Gpc February 1997

gpc@gnu.de

12 participants
34 discussions

Re: GPC in an instructional envionment
by Peter Gerwinski 24 Feb '97

24 Feb '97

According to Phil Nelson: > > For those who really want some sort of a programming envoronment to go > with gpc and are running X, look at xwpe. (X window Programming Environment) > The environment is quite similar to the Turbo-Pascal environment. It also runs without X in an (n?)curses text environment. Those who are running DOS (or OS/2 DOS box or Linux DOSemu or similar things) we can also recommand the RHIDE which is even more similar to the Turbo-Pascal environment. Hope this helps, Peter Dipl.-Phys. Peter Gerwinski, Essen, Germany, free physicist and programmer peter.gerwinski(a)uni-essen.de - http://home.pages.de/~peter.gerwinski/ [970201] maintainer GNU Pascal - http://home.pages.de/~gnu-pascal/ [970125]

1 0

... and the next alpha ...
by Peter Gerwinski 24 Feb '97

24 Feb '97

Hello again! The next alpha GPC is on agnes, as source (diff against gpc-970215) and binaries for DOS (DJGPP) and Linux (elf). Changes from gpc-970219 to gpc-970221: * Pointers to functions did not survive transport through a GPI file. Now they do. * Now String variables can be now can be initialized with a String value, too - and not only `array of char' variables. :) * Sets of Chars work again. (Was broken in gpc-970215.) * Extended Pascal's "Index" function works again. (Was broken in gpc-970215.) Many thanks to everybody who helped me to find all these bugs. I am especially grateful to Miklos Cserzo, Scott A. Moore, and Sven Hilscher for their short, but very useful test programs. Peter Dipl.-Phys. Peter Gerwinski, Essen, Germany, free physicist and programmer peter.gerwinski(a)uni-essen.de - http://home.pages.de/~peter.gerwinski/ [970201] maintainer GNU Pascal - http://home.pages.de/~gnu-pascal/ [970125]

3 3

constant arrays of type string
by Sven＠rufus.central.de 23 Feb '97

23 Feb '97

Hello gpc, this is from info file gpc.8 but I can't compile this, I need string array constants: program str_arr; const MyStringsCount = 5; type Ident = string(20); var MyStrings : array [1..MyStringsCount] of Ident value [ 1:'EXPORT'; 2:'IMPLEMENTATION'; 3:'IMPORT'; 4:'INTERFACE'; 5:'MODULE']; begin end. ___, ((__ o ,____)) V E N I

1 0

New alpha & binaries
by Peter Gerwinski 19 Feb '97

19 Feb '97

Hello, everybody! I was asked for binaries of the GPC alpha version because some people (including myself :-) had trouble to compile it, especially on non-UNIX systems. I am now uploading binaries of the current alpha GPC for Linux (elf) and for DOS (DJGPP) to ftp://agnes.dida.physik.uni-essen.de/gnu-pascal/alpha/. The "current alpha GPC" means of course gpc-970219, not the obsolete and buggy gpc-970215. ;-) Changes from gpc-970215 to gpc-970219: * Everything which treads `Char' as an ordinal type (e.g. `ord', `min', `succ', `case') works again. (It was broken in gpc-970215.) * Initialized `Char' variables work again; `array of char' variables now can be initialized with a String value. :) * Chars above chr ( 127 ) did not work in some situations, e.g. as `case' labels. Now they do. * According to Extended Pascal, you can assign String constants to Chars. The empty String yields a space character. That's it for the moment. Yours, Peter Dipl.-Phys. Peter Gerwinski, Essen, Germany, free physicist and programmer peter.gerwinski(a)uni-essen.de - http://home.pages.de/~peter.gerwinski/ [970201] maintainer GNU Pascal - http://home.pages.de/~gnu-pascal/ [970125]

1 0

Changing e-mail address in the list
by Peter Gerwinski 18 Feb '97

18 Feb '97

According to adamss3(a)nationwide.com: > Does anyone know how I can get my old email ID dropped off of this list? Write to gpc-request(a)hut.fi. Since it is a human, no program, who takes care of the list, he will understand your request. > PS - please reply to email ID The e-mail reply did not get through, sorry. Hope this helps, Peter Dipl.-Phys. Peter Gerwinski, Essen, Germany, free physicist and programmer peter.gerwinski(a)uni-essen.de - http://home.pages.de/~peter.gerwinski/ [970201] maintainer GNU Pascal - http://home.pages.de/~gnu-pascal/ [970125]

1 0

None
by adamss3＠nationwide.com 17 Feb '97

17 Feb '97

X400-Originator: adamss3(a)lms01.ent.nwie.net X400-Recipients: gpc(a)hut.fi X400-Mts-Identifier: [/PRMD=NATIONWIDE/ADMD=IBMX400/C=US/;0012000001019004000004] X400-Content-Type: P2-1988 (22) Message-Id: <0012000001019004000004*@MHS> To: "gpc(a)hut.fi" <gpc(a)hut.fi> Date: Mon, 17 Feb 1997 12:19:36 -0500 Does anyone know how I can get my old email ID dropped off of this list? It receives internet mail, but cannot send anymore. Thanks. Steve PS - please reply to email ID

