[fpc-pascal] FPC for Scratchbox-ARM (compiling with errors)
Hi EveryOne, I'm tryng to compiling fpc compiler and units under scratchbox for ARM I have crosscompile, with fpcarm in linux i386, fpcmake and pp for ARM, I have copy to scratchbox system and I have used both program to generate a complete build in scratchbox using # export FPC=pp # make CPU_TARGET=arm OS_TARGET=linux build but it's fail in math.pp compiling (division by zero error) . . . /trabajo/temp/pp -Ur -Ur -Xs -n -Fi../inc -Fi../arm -Fi../unix -Fiarm -FE. -FU/trabajo/fpcbuild_2.0.4_exp/fpcsrc/rtl/units/arm-linux -darm -dRELEASE -Sg ../objpas/typinfo.pp /trabajo/temp/pp -Ur -Ur -Xs -n -Fi../inc -Fi../arm -Fi../unix -Fiarm -FE. -FU/trabajo/fpcbuild_2.0.4_exp/fpcsrc/rtl/units/arm-linux -darm -dRELEASE ../objpas/math.pp math.pp(125,22) Fatal: Compilation aborted An unhandled exception occurred at $B7F0DC04 : EInvalidOp : Invalid floating point operation $B7F0DC04 $00087A00 $00087BAC $0013E458 $0014BF1C $0014C4C0 $000FBD4C $000F5F14 $000781A4 $00032BC8 $8368 make[8]: *** [math.ppu] Error 217 . . . any suggestion to resolve it? Jose Pascual ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] FPC for Scratchbox-ARM (compiling with errors)
Op Tue, 20 Feb 2007, schreef Jose Pascual: > Hi EveryOne, > > I'm tryng to compiling fpc compiler and units under scratchbox for ARM > > I have crosscompile, with fpcarm in linux i386, fpcmake and pp for ARM, I > have copy to scratchbox system > > and I have used both program to generate a complete build in scratchbox > using > > # export FPC=pp > # make CPU_TARGET=arm OS_TARGET=linux build > > but it's fail in math.pp compiling (division by zero error) Is /trabajo/temp/pp still an i386 compiler? What compiler version is it? A possible solution could be to cycle an i386 compiler first (include debug info with -g in case it'll have the same problem), and use that as bootstrap compiler. Daniël > /trabajo/temp/pp -Ur -Ur -Xs -n -Fi../inc -Fi../arm -Fi../unix -Fiarm -FE. > -FU/trabajo/fpcbuild_2.0.4_exp/fpcsrc/rtl/units/arm-linux -darm -dRELEASE > -Sg ../objpas/typinfo.pp > /trabajo/temp/pp -Ur -Ur -Xs -n -Fi../inc -Fi../arm -Fi../unix -Fiarm -FE. > -FU/trabajo/fpcbuild_2.0.4_exp/fpcsrc/rtl/units/arm-linux -darm -dRELEASE > ../objpas/math.pp > math.pp(125,22) Fatal: Compilation aborted > An unhandled exception occurred at $B7F0DC04 : > EInvalidOp : Invalid floating point operation > $B7F0DC04 > $00087A00 > $00087BAC > $0013E458 > $0014BF1C > $0014C4C0 > $000FBD4C > $000F5F14 > $000781A4 > $00032BC8 > $8368 > > make[8]: *** [math.ppu] Error 217 > . > . > . > > > any suggestion to resolve it? > > > Jose Pascual > > > > ___ > fpc-pascal maillist - fpc-pascal@lists.freepascal.org > http://lists.freepascal.org/mailman/listinfo/fpc-pascal > ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
RE: [fpc-pascal] FPC for Scratchbox-ARM (compiling with errors)
> > Hi EveryOne, > > > > I'm tryng to compiling fpc compiler and units under scratchbox for > ARM > > > > I have crosscompile, with fpcarm in linux i386, fpcmake and pp for > > ARM, I have copy to scratchbox system > > > > and I have used both program to generate a complete build in > > scratchbox using > > > > # export FPC=pp > > # make CPU_TARGET=arm OS_TARGET=linux build > > > > but it's fail in math.pp compiling (division by zero error) > > Is /trabajo/temp/pp still an i386 compiler? What compiler version is > it? /trabajo/temp/pp is a ARM binary. version is 2.0.4 #/trabajo/temp/pp Free Pascal Compiler version 2.0.4 [2007/02/19] for arm Copyright (c) 1993-2006 by Florian Klaempfl . . . #readelf -h /trabajo/temp/pp ELF Header: Magic: 7f 45 4c 46 01 01 01 61 00 00 00 00 00 00 00 00 Class: ELF32 Data: 2's complement, little endian Version: 1 (current) OS/ABI:ARM ABI Version: 0 Type: EXEC (Executable file) Machine: ARM Version: 0x1 Entry point address: 0x80c0 Start of program headers: 52 (bytes into