[hxp]

Jae --

Glad to see it!

I intend to give yr chgs a good testing tomorrow on both mysql/linux and hsql/windows.

+more to say re your last 2 msgs, but it can wait till later.

--Howard

[hxp]

Hi Julien --

As I wrote at the beginning of this thread...

"hxp" wrote:
I have the backends of 2 modules -- FAQ and Quotes -- revamped to a pure-CMP approach. ... I've successfully upgraded ... to pure-CMP so that it doesnt use sql/ddl anymore at all. I've recently gotten xdoclet, unknown-pk, and CMR to coexist without problems.

The final step for this phase is making sure that this works without a hitch on both HSQL and MySQL. With the fixes Jae committed a few days ago, I'm testing and expecting things to go smoothly. (After I commit, it would be good if other people can test on postgres, oracle, etc)

"hxp" wrote:

If you want me to go the CVS route, I'm willing to become the maintainer of the FAQ module, and begin reworking the rest of it (some for 1.1; more for 1.2).
...
I will be travelling from mid-May to early-July (and unavailable for coding), so I don't want to raise expectations too much.

From July onwards, I would foresee revamping & enhancing the entire FAQ module, consistent with JSR-168 and whatever frontend architecture is optimal for Nukes at that point.

Because I'm running out of time to prepare for 2 months of travel, I will commit my final contribution for 1.1 this week. It will contain only backend improvements.

I will overhaul the frontend completely for 1.2, after my travels are over ~ mid-July... and then continue to be maintainer of the FAQ module, as offered in my original post.

Thanks for your email and offer of help :) I'll be back to you when I need specific help, such as with any CVS hitches.

-- Howard

[jae77]

howard -

i just saw your cvs commits for the quotes code (cool stuff).

i've got a question about the unknown primary key stuff...how exactly are you using it?

in the news code, when i define the getter for the primary key field and add all the corresponding xdoclet code, i just state that it should be auto-incremented. i'm able to insert, etc w/o issue.

i understand the need for the unknown-pk stuff if you are not using an integer as the pk for the database, but i'm not really sure what it gets you in this situation (and i'm worried that i may be doing something incorrect by not having it myself).

[hxp]

Jae --

"jae77" wrote:
howard -

i just saw your cvs commits for the quotes code (cool stuff).

:)

"jae77" wrote:


i've got a question about the unknown primary key stuff...how exactly are you using it?

in the news code, when i define the getter for the primary key field and add all the corresponding xdoclet code, i just state that it should be auto-incremented. i'm able to insert, etc w/o issue.

i understand the need for the unknown-pk stuff if you are not using an integer as the pk for the database, but i'm not really sure what it gets you in this situation (and i'm worried that i may be doing something incorrect by not having it myself).

My earlier plan was that I was going to write 2 big long hairy Wikis on these 2 subjects you ask about --- "Why?" + "How?"

But now all of a sudden I'm committing :) working code :), and never got around to explaining the strategy & architecture & thinking behind this whole pure-CMP architecture :-(

Cant do it justice right now; but here're a few stabs at it:

------------------------------------------------------------------------------

Correctness vs Strategy

"jae77" wrote:
(and i'm worried that i may be doing something incorrect by not having it myself).

* if it works, it's not "incorrect".

* But... just becasue 2 different code approaches produce the same immediate result doesnt mean that they are equivalently good in the bigger picture.


Why --

Bigger question:"Why Pure-CMP?" .... (scattered answers to this one all other the past 3 months of Forums, and I have to gather into a Wiki... )

You know many of the reasons, but just 1 to highlight here: Abstraction makes a real difference. Clean (well architected) abstractions lead to much more successful software and systems.

Smaller question: "Why Pure-CMP using unknown-PK?"...

* Using unknown-pk allows you to have all the benefits you want from a PK --- unique identifier, not null, automatically incremented, automatically indexed ----- and getting all this benefit requires zero java code in the Entity EJB class! A small amount of xdoclet tagging, and a few conventions to follow in the java code that uses the PK... that's all it takes.

* code is shorter. easier to read. easier to maintain. fewer places for bugs to hide.

* recognizes that PK & FK are part of the workings of the DBMS layer, and should be confined to that layer.

* lets each underlying DBMS generate and manage keys in its own optimum way.

* When it comes to what is the optimal way for each DBMS to uniquely identify and index rows in its tables, the DBMS knows best..... but from the Java/OO/Pure-CMP perspective we dont know and we dont care.

ergo, "unknown"-PK

* At the Pure-CMP layer, we get what we want: fully-abstract automatic "Entity Identity" generation and management.

--- unknown-pk gives us the "PK" side of this fully-abstract automatic "Entity Identity" generation and management

--- CMR gives us the "FK" side of this fully-abstract automatic "Entity Identity" generation and management



How --

OK, to keep things sorted, the "HOW" part is over on our "nuts & bolts" pure-CMP thread ---

MySQL & Pure-CMP --- issues w News & FAQ modules
http://www.jboss.org/index.html?module=bb&op=viewtopic&t=49132


~~~~~~~~~~~~~~~~~~~


Now, check back to the first msg on this thread, and see where I mentioned "converging" our approaches... for now, you dont really have to do anything, but if you want to switch to the unknown-pk approach, that would be the only convergence that we need to do before 1.1.

We can converge formatting/style later; lets focus on getting substance converged for now. (I actually like your formatting style conventions -- such as lining up all the "=" vertically -- better than what I used. The alignment makes it easier to read at a glance (back in the Fortran days, such symmetry helped make it easier to find typo-bugs, before waiting 15 minutes for the cards to compile and then finding out 8-) ))

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Really tired; got a million things to do before going into travel mode for 2 months starting nxt wk, so doing Wikis fell off the bottom of the list... (for now). Hope these notes are good enuf for now.

But still must make sure I follow thru on what I did promise, and make sure nothing i commit to cvs breaks the build on any of the platform combinations. (so let me know if you see any issues or experience any problems, before I teardown and put everything in hyberspace.)

--- Howard

[hxp]

To anybody who's reading this thread ---


The new pure-CMP Quotes and FAQ modules are now checked into CVS :)

Please grab a copy, try em out, and post some bugs/critique/feedback.

Thanks!

-- Howard

[sbartone]

"hxp" wrote:
Jae --

"jae77" wrote:

are you sure you're not missing something w/ the xdoclet/ant "config/setup?"

No, I'm not sure. Please say more about this when you get a chance.

"jae77" wrote:

i ask b/c we are generating successful deployments for other components and mysql.

Careful, careful... those were not pure-CMP deployments... weren't relying on create-table....

I suspect we'll find something wrong with the interaction of create-table and nukes and mysql....

"jae77" wrote:
i'm curious to know what the issue w/ the entity command is and how it relates to creating the tables themselves.

Just a suspicion at this point; haven't had time to investigate to determine root cause.... will post more ~Sunday/Monday.

I have more to say on this that's relevant to News Module as well --- so I'm forking this thread, and continuing implem/debugging discussion at:

MySQL & Pure-CMP --- issues w News & FAQ modules

http://jboss.org/index.html?module=bb&op=viewtopic&t=49132

Bye for a few days.

-- Howard