pcp
[Top] [All Lists]

Re: Java JNI PMAPI

To: Ken McDonell <kenj@xxxxxxxxxxxxxxxx>
Subject: Re: Java JNI PMAPI
From: Mark Goodwin <markgw@xxxxxxx>
Date: Thu, 24 Nov 2005 08:42:17 +1100 (EST)
Cc: Alan Hoyt <ahyt@xxxxxxxxxxxxx>, pcp@xxxxxxxxxxx
In-reply-to: <1132774670.4368.5.camel@xxxxxxxxxxxxxxxxxxxxx>
References: <20050411093715.A15051@xxxxxxxxxxxxxxx> <Pine.LNX.4.61.0504220823110.9546@xxxxxxxxxxxxxxxxxxxxxxxxx> <436A02CA.3020202@xxxxxxxxxxxxx> <436ADFC1.4000403@xxxxxxxxxxxxx> <1132774670.4368.5.camel@xxxxxxxxxxxxxxxxxxxxx>
Sender: pcp-bounce@xxxxxxxxxxx
On Thu, 24 Nov 2005, Ken McDonell wrote:

> Alan (and others),
> 
> While I'm no longer at SGI, I remain active (and passionate) about PCP.
> 
> Mark has a new PCP open source release that is pending, this will
> include the pieces necessary for the Java stuff.  

I know I have been promising for a while, but yes this release will
happen real soon now :-)

We have a late-breaking change that I'd like to try and include.
Basically, it provides a callback hook for PMAPI functions such
as pmFetch() to augment the select mask when blocked in __pmGetPDU().
One thing this would allow is a PMDA to also be a client (deadlock
avoidance assumed). If anyone in open source land would be interested
in helping out with this, please let me know.

> Once that happens, I
> believe interested parties (which now includes me) could petition SGI to
> make the unreleased and unsupported PCP components (Java and DBMS chunks
> in particular) available under the GPL to anyone who's willing to take
> on the gatekeeper role.

I do not anticipate any problems with this. I think it will happen
provided we can get interested parties moving on a new sourceforge
project ...

> On another front, I'm well down the track to an open source pmchart
> replacement (kmchart), based on Qt and qwt ... send me e-mail if you'd
> be interested in participating in this development effort.

... might be a good first chunk of source for a new sourceforge project ;-)

Thanks
-- Mark

<Prev in Thread] Current Thread [Next in Thread>