|
|
I like that one better, too. But the one i sent applied better :P
|
|
|
|
|
You use the tools that fit the job you're trying to do. If you did try to make one tool to fit all problems (i.e. the web, AI, heavy calculations, intuative UI, etc) it would be large and unwieldy. And you would still have times when you need something else.
|
|
|
|
|
I'll follow up on your tool analogy with the way modern automobiles are constructed. Most, if not all, require the purchase of special tools to do certain jobs. No standard tool can do certain jobs because of location, space constraints, or just plain weird configurations - and this was done by design.
In my opinion (at least) this sucks.
It's ridiculous to create a tool for each job. Whether it be automotive repair or coding.
If you accept this concept, then where does one draw the line and say work a bit harder this once and be done with it.
There are an awful lot of languages out there that seem to be about as generally useful as a wrench for a 7-sided nut. A combination of a dust-collecting wrench and a *&$($& pain of a obstacle. Personally, I'd find it more sensible to get a 6-sided nut to replace it . . .
Should there be no more new languages. No - I'm not saying that. But making some subset of programming easier at the cost of making everything else about the project (maintenance and maintainers) more difficult isn't very good.
Real life example: There's a major app where I am, written in Visual FoxPro. It certainly made dbase access easier - managed all sorts of stuff - and is also incompatible with the universe, VFP will soon be deprecated, and except for the developer who created it, is barely maintainable by current staff. SO much for diversity. At least the developer has job security.
I'm not necessarily prepared to sketch the lines that mark the limits wherein one should use an existing language and outside of which one needs a new solution. The solution, though, can often be achieved by reframing the problem and using what one has. There are times where, as you stated,
Joe Q wrote: when you need something else.
They are relatively rare.
The current situation, where we have a Language-of-the-Month-Club scenario is neither useful nor acceptable.
"The difference between genius and stupidity is that genius has its limits." - Albert Einstein
"How do you find out if you're unwanted if everyone you try to ask tells you to stop bothering them and just go away?" - Balboos HaGadol
"It's a sad state of affairs, indeed, when you start reading my tag lines for some sort of enlightenment. Sadder still, if that's where you need to find it." - Balboos HaGadol
|
|
|
|
|
I agree with the concepts of what you are saying. There should not be a language of the month club as it seems like there is for web development.
<blockquote class="FQ"><div class="FQA">Balboos wrote:</div>The current situation, where we have a Language-of-the-Month-Club scenario is neither useful nor acceptable. </blockquote>
However, programs do need to be individually tailored for what is going on. I've done user interfaces where there's a lot of input validation, work with strings, things like that. I've also written missile guidance software that is very mathmatically intensive and dealt a lot with hardware interaction. Both are backed by different trains of thought, both needed differt tools.
It does seem, especially in web development, that a lot of languages could have been developed by simply adding extension to other languages. However, someone wouldn't get the credit for coming up with a new language if you did that.
Have a great day!
|
|
|
|
|
See Genesis, chapter 11, and remember ALGOL, ADA and many others.
|
|
|
|
|
|
I think we're rapidly approaching the point where the language itself is all but irrelevant. We're moving more and more towards calling into huge frameworks that pretty much cover everything that you could possibly wish for already. It won't be long before we start seeing what used to be entire apps in specific domains just being part of the library. For example if you need to add accounting functionality to a business program I can see a day where an entire accounting module will be part of the framework, or an inventory module etc etc. Increasing abstraction at higher and higher levels until we reach the point where a person with limited computer skills will buy a computer and drag and drop the chunks they need to make any kind of app for their needs.
I've been coding for a long time and we've long since reached the point where nothing new in hardware or programming technology is adding anything *truly* needed any more to satisfy real world requirements. The entire computer industry from hardware to software have been doing nothing but gilding the lilly since at least the last 10 years, arguably a lot longer since the first modern computer the Xerox Alto[^] was released in 1973 with a mouse and GUI and everything since has only been refinements of very little real world benefit.
A language is irrelevant to me these days, the underlying framework is *far* more important a factor in getting things done.
|
|
|
|
|
John C wrote: For example if you need to add accounting functionality to a business program I can see a day where an entire accounting module will be part of the framework, or an inventory module etc etc. Increasing abstraction at higher and higher levels until we reach the point where a person with limited computer skills will buy a computer and drag and drop the chunks they need to make any kind of app for their needs.
