db-derby-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Mamta Satoor <msat...@gmail.com>
Subject Re: Google Summer of Code thoughts
Date Thu, 29 Jan 2009 14:26:12 GMT
Bryan, thanks for thinking far ahead in Summer already. If we start
discussing the project list now, we will have a good list of projects
when the program actually starts. I like all the project ideas and
Kathey's Derby Test and Fix idea. But like Kathey, I am not too sure
about the optimizer related fixes. It might be a little bit too much
to bite for the short time that interns will be involved. Just my


On Wed, Jan 28, 2009 at 4:00 PM, Kathey Marsden
<kmarsdenderby@sbcglobal.net> wrote:
> Bryan Pendleton wrote:
>> I've been thinking about next summer's Google Summer of Code
>> (http://code.google.com/soc). In particular, I've been thinking
>> about possible Derby-related projects that we might suggest to
>> potential SoC interns.
>> I've listed several possibilities below, and am interested in
>> getting some feedback, including:
>>  - do these projects seem appropriate (not too big, not too small,
>>   can be sub-divided into smaller projects and smaller steps
>>   successfully, have a coherent theme, would be appealing to an
>>   intern, would be well-received by the Derby community, etc.)
>>  - are there other projects like these that are worth considering?
>>  - how can we attract interns motivated to succeed at efforts
>>   like these?
>> Would you be interested in getting involved as a mentor for
>> one of these projects, or perhaps for another Derby project?
>> The more mentors we can get, the more students we can host, and
>> the more the Derby community benefits.
>> What is the next step? How should we capture and develop project
>> ideas such as these, and how can we start building interest for
>> interns who might want to apply for these projects? Should we
>> put this information somewhere in the wiki? How could we organize it?
>> thanks,
>> bryan
>> ========================
>> Project: Derby Query Optimizer.
>> Join the Derby community and contribute to the ongoing development
>> of Derby's powerful and sophisticated query optimizer. We'll fix
>> optimizer bugs such as DERBY-4007, DERBY-3753, DERBY-3892 or
>> DERBY-2752; we'll improve servicability by enhancing tracing
>> (DERBY-837) or building tools for capturing and analyzing query plans;
>> if time permits, we could look to add new features such as
>> common sub-expression elimination (DERBY-890) or improved
>> sub-query processing (DERBY-2231). Work is also needed on optimizer
>> performance (DERBY-1905, DERBY-1906, DERBY-1907, DERBY-1908,
>> DERBY-2130) , optimizer regression testing, and optimizer documentation.
>> Project: Derby XML support.
>> Join the Derby community and contribute to the ongoing development
>> of Derby's standards-compliant XML support. We'll fix XML bugs such
>> as DERBY-3870, DERBY-2739, and DERBY-2106; we'll work on additional
>> XML support such as import/export (DERBY-1892), JDBC XML apis and the
>> java.sql.SQLXML data type, and of course we'll need to add to the
>> Derby XML test suites and documentation set.
>> Project: Derby Data Mining support.
>> Join the Derby community and contribute to the ongoing development
>> of Derby's standards-compliant data mining support (DERBY-581).
>> We'll fix bugs in the current support (DERBY-3634, DERBY-3635,
>> DERBY-3505); we'll add additional features such as ROLLUP (DERBY-3002)
>> and CUBE grouping functions, additional window functions, and
>> window behaviors such as ordering and partitioning; we'll need
>> to verify that these new features perform adequately, and we'll
>> also work on regression testing and documentation in this area of Derby.
> Thanks Bryan for getting us thinking about GSOC.
> These projects look good if we can get mentors that are well versed in the
> area in question. Optimizer work is particularly risky IMHO and so we should
> make sure we have a strong mentor for careful review and guidance. I think
> the bug lists are ambitious for a student for the summer, but it is good to
> have the longer lists of issues to choose from.
> I would like to offer the general "Derby Test and Fix" project for JUnit
> conversions and newcomer bugs.  I will mentor  that project. I think it is a
> good fit for students with us for such a short time and  I think we have had
> success with it in the past.  Do you think this project is too general?
> Kathey

View raw message