newLISP Fan Club

Forum => newLISP newS => Topic started by: Lutz on July 23, 2009, 11:47:07 AM

Title: Maintenance Release newLISP v.10.1.1
Post by: Lutz on July 23, 2009, 11:47:07 AM
• This stable maintenance release has minor feature enhancements and bug fixes



Release notes: http://www.newlisp.org/downloads/newLISP-10.1-Release.html

Downloads: http://www.newlisp.org/index.cgi?page=Downloads



Ps: the release notes contain 10.1.1 tags for the new and fixed stuff
Title:
Post by: newdep on July 24, 2009, 12:05:27 AM
Thats actualy quite an addon this 10.1.1 release, nice enhancements..thanks..



Also the manual in the new format it looks very good..

.. nice character-set and clean layout.. smooth!
Title: NAN and/or INF
Post by: newdep on July 24, 2009, 01:42:33 AM
consistent it is ;-)



> (sqrt -1)

nan

> (sqrt (div -1 0))

inf







> (inf? (sqrt -1))

nan

> (inf? (div -1 0))

true

>
Title:
Post by: newdep on July 24, 2009, 05:48:08 AM
PS: Actualy



> (inf? (sqrt -1))

nan



should be nil ! and not nan..As "Infinite" isn't equal to "Not A Number"
Title:
Post by: Lutz on July 24, 2009, 07:13:21 AM
Its seems to happen only on Windows, but is fine on MAC OSX, FreeBSD and Linux returning 'nil', on Windows all operations on NaN return NaN ... will investigate.
Title:
Post by: newdep on July 24, 2009, 07:57:14 AM
Aaa yes your right.. I was indeed working on Windows with this one..



But then Its probably OS dependent and not easy to change because the

OS itself in some cases returns NaN instead of Inf...
Title:
Post by: Lutz on July 24, 2009, 10:05:51 AM
But this will be fixed in the next version.



There is currently an old catch-all #ifdef WIN_32 for all operations on NaN values, which can now be eliminated. It was necessary for older compilers on Windows.
Title:
Post by: cormullion on July 24, 2009, 12:20:31 PM
Talking of NaN, when I went round the supermarket yesterday I noticed that a few of the printed labels showed £Nan.p for the price per 100gm value. I like the way the error worked all the way through to the label printer OK...



Lutz: that new find option looks cool!



Also, on the documentation page of newlisp.org, could you consider slowly migrating some of the entries under Tutorials to point to the docuwiki version, since I probably won't be updating the older versions now, and they're already showing their age... :)
Title:
Post by: newdep on July 24, 2009, 12:23:29 PM
Now I keep seeing NaN's everywhere..

I just ordered 2 Garlic NaN's at the Indian Takeaway ;-)
Title:
Post by: pjot on August 01, 2009, 11:07:49 AM
The Makefile for 64bit Linux 'makefile_linuxLP64' seems to have a problem... below an changed version so it works with my 64bit Ubuntu 9.04 Linux.
Quote
# makefile for newLISP 64bit v.10.x.x on 64 bit LINUX tested on Intel Core Duo 2

#

# Note, that readline support may require different libraries on different OSs

#



OBJS = newlisp.o nl-symbol.o nl-math.o nl-list.o nl-liststr.o nl-string.o nl-filesys.o

   nl-sock.o nl-import.o nl-xml.o nl-web.o nl-matrix.o nl-debug.o pcre.o



CFLAGS = -m64 -Wall -pedantic -Wno-uninitialized -Wno-strict-aliasing -Wno-long-long -c -O2 -g -DNEWLISP64 -DLINUX



CC = gcc



default: $(OBJS)

#   $(CC) $(OBJS) -m32 -g -lm -ldl -lreadline -o newlisp # for UBUNTU Debian

#   $(CC) $(OBJS) -m32 -g -lm -ldl -lreadline -ltermcap -o newlisp # slackware

#   $(CC) $(OBJS) -m32 -g -lm -ldl -lreadline -lncurses -o newlisp # other Linux Dist

   $(CC) $(OBJS) -g -o newlisp -ldl -lm

   strip newlisp



.c.o:

   $(CC) $(CFLAGS) $<



$(OBJS): primes.h protos.h makefile_linuxLP64


For the other targets you can see a problem as the Makfeile is using both -m32 and -m64. In that situation, linking never succeeds.