Why stop there? Why still require "programming" at all? IMHO, the leap from programmer-accessible to non-programmer-accessible is made when a task that formerly required basic programming skill no longer requires any: spreadsheets and GUIs date back decades, but those are just the big flashy examples; plenty of kids with zero programming experience out there putting together terrible - but complex - web pages for instance.
I hold that an API (framework/library/module whatever you want to call it) is merely a rest stop on the path toward full-blown application... and the next logical step between the two is a domain-specific language that does not require familiarity with its target language.
John C wrote: The entire computer industry from hardware to software have been doing nothing but gilding the lilly since at least the last 10 years, arguably a lot longer since the first modern computer the Xerox Alto[^] was released in 1973 with a mouse and GUI and everything since has only been refinements of very little real world benefit.
Sure - we got C in 1972 and GUIs in 1973; what more do we need? 
You're sounding like a crotchety old man... yeah, most of what we use today is built on foundations laid decades ago; so what? At some point, you've gotta move from revolution to refinement or you'll never have more than a rough approximation of what you really wanted.
|
|
|
|
|
A long time ago, it was said:*
"Everything which can be said has been said.
Everything which can be done has been done
And there's nothing new under the sun."
The question is - which of you two does this support?
* Ecclesiastes 1:9-14
"The difference between genius and stupidity is that genius has its limits." - Albert Einstein
"How do you find out if you're unwanted if everyone you try to ask tells you to stop bothering them and just go away?" - Balboos HaGadol
"It's a sad state of affairs, indeed, when you start reading my tag lines for some sort of enlightenment. Sadder still, if that's where you need to find it." - Balboos HaGadol
|
|
|
|
|
Folks,
IMHO, a good modern programming language can be learned in a month and a half by a person who already knows 3-4 languages (C/C++, Java, VB.NET, Pascal). Frameworks, architecture, component base is a different story but that's largely language-independent.
- Nick
|
|
|
|
|
I agree - it's the principles of programming that matter, not the specific expression. I only wish they would standardize a few things like how to create comments and end of lines.
Melting Away
www.deals-house.com
www.innovative--concepts.com
|
|
|
|
|
Nick Alexeev wrote: IMHO, a good modern programming language can be learned in a month and a half by a person who already knows 3-4 languages (C/C++, Java, VB.NET, Pascal)
Not fully. After 15 years and nearly 1,000,000 lines of code I have written I am still learning new concepts in C++.
Here is one concept I did not know after 14.5 years of C++ but now I use every day.
http://www.codeproject.com/Lounge.aspx?msg=2997457#xx2997457xx[^]
John
modified on Monday, April 13, 2009 12:22 PM
|
|
|
|
|
Right, off back to school to learn a truly different language then - I suspect you may form a different opinion if you try to learn, for example, Haskell.
On the other hand, you'll pick up lots of insight that'll carry over to the languages you already know.
|
|
|
|
|
Use the language that's most suited to the job at hand. Every language has it's strengths and it's weaknesses. Trying to make one-fits-all...
rgds /Jonas [using five languages in day to day work]
|
|
|
|
|
Jonas Hammarberg said: "Use the language that's most suited to the job at hand."
Absolutely agree. Plenty of languages to choose from nowadays, and each has its strong points. Besides that, the very idea of learning another one gives me heartburn.
|
|
|
|
|
But you can't learn them all 
|
|
|
|
|
The first book i ever read on programming languages (vs. one specific language) had a "tower of babel" drawn on the cover. That book is decades old now, and there must have been hundreds, if not thousands, of new languages created since it was published.
Build! Build! Ever higher...
|
|
|
|
|
I believe having thousands of computer languages is a huge detriment to innovation in the whole computer industry, just like I believe having 1000+ different variants of linux is to the linux os. I mean it causes programmers to waste time learning a new language that does basically the same thing and the wheel gets reinvented over and over again...
John
|
|
|
|
|
Yeah, right ...and while we're at it, how about elimination of all those different types of cars, wines, clothes...
In my 30-year career as a programmer, I have written production code in over a dozen assembly languages, and over 20 high-level languages. Often using two or more on a single project. That's because different languages have different tradeoffs, so that the "best" language depends on several factors -- like 1) what's available for the platform, and 2) what kind of problem is being solved. And sometimes 3) whether you can find programmers that know it (that's what's currently killing Forth and Delphi).
"Should we move to a single programming language?" Get real. This can't even be a serious question. Although one of the things I tell all of my students is that there is no such thing as a stupid question, I might consider this one an exception.
|
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
