From: UnixOS2 Archive To: "UnixOS2 Archive" Date: Thu, 16 Jan 2003 04:48:26 EST-10EDT,10,-1,0,7200,3,-1,0,7200,3600 Subject: [UnixOS2_Archive] No. 15 ************************************************** Wednesday 15 January 2003 Number 15 ************************************************** Subjects for today 1 Re: Testing for LINK386 on path : Csaba" 2 Re: Testing for LINK386 on path : Thomas E. Dickey" 3 Re: Building NCURSES apps : Thomas E. Dickey" 4 Re: Testing for LINK386 on path : illya at vaeshiep.demon.nl 5 Re: Building NCURSES apps : John Poltorak 6 Re: Testing for LINK386 on path : John Poltorak 7 Re: Testing for LINK386 on path : illya at vaeshiep.demon.nl 8 ZLIB : John Poltorak 9 Re: ZLIB : John Poltorak 10 Re: ZLIB : Franz Bakan" 11 Re: Building NCURSES apps : Thomas Dickey 12 UnixOS/2 Build System - mailing list : John Poltorak 13 How to identify OS/2 from shell script : John Poltorak 14 Re: Testing for LINK386 on path : Andreas Buening 15 Re: Building NCURSES apps : John Poltorak 16 Re: How to identify OS/2 from shell script : Sebastian Wittmeier (ShadoW)" 17 Re: UnixOS/2 Build System - mailing list : Yuri Dario" **= Email 1 ==========================** Date: Thu, 16 Jan 2003 00:40:33 -0000 From: "Csaba" Subject: Re: Testing for LINK386 on path On 9 Jan 2003, at 12:50, John Poltorak wrote: > > How would I test for the presence of LINK386 being on the path from within > a shell script? 1) Native way: check that it's not in the current directory, then try to run it. 2) "configure way" : chop the path at the path separators (which is semicolon for OS/2 but colon for Unix), then for each section check that it's a directory ( -t ), then concatenate "link386" to the directory name and check that it's executable ( -x ) or at least exists ( -e ). -- Ceci n'est pas un .signature **= Email 2 ==========================** Date: Thu, 16 Jan 2003 05:00:07 -0500 (EST) From: "Thomas E. Dickey" Subject: Re: Testing for LINK386 on path On Thu, 16 Jan 2003 illya at vaeshiep.demon.nl wrote: > Csaba: > >On 9 Jan 2003, at 12:50, John Poltorak wrote: > >>How would I test for the presence of LINK386 being on the path from within > >>a shell script? > >1) Native way: check that it's not in the current directory, then try > >to run it. > >2) "configure way" : chop the path at the path separators (which is > >semicolon for OS/2 but colon for Unix), then for each section check > >that it's a directory ( -t ), then concatenate "link386" to the > >directory name and check that it's executable ( -x ) or at least > >exists ( -e ). > > 3) Use "which", which ofcourse must be present on the system. If it says > "no link386 on (...)" (or something like that), it's not on the > path. "which" is csh-specific. ksh provides "type" -- T.E.Dickey http://invisible-island.net ftp://invisible-island.net **= Email 3 ==========================** Date: Thu, 16 Jan 2003 06:12:38 -0500 (EST) From: "Thomas E. Dickey" Subject: Re: Building NCURSES apps On Thu, 16 Jan 2003, John Poltorak wrote: > On Wed, Jan 15, 2003 at 05:46:08PM -0500, Thomas Dickey wrote: > > > did you do > > configure --with-ncurses > > If not, it doesn't look for ncurses-specific features, but only for generic > > curses stuff. > > *That* is what I had overlooked. > > Presumably I should use this option with all your apps... The two options --with-ncurses --with-screen=ncurses should cover all of those. -- T.E.Dickey http://invisible-island.net ftp://invisible-island.net **= Email 4 ==========================** Date: Thu, 16 Jan 2003 07:07:54 GMT From: illya at vaeshiep.demon.nl Subject: Re: Testing for LINK386 on path Csaba: >On 9 Jan 2003, at 12:50, John Poltorak wrote: >>How would I test for the presence of LINK386 being on the path from within >>a shell script? >1) Native way: check that it's not in the current directory, then try >to run it. >2) "configure way" : chop the path at the path separators (which is >semicolon for OS/2 but colon for Unix), then for each section check >that it's a directory ( -t ), then concatenate "link386" to the >directory name and check that it's executable ( -x ) or at least >exists ( -e ). 