1 0

Re: gpc linking problem cd.
by Peter Gerwinski 17 Feb '97

17 Feb '97

According to Krzysztof Kwapien: > Reading specs from d:/djgpp/lib\specs.gpc > gpc version 1.2(2.7.2) That's it! You are using the beta version gpc-1.2(2.7.2) of GNU Pascal which had a bug concerning linking. If you upgrade to gpc-2.0(2.7.2.1), the bug will vanish. Please download and install ftp://kampi.hut.fi/jtv/gnu-pascal/djgpp/gpc20b.zip It might be necessary to upgrade the whole DJGPP package (or at least the binutils) from v2.0 to v2.01 - which might be reasonable anyway. The current version of RHIDE for instance is much more stable than that which was available in the days of gpc-1.2(2.7.2). Hope this helps, Peter Dipl.-Phys. Peter Gerwinski, Essen, Germany, free physicist and programmer peter.gerwinski(a)uni-essen.de - http://home.pages.de/~peter.gerwinski/ [970201] maintainer GNU Pascal - http://home.pages.de/~gnu-pascal/ [970125]

1 0

gpc linking problem
by Krzysztof Kwapien 17 Feb '97

17 Feb '97

I wrote the hello world program. begin writeln('Hello World'); end. I compiled it with gpc -o hello hello.pas and it Worked!! But when i tried to compile it with gpc -c hello.pas <- this worked gpc -o hello hello.o <- this crashed with a SIGSEGV ..... it crashed when linking. What's the problem? ( I tried the -v option, and it seems as if some of the temporary files, where not linked in ) Because RHIDE also does it in the same manner it crashes too Help: kkwapien(a)mimuw.edu.pl

2 2

Again: new alpha GPC on agnes
by Peter Gerwinski 15 Feb '97

15 Feb '97

Hello, GPC testers! ;-) Here is the next alpha version of GPC. You can download it at ftp://agnes.dida.physik.uni-essen.de/gnu-pascal/alpha/gpc-2.0-970215.diff.gz Although the changes since gpc-970207 are crucial they are almost invisible. What I did was to "sort" the different language dialects and to implement options which allow a clear distinction between them. For example, GPC now defaults to allow nested comments { like this one (* you see? *) }, whereas ISO Standard allows comments starting with `(*' and ending with '}'. For example, GPC now understands Borland-style character constants like `#10' (* as an alternative to `chr ( 10 )'; the notation `^J' is *not* yet implemented *). Data type like `Word' or `Pointer' are built into GPC now, but you still cannot `writeln' them. If you specify one of `--standard-pascal-level-0', `--standard-pascal', `--extended-pascal', or `--object-pascal', nested comments are switched off by default { does anybody really use comments starting with `(*' and ending with `}'? *), and all Borland extensions (including the `#' character constants and the above data types) which contradict ISO are switched off. When you combine the above switches with `--pedantic', even those extensions which do not contradict the standard trigger warnings, with `--pedantic-errors' they trigger error messages. To make things more symmetric, `--borland-pascal' yields the opposite effect. With this switch, warnings about a missing program header or misuse of "typed constants" are omitted; together with `--pedantic' you get warnings about Extended Pascal features not in Borland Pascal, and so on. I heavily modified the parsing of String-related stuff. Now everything is prepared for the implementation of general Schema types ... Now it's your turn: I would like to know how much is broken due to the above improvements. * Please try to recompile all your GNU Pascal programs with this alpha compiler. There should be no problems. * Please try to compile your Standard Pascal programs with GPC with and without the switches `--standard-pascal' and/or `--pedantic'. There should be less problems than before, and there should be less problems with `--standard-pascal' than without. * The same for not too sophisticated Extended Pascal programs. (For example, Schema types are still not implemented.) * The same for not too sophisticated Borland Pascal programs. There should be less problems with `--borland-pascal' than without. That's it. Please post your "everything works" reports to this list. (* But bug reports are welcome, too. :*) Yours, Peter Dipl.-Phys. Peter Gerwinski, Essen, Germany, free physicist and programmer peter.gerwinski(a)uni-essen.de - http://home.pages.de/~peter.gerwinski/ [970201] maintainer GNU Pascal - http://home.pages.de/~gnu-pascal/ [970125]

1 0

Re: Restrictions on packed types?
by Peter Gerwinski 10 Feb '97

10 Feb '97

According to Jakob Heinemann: > I read somewhere that GPC actually doesn't pack things in types which > are > declared as PACKED. Is this true? What about packed array's of char, > i.e. > non-varying strings? It is true, but it only affects things which are smaller than one byte or structures containings things which should be aligned on a 2- or 4-byte boundary. One example: Type ByteInt = __byte__ Integer; Byte = __unsigned__ ByteInt; Var AOC: array [ 1..100 ] of Char; PAOC: packed array [ 1..100 ] of Char; AOBoo: array [ 1..100 ] of Boolean; PAOBoo: packed array [ 1..100 ] of Boolean; AOBy: array [ 1..100 ] of Byte; PAOBy: packed array [ 1..100 ] of Byte; all have 100 bytes (while PAOBoo should have ( 100 + 7 ) div 8 bytes), and MyRec: record B: Byte; C: Char; I: Integer; end (* MyRec *); has 8 bytes, not 6. \begin{sigh} When we say that GPC "does not pack" this means about the same as what Borland means when they claim that Borland Pascal "always packs". The only difference is that Borland Pascal does not care about 32-bit alignment and thus often produces records of smaller size - at the expense of loss of speed. \end{sigh} Hope this helps, Peter Dipl.-Phys. Peter Gerwinski, Essen, Germany, free physicist and programmer peter.gerwinski(a)uni-essen.de - http://home.pages.de/~peter.gerwinski/ [970201] maintainer GNU Pascal - http://home.pages.de/~gnu-pascal/ [970125]

1 0

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

2001

2000

1999

1998

1997

1996

1995

Gpc February 1997