Hello,
We finally have a definitive position from Oracle - thanks for posting
Serguei.
Serguei states Oracle's position in a number of points, which I'll
brutally summarize here:
1. While the API is a good idea, there is no demand from the community.
2. The target audience is very limited.
3. Oracle's resources are limited, and the JSR is targeted at a limited
audience.
4. Would need more feedback from the community - but wouldn't be done in
the JDK 8 time frame.
We've been through the process of getting community feedback. The
response was generally positive. Steve can fill us in further.
The target audience for the API was limited - tool vendors and JDK
support. The target audience for the tools that were to be based on the
Kato API were Java developers, support, and operations.
My view is that while the API is an interesting concept, it is too low a
level for there ever to be community interest in it's development -
there are so few vendors. When we started this there was at least Sun,
Oracle (who had just bought BEA), Intel and IBM. Now we are reduced to
Oracle and IBM, which is a very small community indeed, and as they are
both cooperating somewhat through OpenJDK that is even smaller.
I see us continuing in the following form:
1. JSR-326 continues on its standardization of the API. The API as it is
no longer considered part of the Kato project, and isn't actively
developed here.
2. Kato continues with the development of API implementations for the
HotSpot JVM of whatever formats it makes available, or we can make
available.
3. IBM makes available its implementations of JSR 326.
4. We write useful tooling in Kato that make use of JSR-326.
5. Once we have something compelling for the community to use and build
on, demand becomes so great that efforts in OpenJDK snowball and a
post-mortem diagnostics community becomes self sustaining.
Can we produce something that is compelling?
Allowing people to attach their existing debuggers to JVM dumps would be
a good improvement on the existing situation. In the world of 10's of
GBs of heap it is less compelling, but for desktop and smaller
applications, it is more practicable. Java is still behind the likes of
gdb and dbx in this respect. I imagine the ordinary Java developer could
benefit from this.
But is there anything else?
Goodnight,
Stuart
On 26/01/12 16:53, Steve Poole wrote:
Hi all.
Apache Kato was initially created to develop the specification and
reference implementation for JSR 326. That is by definition a
cross-industry activity. To complete the JSR and create an industry
standard API we need a community which has participants on both sides of
the API: those who would use the API to access data and those who would
provide the data to be read. Without a broad consumer/provider community
we are not going to be able to complete the JSR any time soon.
The rebooting of the OpenJDK community and new input from Oracle on their
situation leads me to suggest a new direction for Kato at this time.
I hope that Oracle will post their own statement but I think it's fair
for me to say that, at this point, they have other business concerns that
are more pressing than helping us out right now. That may change later.
The good news is that with the OpenJDK reboot we have a real opportunity
to address some of our technical challenges directly without requiring
Oracles assistance. I'm already an active participant in OpenJDK and I do
intend to scratch a few itches over time that way.
Given all that I'd like to propose the following simple plan to revitalise
Apache Kato
1 - We ignore or put on hold the JSR work. With limited involvement from
Oracle at the moment we have to assume that its not going to complete
anytime soon and maybe never.
2 - We refocus Kato towards post-mortem diagnostics tools. We already
have a good set of tools and I'd like to improve them and add more. For
instance I have some ideas for a Java serialization diagnostics tool that
would fit well here.
3 - Where we find a need for new data from the JVM we (as individuals) work
with the OpenJDK community to make that happen.
On Wed, Jan 18, 2012 at 10:39 AM, Stuart Monteith<stuk...@stoo.me.uk>wrote:
-8 Snip!
