etch-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From scott comer <>
Subject [discuss] release 1.0.2
Date Thu, 22 Jan 2009 21:37:33 GMT
here is a link to my original 1.0.2 release plan:

We've made good progress on finding and fixing bugs.

Bug 	ETCH-6 <> 	if free pool 
exceeds maximum number of threads the exception thrown will kill the 
Bug 	ETCH-8 <> 	canceling 
mailbox notification registration throws exception if not registered
Bug 	ETCH-11 <> 
AlarmManager deadlocks during shutdown()
Bug 	ETCH-18 <> 	KeepAlive 
throws exception while trying to shutdown connection
Bug 	ETCH-23 <> 	FreePool 
has synchronization issues with quick back to back calls to Run
Bug 	ETCH-24 <> 	The 
message direction is always "server" for the result_method messages in 
the java-binding valuefactory
Bug 	ETCH-25 <> 
DefaultDeliveryService ends call with mb.CloseDelivery() and it should 
end it with mb.CloseRead()

We've also made a few improvements to the source and distribution.

Improvement 	ETCH-13 <> 
Copy the rest of the example sources into the distribution
Improvement 	ETCH-17 <> 
SessionListener needs to be more independent
Improvement 	ETCH-19 <> 
Break out common tcp transport options so that they may be shared among 
various transports.

These next items are not in the 1.0.2 bucket but patches are ready and 
are voting to become so.

Bug 	ETCH-27 <> 	mixins 
cause trouble when two or more mixed in files define a type with the 
same name.
Bug 	ETCH-28 <> 	Examples 
do not build due to outdated ant files.
Improvement 	ETCH-29 <> 
Installing all Maven artifacts along with their sources.

These next items are currently in the 1.0.2 bucket but I want to move 
them out. Java code has been committed to support etch-10, but is still 
experimental in nature and not ready for prime time. A proper 
implementation of etch-10 depends upon etch-20. There is also a 
prototype of etch-15, but I do not want to delay what has essentially 
become a bug fix release to wait for it.

Improvement 	ETCH-10 <> 
runtimes cannot support a large number of connections
New Feature 	ETCH-14 <> 
Design and provide example of using forrest to deploy etch documentation
New Feature 	ETCH-15 <> 
design and implement a framework for interoperability testing, plus a 
simple demonstration test
Improvement 	ETCH-16 <> 
reinvigorate the xml format of the basic etch transport
Improvement 	ETCH-20 <> 
need a more flexible buffering scheme to handle different transports
Task 	ETCH-21 <> 	make the 
official build process work again

So, what do y'all think? Make 1.0.2 a primarily bug fixing release and 
throw out the extra new features and improvements?

I will call a vote on monday if there is no substantial changes per 

scott out

  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message