Re: [dev] Netbeans 6 plugin

2007-09-28 Thread Juergen Schmidt

Craig Tataryn wrote:

Yes!  Thanks.  I was getting distracted with the test package
dependencies problems but once I applied your setting to my
private.properties file and right clicked on the project and chose
Create NBM it worked!

good to hear that



Thanks again,

P.S. How do you subscribe to the api list?  I couldn't find a
subscribe address anywhere.  I guessed at [EMAIL PROTECTED]
but got a bounce.
you have to login on the api project site and than you find on the left 
site a link Mailing lists. There you find everything you need. Don't 
ask me why it is so ;-)


Juergen




Craig.

On 9/27/07, Steffen Grund [EMAIL PROTECTED] wrote:

Hi Craig,

try checking out the latest changes I made this morning. You have a
chance to compile the plug-in with the following line added to
nbproject/private/private.properties:

jhall.jar=netbeans_installation_directory/harness/jsearch-2.0_04.jar

I just presume, that the jsearch.jar version is the same in your
NetBeans. You'd better check this.

Please let me know if you run into any other issues.

Regards, Steffen


Craig Tataryn wrote:

I read in the archives about someone trying to get OO plugin working
for Netbeans 6, the solution was to download and build from source.

I did that, but have been having trouble.  It seems the default build
script in HEAD is NB 5.x specific (it relies on certain dependencies
being available that presumably you've added through the update
centre).

Is there perhaps a certain branch I should be checking out that has an
NB 6 build script?

Thanks,

p.s. if I've gotten the wrong mailing list, could you kindly forward
me the address to subscribe to the api list?

Craig.

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]


-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]




-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]



-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]



Re: [dev] Netbeans 6 plugin

2007-09-28 Thread Laurent Godard

HI


P.S. How do you subscribe to the api list?  I couldn't find a
subscribe address anywhere.  I guessed at [EMAIL PROTECTED]
but got a bounce.
you have to login on the api project site and than you find on the left 
site a link Mailing lists. There you find everything you need. Don't 
ask me why it is so ;-)


[EMAIL PROTECTED]
and then reply to confirmation message

Laurent

--
Laurent Godard [EMAIL PROTECTED] - Ingénierie OpenOffice.org - 
http://www.indesko.com
Nuxeo Enterprise Content Management  http://www.nuxeo.com - 
http://www.nuxeo.org

Livre Programmation OpenOffice.org, Eyrolles 2004-2006

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]



Re: [dev] Proposed Thread / Process Lifecycle

2007-09-28 Thread Kay Ramme - Sun Germany - Hamburg

Hi Joerg,

Joerg Budischewski wrote:

Hi Kay,

hmm, your proposal looks more like a concept rather than an API proposal .
exactly! Everybody may implement it as appropriate, the point being that 
all daemon threads eventually get joined, e.g when de-initializing a 
shared library.




Will there be an API somewhere in sal for this ? If yes, I would rather 
see the API docs before actually putting comments.
See above. I may make some helper functions available in SAL later on, 
certainly along the lines of the example code.
Only critical point is the proper synchronization of the activies (AKA 
non-daemon threads), as only the last one is allowed to call exit.


I can't see how your concept should work work for daemon threads, when 
shared libraries are dynamically loaded. From man atexit


--snip--
The  atexit() function registers the given function to be called at 
normal program termination, either via exit(3) or via return from the 
program's main().  Functions so registered are called in the reverse 
order of their registration; no arguments are passed.

--snap--


Dynamically using a shared library leads to registered atexit handlers 
being called when dlcloseing.


See man atexit:
 Linux Notes
   Since glibc 2.2.3, atexit() (and on_exit(3)) can be used within 
a shared library  to  establish  functions  that  are

   called when the shared library is unloaded.

Though I don't know about other platforms. But atexit used by 
dynamically opened libraries may be a problem anyway.


.. AFAIK, the static destructors of objects in shared libraries are 
called by atexit functions, arent they ? So your daemon_join_ would be 
called too late.
They are called at atexit time only in case of process termination, 
otherwise they are called indirectly by dlclose, so this is not a problem.




I'd rather prefer an explicit daemon_join in main() to avoid those kind 
of problems ...
That would be wrong for daemons, as daemons are implementation details, 
being implemented as needed.
Something similar is needed for activities (please see my example code), 
to properly synchronize calling exit.