Peter
Title:
Post by: TedWalther on August 01, 2009, 11:53:16 AM
Thanks for the bug report.  I'm looking at incorporating automatic flag setting in the alternative configure file.



Can you post a diff, so I can see which lines you changed?



The following command should work:



diff oldfile newfile > file.patch
Title:
Post by: pjot on August 01, 2009, 12:13:33 PM
The only thing you have to do is removing all '-m32' entries.



Peter
Title:
Post by: TedWalther on August 01, 2009, 04:22:28 PM
Ok.  Can you try using configure-alt, and then type make?



Let me know if the alternative configure script generates a correct Makefile for you.



Ted
Title:
Post by: newdep on August 02, 2009, 02:09:56 AM
newlisp 10.1.1 compiles in the gcc 335 evironment with readline support on OS2

The 'spawn part doesnt work deu to use of mmap inside newlisp which isnt

still not supported in the OS2 port of gcc.





configure and configure-alt return errors.



makefile_os2 contains no tabs



primes.c needs this adjustement but its in the way of win32.



#ifndef OS2

     {"spawn",      p_spawn,  0},

     {"sync",       p_sync,        0},

     {"abort",      p_abort,  0},

     {"send",       p_send,        0},

     {"receive",    p_receive,     0},

     {"share",      p_share,  0},

#endif



Ill check gcc440 and publish on my page the precompiled version.
Title:
Post by: pjot on August 02, 2009, 04:03:14 AM
Quote
Can you try using configure-alt, and then type make?


This 'configure-alt' works fine.


Quote
peter@solarstriker:~/installation/newlisp-10.1.1$ ./configure-alt



Detected Operating System LINUX

Detected memory model LP64, using memory model LP64

UTF-8 is enabled

readline support is disabled



To enable readline support, use the --enable-readline option, and specify the

proper libraries to link by setting the RLFLAGS variable when running

configure.  Also if necessary, use the INCLUDES variable to specify where the

readline header files are."





Type "make" to build your newlisp binary.



If that doesn't work, type "make help"

peter@solarstriker:~/installation/newlisp-10.1.1$ vim Makefile

peter@solarstriker:~/installation/newlisp-10.1.1$ make

make -f makefile_configure

make[1]: Entering directory `/home/peter/installation/newlisp-10.1.1'

gcc -m64  -Wall -pedantic -Wno-long-long -Wno-strict-aliasing -O2 -c  -DNEWCONFIG newlisp.c

gcc -m64  -Wall -pedantic -Wno-long-long -Wno-strict-aliasing -O2 -c  -DNEWCONFIG nl-symbol.c

gcc -m64  -Wall -pedantic -Wno-long-long -Wno-strict-aliasing -O2 -c  -DNEWCONFIG nl-math.c

gcc -m64  -Wall -pedantic -Wno-long-long -Wno-strict-aliasing -O2 -c  -DNEWCONFIG nl-list.c

gcc -m64  -Wall -pedantic -Wno-long-long -Wno-strict-aliasing -O2 -c  -DNEWCONFIG nl-liststr.c

gcc -m64  -Wall -pedantic -Wno-long-long -Wno-strict-aliasing -O2 -c  -DNEWCONFIG nl-string.c

gcc -m64  -Wall -pedantic -Wno-long-long -Wno-strict-aliasing -O2 -c  -DNEWCONFIG nl-filesys.c

gcc -m64  -Wall -pedantic -Wno-long-long -Wno-strict-aliasing -O2 -c  -DNEWCONFIG nl-sock.c

gcc -m64  -Wall -pedantic -Wno-long-long -Wno-strict-aliasing -O2 -c  -DNEWCONFIG nl-import.c

gcc -m64  -Wall -pedantic -Wno-long-long -Wno-strict-aliasing -O2 -c  -DNEWCONFIG nl-xml.c

gcc -m64  -Wall -pedantic -Wno-long-long -Wno-strict-aliasing -O2 -c  -DNEWCONFIG nl-web.c

gcc -m64  -Wall -pedantic -Wno-long-long -Wno-strict-aliasing -O2 -c  -DNEWCONFIG nl-matrix.c

gcc -m64  -Wall -pedantic -Wno-long-long -Wno-strict-aliasing -O2 -c  -DNEWCONFIG nl-debug.c