file) Start of section headers: 1880176 (bytes into file) Flags: 0x2, has entry point, GNU EABI Size of this header: 52 (bytes) Size of program headers: 32 (bytes) Number of program headers: 3 Size of section headers: 40 (bytes) Number of section headers: 7 Section header string table index: 6 > A possible solution could be to cycle an i386 compiler first (include > debug info with -g in case it'll have the same problem), and use that > as bootstrap compiler. But I'm going to geneate i386 binary compiler and tools so I can't to execute in scratchbox. I think your suggestion is useful to generate precompiled units in i386 linux, but I'd like to build all fpc inside sratchbox using arm binary compiler and tools, a arm not-crosscompiling fpc system. Jose Pascual. > Daniël > > > > /trabajo/temp/pp -Ur -Ur -Xs -n -Fi../inc -Fi../arm -Fi../unix - > Fiarm -FE. > > -FU/trabajo/fpcbuild_2.0.4_exp/fpcsrc/rtl/units/arm-linux -darm > > -dRELEASE -Sg ../objpas/typinfo.pp /trabajo/temp/pp -Ur -Ur -Xs -n > > -Fi../inc -Fi../arm -Fi../unix -Fiarm -FE. > > -FU/trabajo/fpcbuild_2.0.4_exp/fpcsrc/rtl/units/arm-linux -darm > > -dRELEASE ../objpas/math.pp > > math.pp(125,22) Fatal: Compilation aborted An unhandled exception > > occurred at $B7F0DC04 : > > EInvalidOp : Invalid floating point operation > > $B7F0DC04 > > $00087A00 > > $00087BAC > > $0013E458 > > $0014BF1C > > $0014C4C0 > > $000FBD4C > > $000F5F14 > > $000781A4 > > $00032BC8 > > $8368 > > > > make[8]: *** [math.ppu] Error 217 > > . > > . > > . > > > > > > any suggestion to resolve it? > > > > > > Jose Pascual > > > > > > > > ___ > > fpc-pascal maillist - fpc-pascal@lists.freepascal.org > > http://lists.freepascal.org/mailman/listinfo/fpc-pascal > > > > __ Información de NOD32, revisión 2071 (20070220) __ > > Este mensaje ha sido analizado con NOD32 antivirus system > http://www.nod32.com ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] sqldb firebird on linux can't find client lib
The fact that it still reports it as 'not loaded', I conclude that the dlopen operation failed. Why this would be so I cannot say. Has found any hints about this, Michael? I'm sorry, but I had to give up on this because there are many other parts of the project which are also need to be written and finished. So I got to use FB 1.5.4 for the moment until this problem is found and fixed. :( -Bee- has Bee.ography at: http://beeography.wordpress.com ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
RE: [fpc-pascal] FPC for Scratchbox-ARM (compiling with errors)
Op Tue, 20 Feb 2007, schreef Jose Pascual: > /trabajo/temp/pp is a ARM binary. version is 2.0.4 > > #/trabajo/temp/pp > Free Pascal Compiler version 2.0.4 [2007/02/19] for arm > Copyright (c) 1993-2006 by Florian Klaempfl > . That is then the problem. You cannot cycle on ARM with a 2.0 compiler. With recent 2.1 this is possible. > > A possible solution could be to cycle an i386 compiler first (include > > debug info with -g in case it'll have the same problem), and use that > > as bootstrap compiler. > > But I'm going to geneate i386 binary compiler and tools so I can't to > execute in scratchbox. > > I think your suggestion is useful to generate precompiled units in i386 > linux, but I'd like to build all fpc > inside sratchbox using arm binary compiler and tools, a arm > not-crosscompiling fpc system. I recommend 2.1 for that, 2.0 is not realiable enough on ARM for that. Just to be clear: 2.1 needs to be bootstrapped with a 2.0 compiler, so you need to start with a 2.0, then bootstrap a 2.1 crosscompiler for ARM, and use that to bootstrap a native ARM compiler. Daniël___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] FPC for Scratchbox-ARM (compiling with errors)
