Wouldn't we also be looking at, at least three different versions? AIX uses
SRC to control system services, Mac OS X uses launchd, while Linux uses
init/inetd.
On Feb 21, 2011, at 9:44 AM, David Ashley wrote:
> Hummm, I am not sure this would work without doing some weird stuff.
>
> Yes, SysFi
Hummm, I am not sure this would work without doing some weird stuff.
Yes, SysFile has two implementations, but it also has two different make files
(Windows and *nix). I don't think we currently have any conditional products
for
the Windows or *nix versions of ooRexx (considered separately). Fo
On Mon, Feb 21, 2011 at 6:42 AM, Rick McGuire wrote:
> One approach we've used a lot is to have the build for a particular
> platform determine which code gets linked in to an executable. For
> example, the SysFile class has two distinct implementations, and the
> makefiles determine which one g
Oh, and someone mentioned that having a command line option to not daemonize
would benefit in debugging.
Bruce
On Feb 21, 2011, at 6:38 AM, David Ashley wrote:
> While I am not actively working on rxapi, it is on my todo list. Currently, I
> am
> leaning in the direction of using command line
I think a command line switch would work just fine on OSX.
On Feb 21, 2011, at 6:38 AM, David Ashley wrote:
> While I am not actively working on rxapi, it is on my todo list. Currently, I
> am
> leaning in the direction of using command line arguments to determine whether
> or
> not to deamoni
One approach we've used a lot is to have the build for a particular
platform determine which code gets linked in to an executable. For
example, the SysFile class has two distinct implementations, and the
makefiles determine which one gets compiled and linked. For the rxapi
executable, we have the
While I am not actively working on rxapi, it is on my todo list. Currently, I
am
leaning in the direction of using command line arguments to determine whether
or
not to deamonize the client. But I am open to other ideas about how to
implement
the needed functionality. I will probably start on
Ok, I may have misunderstood the message.
Bruce
On Feb 20, 2011, at 7:33 AM, Rick McGuire wrote:
> On Sun, Feb 20, 2011 at 10:23 AM, CVBruce wrote:
>> Rony
>> This is a known problem. I've brought it up before. The problem is that
>> Mac OS X daemons that are controlled by launchd, are not all
On Sun, Feb 20, 2011 at 10:23 AM, CVBruce wrote:
> Rony
> This is a known problem. I've brought it up before. The problem is that
> Mac OS X daemons that are controlled by launchd, are not allowed to
> daemonize themselves. Since the current rxapi daemonizes a child process,
> and then exits, l
Rony,
The behavior is the same in all ooRexx 4.* versions.
Bruce
On Feb 20, 2011, at 6:19 AM, Rony G. Flatscher wrote:
> Hi there,
>
> while looking at the Console's log entries ("Application - Console") it seems
> that launchd tries restarting the rxapi daemon in 10 second intervals. Is
> th
Rony
This is a known problem. I've brought it up before. The problem is that Mac
OS X daemons that are controlled by launchd, are not allowed to daemonize
themselves. Since the current rxapi daemonizes a child process, and then
exits, launchd looses all communication/control of the child pro
Hi there,
while looking at the Console's log entries ("Application - Console") it
seems that launchd tries restarting the rxapi daemon in 10 second
intervals. Is this intentional or a potential problem?
This is how the rxapi-daemon gets started (copied from Bruce's script):
launchctl load -w
12 matches
Mail list logo