3) Use "which", which ofcourse must be present on the system. If it says "no link386 on (...)" (or something like that), it's not on the path. **= Email 5 ==========================** Date: Thu, 16 Jan 2003 08:58:04 +0000 From: John Poltorak Subject: Re: Building NCURSES apps On Wed, Jan 15, 2003 at 05:46:08PM -0500, Thomas Dickey wrote: > did you do > configure --with-ncurses > If not, it doesn't look for ncurses-specific features, but only for generic > curses stuff. *That* is what I had overlooked. Presumably I should use this option with all your apps... > -- > Thomas E. Dickey > http://invisible-island.net > ftp://invisible-island.net -- John **= Email 6 ==========================** Date: Thu, 16 Jan 2003 10:12:44 +0000 From: John Poltorak Subject: Re: Testing for LINK386 on path On Thu, Jan 16, 2003 at 05:00:07AM -0500, Thomas E. Dickey wrote: > On Thu, 16 Jan 2003 illya at vaeshiep.demon.nl wrote: > > > Csaba: > > >On 9 Jan 2003, at 12:50, John Poltorak wrote: > > >>How would I test for the presence of LINK386 being on the path from within > > >>a shell script? > > >1) Native way: check that it's not in the current directory, then try > > >to run it. > > >2) "configure way" : chop the path at the path separators (which is > > >semicolon for OS/2 but colon for Unix), then for each section check > > >that it's a directory ( -t ), then concatenate "link386" to the > > >directory name and check that it's executable ( -x ) or at least > > >exists ( -e ). > > > > 3) Use "which", which ofcourse must be present on the system. If it says > > "no link386 on (...)" (or something like that), it's not on the > > path. > > "which" is csh-specific. ksh provides "type" I guess in the context that I want to use this test I can simply use:- test -f $OSRT\os2\link386.exe Is LINK386.EXE considered to be an absolute requirement for a build environment? > -- > T.E.Dickey > http://invisible-island.net > ftp://invisible-island.net -- John **= Email 7 ==========================** Date: Thu, 16 Jan 2003 11:57:23 GMT From: illya at vaeshiep.demon.nl Subject: Re: Testing for LINK386 on path Thomas Dickey: >>3) Use "which", which ofcourse must be present on the system. If it says >>"no link386 on (...)" (or something like that), it's not on the >>path. >"which" is csh-specific. ksh provides "type" I didn't mean a shell builtin, but a separate binary (which.exe), thats why I added "which ofcourse must be present on the system". **= Email 8 ==========================** Date: Thu, 16 Jan 2003 15:21:04 +0000 From: John Poltorak Subject: ZLIB Do we currently have an OS/2 maintainer for ZLIB? I saw v1.1.4 recently but don't reacall where... -- John **= Email 9 ==========================** Date: Thu, 16 Jan 2003 15:55:20 +0000 From: John Poltorak Subject: Re: ZLIB On Thu, Jan 16, 2003 at 04:44:04PM +0100, Franz Bakan wrote: > On Thu, 16 Jan 2003 15:21:04 +0000, John Poltorak wrote: > > > > >Do we currently have an OS/2 maintainer for ZLIB? > > > >I saw v1.1.4 recently but don't reacall where... > > There is no maintainer, There appears to be at least three !!! :-)... > but I uploaded a package > which is compatible with gimp/gtk > to os2unix or unixos2 and asked here on the list > to check and comment. > But there was no reply to that. That's often the case. People don't look at it if there isn't an immediate requirement to do so... > So I don't know if I should upload my version to > HOBBES and replace the 1.1.4 version which doesn't > work with