Daniël Mantione schrieb: > > Op Tue, 20 Feb 2007, schreef Jose Pascual: > >> /trabajo/temp/pp is a ARM binary. version is 2.0.4 >> >> #/trabajo/temp/pp >> Free Pascal Compiler version 2.0.4 [2007/02/19] for arm >> Copyright (c) 1993-2006 by Florian Klaempfl >> . > > That is then the problem. You cannot cycle on ARM with a 2.0 compiler. > With recent 2.1 this is possible. > >>> A possible solution could be to cycle an i386 compiler first (include >>> debug info with -g in case it'll have the same problem), and use that >>> as bootstrap compiler. >> But I'm going to geneate i386 binary compiler and tools so I can't to >> execute in scratchbox. >> >> I think your suggestion is useful to generate precompiled units in i386 >> linux, but I'd like to build all fpc >> inside sratchbox using arm binary compiler and tools, a arm >> not-crosscompiling fpc system. > > I recommend 2.1 for that, 2.0 is not realiable enough on ARM for that. > Just to be clear: 2.1 needs to be bootstrapped with a 2.0 compiler, so you > need to start with a 2.0, then bootstrap a 2.1 crosscompiler for ARM, and > use that to bootstrap a native ARM compiler. The 2.1.x compiler has also a problem: make all doesn't work, only make all OPT=-O- -Ooregvar I tried to debug it for several hours but I didn't find what's wrong :( ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
[fpc-pascal] Pascal is alive!!??
In the ACM-ICPC International Collegiate Programming Contest dropped pascal for this year final competitions and for next year, the students are not allowed to use pascal as their programming language. Did you know this? What is your idea? Also, in TopCoder site, you can not develop your programs with Pascal while as I know, Pascal has all the properties that C++ or JAVA have. Any idea? Is there any technical problem with Pascal to be used in TopCoder's contest? ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Pascal is alive!!??
Op Tue, 20 Feb 2007, schreef [EMAIL PROTECTED]: > In the ACM-ICPC International Collegiate Programming Contest dropped pascal > for this year final competitions and for next year, the students are not > allowed to use pascal as their programming language. Did you know this? What > is your idea? Also, in TopCoder site, you can not develop your programs with > Pascal while as I know, Pascal has all the properties that C++ or JAVA have. > > Any idea? Is there any technical problem with Pascal to be used in TopCoder's > contest? I saw this coming for a few years. Free Pascal has most definately been able to delay this for a few years, it was the de facto standard in competitions in recent years. But the trend was there. When I participated myself the amount of people using Pascal was in decline, usually in the end it were two teams of 50 or so who used Pascal. Note that only the *requirement* to offer a Pascal environment has been scrapped, organisers of contests can still provide it if they want. So, please lobby at the contest organisation for a Pascal environment. The only fix here is to strengthen our position in education. Most people today participate in Java, which is silly as it puts you in a clear disadvantage. I once submitted the first non-Java implementation of a problem after an hour into the contest. The reference implementation of the jury took 10 minutes and 60 MB of memory. My Pascal implementation gave the result instantly, while using 300 kilobytes of memory. The jury was totally blown away; after the contest we did investigate and it turned out that Pascal text I/O versus Java text I/O was 100% responsible for the difference. In short, Pascal still rocks in contests. One thing is very important: a rock solid text mode IDE under both Windows and Linux. This makes a difference in a contest. FPC has never been able to live up to the Turbo Pascal level here. I'd say especially the IDE in Linux was only useable for people knowledged with FPC to work around the limitations. Perhaps we should ask Tom Verhoeff, he is very involved in the icpc's. Daniël___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Pascal is alive!!??