Bye,

Joerg


Thanks for your help :-)


   Kay



Kay Ramme - Sun Germany - Hamburg wrote:


-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]



[dev] WARNING: Do Not Install IBM Lotus Symphony (Beta)

2007-09-28 Thread Allen Pulsifer
To all the OpenOffice users who might be interesting in trying out IBM Lotus
Symphony (Beta):

   DO NOT INSTALL IBM Lotus Symphony (Beta) !!!

I tried it out.  It took over all of my OpenDoc file associations, without
warning, and installed itself into a handful of other file associations.  IT
HAS NO UNINSTALLER, SO THESE CHANGES CANNOT BE EASILY REVERSED!

Basically, it made a mess out of my registry.  Its taking me about an hour
to fix it.

WARNING: INSTALL IBM Lotus Symphony (Beta) AT YOUR OWN PERIL!

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]



Re: [dev] Proposed Thread / Process Lifecycle

2007-09-28 Thread Tor Lillqvist
 . AFAIK, the static destructors of objects in shared libraries are 
 called by atexit functions, arent they ? 

The semantics of functions registered with atexit() when dynamically loaded 
shared libraries are involved is completely system-dependent and a big mess. 
atexit() is close to unusable because what it actually does, exactly, is so 
underspecified. Im my not so humble opinion atexit() should be banned. I 
seriously doubt any C++ implementation directly uses the C atexit() function to 
handle destruction of static objects.

--tml


-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]



Re: [dev] WARNING: Do Not Install IBM Lotus Symphony (Beta)

2007-09-28 Thread Benjamin Huot
Its in the Symphony Program folder on your hard drive where the  
program is located at.


On Sep 28, 2007, at 9:34 AM, Allen Pulsifer wrote:


Well I tried it out on Windows too and I could install and
de-install it, using the Software option in the System
folder. However, some associations (I think ODT or ODS) did
not get restored correctly.


Hello Rony,

Kindly forgive my confusion.  What is the software option in the  
system
folder?  I looked in Control Panel | Add/Remove Programs and could  
not find
an entry for Symphony.  Then I tried rerunning the installer and it  
only
offered to install, as if it had never been installed.  It had no  
option for

repair or remove.

Allen

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]



-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]



RE: [dev] WARNING: Do Not Install IBM Lotus Symphony (Beta)

2007-09-28 Thread Allen Pulsifer
 Well I tried it out on Windows too and I could install and 
 de-install it, using the Software option in the System 
 folder. However, some associations (I think ODT or ODS) did 
 not get restored correctly.

Hello Rony,

Kindly forgive my confusion.  What is the software option in the system
folder?  I looked in Control Panel | Add/Remove Programs and could not find
an entry for Symphony.  Then I tried rerunning the installer and it only
offered to install, as if it had never been installed.  It had no option for
repair or remove.

Allen

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]



Re: [dev] WARNING: Do Not Install IBM Lotus Symphony (Beta)

2007-09-28 Thread Rony G. Flatscher

Allen Pulsifer wrote:
 To all the OpenOffice users who might be interesting in trying out IBM Lotus
 Symphony (Beta):

DO NOT INSTALL IBM Lotus Symphony (Beta) !!!

 I tried it out.  It took over all of my OpenDoc file associations, without
 warning, and installed itself into a handful of other file associations.  IT
 HAS NO UNINSTALLER, SO THESE CHANGES CANNOT BE EASILY REVERSED!

 Basically, it made a mess out of my registry.  Its taking me about an hour
 to fix it.

 WARNING: INSTALL IBM Lotus Symphony (Beta) AT YOUR OWN PERIL!
   
Well I tried it out on Windows too and I could install and de-install
it, using the Software option in the System folder. However, some
associations (I think ODT or ODS) did not get restored correctly.

Start-up times are abmysal in this beta drop, not sure whether IBM is
able to improve that.

---rony

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]



Re: [dev] Proposed Thread / Process Lifecycle

2007-09-28 Thread Joerg Budischewski

Hi Kay,

hmm, your proposal looks more like a concept rather than an API proposal .

Will there be an API somewhere in sal for this ? If yes, I would rather 
see the API docs before actually putting comments.


I can't see how your concept should work work for daemon threads, when 
shared libraries are dynamically loaded. From man atexit


--snip--
The  atexit() function registers the given function to be called at 
normal program termination, either via exit(3) or via return from the 
program's main().  Functions so registered are called in the reverse 
order of their registration; no arguments are passed.

--snap--

. AFAIK, the static destructors of objects in shared libraries are 
called by atexit functions, arent they ? So your daemon_join_ would be 
called too late.


I'd rather prefer an explicit daemon_join in main() to avoid those kind 
of problems ...



