Date: Sun, 16 Jan 2005 00:04:20 EST-10EDT,10,-1,0,7200,3,-1,0,7200,3600 Subject: [UnixOS2_Archive] No. 503 ************************************************** Saturday 15 January 2005 Number 503 ************************************************** Subjects for today 1 logrotate : John Poltorak 2 Re: logrotate : IanM" 3 Re: logrotate : John Poltorak 4 Re: zlib : Dave Yeo" 5 Re: zlib : John Poltorak 6 Re: OpenSSL update : Dave Yeo" 7 Re: OpenSSL update : John Poltorak 8 Re: zlib : Dave Yeo" 9 Re: OpenSSL update : Stefan.Neis at t-online.de 10 Re: OpenSSL update : Dave Yeo" 11 Re: OpenSSL update : Stefan.Neis at t-online.de 12 Re: fd_set in DIALOG : Stefan.Neis at t-online.de 13 Re: SW: postgreSQL V8.0 RC5 release] : lsunley at mb.sympatico.ca 14 Re: OpenSSL update : Brian Havard" 15 Re: SW: postgreSQL V8.0 RC5 release] : Knut St. Osmundsen" 16 Re: SW: postgreSQL V8.0 RC5 release] : Dave Yeo" 17 Re: SW: postgreSQL V8.0 RC5 release] : lsunley at mb.sympatico.ca 18 Re: SW: postgreSQL V8.0 RC5 release] : lsunley at mb.sympatico.ca 19 Re: SW: postgreSQL V8.0 RC5 release] : John Poltorak 20 Date coversion using GNU Date : John Poltorak **= Email 1 ==========================** Date: Fri, 14 Jan 2005 12:56:04 +0000 From: John Poltorak Subject: logrotate Is logrotate available on OS/2? I'm sure I have come across it somewhere but can't find it on Hobbes. -- John **= Email 2 ==========================** Date: Sat, 15 Jan 2005 00:32:23 +1100 (EDT) From: "IanM" Subject: Re: logrotate Hi John > Is logrotate available on OS/2? I'm sure I have come across it somewhere > but can't find it on Hobbes. You might be thinking of zwrotlogs (rotate logs) from Original by Ben Laurie Enhanced by Z. Wagner - Ice Bear Soft http://os2site.com/sw/internet/www/apache/ There is also rotatelogs.exe which should be in your apache\src\support directory :-) Cheers IanM http://www.os2site.com/ Cutler's Law: Computers have a reset button -- for a reason. **= Email 3 ==========================** Date: Fri, 14 Jan 2005 14:32:49 +0000 From: John Poltorak Subject: Re: logrotate On Sat, Jan 15, 2005 at 12:32:23AM +1100, IanM wrote: > There is also rotatelogs.exe which should be in your apache\src\support > directory :-) That's probably the one I'm thinking of... It seems that logrotate is actually just a shell script so maybe it 'just works' on OS/2. The docs I've seen says it needs BASH - can't see why it wouldn't work withe PDKSH. Maybe I'll give it a a try. > Cheers > IanM > http://www.os2site.com/ > > Cutler's Law: Computers have a reset button -- for a reason. -- John **= Email 4 ==========================** Date: Fri, 14 Jan 2005 08:08:10 -0800 From: "Dave Yeo" Subject: Re: zlib On Fri, 14 Jan 2005 10:24:18 +0000, John Poltorak wrote: >Has anyone managed to build the latest version of ZLIB? > >It's available here:- > >http://www.gzip.org/zlib/zlib-1.2.2.tar.gz > >although you wouldn't realise this was the latest version if you looked at >the ZLIB homepage. Couldn't find it there. Did find it here http://www.zlib.net/zlib-1.2.2.tar.gz Dave **= Email 5 ==========================** Date: Fri, 14 Jan 2005 16:22:11 +0000 From: John Poltorak Subject: Re: zlib On Fri, Jan 14, 2005 at 08:08:10AM -0800, Dave Yeo wrote: > On Fri, 14 Jan 2005 10:24:18 +0000, John Poltorak wrote: > > >Has anyone managed to build the latest version of ZLIB? > > > >It's available here:- > > > >http://www.gzip.org/zlib/zlib-1.2.2.tar.gz > > > >although you wouldn't realise this was the latest version if you looked at > >the ZLIB homepage. > > Couldn't find it there. Did find it here http://www.zlib.net/zlib-1.2.2.tar.gz But can you build it? > Dave -- John **= Email 6 ==========================** Date: Fri, 14 Jan 2005 08:25:22 -0800 From: "Dave Yeo" Subject: Re: OpenSSL update On Fri, 14 Jan 2005 11:50:20 +0100, Stefan.Neis at t-online.de wrote: > Hi, > >> Can anyone comment on these code changes for OpenSSL:- ? >> >> http://cvs.openssl.org/chngview?cn=12821 > >"Interesting". That should work just fine. Well the latest cvs gives these errors on OS/2 now (building shared libs) crypto\bn\asm\bn-os2.obj(bn-os2.obj) : error L2025: bn_sub_part_words : symbol defined more than once LINK386 : error L2022: # (alias #) : export undefined LINK386 : error L2022: ENGINE_load_gmp (alias ENGINE_load_gmp) : export undefined tmp_dll\o_str.obj(o_str.obj) : error L2029: 'strncasecmp' : unresolved external tmp_dll\o_str.obj(o_str.obj) : error L2029: 'strcasecmp' : unresolved external There were 5 errors detected make: *** [out_dll/crypto.dll] Error 1 I'm not sure if cvs (actually rsync) is working right. Openssl developers want us to rsync the whole cvs repositry then use cvs locally to update. Unluckily with the rsync from ux2bs I see lots of error messages. eg delete_one: unlink CVSROOT/..