What about Topcoder's and its Arena? Daniël Mantione wrote: Op Tue, 20 Feb 2007, schreef [EMAIL PROTECTED]: In the ACM-ICPC International Collegiate Programming Contest dropped pascal for this year final competitions and for next year, the students are not allowed to use pascal as their programming language. Did you know this? What is your idea? Also, in TopCoder site, you can not develop your programs with Pascal while as I know, Pascal has all the properties that C++ or JAVA have. Any idea? Is there any technical problem with Pascal to be used in TopCoder's contest? I saw this coming for a few years. Free Pascal has most definately been able to delay this for a few years, it was the de facto standard in competitions in recent years. But the trend was there. When I participated myself the amount of people using Pascal was in decline, usually in the end it were two teams of 50 or so who used Pascal. Note that only the *requirement* to offer a Pascal environment has been scrapped, organisers of contests can still provide it if they want. So, please lobby at the contest organisation for a Pascal environment. The only fix here is to strengthen our position in education. Most people today participate in Java, which is silly as it puts you in a clear disadvantage. I once submitted the first non-Java implementation of a problem after an hour into the contest. The reference implementation of the jury took 10 minutes and 60 MB of memory. My Pascal implementation gave the result instantly, while using 300 kilobytes of memory. The jury was totally blown away; after the contest we did investigate and it turned out that Pascal text I/O versus Java text I/O was 100% responsible for the difference. In short, Pascal still rocks in contests. One thing is very important: a rock solid text mode IDE under both Windows and Linux. This makes a difference in a contest. FPC has never been able to live up to the Turbo Pascal level here. I'd say especially the IDE in Linux was only useable for people knowledged with FPC to work around the limitations. Perhaps we should ask Tom Verhoeff, he is very involved in the icpc's. Daniël ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Pascal is alive!!??
On 2/20/07, Daniël Mantione <[EMAIL PROTECTED]> wrote: Op Tue, 20 Feb 2007, schreef [EMAIL PROTECTED]: > In the ACM-ICPC International Collegiate Programming Contest dropped pascal > for this year final competitions and for next year, the students are not > allowed to use pascal as their programming language. Did you know this? What > is your idea? Also, in TopCoder site, you can not develop your programs with > Pascal while as I know, Pascal has all the properties that C++ or JAVA have. > > Any idea? Is there any technical problem with Pascal to be used in TopCoder's > contest? This is sad because Object Pascal is a nice language and fpc is a great compiler. Most C programs can be translated without too much trouble. And ofcourse, things like strings are so easy. In short, Pascal still rocks in contests. One thing is very important: a rock solid text mode IDE under both Windows and Linux. This makes a difference in a contest. FPC has never been able to live up to the Turbo Pascal level here. Editors like (g)Vim and (X)Emacs can be easily customized to invoke the compiler. With fpc, you don't need a makefile as the compiler is so fast. So, all is required is a keybinding to fire up the compiler. I'd say especially the IDE in Linux was only useable for people knowledged with FPC to work around the limitations. If you are talking about the debugger, then I guess you have a point but with Object Pascal, the need for requiring the services of a debugger is much less say when using C or C++. But then, Emacs comes with an interface to gdb and its delphi-mode is good too. Cheers, Krishna -- We will never run out of things to program as long as there is a single program around. - Alan Perlis ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Pascal is alive!!??
Op Tue, 20 Feb 2007, schreef [EMAIL PROTECTED]: > What about Topcoder's and its Arena? Well, I don't know anything about a Topcoders contest. Daniël___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
[fpc-pascal] Re: Pascal is alive!!??
