Fly Away wrote:
> On Aug 22, 3:15 pm, "James Giles"
worldnet.att.net> wrote:...
>> I've written support for them. The code belongs to a former
>> employer, or I'd just post it. It's not hard. But - it's a library
>> issue, not a language issue. Or do you think libraries are
>> part of the language? That's *really* wierd.
>
> Libraries are a huge part of the reason why one language would be
> chosen over another.
And which library is it that Fortran can't call? None? That's
what I thought too.
>> I've not
>> recently found a Fortran from which I couldn't call such
>> library routines if I needed them. (You would probably be
>> surprised at how little value such things have for tasks such
>> as compiler construction.
> Not all of us write compilers. I just like the power of sed for
> various things, but I don't always have it installed.
Sed is not something I would use Fortran to do. 99.99%% of
all sed-like operations I do ar in a text editor. No language
involved at all. You can't oppose Fortran for things no one
recommends it for. That's silly.
>>> Task - file manipulation. Can you create a folder or copy a file in
>>> a portable standard conforming way?
>>
>> Not in Fortran. Not in any language.
> See "Scripting languages".
I've seen them. They aren't portable. The don't run at all on
systems that don't support the environment they assume. Were'nt
you the one that dismissed Fortran because it couldn't do things
portably?
Besides, no one uses Fortran as a scripting language. Again, you
can't criticize Fortran users (like you have repeatedly done) for
things that they aren't even doing.
>> The concept of "folder"
>> (or "directory", "path", etc) is not portable. Some systems haven't
>> got them. Some systems haven't got files at all.
> The concept of "folder" and "file" is way too common now to ignore it.
> How often does one write software in Fortran for such systems?
> Even Fortran reads from and writes to files. That's why you code:
> open(unit=100, FILE='file.txt')
And it's also why Fortran (and C, and C++, etc.) have explicit caveats
that the features involved aren't necessarily portable.
>> Never met a sorting algorithm I couldn't code *more* legibly in
>> Fortran.
> I consider coding something that's been implemented a million times
> before me by a lot smarter people a waste of time. See "reinventing
> the wheel". Besides, what are the chances my implementation could be
> better, if not much worse?
Which is also why I also use off the shelf components whenever
I can - and *from* Fortran, if that's the language I'm using. How
does this demonstrate and deficiency in Fortran?
...
>> I have no I idea what problem you're addressing. I use tools, not
>> languages, to write web pages.
>
> A language is a tool too.
It's a tool building tool, not a web page builting tool.
>> If you are
>> not presented with choices, that fact doesn't demonstrate
>> any weakness of the languages that you couldn't choose.
>
> If I am presented with a choice I will have to know beforehand what
> tool I have to use. I can't afford to find out the tool was not
> suitable for the task post factum. It actually has happened to me...
Wait a minute, you were the one that criticized Fortran users for
not moving to more modern things. Now you're recommending that
choices be disregarded that don't match your preconcieved notions?
And what do you do if *none* of the choices match you prior
experience? That happend a lot in computing.
>> You're the emotional one. I'm not particularly interested in C or
>> C++ anymore because I no longer use them. When I suppored
>> C compilers, I knew more about it than any three random users
>> you would care to name. If I have a disrespect for the language,
>> it's because I know it very well indeed.
>
> You don't have to be _interested_ in a language to avoid cursing it.
I don't have encyclopedic knowledge of the Harry Potter
universe: so what curse have I uttered? I have direct knowledge
of language design issues that I'm willing to share. C happens
to be a convenient counterexample to proper design principles
that I know sespecially well. I suppose other badly designed
languages would suit the purpose as well. I don't feel like
learning new bad languages just to have an additional counter-
example.
> Another point - I am not telling anyone what they should or shouldn't
> use. I have an opinion, I express it. You respond. Noone has to make
> decisions based on it. If I learn something useful I will consider it.
> I am completely open for new knowledge. It's just that I prefer a more
> peaceful discussions without personal references :)
Then you should adopt that attitude in your own articles.
>> Especially telling people that *have* successfully
>> written just such applications in Fortran. Many of whom know
>> and have used dozens of languages and who chose Fortran
>> because they found it *better* than the alternatives.
>
> That's the key point. They know it better, not because it's superior
> to all other languages.
What part of the preceeding paragraph didn't you read? I said
nothing about what language people knew better. In fact,
I said that many have used a large variety of languages. It
might be that I know C better that Fortran (I've certainly
spent more effort and time trying to interpret the standard).
I said that such people chose Fortran because they foud it
to be *better*, not because the found it more familiar. Many
people may do that too, but it's not valid to assume that
familiarity is the basis of their choice.
...
>> Programming languages are multiplex tools. Almost all of them have
>> the same collection of capabilities. Fortran and C, for example, are
>> practically isomorphic. The set of applications each is suited to is
>> the *same* set. What distinguishes them is which makes code easier
>> to write, read, verify, correct, extend, or maintain.
>
> That's why some languages become popular and widely used and others
> fade and become a small niche language no matter how perfect in theory
> they are. ...I am not pointing fingers here.
Popularity and quality are not in any way correlated in computing.
If you believe they are, you have to explain the popularity of
Windows.
>> There have been objective experiments to measure the effect of
>> language features on programmer productivity. (There haven't
>> been enough of them, and since the results usually gore some
>> important sacred cows they aren't widely reported. And people
>> with vested interests to protect lobby against funding them.
>
> So it's a conspiracy against Fortran, right?
So, *you're* the one that prefers "more peaceful discussions
without personal references". Sure. I think you may be the
guy trying to sell bridges in New York City we always hear
about. Sometimes you almost sound like you have serious
intent - then you resort to statements that almost justify
killfiling.
Make up your mind. Are you really interested in a discussion,
or are you really trolling. Stop pretending to both.
--
J. Giles
"I conclude that there are two ways of constructing a software
design: One way is to make it so simple that there are obviously
no deficiencies and the other way is to make it so complicated
that there are no obvious deficiencies."
-- C. A. R. Hoare
"Simplicity is prerequisite for reliability" -- E. W. Dijkstra