Bye,

Joerg

Kay Ramme - Sun Germany - Hamburg wrote:

Hi again,

if no one objects, this is going to become the standards policy 
regarding thread life cycles soon.


First thing I am going to change accordingly is the pyuno bridge ...

   Kay

Kay Ramme - Sun Germany - Hamburg wrote:


Hi OOo developers,

once a while we face the situation, that it is unclear how a long a 
thread should live, if it should be cancellable, and when the 
containing process is going to terminate (e.g. 
http://udk.openoffice.org/issues/show_bug.cgi?id=80300).


Therefor I wrote a short proposal about controlling the life time, 
supplemented by some example code to illustrate the behavior.


If you are interested, please have a look at
  http://wiki.services.openoffice.org/wiki/User:Kr/A_Thread%27s_Life
and comment on the discussion page.

Ideally the proposal becomes a specification, being mandatory for 
thread implementors.


Thanks for your help and feedback


 Kay

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]



-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]




-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]



Re: [dev] Proposed Thread / Process Lifecycle

2007-09-28 Thread Kay Ramme - Sun Germany - Hamburg

Tor,

Tor Lillqvist wrote:
. AFAIK, the static destructors of objects in shared libraries are 
called by atexit functions, arent they ? 


The semantics of functions registered with atexit() when dynamically loaded 
shared libraries are involved is completely system-dependent and a big mess. 
atexit() is close to unusable because what it actually does, exactly, is so 
underspecified. Im my not so humble opinion atexit() should be banned. I 
seriously doubt any C++ implementation directly uses the C atexit() function to 
handle destruction of static objects.
you are reading my thoughts, especially wrt atexit vs. C++ (see 
http://wiki.services.openoffice.org/wiki/Uno/Cpp/Tutorials/Global_References 
for some analysis :-)




--tml


Kay

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]



Re: [dev] WARNING: Do Not Install IBM Lotus Symphony (Beta)

2007-09-28 Thread Benjamin Huot

go into the _uninst folder and launch uninstaller.bin

On Sep 28, 2007, at 9:49 AM, Allen Pulsifer wrote:


Its in the Symphony Program folder on your hard drive where the
program is located at.


I also looked for a Symphony folder under the Start menu (in case  
it had an
uninstaller) but Symphony does not install its Start items into a  
folder.


So I assume you mean I should have looked in

   C:\Program Files\IBM\Lotus\Symphony\ ?

What would I have been looking for, if I had not just spent the  
last hour

cleaning the registry by hand?

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]



-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]



Re: [dev] WARNING: Do Not Install IBM Lotus Symphony (Beta)

2007-09-28 Thread Caio Tiago Oliveira

Allen Pulsifer, 28-09-2007 13:49:
Its in the Symphony Program folder on your hard drive where the  
program is located at.


I also looked for a Symphony folder under the Start menu (in case it had an
uninstaller) but Symphony does not install its Start items into a folder.

So I assume you mean I should have looked in

   C:\Program Files\IBM\Lotus\Symphony\ ?

What would I have been looking for, if I had not just spent the last hour
cleaning the registry by hand?


Even OOo won't restore the file associations to the older program...

the easier way to fix it is to repair the OOo setup or reinstall.



-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]



RE: [dev] WARNING: Do Not Install IBM Lotus Symphony (Beta)

2007-09-28 Thread Allen Pulsifer
 Even OOo won't restore the file associations to the older program...
 the easier way to fix it is to repair the OOo setup or reinstall.

Yes, I tried that.  OOo very politely respected the Symphony file
associations and did no repair.  The only way I got it to work was to delete
all traces of Symphony from the file associations.  If the file association
was Symphony-only or Symphony and OpenOffice, then I simply deleted it and
fixed it later with a reinstall of OOo.  For all other associations, I only
deleted the Symphony entry.  Hint for anyone suffering from the same
problem: I eventually figured they could be easily located by searching for
the partial string standalone_ (without quotes).  Better solution: don't
install Symphony until IBM gets its act together.  Its particularly
unforgivable IMO for beta software not to uninstall without a trace because
that's what its for: to install, try out, and then delete.

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]



Re: [dev] WARNING: Do Not Install IBM Lotus Symphony (Beta)

2007-09-28 Thread Hubert Figuiere

On Fri, 2007-09-28 at 11:56 -0400, Allen Pulsifer wrote:
 To all the OpenOffice users who might be interesting in trying out IBM
 Lotus
 Symphony (Beta):
 
DO NOT INSTALL IBM Lotus Symphony (Beta) !!!

Here the TnstallShield crash and let a process running with a high CPU
load from /tmp

I'm not running Windows.

Hub

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]