Note that only the *requirement* to offer a Pascal environment has been scrapped, organisers of contests can still provide it if they want. So, please lobby at the contest organisation for a Pascal environment. I did lobby them since the first time they had intention to drop pascal, but they simply didn't listen to me. :P The regular online contest still allows pascal though, they just don't want to see pascal on the final stage. Another thing that made me disappointed is they still use the old FPC v.1.0.6! They didn't care even when I told them to upgrade to FPC v.2.0.2 (the last version available at the time). :( The only fix here is to strengthen our position in education. Most people today participate in Java, which is silly as it puts you in a clear disadvantage. I once submitted the first non-Java implementation of a problem after an hour into the contest. The reference implementation of the jury took 10 minutes and 60 MB of memory. My Pascal implementation gave the result instantly, while using 300 kilobytes of memory. The jury was totally blown away; after the contest we did investigate and it turned out that Pascal text I/O versus Java text I/O was 100% responsible for the difference. Absolutely correct! The funny thing is, most of top rank solutions are done using C or C++. Very little of them is using Java, but still they insist to allow Java! Pascal didn't do as much as C or C++ (yet) just because pascal isn't very popular in their community. But if we want to involve ourselves in, I believe pascal is able to beat C or C++, as some of my friends (from Indonesia) did it. In short, Pascal still rocks in contests. One thing is very important: a rock solid text mode IDE under both Windows and Linux. This makes a difference in a contest. FPC has never been able to live up to the Turbo Pascal level here. I'd say especially the IDE in Linux was only useable for people knowledged with FPC to work around the limitations. FPC text mode IDE is indeed very good and easy to use for common programming purposes, both on win32 console and linux terminal. I use it all the time when I code on remote servers. Of source you can't expect it to be as complete as or as easy as GUI IDE like Lazarus since it works on text mode, but it does the best thing it could for a text based IDE. -Bee- has Bee.ography at: http://beeography.wordpres.com ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Re: Pascal is alive!!??
Op Tue, 20 Feb 2007, schreef [EMAIL PROTECTED]: > > Note that only the *requirement* to offer a Pascal environment has been > > scrapped, organisers of contests can still provide it if they want. So, > > please lobby at the contest organisation for a Pascal environment. > > I did lobby them since the first time they had intention to drop pascal, but > they simply didn't listen to me. :P The regular online contest still allows > pascal though, they just don't want to see pascal on the final stage. Another > thing that made me disappointed is they still use the old FPC v.1.0.6! They > didn't care even when I told them to upgrade to FPC v.2.0.2 (the last version > available at the time). :( Well, I still have many contacts in the ICPC world. It can still use them, but I need backup, since I already lobbied hard in the past. So please continue the lobby. Each year, also for the finals, each contest orginaising comittee decides what languages to offer. So, each year is a new opportinity to convince people to offer Pascal. > > The only fix here is to strengthen our position in education. Most people > > today participate in Java, which is silly as it puts you in a clear > > disadvantage. I once submitted the first non-Java implementation of a > > problem after an hour into the contest. The reference implementation of > > the jury took 10 minutes and 60 MB of memory. My Pascal implementation > > gave the result instantly, while using 300 kilobytes of memory. The jury > > was totally blown away; after the contest we did investigate and it > > turned out that Pascal text I/O versus Java text I/O was 100% responsible > > for the difference. > > Absolutely correct! The funny thing is, most of top rank solutions are done > using C or C++. Very little of them is using Java, but still they insist to > allow Java! Pascal didn't do as much as C or C++ (yet) just because pascal > isn't very popular in their community. But if we want to involve ourselves in, > I believe pascal is able to beat C or C++, as some of my friends (from > Indonesia) did it. Yes, Pascal perfectly suitable for contest, and due to lower LOC counts preferable over C++. A problem is the STL though, C++ coders can sometimes take shortcuts here; while juries somehow seem tempted to forbid the use of units in Pascal (an issue I made myself angry about many times in the past). > > In short, Pascal still rocks in contests. One thing is very important: a > > rock solid text mode IDE under both Windows and Linux. This makes a > > difference in a contest. FPC has never been able to live up to the Turbo > > Pascal level here. I'd say especially the IDE in Linux was only useable > > for people knowledged with FPC to work around the limitations. > > FPC text mode IDE is indeed very good and easy to use for common programming > purposes, both on win32 console and linux terminal. I use it all the time when > I code on remote servers. Of source you can't expect it to be as complete as > or as easy as GUI IDE like Lazarus since it works on text mode, but it does > the best thing it could for a text based IDE. In the past, Turbo Pascal gave people the biggest advantage, since you develop and debug your program a lot faster than with gcc or Java. FPC has never given these benefits. I've used the FPC IDE in the past, but it always working around the limitations of FPC IDE in xterm. Debugging with gdb wasn't worse than using the FPC IDE. This affects the decisions for teams what language to use. Lazarus, with all its RAD stuff, is terribly in the way. I have done two contests with Delphi as IDE, and while it works very well, for this kind of development an IDE that is focussed just code really helps. Further, few contest organizers have been installing Delphi in the past, but many installed the text mode IDE, so this is to focus on. Daniël___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] sqldb firebird on linux can't find client lib
Am Dienstag, den 20.02.2007, 19:16 +0700 schrieb Bisma Jayadi: > > The fact that it still reports it as 'not loaded', I conclude that the > > dlopen operation failed. Why this would be so I cannot say. > > Has found any hints about this, Michael? I'm sorry, but I had to give up on > this > because there are many other parts of the project which are also need to be > written and finished. So I got to use FB 1.5.4 for the moment until this > problem > is found and fixed. :( Try recompiling the client yourself, do not use binary prepacked things. As a last resort you could try using the libraries contained in the server package of FB and install these manually on the client. HTH, Marc ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
[fpc-pascal] Re: Pascal is alive!!??
