Re: [HACKYSTAT-DEV-L] Reduction Function parameter design and documentation

[(Cedric) Qin ZHANG]

When there is no parameter supplied, ActiveTimeReducer return one single
stream with the sum of all member's active time, while
MemberActiveTimeReducer returns multiple streams, one for each member.

Of course, you can add another parameter to combine the two reducers,
but I would say keeping two reducers make things clearer.

Cheers,

Cedric

Aaron Kagawa wrote:

Hey Philip and Cedric,

I have another question..

Shouldn't the ActiveTimeReducer and MemberActiveTimeReducer be combined?
Likewise, the CodeChurn, Commit,  and UnitTest. It seems that the
BuildReducer already took that step.

thanks, aaron

At 02:28 AM 3/29/2005, you wrote:

Hey Philip and Cedric,

Questions about the reducer revisions and the filePattern parameter.

1) could/should the Build Reducer have a filePattern parameter? I'm not
sure but couldn't other Projects have different build directories?
(Probably not.. the other two are better questions.)

2) could/should the CppUnitTest Reducer have a filePattern parameter?
The
UnitTest reducer has one. Its not clear why CppUnitTest shouldn't
have one.

3) could/should the Issue Reducer have a filePattern parameter?
Issues can
be generated for different workspaces: hackyKernel, hackyStdExt, or even
hackyAnt/src/org/hackystat/stdext/sensor/ant/jira. At the very least, I
think we should be able to show different Issue streams for different
top
level modules.

thanks, aaron

At 09:20 AM 3/22/2005, you wrote:

Greetings, all,

I have just posted a new Jira issue that proposes a comprehensive
redesign of the current set of Reduction Functions to improve their
utility and usability:

<http://hackydev.ics.hawaii.edu:8080/browse/HACK-223>

Please take a look at this and let us know what you think.

I have assigned this issue to Cedric, because the implementation of
these changes has a ripple effect into the current set of Streams
definitions on the public server. Thus, once a Reduction Function's
parameters are standardized according to this proposal, the Streams
associated with it may need to be modified.

However, my hope is that some of you will help Cedric with this effort
by volunteering to upgrade your Reduction Functions to conform to the
new standard.  Please email him directly if you are willing.

I would like this to be accomplished by the next stable release, which
is likely to occur in mid-April.

Cheers,
Philip