#admininfo.a30366: Permission denied failed to set permissions on . : Permission denied rsync error: timeout in data send/receive (code 30) at io.c(153) failed to set permissions on CVSLOGS : Permission denied rsync: connection unexpectedly closed (76264 bytes read so far) rsync error: error in rsync protocol data stream (code 12) at io.c(150) failed to set permissions on CVSROOT : Permission denied failed to set permissions on CVSROOT/Emptydir : Permission denied rsync: connection unexpectedly closed (76264 bytes read so far) rsync error: error in rsync protocol data stream (code 12) at io.c(150) so not totally sure yet if updating Dave **= Email 7 ==========================** Date: Fri, 14 Jan 2005 17:20:09 +0000 From: John Poltorak Subject: Re: OpenSSL update On Fri, Jan 14, 2005 at 08:25:22AM -0800, Dave Yeo wrote: > On Fri, 14 Jan 2005 11:50:20 +0100, Stefan.Neis at t-online.de wrote: > > > Hi, > > > >> Can anyone comment on these code changes for OpenSSL:- ? > >> > >> http://cvs.openssl.org/chngview?cn=12821 > > > >"Interesting". That should work just fine. > > Well the latest cvs gives these errors on OS/2 now (building shared libs) > > crypto\bn\asm\bn-os2.obj(bn-os2.obj) : error L2025: bn_sub_part_words : symbol defined more than once > LINK386 : error L2022: # (alias #) : export undefined > LINK386 : error L2022: ENGINE_load_gmp (alias ENGINE_load_gmp) : export undefined > > tmp_dll\o_str.obj(o_str.obj) : error L2029: 'strncasecmp' : unresolved external > > tmp_dll\o_str.obj(o_str.obj) : error L2029: 'strcasecmp' : unresolved external > > There were 5 errors detected > make: *** [out_dll/crypto.dll] Error 1 Hmmm... I thought the idea of the patch was to eliminate the str[n]casecmp errors... The other errors may well be due to the introduction of new code since the release of 0.9.7e... Any chance of reporting your results here:- mailto:openssl-dev at openssl.org > I'm not sure if cvs (actually rsync) is working right. Openssl developers want us to rsync the whole cvs repositry then use cvs locally to update. Unluckily with the rsync from ux2bs I see lots of error messages. eg There is a newer OS/2 build of RSYNC here:- http://unix.os2site.com/sw/pub/binary/autoconf/rsync-2_6_2-bin.zip I haven't had chance to try that one yet. > Dave -- John **= Email 8 ==========================** Date: Fri, 14 Jan 2005 09:56:55 -0800 From: "Dave Yeo" Subject: Re: zlib On Fri, 14 Jan 2005 16:22:11 +0000, John Poltorak wrote: >But can you build it? Yes Dave **= Email 9 ==========================** Date: Fri, 14 Jan 2005 19:59:03 +0100 From: Stefan.Neis at t-online.de Subject: Re: OpenSSL update Hi, > Well the latest cvs gives these errors on OS/2 now (building shared libs) > > crypto\bn\asm\bn-os2.obj(bn-os2.obj) : error L2025: bn_sub_part_words : symbol defined more than once > LINK386 : error L2022: # (alias #) : export undefined > LINK386 : error L2022: ENGINE_load_gmp (alias ENGINE_load_gmp) : export undefined > > tmp_dll\o_str.obj(o_str.obj) : error L2029: 'strncasecmp' : unresolved external > > tmp_dll\o_str.obj(o_str.obj) : error L2029: 'strcasecmp' : unresolved external > > There were 5 errors detected > make: *** [out_dll/crypto.dll] Error 1 Hm, I'm pretty confident that at least str(n)casecmp shouldn't give any more problems, so given the problems with rsync/cvs you describe, I'd rather suppose that something went wrong there... To simplify life, I'd recommend to just download a daily snapshot from one of the mirrors listed below all those more or less current tarballs of official releases on http://www.openssl.org/source/ Compiling today's snapshot in DLL mode. I get (only): crypto\bn\asm\bn-os2.obj(bn-os2.obj) : error L2025: bn_sub_part_words : symbol defined more than once LINK386 : error L2022: ENGINE_load_gmp (alias ENGINE_load_gmp) : export undefined There were 2 errors detected Those in fact seem to indicate newly added problems ... Regards, Stefan **= Email 10 ==========================** Date: Fri, 14 Jan 2005 12:46:13 -0800 From: "Dave Yeo" Subject: Re: OpenSSL update On Fri, 14 Jan 2005 19:59:03 +0100, Stefan.Neis at t-online.de wrote: >Hm, I'm pretty confident that at least str(n)casecmp shouldn't give >any more problems, so given the problems with rsync/cvs you describe, >I'd rather suppose that something went wrong there... > >To simplify life, I'd recommend to just download a daily snapshot >from one of the mirrors listed below all those more or less current >tarballs of official releases on http://www.openssl.org/source/ > >Compiling today's snapshot in DLL mode. I get (only): > >crypto\bn\asm\bn-os2.obj(bn-os2.obj) : error L2025: bn_sub_part_words : symbol defined more than once >LINK386 : error L2022: ENGINE_load_gmp (alias ENGINE_load_gmp) : export undefined > >There were 2 errors detected Actually I made a mistake. Some of those errors showed up after I tried to tweek the DEFs. I had the same errors as above as well as the str* errors. Unluckily it seems that rsync (even the new one) is too broken to keep OpenSSL current and I don't have the bandwidth and internet time to keep downloading the tarballs (takes over a hour here). Just wish the phone company would update the lines out here. Still when I first moved to this area the only choice was party lines so I can't complain too much Dave **= Email 11 ==========================** Date: Fri, 14 Jan 2005 22:10:01 +0100 From: Stefan.Neis at t-online.de Subject: Re: OpenSSL update Hi, > Compiling today's snapshot in DLL mode. I get (only): > > crypto\bn\asm\bn-os2.obj(bn-os2.obj) : error L2025: bn_sub_part_words : symbol defined more than once That one is due to a new assembler routine which requires -DOPENSSL_BN_ASM_PART_WORDS as an additional flag for compilation - the Makefile(s) for OS/2 apparently need to be adapted. > LINK386 : error L2022: ENGINE_load_gmp (alias ENGINE_load_gmp) : export undefined Hm. the settings for compiling e_gmp.c apparently don't quite match the list of defined symbols. Since the later seems to be generated, I guess somebody (Brian?) forgot to include updates for util\pl\OS2-EMX.pl (or a similar file) when submitting some other changes - I suppose those problems will be updated over the next couple of days... Regards, Stefan **= Email 12 ==========================** Date: Fri, 14 Jan 2005 23:15:50 +0100 From: Stefan.Neis at t-online.de Subject: Re: fd_set in DIALOG > gcc -Zmt -D__ST_MT_ERRNO__ -D_XOPEN_SOURCE=500 -D_POSIX_C_SOURCE > -I/usr/include/ncurses -DHAVE_CONFIG_H -I. -I. > -DLOCALEDIR=\"g:/usr/share/locale\" -c ui_getc.c > In file included from g:\ux2bs\posix2\include\unistd.h:232, > from dialog.h:31, > from ui_getc.c:23: > g:\ux2bs\posix2\include\sys/unistdx.h:149: warning: `struct _fd_set' > declared inside parameter list > g:\ux2bs\posix2\include\sys/unistdx.h:149: warning: its scope is only this > definition or declaration, > g:\ux2bs\posix2\include\sys/unistdx.h:149: warning: which is probably not > what you want. > ui_getc.c: In function `dlg_getc_ready': > ui_getc.c:95: `fd_set' undeclared (first use in this function) > ui_getc.c:95: (Each undeclared identifier is reported only once > ui_getc.c:95: for each function it appears in.) > ui_getc.c:95: parse error before `read_fds' > ui_getc.c:99: `read_fds' undeclared (first use in this function) > ui_getc.c:104: parse error before `)' > make: *** [ui_getc.o] Error 1 > > Does this mean anything to anyone? Posix/2's sys/unistdx.h believes that it should not include a definition of select if either _POSIX_SOURCE or _XOPEN_SOURCE are defined. Here, even both are defined, but the source code apparently still expects to get a definition of select. I'm quite confused as to what the headers really should do, but anyway, replacing #ifndef _POSIX_SOURCE # include # ifndef _XOPEN_SOURCE # include /* structure timeval required for select() */ # include # endif # include #endif (around line 10 of posix2\include\sys\unistdx.h) by #include #include should get the compilation going... Regards, Stefan **= Email 13 ==========================** Date: Fri, 14 Jan 2005 19:36:26 -0500 From: lsunley at mb.sympatico.ca Subject: Re: SW: postgreSQL V8.0 RC5 release] In <20050114115941.D18360 at warpix.org>, on 01/14/05 at 11:59 AM, John Poltorak said: >Just forwarding news about the latest RC for postgreSQL. >Looks like someone has been busy ;-)... >Is there any chance that it will run on machines it wouldn't run on >before or is that a gcc issue? That's a GCC issue. I am not going to touch the fork() code unless Knut desperately wants some help :-) I looked at it, but did not think that it would be an easy thing to fix up unless I really understood what it is doing (and right now I do not :-). GA release of postgreSQL should be out next week. Right after that the patches I have will be submitted to the postgreSQL folks for inclusion in the main code repository, then anyone should be able to build it "out of the box" - given that you have all the required stuff - GCC 3.3.5 with a fixed LIBC (which is not required to build, just to run it for more than 50 query sessions), the Innotek ports for all the *nix stuff, including Perl, gettext, ash, and you will need the bison 1.875 port, and flex. To build the documentation you will need the openJade and openSP ports. Lorne -- ----------------------------------------------------------- lsunley at mb.sympatico.ca ----------------------------------------------------------- **= Email 14 ==========================** Date: Sat, 15 Jan 2005 13:10:13 +1000 (EST) From: "Brian Havard" Subject: Re: OpenSSL update On Fri, 14 Jan 2005 22:10:01 +0100, Stefan.Neis at t-online.de wrote: > Hi, > >> Compiling today's snapshot in DLL mode. I get (only): >> >> crypto\bn\asm\bn-os2.obj(bn-os2.obj) : error L2025: bn_sub_part_words : symbol defined more than once > >That one is due to a new assembler routine which requires >-DOPENSSL_BN_ASM_PART_WORDS as an additional flag for compilation >- the Makefile(s) for OS/2 apparently need to be adapted. > >> LINK386 : error L2022: ENGINE_load_gmp (alias ENGINE_load_gmp) : export undefined > >Hm. the settings for compiling e_gmp.c apparently don't quite >match the list of defined symbols. Since the later seems to be >generated, I guess somebody (Brian?) forgot to include updates >for util\pl\OS2-EMX.pl (or a similar file) when submitting some >other changes - I suppose those problems will be updated over >the next couple of days... ???? I just built a clean checkout of the OpenSSL_0_9_7-stable branch & it built without error. Are you building that or HEAD (0.9.8)? -- ______________________________________________________________________________ | Brian Havard | "He is not the messiah! | | brianh at kheldar.apana.org.au | He's a very naughty boy!" - Life of Brian | ------------------------------------------------------------------------------ **= Email 15 ==========================** Date: Sat, 15 Jan 2005 03:41:34 +0100 From: "Knut St. Osmundsen" Subject: Re: SW: postgreSQL V8.0 RC5 release] lsunley at mb.sympatico.ca wrote: > In <20050114115941.D18360 at warpix.org>, on 01/14/05 > at 11:59 AM, John Poltorak said: > > > > > >>Just forwarding news about the latest RC for postgreSQL. > > >>Looks like someone has been busy ;-)... > > >>Is there any chance that it will run on machines it wouldn't run on >>before or is that a gcc issue? > > > That's a GCC issue. I am not going to touch the fork() code unless Knut > desperately wants some help :-) I looked at it, but did not think that it > would be an easy thing to fix up unless I really understood what it is > doing (and right now I do not :-). > > > GA release of postgreSQL should be out next week. Right after that the > patches I have will be submitted to the postgreSQL folks for inclusion in > the main code repository, then anyone should be able to build it "out of > the box" - given that you have all the required stuff - GCC 3.3.5 with a > fixed LIBC (which is not required to build, just to run it for more than > 50 query sessions), the Innotek ports for all the *nix stuff, including > Perl, gettext, ash, and you will need the bison 1.875 port, and flex. To > build the documentation you will need the openJade and openSP ports. > I'm putting this here so everyone which potentially would do their own libc builds get the message. Rules when building your own libc: 1. Don't name it similar to an existing or future release from me. Just keep the name in the public sources (libc06xx.dll ATM). 