Well, I still have many contacts in the ICPC world. It can still use them, but I need backup, since I already lobbied hard in the past. So please continue the lobby. Each year, also for the finals, each contest orginaising comittee decides what languages to offer. So, each year is a new opportinity to convince people to offer Pascal. I'll try but can't promise you the result as I think those people -mostly- underestimate pascal. You know, very C/C++ programmers typical. They just simply ignore what pascal can do nowadays. :P Yes, Pascal perfectly suitable for contest, and due to lower LOC counts preferable over C++. A problem is the STL though, C++ coders can sometimes take shortcuts here; while juries somehow seem tempted to forbid the use of units in Pascal (an issue I made myself angry about many times in the past). Agree. I think they should allow common/standard modern pascal units, such as sysutils and classes. Without them, we almost need to rewrite all usable classes and functions/procedures by ourselves. In the past, Turbo Pascal gave people the biggest advantage, since you develop and debug your program a lot faster than with gcc or Java. FPC has never given these benefits. I've used the FPC IDE in the past, but it always working around the limitations of FPC IDE in xterm. Debugging with gdb wasn't worse than using the FPC IDE. This affects the decisions for teams what language to use. Ah, the debugger! You have a point here. I forgot to mention it because I rarely use it. :D Lazarus, with all its RAD stuff, is terribly in the way. I have done two contests with Delphi as IDE, and while it works very well, for this kind of development an IDE that is focussed just code really helps. Further, few contest organizers have been installing Delphi in the past, but many installed the text mode IDE, so this is to focus on. Just forget the GUI IDE, programming contests don't need a bloated hard to configure IDE. No offense for Lazarus. But IMO a GUI IDE is just overkill for algorithm/logic programming purposes. A fully working text mode IDE is more than enough. -Bee- has Bee.ography at: http://beeography.wordpress.com ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
[fpc-pascal] Re: sqldb firebird on linux can't find client lib
Try recompiling the client yourself, do not use binary prepacked things. I did, but the problem still persists. :( As a last resort you could try using the libraries contained in the server package of FB and install these manually on the client. I did it as well. FYI, the client, the server, and the app is on the same machine, at least on testing stage. -Bee- has Bee.ography at: http://beeography.wordpress.com ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Re: sqldb firebird on linux can't find client lib
Am Dienstag, den 20.02.2007, 23:51 +0700 schrieb [EMAIL PROTECTED]: > > Try recompiling the client yourself, do not use binary prepacked things. > > I did, but the problem still persists. :( > > > As a last resort you could try using the libraries contained in the > > server package of FB and install these manually on the client. > > I did it as well. FYI, the client, the server, and the app is on the same > machine, at least on testing stage. Sounds fishy. Are you sure the access rights for the user running your app are okay (mode 755 on the libs)? Or the startup scripts for the user (.login, .profile, ...) modify the library search path? I can't imagine anything else ... something must be different in installing client or server that stops your program from working correctly ... you've got to sort that out. Running out of ideas, Marc ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
[fpc-pascal] Advantages of Pascal
Dear FreePascal adepts, when I was younger (about seven years ago), I wrote several programs in Borland's Turbo Pascal version 7 (IIRC). I then learned C and haven't looked much at Pascal ever since. However, I'm still fond of Pascal's clear syntax, so I would like to know more about its advantages. I would be very grateful if you could answer the following questions, loosely based on the "Advantages" section of the FreePascal homepage. 1) "No Makefiles" ...and yet you ship a Makefile generator with FPC. Why? 2) "Pascal compilers are Fast with a big F [...]" Why is that? 3) "Great integration with assembler" Is being able to use Assembler easily inline really that much better as opposed to having the code in separate object files? 4) "Smartlinking" Ah, I think you can have that one in C with some obscure linker flags, too. 5) Why use Pascal instead of Ada? 6) Can I have run-time range checking of arrays/strings and *still* use functions that accept arrays of any length? Thank you so much in advance for helping me. Leslie -- "Feel free" - 5 GB Mailbox, 50 FreeSMS/Monat ... Jetzt GMX ProMail testen: www.gmx.net/de/go/mailfooter/promail-out ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Advantages of Pascal
