Re: suggestion to change in Makefile

2009-12-10 Thread Alexander Burger
Hi Boh Yap, thanks for the info! > .and $ ls /usr/libexec/gcc/darwin/ shows: > i386/ ppc/ppc64/ x86_64/ So this looks quite promising :-) > 2. In Apple's Developer Docs they admit its not updated... > ... > ..and in the addressing mode chapter, it doesn't mention 64 bit > registers fo

Re: suggestion to change in Makefile

2009-12-10 Thread Boh Yap
hi Alex, re: Mac OS X assembler, a quick check shows the followg: (I have also include the output of 'man as' as an attachment) BTW I'm running OS X 10.6.2, Developer Tools (XCode) 3.2 The working build of picolisp 32bit was from the testing version, the stable version still gave problem as off D

Re: suggestion to change in Makefile

2009-12-10 Thread Alexander Burger
Hi Boh Yap, > I can also confirm it builds and passes test OK on Snow Leopard Mac OSX. Great! Thanks. > backgrnd... should I wait for the 64 bit version? And if so how much > longer? While the 64-bit version itself is finished, and quite good tested on Linx meanwhile, building it on Snow Leopa

Re: suggestion to change in Makefile

2009-12-10 Thread Boh Yap
hi all, I can also confirm it builds and passes test OK on Snow Leopard Mac OSX. I am just starting to learn picolisp, comming from a Python backgrnd... should I wait for the 64 bit version? And if so how much longer? thnx On Wed, Dec 9, 2009 at 11:55 PM, Jon Kleiser wrote: > Hi Alex, > > You

Re: suggestion to change in Makefile

2009-12-09 Thread Alexander Burger
Hi Jon, > Your new testing release works fine on my Intel Mac. The test gives OK. Good. I successfully tested on a 32-bit machine, and a 32-bit build on a 64-bit machine. Looks also all right. So let's keep it until somebody complains ;-) Cheers, - Alex -- UNSUBSCRIBE: mailto:picol...@software

Re: suggestion to change in Makefile

2009-12-09 Thread Jon Kleiser
Hi Alex, Your new testing release works fine on my Intel Mac. The test gives OK. /Jon > Hi Jon, hi Mansur, > >> When I do ... >> $(/bin/pwd)/p lib/test.l -bye >> .. then I get OK. > > Yep, the explicit path is essential, because otherwise the path-checking > routines have no path to check for ;-

Re: suggestion to change in Makefile

2009-12-09 Thread Alexander Burger
Hi Jon, hi Mansur, > When I do ... > $(/bin/pwd)/p lib/test.l -bye > .. then I get OK. Yep, the explicit path is essential, because otherwise the path-checking routines have no path to check for ;-) So, now I edited Mansur's suggestion into "src/Makefile". I hope I got it right. Anybody who lik

Re: suggestion to change in Makefile

2009-12-09 Thread Jon Kleiser
> Hi Jon, > >> This seemed to work OK on my Intel Mac (OSX 10.5.8). I've spotted no >> problem this far. Maybe I should run a complete test ... How do I do it? > > Great. Thanks! > > You could run the unit tests: > >cd /picoLisp/installation/directory > >$(/bin/pwd)/p lib/test.l -bye > > If

Re: suggestion to change in Makefile

2009-12-09 Thread Jon Kleiser
> Hi, > > I just downloaded the testing version of Pico Lisp, and tried Mansur's > Makefile patch, just modified for Mac/Darwin: > > --- > OS = Darwin > - PICOLISP-FLAGS = -m32 -rdynamic -lc -lm -ldl > + PICOLISP-FLAGS = -m32 -rdynamic > DYNAMIC-LIB-FLAGS = -m32 -shared -export-dynamic >

Re: suggestion to change in Makefile

2009-12-09 Thread Alexander Burger
Hi Jon, > This seemed to work OK on my Intel Mac (OSX 10.5.8). I've spotted no > problem this far. Maybe I should run a complete test ... How do I do it? Great. Thanks! You could run the unit tests: cd /picoLisp/installation/directory $(/bin/pwd)/p lib/test.l -bye If it just says 'OK' e

Re: suggestion to change in Makefile

2009-12-09 Thread Jon Kleiser
> On Wed, Dec 09, 2009 at 04:37:00PM +0500, Mansur Mamkin wrote: >> default on that Linux. So AltLinux team suggests to put -l >> arguments after obj-files as this is more standard-compliant >> (see also http://www.gentoo.org/proj/en/qa/asneeded.xml) > > This proposal makes sense. > > It would be n

Re: suggestion to change in Makefile

2009-12-09 Thread Alexander Burger
On Wed, Dec 09, 2009 at 04:37:00PM +0500, Mansur Mamkin wrote: > default on that Linux. So AltLinux team suggests to put -l > arguments after obj-files as this is more standard-compliant > (see also http://www.gentoo.org/proj/en/qa/asneeded.xml) This proposal makes sense. It would be nice if us

suggestion to change in Makefile

2009-12-09 Thread Mansur Mamkin
Hi all, what do you think about to do small modification in the Makefile I propose this because of link error on my Linux box (AltLinux Desktop). As I undestood, this is because of using --as-needed linker flag as default on that Linux. So AltLinux team suggests to put -l arguments after obj-fil