From my experience with Stack Overflow: You get “rewarded” for the most simple questions answered, stuff so basic that it scream “Please learn the language before asking question”. That’s great for people who want to get these questions answered.
But I don’t have the time for that. I would rather answer some more complicated or at least arcane questions, but the system doesn’t let me, because I didn’t acquire enough credentials to reply.
And answers to questions of that category I tend to find via Web Searches on blogs.
It takes zero reputation to reply to any question. Try it yourself in an anonymous browser. I just did and yep still works. You don’t even need to log in.
(Now if you want to leave comments, that requires a tiny bit of rep, because the purpose of the system is structured Q&A not kibitzing from the peanut gallery…)
These criticisms would be a whole lot more credible if the people making them had literally any idea what they were talking about.
Which is flat out untrue. Try answering any Stack Overflow question yourself in an anonymous browser window. I just did. It worked. In fact that’s also how it worked on day one when we launched it in 2008.
I am fine with criticism of Stack Overflow, there is plenty to criticize. Nothing is so perfect it cannot be improved. But what grinds my gears is criticism based on that isn’t even correct. Such as the ↑ above.
There are questions and non-obvious errors in existing answers and their comments. I consider replying to those to be answers. I won’t quibble about the terminology and concede the point.
I will still not put my answers and and suggestions as a separate answer where I think it won’t be found by the commenter and I will not engage in like-farming just to gain credentials.
I was just describing my experience with Stack Overflow. If this is irrelevant because I’m an outlier, I don’t mind.
Not understanding as in “I can’t possibly answer those” or as in “What is the socialist-sub-conetxt of those queries, as defined in a Diruvian hermenteutic approach in a general Cartesian theology?”
Well large for me would be 100 megabytes of C and I would find it difficult to work on that scale in python. I have seen it tried, where we replaced a large C and Motif GUI with a new Java application and subverted the type system by putting the domain data in <String, Object> maps. It was a disaster. For large scale systems you need strong typing to hold it together and python doesn’t really have that.
I think it’s more because you can easily write obfuscated code other people hate to maintain.
There have been many times where I wrote maliciously clever lines only to delete them in favor of something longer, but far more readable. Because I think of the poor guy who will have to make some changes in the far future. Admittedly, I have only learned to do this because that poor guy was me two years later.
You can write new methods for things like numbers and arrays and even create subclasses of them in Ruby (as again, like in Smalltalk). You can’t in Python – if these things were true objects in Python you could.
WATFOR is short for Waterloo Fortran. You can think of it as sort of a Fortran 66 and it was a very common standard for scientific Fortran programming in the 70’s.
Of course, I was exposed to it in the 90’s, and was pretty horrified, but everyone kept telling me how big a leap it was from the original Fortran (Fortran 55?).
Is vi pronounced “vih” or “vee eye?” I worked with a guy who called it the former but I always heard it as the latter. Same guy preferred vi so much that he downloaded vim, which emulates that editing environment in Windows.
When I took 2 semesters of C programming (in 1994, and which to this day I’ve never used), I really felt at first like we had been thrown into the deep end: we not only had to learn C, but how to get around (at least a little bit) in Unix and use vi . The computer lab ran on AT&T Unix, using terminals (not PCs) which I’d figured, even then, were old.
(I’m sure there are infinite stories better than this but my $0.02; I figure this may have been on the tail end of people having to work that way)
It’s still the fastest* language for one-use CLI pipelines and quick text processing jobs with only one input and one output. (Perl’s better if you have more than one of each, though, and PHP is the fastest for web stuff.)
I rewrote an Active Directory db from linux using gawk, for example. ldapsearch pull 10,000 records | gawk change 'em | ldapmodify put them back.
I think you meant awk has an elegant and concise grammar except for that weirdness about using whitespace as a string concatenation operator.
Most profit-generating programs either translate data from one form to another, or interact with users on the web, or control physical processes making stuff. In the hands of an expert, the fastest and most cost-effective and secure way to get the job done is often GNU awk for the first case, PHP for the second, and pure C for the last.
But I don’t really give a damn, I use whatever’s on the machine in front of me. If I had to choose a language, I’d probably choose GNU awk, because it seems to have the smallest grammar of any readable language. But it’s not for everyone or everything.
* by “fastest” I mean the quickest to get written. “Code development time”.
I remember WATFOR and WATFIV. I haven’t thought of those names in years.
I originally learned Fortran 4. I used a Fortran 2 compiler in the very early microcomputer days. Fortran 2 was an earlier standard that they briefly went back to when writing a full Fortran 4 compiler from scratch was too hard. Any proper Fortran 2 programmers out there?
I first started on Algol 1900 in the sixties. Don’t see a lot of that around these days.
I learned Perl on my own about ten years ago (with the Llama Book) and I use it regularly for the administration of my machines, be it with one-liners or for structured scripting (basically, anytime a loop is needed). I’m fond of it, especially its regular expressions, but I understand how it can become a mess if one isn’t cautious.
When I was beginning, I wrote a CLI calculator. I still use it, but I haven’t added functionalities in years because the code is dirty. The problem is I had too many C coding habits back then, coding in pure Perl style would be more legible. I should start afresh someday.
that isn’t even correct. Such as the ↑ above.
If I had to choose a language, I’d probably choose GNU awk, because it seems to have the smallest grammar of any readable language. But it’s not for everyone or everything.