Patrick Leslie Polzer schrieb: Dear FreePascal adepts, when I was younger (about seven years ago), I wrote several programs in Borland's Turbo Pascal version 7 (IIRC). I then learned C and haven't looked much at Pascal ever since. However, I'm still fond of Pascal's clear syntax, so I would like to know more about its advantages. I would be very grateful if you could answer the following questions, loosely based on the "Advantages" section of the FreePascal homepage. 1) "No Makefiles" ...and yet you ship a Makefile generator with FPC. Why? You're project don't need them. But the fpc libraries are something very complex being build for different target from the same source, containing also informations how to install things, build releases etc. But we're just getting rid of make :) 2) "Pascal compilers are Fast with a big F [...]" Why is that? Module support, properly designed language. 3) "Great integration with assembler" Is being able to use Assembler easily inline really that much better as opposed to having the code in separate object files? Yes, you don't need an external assembler. You can access parameters literally etc. 4) "Smartlinking" Ah, I think you can have that one in C with some obscure linker flags, too. Not as easy as in pascal. 5) Why use Pascal instead of Ada? Why Ada and not Pascal? 6) Can I have run-time range checking of arrays/strings and *still* use functions that accept arrays of any length? Yes. Thank you so much in advance for helping me. Leslie ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Advantages of Pascal
On 20 feb 2007, at 19:46, Patrick Leslie Polzer wrote: 1) "No Makefiles" ...and yet you ship a Makefile generator with FPC. Why? Because you cannot tell the compiler to compiler run time library, then itself with the rtl that was just compiled, then recompile the run time library with the newly compiled compiler, then itself with the newly compiled compiler etc. fpcmake is a tool developed for internal use, but everyone is free to use it. You don't strictly need it except in complex cases like the above, though. 2) "Pascal compilers are Fast with a big F [...]" Why is that? Because you do not need a preprocessor pass, and compiled unit interfaces can be reused much more often than e.g. C pre-compiled headers because no external defines can influence them (and therefore you also need much less checking to see whether or not a unit needs to be recompiled). 3) "Great integration with assembler" Is being able to use Assembler easily inline really that much better as opposed to having the code in separate object files? Personal preference. 4) "Smartlinking" Ah, I think you can have that one in C with some obscure linker flags, too. Nowadays you can indeed, yes. 5) Why use Pascal instead of Ada? Personal preference. 6) Can I have run-time range checking of arrays/strings and *still* use functions that accept arrays of any length? Yes. Jonas ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Advantages of Pascal
Op Tue, 20 Feb 2007, schreef Patrick Leslie Polzer: > Dear FreePascal adepts, > > when I was younger (about seven years ago), I wrote several programs > in Borland's Turbo Pascal version 7 (IIRC). > > I then learned C and haven't looked much at Pascal ever since. > However, I'm still fond of Pascal's clear syntax, so I would like > to know more about its advantages. Please try it. There is no better way to learn about its advantages than actual programming. This can better explain it to you than I can do in 1000 words. For many C programmers, getting into Free Pascal is a very enjoyable experience. > I would be very grateful if you could answer the following questions, > loosely based on the "Advantages" section of the FreePascal homepage. > > > 1) "No Makefiles" > ...and yet you ship a Makefile generator with FPC. Why? Since the compiler is able to find the dependencies of files automatically, do this is not something you need a Makefile for. However, there are several other things, like creating a tar or zip of your program, installing it and some more things. A makefile is one way to provide this functionality, which we make easy and user friendly by providing a makefile generator. Still we are going to kick fpcmake, since makefiles are hard to port. As C programmer you probably know this, to be able to execute a configure script on Windows you need an Unix emulator like Cygwin or Msys. Building the compiler using fpcmake on Windows requires people to several Unix utilities, like cp, mv. In near the future we will move to a build tool which doesn't depend on external tools to copy files etc., increasing the portability of the compiler. > 