2. Don't use it for GA releases if you wanna be nice to be users. If you're desperate about getting your cool-piece-of-software out just bug me enough about it and I'll find some time to put out a bugfix release. 3. State explicitly in the docs that it is a self built libc and that it may conflict with a libc the user have built himself or someone else have released for their alpha of this-piece-of-cool-software. LIBC is a moving target, it is *required* to use the headers and libs of the libc version you're building for when building something. I don't recall if I did any incompatible changes after beta 2 yet. I've done between most of the 0.6 releases. Lorne, please make sure that libc06b3.dll disappear once you've got the next build out. If you wanna do a GA next week, well, I'll *try* find some time this weekend to fix that fork() issue and integrate some patches so a beta 4 (since you've done the beta 3 name) can be squeezed out the door. And btw. dlls are libraries (dynamic link libraries, right). Libraries belong in /lib not /bin. Kind Regards, knut PS. The final 0.6 will not appear for yet a little while because I'm working night and day at some other stuff atm. Hope it will cool down a bit next month, but I wouldn't bet on it... **= Email 16 ==========================** Date: Fri, 14 Jan 2005 18:47:38 -0800 From: "Dave Yeo" Subject: Re: SW: postgreSQL V8.0 RC5 release] On Fri, 14 Jan 2005 19:36:26 -0500, lsunley at mb.sympatico.ca wrote: > the Innotek ports for all the *nix stuff, including >Perl, gettext, ash, and you will need the bison 1.875 port, and flex. Is there an Innotek port of Perl? and where located? Dave **= Email 17 ==========================** Date: Fri, 14 Jan 2005 21:39:28 -0500 From: lsunley at mb.sympatico.ca Subject: Re: SW: postgreSQL V8.0 RC5 release] In <41E882DE.8060000 at anduin.net>, on 01/15/05 at 03:41 AM, "Knut St. Osmundsen" said: >lsunley at mb.sympatico.ca wrote: >> In <20050114115941.D18360 at warpix.org>, on 01/14/05 >> at 11:59 AM, John Poltorak said: >> >> >> >> >> >>>Just forwarding news about the latest RC for postgreSQL. >> >> >>>Looks like someone has been busy ;-)... >> >> >>>Is there any chance that it will run on machines it wouldn't run on >>>before or is that a gcc issue? >> >> >> That's a GCC issue. I am not going to touch the fork() code unless Knut >> desperately wants some help :-) I looked at it, but did not think that it >> would be an easy thing to fix up unless I really understood what it is >> doing (and right now I do not :-). >> >> >> GA release of postgreSQL should be out next week. Right after that the >> patches I have will be submitted to the postgreSQL folks for inclusion in >> the main code repository, then anyone should be able to build it "out of >> the box" - given that you have all the required stuff - GCC 3.3.5 with a >> fixed LIBC (which is not required to build, just to run it for more than >> 50 query sessions), the Innotek ports for all the *nix stuff, including >> Perl, gettext, ash, and you will need the bison 1.875 port, and flex. To >> build the documentation you will need the openJade and openSP ports. >> >I'm putting this here so everyone which potentially would do their own >libc builds get the message. >Rules when building your own libc: > 1. Don't name it similar to an existing or future release from me. > Just keep the name in the public sources (libc06xx.dll ATM). > 2. Don't use it for GA releases if you wanna be nice to be users. > If you're desperate about getting your cool-piece-of-software out > just bug me enough about it and I'll find some time to put out a > bugfix release. > 3. State explicitly in the docs that it is a self built libc and that > it may conflict with a libc the user have built himself or someone > else have released for their alpha of