gcc -m64  -Wall -pedantic -Wno-long-long -Wno-strict-aliasing -O2 -c  -DNEWCONFIG pcre.c

gcc -m64  -Wall -pedantic -Wno-long-long -Wno-strict-aliasing -O2 -c  -DNEWCONFIG nl-utf8.c

gcc newlisp.o nl-symbol.o nl-math.o nl-list.o nl-liststr.o nl-string.o nl-filesys.o    nl-sock.o nl-import.o nl-xml.o nl-web.o nl-matrix.o nl-debug.o pcre.o nl-utf8.o -m64          -lm -ldl  -o newlisp

strip newlisp

make[1]: Leaving directory `/home/peter/installation/newlisp-10.1.1'


Thanks,

Peter
Title:
Post by: TedWalther on August 02, 2009, 08:33:17 AM
I see in your log you ran vim before running the makefile.  Did you have to change anything?
Title:
Post by: pjot on August 02, 2009, 09:09:03 AM
No, just checking how the generated Makefile looks like. It is an old habit of me to use 'vim' for every tiny task. My bad.



Peter
Title:
Post by: HPW on August 03, 2009, 04:23:37 AM
After testing a while the latest 10.1.1 I ran into problem with (source ...) or (print ...)



(device(open CallBackFilename "a"))
(println)
(setq MbiVarStr(source 'CADT_WOINT_BG 'CADT_WOINT_HG 'CADT_WOINT_TG 'CADT_WOINT_NIS_DAT 'CADT_WOINT_SCI_DAT 'CADT_WOINT_WPL_DAT 'CADT_WOINT_WPR_DAT 'CADT_WOINT_BLV_DAT 'CADT_WOINT_BLL_DAT 'CADT_WOINT_BLR_DAT 'CADT_WOINT_BLH_DAT 'CADT_WOINT_FRO_DAT 'CADT_WOINT_ZUB_DAT 'CADT_WOINT_WUF_DAT 'CADT_WOINT_NL_VAR 'CADT_WOINT_NB_VAR 'CADT_WOINT_NB_RAD 'CADT_WOINT_NB_VIS 'CADT_WOINT_NB_ENA))
(print MbiVarStr)
(close (device))


When reaching symbol CADT_WOINT_FRO_DAT the contained assoc-list is broken and source-string generating stopped.

So either in (source ..) or in (print ...) is a problem.

Switching back to 10.1.0 does work OK.



Any changes there?
Title:
Post by: HPW on August 03, 2009, 04:27:57 AM
And why is the size of the DLL so different?



newLISP 10.1.0 238 KB

newLISP 10.1.1 254 KB



??
Title:
Post by: Lutz on August 03, 2009, 09:21:06 AM
None of the routines involved, including internal helpers, have changed for 'source' and 'print' between 10.1.0 and 10.1.1.



My MinGW MSYS and GCC version have been updated from 10.1.0 to 10.1.1 which probably is responsible for the size of the DLLs and may be responsible for the problem you have with 'source' or 'print'.



I made Win32 installer of 10.1.1 with the old environment and have send you a private message with the address, where you can pick it up.
Title:
Post by: HPW on August 03, 2009, 09:26:23 AM
I will try it out.

Thanks.



Hans-Peter
Title:
Post by: HPW on August 03, 2009, 09:52:45 AM
This time I test with the EXE.

My code to reproduce the problem:



(set 'CADT_WOINT_BG "200000")

(set 'CADT_WOINT_HG "250000")

(set 'CADT_WOINT_TG "7800")

(set 'CADT_WOINT_NIS_DAT '(
  (("Artnr" "O0005") ("RaumAr" "INNIS") ("ErfArt" "CADAS") ("HoeheL" "250000") ("HoeheM"
    "250000")
   ("HoeheR" "250000")
   ("HoeheF" "250000")
   ("BreitU" "200000")
   ("BreitM" "200000")
   ("BreitO" "200000")
   ("BreitF" "200000")
   ("Tiefe" "7800")
   ("AufscL" "0")
   ("AufscR" "0")
   ("OffsetX" "0"))))

(set 'CADT_WOINT_SCI_DAT '(
  (("Artnr" "O9402") ("Breite" "200000") ("AnzTu" "2TUER") ("AnzES" "4ESTO") ("AnzMS"
    "OHNE")
   ("AnscLi" "WAND")
   ("AnscRe" "WAND")
   ("OffsetX" "0")
   ("OffsetZ" "0")
   ("UPOS_LISP" "((setq CADT_WOINT_WANDZOFF "2500"))"))))