gimp/gtk. I have never understood all the issues surrounding ZLIB and why it is so difficult to have a single distribution of the package. Why can't we just have one? > Franz -- John **= Email 10 ==========================** Date: Thu, 16 Jan 2003 16:44:04 +0100 (MEZ) From: "Franz Bakan" Subject: Re: ZLIB On Thu, 16 Jan 2003 15:21:04 +0000, John Poltorak wrote: > >Do we currently have an OS/2 maintainer for ZLIB? > >I saw v1.1.4 recently but don't reacall where... There is no maintainer, but I uploaded a package which is compatible with gimp/gtk to os2unix or unixos2 and asked here on the list to check and comment. But there was no reply to that. So I don't know if I should upload my version to HOBBES and replace the 1.1.4 version which doesn't work with gimp/gtk. Franz **= Email 11 ==========================** Date: Thu, 16 Jan 2003 17:56:32 -0500 From: Thomas Dickey Subject: Re: Building NCURSES apps On Thu, Jan 16, 2003 at 10:02:01PM +0000, John Poltorak wrote: > On Thu, Jan 16, 2003 at 06:12:38AM -0500, Thomas E. Dickey wrote: > > On Thu, 16 Jan 2003, John Poltorak wrote: > > > > > On Wed, Jan 15, 2003 at 05:46:08PM -0500, Thomas Dickey wrote: > > > > > > > did you do > > > > configure --with-ncurses > > > > If not, it doesn't look for ncurses-specific features, but only for generic > > > > curses stuff. > > > > > > *That* is what I had overlooked. > > > > > > Presumably I should use this option with all your apps... > > > > The two options --with-ncurses --with-screen=ncurses should cover all of > > those. > > How about TD_LIB? I'll load a new copy. Checking my reference, I was working on changes to the curses-library lookup around 21 December, and made a correction on 29 December. (I noticed the bug while working on lynx, but forgot this one) > Here's what I got:- > > chmod: configure: Permission denied > ./configure --with-ncurses --with-screen=ncurses > loading site script u:/unixos2/lib/config.site > creating cache ./config.cache > checking host system type... i386-pc-os2-emx > Configuring for os2-emx > ... > ... > checking for bcopy in -lbsd... yes > checking for ncurses header in include-path... no > checking for ncurses include-path... configure: error: not found > make > > >From config.log:- > > > configure:3823: gcc -c -Zmt -D__ST_MT_ERRNO__ conftest.c 1>&5 > In file included from configure:3805: > u:\usr\include\ncurses/ncurses.h:58: ncurses_dll.h: No such file or directory > In file included from configure:3805: > u:\usr\include\ncurses/ncurses.h:102: unctrl.h: No such file or directory > configure: failed program was: > #line 3802 "configure" > #include "confdefs.h" > > > > > -- > > T.E.Dickey > > http://invisible-island.net > > ftp://invisible-island.net > > -- > John > > -- Thomas E. Dickey http://invisible-island.net ftp://invisible-island.net **= Email 12 ==========================** Date: Thu, 16 Jan 2003 18:35:24 +0000 From: John Poltorak Subject: UnixOS/2 Build System - mailing list A new mailing list has been set up specifically to test out and extend the UnixOS/2 Build System and everyone is invited to join. You can get further details about the mailing list from:- http://powerusersbbs.net/mailman/listinfo/ux2bs This is a Mailman mailing list set up courtesy of Ted Sikora. Hats off to Ted and Andrew MacIntyre for getting Mailman working on OS/2. The main aim of the list is to develop a system which will get Unix apps built on OS/2 without jumping through countless hoops to get the job done. It is intended that the systmem will provide a global wrapper for building as many apps as possible, simply by running:- build APP It is hoped that subscribers will try building apps for themselves and provide some feedback. -- John **= Email 13 ==========================** Date: Thu, 16 Jan 2003 21:01:22 +0000 From: John Poltorak