Re: [Oorexx-devel] Odd behaviour

2019-02-01 Thread Erich Steinböck
> > * !!! 'lib94 *-* allInstances[.Context~stackFrames[1]] = > .Context~stackFrames[1]~hasMethod("makeString").dylib' > works as designed/expected or something is fishy ? > Hi Enrico, this is to be expected. It's from a test case that runs (among many other methods) package~loadLibrary(*i

[Oorexx-devel] Odd behaviour

2019-01-31 Thread Enrico Sorichetti via Oorexx-devel
Hi all I was trying to understand the flow of the library loading so I modified common/platform/unix/SysLibrary.cpp adding one printf after line 93 91 >>// if not found, then try from /usr/lib 92 >>if (libraryHandle == NULL) 93 >>{ ADDED printf("* !!! '%s'\n",nameBuffer);

Re: [Oorexx-devel] odd behaviour

2015-02-10 Thread Enrico Sorichetti
Thank You Rick, I will stick to clang ! Enrico > On 10 Feb 2015, at 16:25, Rick McGuire wrote: > > There is a known bug in older versions of gcc with namespace support. > Upgrading the compiler level is the only known fix. > > Rick > ---

Re: [Oorexx-devel] odd behaviour

2015-02-10 Thread David Ashley
The development version ooRexx 5.0 will not link on RHEL 6 or CentOS 6 (or anything earlier) due to a linker bug in those releases. This also includes some Ubuntu releases. That is why they are not being built by the nightly build process. David Ashley On 02/10/2015 09:11 AM, Enrico Sorichetti

Re: [Oorexx-devel] odd behaviour

2015-02-10 Thread Rick McGuire
There is a known bug in older versions of gcc with namespace support. Upgrading the compiler level is the only known fix. Rick On Tue, Feb 10, 2015 at 10:11 AM, Enrico Sorichetti wrote: > I just pulled svn Rev: 10882 > > it builds and works on APPLE Yosemite using > — Apple LLVM version 6.0 (cl

[Oorexx-devel] odd behaviour

2015-02-10 Thread Enrico Sorichetti
I just pulled svn Rev: 10882 it builds and works on APPLE Yosemite using — Apple LLVM version 6.0 (clang-600.0.56) (based on LLVM 3.5svn) it builds and seems to work on CENTOS 7 using — c++ (GCC) 4.8.2 20140120 (Red Hat 4.8.2-16) it fails on CENTOS 6 using — c++ (GCC) 4.4.7 20120313 (Red Hat 4

Re: [Oorexx-devel] Odd behaviour in native code, probably a bug ?

2012-06-07 Thread Rony G. Flatscher
On 07.06.2012 13:47, Rony G. Flatscher wrote: > On 07.06.2012 13:20, Rick McGuire wrote: >> It is a bug in your code. NULL is not a valid item value. > OK, but all directory methods seem to work (e.g. returning .nil when fetching > the entry), except > that one cannot REMOVE the entry, although

Re: [Oorexx-devel] Odd behaviour in native code, probably a bug ?

2012-06-07 Thread Rony G. Flatscher
On 07.06.2012 13:20, Rick McGuire wrote: > It is a bug in your code. NULL is not a valid item value. OK, but all directory methods seem to work (e.g. returning .nil when fetching the entry), except that one cannot REMOVE the entry, although it returns .nil, as if the entry got removed successfu

Re: [Oorexx-devel] Odd behaviour in native code, probably a bug ?

2012-06-07 Thread Rick McGuire
It is a bug in your code. NULL is not a valid item value. Rick On Thu, Jun 7, 2012 at 7:14 AM, Rony G. Flatscher wrote: > To make a long story (two days of debugging) short: in native code a > directory object gets created. > > Then using: > > SendMessage2(dir, "PUT", item, index) > > where ite

[Oorexx-devel] Odd behaviour in native code, probably a bug ?

2012-06-07 Thread Rony G. Flatscher
To make a long story (two days of debugging) short: in native code a directory object gets created. Then using: SendMessage2(dir, "PUT", item, index) where item is NULL/NULLOBJECT works without an error and an entry in the directory gets created. >From this moment on it is not possible to