non portable - a problem with direct access
in [Fortran]
|
Prev: McFor: a MATLAB-to-Fortran 95 Compiler
Next: 100% free dell laptaps offerd by dell and Vodafone companies.
From: Clive Page on 30 Jul 2010 17:32 In message <d2e54e69-e8c3-4ff5-8969-d8dde5e6bf8c(a)f42g2000yqn.googlegroups.com>, e p chandler <epc8(a)juno.com> writes >The keys, the offsets and all the data are in chunks of 32 bit words. >The design spec was Fortran 77, so I used unformatted direct access. >Writing for gfortran, I set RECL=4. The program works just fine. I This cautionary tale is yet another reason for NOT using Fortran77. Had Fortran90 (or later) been used, you can use an INQUIRE by IOLENGTH which solves this problem and allows you to program direct-access unformatted I/O in a completely portable fashion. Given the number of good compilers for Fortran90/95, many of them free, and the fact that they all accept Fortran77 code (since it is for practical purposes a subset), I really do not understand the continued use of an obsolete version of the language. But we see that in this and in so many other posts on this newsgroup. Is there some special feature of Fortran77 that has completely escaped me that makes it so attractive? -- Clive Page
From: Sjouke Burry on 30 Jul 2010 18:51 Clive Page wrote: > In message > <d2e54e69-e8c3-4ff5-8969-d8dde5e6bf8c(a)f42g2000yqn.googlegroups.com>, e p > chandler <epc8(a)juno.com> writes >> The keys, the offsets and all the data are in chunks of 32 bit words. >> The design spec was Fortran 77, so I used unformatted direct access. >> Writing for gfortran, I set RECL=4. The program works just fine. I > > This cautionary tale is yet another reason for NOT using Fortran77. Had > Fortran90 (or later) been used, you can use an INQUIRE by IOLENGTH which > solves this problem and allows you to program direct-access unformatted > I/O in a completely portable fashion. > > Given the number of good compilers for Fortran90/95, many of them free, > and the fact that they all accept Fortran77 code (since it is for > practical purposes a subset), I really do not understand the continued > use of an obsolete version of the language. But we see that in this and > in so many other posts on this newsgroup. Is there some special feature > of Fortran77 that has completely escaped me that makes it so attractive? > A working build in graphics interface for example? (MS version 5.1 with graphics.lib) And simple enough to interface svgacc.lib(c/c++) to it?
From: e p chandler on 30 Jul 2010 19:42 <nmm1(a)cam.ac.uk> wrote in message news:i2ve2j$4av$1(a)smaug.linux.pwf.cam.ac.uk... > In article <8bgr7sFtqoU1(a)mid.individual.net>, > Steve Lionel <steve.lionel(a)intel.invalid> wrote: >>On 7/30/2010 3:56 PM, GaryScott wrote: >>> This is deja vu all over again...how many times does this >>> happen...maybe the IVF manual should put this in 36 point font :) >> >>If users actually read the manual, and thought to look at the proper >>section, that might help. However, this gives me an idea as part of a >>"new user's FAQ" I am working on. > > There is a strong negative correlation between the people who most > need to read documentation and those that do so, and that has been > true for as long as I have been involved. > > That being said, this aspect of Fortran is a particular evil "gotcha". One problem with documentation is that when it reaches a certain size some things are "obvious" only when you know where they are. I remember not knowing a certain switch for gcc that changes the calling convention on Win32. I tip my hat to the person who pointed it out to me. But I _had_ read through the documentation for gcc several times, and what I needed was not at all obvious to me. It was not clear what was important or where I could find what I needed. [No this is not a criticism of gcc or all the hard work that has gone into it. But someone else must find the documentation somewhat daunting.]
From: Uno on 30 Jul 2010 19:53 e p chandler wrote: > > <nmm1(a)cam.ac.uk> wrote in message > news:i2ve2j$4av$1(a)smaug.linux.pwf.cam.ac.uk... >> In article <8bgr7sFtqoU1(a)mid.individual.net>, >> Steve Lionel <steve.lionel(a)intel.invalid> wrote: >>> On 7/30/2010 3:56 PM, GaryScott wrote: >>>> This is deja vu all over again...how many times does this >>>> happen...maybe the IVF manual should put this in 36 point font :) >>> >>> If users actually read the manual, and thought to look at the proper >>> section, that might help. However, this gives me an idea as part of a >>> "new user's FAQ" I am working on. >> >> There is a strong negative correlation between the people who most >> need to read documentation and those that do so, and that has been >> true for as long as I have been involved. >> >> That being said, this aspect of Fortran is a particular evil "gotcha". > > One problem with documentation is that when it reaches a certain size > some things are "obvious" only when you know where they are. I remember > not knowing a certain switch for gcc that changes the calling convention > on Win32. I tip my hat to the person who pointed it out to me. But I > _had_ read through the documentation for gcc several times, and what I > needed was not at all obvious to me. It was not clear what was important > or where I could find what I needed. > [No this is not a criticism of gcc or all the hard work that has gone > into it. But someone else must find the documentation somewhat daunting.] > > > > I'd like to see gfortran.pdf have some screenshots and more examples. I think documentation is simply one of those things that implementors don't want to do so much, yet this scaffolding is necessary. -- Uno
From: nmm1 on 31 Jul 2010 04:32
In article <i2vo1h$m8f$1(a)news.eternal-september.org>, e p chandler <epc8(a)juno.com> wrote: > >One problem with documentation is that when it reaches a certain size some >things are "obvious" only when you know where they are. I remember not >knowing a certain switch for gcc that changes the calling convention on >Win32. I tip my hat to the person who pointed it out to me. But I _had_ read >through the documentation for gcc several times, and what I needed was not >at all obvious to me. It was not clear what was important or where I could >find what I needed. >[No this is not a criticism of gcc or all the hard work that has gone into >it. But someone else must find the documentation somewhat daunting.] I think that we were all speaking generically, and not in this particular case. In my fairly extensive experience, this particular aspect catches a LOT of people out - and I am one of the ones who was caught! Back in the early days of Fortran 77, true :-) And you are correct about the documentation, whether you mean the Fortran standard or gcc's. Even experts on the standard sometimes miss some aspects. But I can assure you that it is easy to follow compared to C, C++ or POSIX :-( Regards, Nick Maclaren. |