(set 'CADT_WOINT_WPL_DAT '(
  (("Artnr" "O9550") ("FueFvo" "M01") ("FueFhi" "M01") ("Varian" "L") ("Hoehe" "246000")
   ("OffsetX" "0")
   ("OffsetY" "-4050")
   ("OffsetZ" "0")
   ("UPOS_LISP" "((setq CADT_WOINT_EINBAU "LAUFV"))"))))

(set 'CADT_WOINT_WPR_DAT '(
  (("Artnr" "O9550") ("FueFvo" "M01") ("FueFhi" "M01") ("Varian" "R") ("Hoehe" "246000")
   ("OffsetX" "200000")
   ("OffsetY" "-250")
   ("OffsetZ" "0")
   ("UPOS_LISP" "((setq CADT_WOINT_EINBAU "LAUFH"))"))))

(set 'CADT_WOINT_BLV_DAT nil)

(set 'CADT_WOINT_BLL_DAT nil)

(set 'CADT_WOINT_BLR_DAT nil)

(set 'CADT_WOINT_BLH_DAT nil)

(set 'CADT_WOINT_FRO_DAT '(
  (("Artnr" "O7154") ("FueFvo" "M01") ("FueFhi" "M01") ("ProArl" "GRILV") ("ProAul"
    "DIPRV")
   ("ProArr" "GRILV")
   ("ProAur" "ABDEP")
   ("Laufar" "OBENL")
   ("Buerst" "OHNE")
   ("StopLi" "OHNE")
   ("StopRe" "OHNE")
   ("FarbSt" "")
   ("Anschl" "NACHR")
   ("Hoehe" "244700")
   ("Breite" "98250")
   ("OffsetX" "3500")
   ("OffsetY" "-4300")
   ("OffsetZ" "1300"))
  (("Artnr" "O7154") ("FueFvo" "M01") ("FueFhi" "M01") ("ProArl" "ABSCH") ("ProAul"
    "ABDEP")
   ("ProArr" "GRILV")
   ("ProAur" "DIPRV")
   ("Laufar" "OBENL")
   ("Buerst" "BUERL")
   ("StopLi" "OHNE")
   ("StopRe" "OHNE")
   ("FarbSt" "")
   ("Anschl" "NACHL")
   ("Hoehe" "244700")
   ("Breite" "98250")
   ("OffsetX" "98250")
   ("OffsetY" "-500")
   ("OffsetZ" "1300"))))

(set 'CADT_WOINT_ZUB_DAT nil)

(set 'CADT_WOINT_WUF_DAT nil)

(set 'CADT_WOINT_NL_VAR '(
  ("AnordnungAuswahl" "O22ZL")
  ("AufschlagLi" 0)
  ("AufschlagRe" 0)
  ("BlendenV" "")
  ("BlendenH" "")
  ("BlendenHoe" "BlendenHoe56")
  ("BlenFv" "M01")
  ("BlenFh" "M01")
  ("FueFvo" "M01")
  ("FueFhi" "M01")
  ("FuelF1" nil)
  ("FuelF2" nil)
  ("FuelF3" nil)
  ("FueF1H" nil)
  ("FueF2H" nil)
  ("FueF3H" nil)
  ("LaufSchiene" "LaufSchieneD2")
  ("LaufUnten" "")
  ("NischBreite" 2000)
  ("NischBreiteU" 2000)
  ("NischBreiteM" 2000)
  ("NischBreiteO" 2000)
  ("NischHoehe" 2500)
  ("NischHoeheL" 2500)
  ("NischHoeheM" 2500)
  ("NischHoeheR" 2500)
  ("TuerAnordnung" "AnordnungNI")
  ("TuerAnzahl" "TuerAnzahl2")
  ("Tuerart" "TuerartS")
  ("Tuervar" "TuervarE")
  ("WandansFh" "M01")
  ("WandansFv" "M01")
  ("WandansL" "Checked")
  ("WandansR" "Checked")))