Subject: How to identify OS/2 from shell script What is the best way of identifying the operating system as OS/2 from within a shell script? One suggestion is uname. Anything else? How about using something like if `ver`="Operating System/2" ? Is there any way to identify that the OS uses drive letters, or ';' as path seperators, or '\' as directory seperators? I wouldn't want to rely on environment variables, but are there any unique to OS/2? -- John **= Email 14 ==========================** Date: Thu, 16 Jan 2003 21:51:36 +0100 From: Andreas Buening Subject: Re: Testing for LINK386 on path illya at vaeshiep.demon.nl wrote: > > Thomas Dickey: > >>3) Use "which", which ofcourse must be present on the system. If it says > >>"no link386 on (...)" (or something like that), it's not on the > >>path. > >"which" is csh-specific. ksh provides "type" Other shells have "type -p", "whence" or whatever. > I didn't mean a shell builtin, but a separate binary (which.exe), thats why I added "which > ofcourse must be present on the > system". The bad thing is that some "which" implementations return 0 even if the program can't be found (and on some OS/2 systems "which" isn't even available). I know of no (simple) portable way to find programs. Bye, Andreas -- One OS to rule them all, One OS to find them, One OS to bring them all and in the darkness bind them In the Land of Mordor where the Shadows lie. **= Email 15 ==========================** Date: Thu, 16 Jan 2003 22:02:01 +0000 From: John Poltorak Subject: Re: Building NCURSES apps On Thu, Jan 16, 2003 at 06:12:38AM -0500, Thomas E. Dickey wrote: > On Thu, 16 Jan 2003, John Poltorak wrote: > > > On Wed, Jan 15, 2003 at 05:46:08PM -0500, Thomas Dickey wrote: > > > > > did you do > > > configure --with-ncurses > > > If not, it doesn't look for ncurses-specific features, but only for generic > > > curses stuff. > > > > *That* is what I had overlooked. > > > > Presumably I should use this option with all your apps... > > The two options --with-ncurses --with-screen=ncurses should cover all of > those. How about TD_LIB? Here's what I got:- chmod: configure: Permission denied ./configure --with-ncurses --with-screen=ncurses loading site script u:/unixos2/lib/config.site creating cache ./config.cache checking host system type... i386-pc-os2-emx Configuring for os2-emx ... ... checking for bcopy in -lbsd... yes checking for ncurses header in include-path... no checking for ncurses include-path... configure: error: not found make From config.log:- configure:3823: gcc -c -Zmt -D__ST_MT_ERRNO__ conftest.c 1>&5 In file included from configure:3805: u:\usr\include\ncurses/ncurses.h:58: ncurses_dll.h: No such file or directory In file included from configure:3805: u:\usr\include\ncurses/ncurses.h:102: unctrl.h: No such file or directory configure: failed program was: #line 3802 "configure" #include "confdefs.h" > -- > T.E.Dickey > http://invisible-island.net > ftp://invisible-island.net -- John **= Email 16 ==========================** Date: Thu, 16 Jan 2003 23:17:47 +0100 (CET) From: "Sebastian Wittmeier (ShadoW)" Subject: Re: How to identify OS/2 from shell script On Thu, 16 Jan 2003 21:01:22 +0000, John Poltorak wrote: >if `ver`="Operating System/2" ? And tomorrow ver says "eComStation x.y" or "OSFree 0.x running on FreeOS 0.y" Sebastian **= Email 17 ==========================** Date: Thu, 16 Jan 2003 23:32:23 +0100 (CET) From: "Yuri Dario" Subject: Re: UnixOS/2 Build System - mailing list Hi John, >A new mailing list has been set up specifically to test out and extend the >UnixOS/2 Build System and everyone is invited to join. I'm sorry, but why another list? isn't this one enough? Bye, Yuri Dario /* * member of TeamOS/2 - Italy * http://www.quasarbbs.net/yuri * http://www.teamos2.it * http://www.opera.com/os2/ */