this-piece-of-cool-software. >LIBC is a moving target, it is *required* to use the headers and libs of >the libc version you're building for when building something. I don't >recall if I did any incompatible changes after beta 2 yet. I've done >between most of the 0.6 releases. >Lorne, please make sure that libc06b3.dll disappear once you've got the > next build out. If you wanna do a GA next week, well, I'll *try* find >some time this weekend to fix that fork() issue and integrate some >patches so a beta 4 (since you've done the beta 3 name) can be squeezed >out the door. That would be great... :-) Will do, like I said, releasing the the name was an accident and I will make it disappear next release of postgreSQL. >And btw. dlls are libraries (dynamic link libraries, right). Libraries >belong in /lib not /bin. OK >Kind Regards, > knut >PS. The final 0.6 will not appear for yet a little while because I'm >working night and day at some other stuff atm. Hope it will cool down a >bit next month, but I wouldn't bet on it... That would be a first, I've never heard of things actually cooling down :-) -- ----------------------------------------------------------- lsunley at mb.sympatico.ca ----------------------------------------------------------- **= Email 18 ==========================** Date: Fri, 14 Jan 2005 21:43:59 -0500 From: lsunley at mb.sympatico.ca Subject: Re: SW: postgreSQL V8.0 RC5 release] In <20050115024718.D84F9F5502 at generation.lgisp.net>, on 01/14/05 at 06:47 PM, "Dave Yeo" said: >On Fri, 14 Jan 2005 19:36:26 -0500, lsunley at mb.sympatico.ca wrote: >> the Innotek ports for all the *nix stuff, including >>Perl, gettext, ash, and you will need the bison 1.875 port, and flex. >Is there an Innotek port of Perl? and where located? >Dave That was poor phrasing... The Perl I am using is 5.8.0 from Hobbes, the Innotek site has ports for ash, autoconf, automake, gettext (***** very important to have *****) and others, but no Perl Lorne -- ----------------------------------------------------------- lsunley at mb.sympatico.ca ----------------------------------------------------------- **= Email 19 ==========================** Date: Sat, 15 Jan 2005 11:48:48 +0000 From: John Poltorak Subject: Re: SW: postgreSQL V8.0 RC5 release] On Fri, Jan 14, 2005 at 09:43:59PM -0500, lsunley at mb.sympatico.ca wrote: > In <20050115024718.D84F9F5502 at generation.lgisp.net>, on 01/14/05 > at 06:47 PM, "Dave Yeo" said: > > >On Fri, 14 Jan 2005 19:36:26 -0500, lsunley at mb.sympatico.ca wrote: > > >> the Innotek ports for all the *nix stuff, including > >>Perl, gettext, ash, and you will need the bison 1.875 port, and flex. > > >Is there an Innotek port of Perl? and where located? > >Dave > > That was poor phrasing... > > The Perl I am using is 5.8.0 from Hobbes, the Innotek site has ports for > ash, autoconf, automake, gettext (***** very important to have *****) and > others, but no Perl The OS/2 build of Perl is firmly designed around EMX so building it under libc is not a trivial task AIUI. Once the Perl build has its EMXisms removed, I'd like to migrate UX2BS to gcc 3.3.5 which should provide a much better standardised build environment, although I'm still struggling to get an Innotek version of PDKSH built. I never did have much success with ASH, although Knut seems to use it without any problems... > Lorne > > -- > ----------------------------------------------------------- > lsunley at mb.sympatico.ca > ----------------------------------------------------------- > -- John **= Email 20 ==========================** Date: Sat, 15 Jan 2005 12:09:46 +0000 From: John Poltorak Subject: Date coversion using GNU Date GNU Date seems like a useful utility for handling dates in various formats and I'm sure it can be used to convert a datestamp in seconds (from 1/1/1970) to something more readable, but I can't get the format correct. Here is what I tried:- date +%Y:%m:%d:%H:%M:%S --date=1105707584 but this returns invalid date. If I run:- date +%Y:%m:%d:%H:%M:%S --date=1 that results in:- 2005:01:15:01:00:00 It's as if the program is taking today as its base date rather than 1/1/1970. Can anyone shed any light on how I get the program to do what I want it to? -- John