(set 'CADT_WOINT_NB_VAR '(
  ("Anordnung" "O22ZL")
  ("AufschlagLi" "0")
  ("AufschlagRe" "0")
  ("BlendenH" "")
  ("BlendenHoeObj" "")
  ("BlendenV" "")
  ("FRO_Breite" "982.5")
  ("FRO_Hoehe" "2447.0")
  ("FrontArtnr" "O7154")
  ("GesamtBreite" "2000.0")
  ("LaufSchieneObj" "LaufSchieneD2")
  ("LaufUntenObj" "")
  ("LaufUntenPicFile" "")
  ("NischBreite" "2000")
  ("NischBreiteM" "2000")
  ("NischBreiteO" "2000")
  ("NischBreiteU" "2000")
  ("NischHoehe" "2500")
  ("NischHoeheL" "2500")
  ("NischHoeheM" "2500")
  ("NischHoeheR" "2500")
  ("TuerAnordnung" "1")
  ("TuerAnordnungL" "1")
  ("TuerAnordnungR" "1")
  ("TuerAnzahlObj" "")
  ("TuerartObj" "TuerartS")
  ("TuervarObj" "TuervarE")
  ("WandansL" "Checked")
  ("WandansR" "Checked")))

(set 'CADT_WOINT_NB_RAD '(
  ("BlendenHoe" 0)
  ("LaufSchiene" 1)
  ("LaufUnten" nil)
  ("TuerAnzahl" 1)
  ("Tuerart" 0)
  ("Tuervar" 0)))

(set 'CADT_WOINT_NB_VIS '(
  ("AufschlagLi" "False")
  ("AufschlagRe" "False")
  ("GWandansFv" "True")
  ("GWandansFh" "True")
  ("GFueFvo" "True")
  ("GFuelFv" "False")
  ("GFueFhi" "False")
  ("GFuelFh" "False")
  ("GBlenFv" "False")
  ("GBlenFh" "False")))

(set 'CADT_WOINT_NB_ENA '(
  ("LaufSchieneD1" "True")
  ("LaufSchieneD2" "True")
  ("LaufSchieneB1" "True")
  ("LaufSchieneB2" "True")
  ("LaufUntenA" "False")
  ("LaufUntenL" "False")
  ("LaufUntenF" "False")
  ("WandansL" "True")
  ("WandansR" "True")
  ("CWandansFh" "False")
  ("TuerAnzahl1" "False")
  ("TuerAnzahl2" "True")
  ("TuerAnzahl3" "False")
  ("TuerAnzahl4" "True")
  ("BArtikelListe" "True")
  ("BExit" "True")
  ("BWeiter" "True")))


  (device(open "c:\tmp\nl_testout.lsp" "a"))
  (println)
  (setq MbiVarStr(source 'CADT_WOINT_BG 'CADT_WOINT_HG 'CADT_WOINT_TG 'CADT_WOINT_NIS_DAT 'CADT_WOINT_SCI_DAT 'CADT_WOINT_WPL_DAT 'CADT_WOINT_WPR_DAT 'CADT_WOINT_BLV_DAT 'CADT_WOINT_BLL_DAT 'CADT_WOINT_BLR_DAT 'CADT_WOINT_BLH_DAT 'CADT_WOINT_FRO_DAT 'CADT_WOINT_ZUB_DAT 'CADT_WOINT_WUF_DAT 'CADT_WOINT_NL_VAR 'CADT_WOINT_NB_VAR 'CADT_WOINT_NB_RAD 'CADT_WOINT_NB_VIS 'CADT_WOINT_NB_ENA))
  (print MbiVarStr)
  (close (device))


So the new version fail with this and the version made with the old enviroment works.

So they are not completly compatible.
Title:
Post by: HPW on August 03, 2009, 10:18:38 AM
The new version seems to fail with a amount of 2050 Bytes of source string.
Title:
Post by: Lutz on August 03, 2009, 10:47:37 AM
The new version of MinGW GCC seems to limit to 2048 characters. You can show this with the following simple program on Win32 (v 10.1.1):


(setq atxt (string (sequence 1 1000)))
(println atxt)

(write-file "text.txt" atxt )

(exit)


it will cut of at about 500 for the display but save it in full length. I may have to go back to the older GCC when on Win32, if I cannot work around this.
Title:
Post by: Lutz on August 03, 2009, 12:25:08 PM
A release update 10.1.2 for Win32 only has been posted here to correct this problem:



http://www.newlisp.org/downloads/newlisp-10102-win-gs-132.exe