2) "Pascal compilers are Fast with a big F [...]" > Why is that? The main reason for this is the unit system. If a C program uses a library, all header files need to be loaded and parsed, for each source file again. A Pascal compiler can simply load a unit from disk, so it doesn't need to parse headers. Further, once loaded, the unit remains in memory, so if the next file is compiled, it is ready to be used again. Some C-compilers provide pre-compiled headers to provide some of the same advantages, but since the content of a header can change with a simple define, this is limited; until now this has not been very successfull in obtaining similar performance to Pascal compilers. > 3) "Great integration with assembler" > Is being able to use Assembler easily inline really that much > better as opposed to having the code in separate object files? Assembler at the Pascal level frees you from the trouble to access parameters. If you would write pure assembler you need to work with offsets on the stack that need to match what the compiler uses. This is very error prone. Further, you can access any Pascal symbol directly from your assembler routines, which saves you all kinds of cubersome imports in your assembler files. > 4) "Smartlinking" > Ah, I think you can have that one in C with some obscure linker > flags, too. Recently it has become possible to do real smartlinking with gcc. FPC has done this since the late 90's; while gcc could only remove unused files, FPC could remove individual variables and procedures. Still, it works better. Try to statically link a hello world in C and then smartlink a hello word in FPC. Compare the executables. > 5) Why use Pascal instead of Ada? Ada is a good language. Pascal is good language too, and modern Object Pascal might be even more advanced than Ada. For certain is that Pascal has better development tools, a much larger treasury chest of existing code and libraries and is much better documented due to the libraries of book been written on it. > > 6) Can I have run-time range checking of arrays/strings and *still* > use functions that accept arrays of any length? Yes. We call this open arrays. > Thank you so much in advance for helping me. You're welcome :) Daniël Mantione___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Wikies
Hello, I implemented a trivial solution for this. Take a look at: http://wiki.freepascal.org/New_Main_Page I simply moved free pascal stuff to a new page on http://wiki.freepascal.org/Free_Pascal_Documentation And added a big entry for this on the main page. So, what do you guys think? Can I move that to the main page? I think that even my trivial solution is much better then the current solution: No links to fpc stuff at all. -- Felipe Monteiro de Carvalho ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Wikies
Felipe Monteiro de Carvalho schrieb: Hello, I implemented a trivial solution for this. Take a look at: http://wiki.freepascal.org/New_Main_Page I simply moved free pascal stuff to a new page on http://wiki.freepascal.org/Free_Pascal_Documentation And added a big entry for this on the main page. So, what do you guys think? Can I move that to the main page? I think that even my trivial solution is much better then the current solution: No links to fpc stuff at all. I added the important fpc stuff already to the main page? ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Wikies
On 2/20/07, Florian Klaempfl <[EMAIL PROTECTED]> wrote: I added the important fpc stuff already to the main page? Sorry, I don´t understand what you mean. Can you explain better? I just went to the main page and couldn´t find fpc things: http://wiki.freepascal.org/ -- Felipe Monteiro de Carvalho ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Wikies
Felipe Monteiro de Carvalho schrieb: On 2/20/07, Florian Klaempfl <[EMAIL PROTECTED]> wrote: I added the important fpc stuff already to the main page? Sorry, I don´t understand what you mean. Can you explain better? I just went to the main page and couldn´t find fpc things: http://wiki.freepascal.org/ ? Just read the about section? ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
[fpc-pascal] open array and 2.1.1
Hello, the following code: program testarray; uses classes; procedure testparam(aarray: array of tclass); begin end; begin testparam([nil, tlist, tstringlist]); end. compiles under 2.0.4 but fails under 2.1.1 (rev 6583), the compiler think that I have an array of const because of the nil argument. Known problem? Something that was changed? Thanks, -- Joao Morais ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] Wikies
On 2/20/07, Florian Klaempfl <[EMAIL PROTECTED]> wrote: ? Just read the about section? Ah,. sorry =) -- Felipe Monteiro de Carvalho ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
