numba is a JIT compiler for Python.
What is the Perl equivalent of Jupyter notebooks? Numba?
How's Perl 6 going?
have yet to find a data processing library/module/implementation
numpy, matplotlib, any of the neural network libraries.
whatever fad language is popular at the moment.
Yeah, Python that fad language from 1991.
2000 if you're counting Python 2. 2008 for Python 3.
Python is hitting critical mass and being used everywhere. Perl is going the way of COBOL. There will always be profit in using it and always be someone doing it. But they're not teaching Perl to middle schoolers, for good reason.
CPAN is twice the size of PyPI measured by discrete modules. Determination of usefulness is left as an exercise for the reader.
Where does npm fit if you want to do shear numbers? How many of those packages are still maintained?
Such as? Python 3.3 which is the first usable of the Python 3's is 5 years old.
All of my Python 2 code I just treat as a separate external application (Like I do Perl, C, Matlab, etc)
Not to mention having to exactly preserve the white-space when copying any code around
The grey beards on Slashdot continually bring this up. I have this yet to be an actual problem, ever. Are you guys that inept at copy and pasting?
If it's ever not perfect it gives a good time to scan the syntax and actually read what you're blindly copy and pasting in.
unlike other languages that don't care about white-space.
Yeah, Makefiles grok whitespace wonderfully.
Clear Case source control system has *massive* swaths of its codebase written in Perl.
For anyone that's used ClearCase that would probably double the amount of hate going towards Perl.
I wish you success in the market with that approach.
It's been great so far. Key is to focus in on a niche in industry and have the 'jack of most trades' approach to it.
You're conflating the language with a library.
No, I pointed out that once you want to move beyond some simple scripts, Python has those libraries while Perl does not.
And once you want to move beyond some simple automation scripts it has a much larger ecosystem than Perl.
You can learn Perl AND Python + libs, or you can just learn Python. It has the benefit of having a larger ecosystem and being readable.
If I wanted to make something that no one else would bother touching I'd write it in Perl.
For the most part Perl is being used as a wrapper for something else. It's a thin script to automate some bulk of work done *elsewhere* in the system.
I want to know where Perl beats Python hands down in performance that doesn't include extra libraries.
When your peers and successors only need to pick up C# to get their work done then C# is all that matters. With Perl being 'write only' that is a near impossible task. Given the choice of fixing existing Perl or re-writing it in Python (or even C# in your case) I'll always chose the latter.
I write wrappers like that all day long. Python wrapping C# for dSpace, ETAS and CRETA. Python wrapping C for Vector CANape. Python wrapping C for Matlab/Simulink. When my peers need to use any of those tools all they need to know is Python. There's even a chance they can fix or improve it. With Perl that's mostly not an option.
This is the PhD gamble. You hope that you learn enough and live long enough for your cutting edge research to find a practical purpose.
Back in 2004 DARPA sponsored a 'small' project to drive cars autonomously. Lots of companies and schools threw warm bodies at the problem and for a few years it some of it was purely theoretical research.
Then it reached a tipping point that a profitable end was in sight.
Uber went in and cleaned out CMU's autonomous vehicle department.
That doesn't matter to the end user. The end user only needs to know Python to work with it.
TensorFlow begs to differ.