This one time, at band camp, Michael Colbert said:

MC>Cool.  I was thinking the existing JUnit tests would be a good place to start,
MC>too.  They'll give us an idea of what each feature does already from the
MC>developers' perspective.  We may need to expand coverage in some areas, and
MC>some of the more demanding/complex requirements will require some functional
MC>tests that encompass several classes and/or features.  I haven't looked at the
MC>Castor tests in a while.  I'll take a look sometime today or tomorrow and see
MC>how much of this stuff is already in there.
MC>
MC>The mailing list will definitely give us useful information from the users'
MC>perspective.

Mike, Daniel, 

I was actually going to suggest that this list should be tightly coupled
with the JDO tests (CTF-JDO). This is a great idea, Mike. This will also
help the CTF-JDO to move forward faster. Currently, I don't check in any
code fixes without a test case proving it. However, this is not to say
that the test cases are comprehensive. But this effort would certainly
move them in that direction.

The idea for a simple doc using a table to establish a matrix of features
is also a good idea. This is exactly how I'm laying out the project page
right now. I suggest that you use one of the other docs in src/docs as a
template to begin. I suggest that the doc be called jdo-features.xml and
that it include a column indicating whether a test case has been written
yet. This column would need to work in conjunction with the project doc
on which I'm currently working. The best thing that you can do is start
the doc and send all updates to me for checkin to CVS.

Bruce

MC>--- Daniel Honig <[EMAIL PROTECTED]> wrote:
MC>> I would be willing to assist Michael in this effort.
MC>> 
MC>> I would suppose that the J-Unit test cases are a good starting point.
MC>> 
MC>> We could use the mailing list to get an idea of what alot of
MC>> the big questions are and ensure they are clearly outlined
MC>> in the matrix.
MC>> 
MC>> 
MC>> -Daniel
MC>> 
MC>> -----Original Message-----
MC>> From: Michael Colbert [mailto:[EMAIL PROTECTED]]
MC>> Sent: Monday, October 07, 2002 3:01 PM
MC>> To: [EMAIL PROTECTED]
MC>> Subject: Re: [castor-dev] Adding new dependent objects to an exiting
MC>> persistent object
MC>> 
MC>> 
MC>> Hi Bruce,
MC>> 
MC>> What I have in mind is a document targeted at Castor JDO users (and
MC>> prospective
MC>> users) that very concisely indicates the development status of each feature.
MC>> Perhaps something as simple as a three column table, with: FEATURE NAME,
MC>> TO-DO,
MC>> and COMPLETED.
MC>> 
MC>> The TO-DO column would contain a bullet list of planned functions/abilities.
MC>> 
MC>> The COMPLETED column would contain a bullet list of completed
MC>> functions/abilities.
MC>> 
MC>> This document would be a quick reference, not an exhaustive resource.
MC>> 
MC>> Seems simple enough.  And that's the point.  How to get there is another
MC>> question, I suppose.  Initially, I guess it could be as easy as writing down
MC>> the list of features and then for each one asking, "What should it do?"  And
MC>> then asking, "What does it do now?"  But, the answers would have to be
MC>> specific, accurate, and provable.  This may mean writing some functional
MC>> tests
MC>> to validate the answers.
MC>> 
MC>> A step forward?
MC>> 
MC>> Mike Colbert
MC>> 
MC>> 
MC>> 
MC>> --- Bruce Snyder <[EMAIL PROTECTED]> wrote:
MC>> > This one time, at band camp, Michael Colbert said:
MC>> >
MC>> > MC>Would it be possible to get someone from the JDO team to assist in
MC>> > creating a
MC>> > MC>status sheet on all the different JDO features (caching, extends,
MC>> depends,
MC>> > long
MC>> > MC>transactions, lazy loading, etc.) and their respective implementation
MC>> > status?
MC>> > MC>
MC>> > MC>I believe this would be helpful to all, since due to the high learning
MC>> > curve
MC>> > MC>associated with figuring out proper castor JDO usage (esp. the
MC>> mapping),
MC>> > it's
MC>> > MC>incredibly difficult to tell whether or not problems like this are due
MC>> to
MC>> > MC>improper usage, or "not quite done" features.
MC>> > MC>
MC>> > MC>I am predicting that Bruce Snyder will have absolutely no reply to Ryan
MC>> > MC>McDonough's question here, other than "This one time at band camp ...
MC>> > please
MC>> > MC>send your client code and mapping."  On the other hand, if Ryan had a
MC>> > place to
MC>> > MC>quickly and painlessly look up the status of a particular feature (in
MC>> this
MC>> > MC>case, dependent objects), he may never resort to such "probing"
MC>> questions.
MC>> >  (I
MC>> > MC>assume he is probing, trying to quickly find out if this is a
MC>> well-known
MC>> > issue,
MC>> > MC>because otherwise he would have sent all his code.)  The mailing list
MC>> > archive
MC>> > MC>is not entirely helpful in this regard, nor is the bugzilla database.
MC>> > What I
MC>> > MC>have in mind is a one page list of all the major features and their
MC>> status
MC>> > MC>("not implemented", "partially implemented", "mostly implemented",
MC>> "fully
MC>> > MC>implemented and tested", etc. along with helpful footnotes along the
MC>> lines
MC>> > of
MC>> > MC>"works while caching is turned off").
MC>> > MC>
MC>> > MC>If there is some way to find this information, quickly, please advise.
MC>> If
MC>> > not,
MC>> > MC>I hereby volunteer to collect and publish in one place all
MC>> authoritative
MC>> > MC>responses to this request as to the status of the JDO features present
MC>> and
MC>> > MC>future.
MC>> >
MC>> > Mike,
MC>> >
MC>> > My apologies for missing your message previously. Currently, there
MC>> > is only one other active developer on Castor JDO and he is somewhat
MC>> > indisposed recently.
MC>> >
MC>> > I think that this is a good idea. Can you elaborate on this further? I'm
MC>> > currently working on a project plan doc for Castor JDO and your idea
MC>> > could definitely influence and supplement this doc.
MC>> >
MC>> > Bruce

-- 
perl -e 'print unpack("u30","<0G)U8V4\@4VYY9&5R\"F9E<G)E=\$\!F<FEI+F-O;0\`\`");'

----------------------------------------------------------- 
If you wish to unsubscribe from this mailing, send mail to
[EMAIL PROTECTED] with a subject of:
        unsubscribe castor-dev

Reply via email to