From: dpb on 21 May 2010 08:50 Rune Allnor wrote: .... > No flames in the anti-fortran discussions from my side. > It's usually enough to point out the facts. Suffice it > to say that such an obvious feature as dynamic memory > handling wasn't added to the language until the mid '90s, > some 40 years after the language was first developed. .... But the point is it has been as has much else and your arguments and "facts" are based on 30+ yr-old "data". --
From: dpb on 21 May 2010 08:55 Steve Amphlett wrote: .... > I really enjoy reading the anti-FORTRAN flame wars here. Of the FORTRAN > being written today, I wonder how much of it is simply adding features > to legacy code and how much of it is new code, using all the bells and > whistles added to it over the years to keep the name alive? In comp.lang.fortran, certainly a high percentage of threads are on new features by folks writing new code although much of what I know they're working on is long-term applications. Others don't provide sufficient background to know for certain the overall history of the code base under development. --
From: Rune Allnor on 21 May 2010 09:12 On 21 Mai, 14:50, dpb <n...(a)non.net> wrote: > Rune Allnor wrote: > > ... > > > No flames in the anti-fortran discussions from my side. > > It's usually enough to point out the facts. Suffice it > > to say that such an obvious feature as dynamic memory > > handling wasn't added to the language until the mid '90s, > > some 40 years after the language was first developed. > > ... > > But the point is it has been as has much else and your arguments and > "facts" are based on 30+ yr-old "data". The language is a hoch-poch of features ad hoc stuck together with string and rubber bands. That's the present-day fact. Compare it to a only sligthly younger language, C++, itself somewhere around 30 years old based on C, which is another 10 years older. C++ allows the programmer to not change the *language*, which is standardized, but by adding *libraries*. The skilled C++ programmers are able to essentially develop their own, application-specific, languages, that easily run way faster than anything fortran has to offer. The present trend in programming is to write human-understandable code. Only amateurs and fools think that source code is written once, never to be seen by human eyes again. The fact of life is that source code that was never intended as anything but a one-off is copied, amended, changed, and even evolves into huge systems. I can't find the references off the top of my head, but I have seen some claims that both the C programming language and the UNIX operating system started out the same was as MS-DOS and later windows; as quick'n dirty ad hoc hacks intended to just get something going. This is how the world *actually* works. The code snippet one hacks up today might be the basis for something you or somebody else expands on years from now. Accept that fact, and learn how to program accordingly. Forgetting all about fortran (except what PITA it is to work with) is a significant first step in that direction. Rune
From: dpb on 21 May 2010 09:33 Rune Allnor wrote: > On 21 Mai, 14:50, dpb <n...(a)non.net> wrote: >> Rune Allnor wrote: >> >> ... >> >>> No flames in the anti-fortran discussions from my side. >>> It's usually enough to point out the facts. Suffice it >>> to say that such an obvious feature as dynamic memory >>> handling wasn't added to the language until the mid '90s, >>> some 40 years after the language was first developed. >> ... >> >> But the point is it has been as has much else and your arguments and >> "facts" are based on 30+ yr-old "data". > > The language is a hoch-poch of features ad hoc stuck together > with string and rubber bands. That's the present-day fact. > > Compare it to a only sligthly younger language, C++, itself > somewhere around 30 years old based on C, which is another > 10 years older. C++ allows the programmer to not change the > *language*, which is standardized, but by adding *libraries*. > The skilled C++ programmers are able to essentially develop > their own, application-specific, languages, that easily run > way faster than anything fortran has to offer. > > The present trend in programming is to write human-understandable > code. Only amateurs and fools think that source code is written > once, never to be seen by human eyes again. The fact of life > is that source code that was never intended as anything but > a one-off is copied, amended, changed, and even evolves into > huge systems. I can't find the references off the top of my > head, but I have seen some claims that both the C programming > language and the UNIX operating system started out the same > was as MS-DOS and later windows; as quick'n dirty ad hoc hacks > intended to just get something going. > > This is how the world *actually* works. The code snippet one hacks > up today might be the basis for something you or somebody else > expands on years from now. > > Accept that fact, and learn how to program accordingly. > > Forgetting all about fortran (except what PITA it is to work with) > is a significant first step in that direction. > > Rune There are arguments between the details of specifications of language Standards as the shortcomings of current and proposed Fortran vis a vis C++ and other languages at clf fairly regularly and there are issues on both sides that have their own winners and losers. I don't believe there's any way one can make such generic claims of superiority of either. Certainly there's no reason one can not write at least as easily and probably more so human-readable Fortran as C++ -- it has all the structured programming constructs of the other; it is again that you're still ranting against the FORTRAN of F66 or thereabouts as opposed to that of today. --
From: Rune Allnor on 21 May 2010 10:04
On 21 Mai, 15:33, dpb <n...(a)non.net> wrote: > I don't believe there's any way one can make such generic claims of > superiority of either. No one have talked about *superiority*. What I have said is that fortran is *inferior* as it has been obsolete for the past 40 years. Rune |