Richard Monson-Haefel & I were discussing (offline) the merits/pitfalls of
using the "updateable" aspects CachedRowSets to flush changes to an
underlying database. We thought it would be worth opening this up for
general discussion.

Here's, the story so far (in reverse order unfortunately)  ....

Perhaps not. But, by employing the RowSetWriter to commit changes we are
altering the very nature and reasons for using EJB in the first place, which
is itself very interesting.  EJB objectifies the data
source providing a component model which ensures safe transactions and the
availability of other value added services (security, passivated, OR
mapping, etc.).  I think your point is valid and I'll have
to give it more thought, but my gut reaction is to avoid using updatable
CachedRowSets.

I think this is a good one for the EJB-INTEREST mailing list.  Maybe they
can flush out some more concrete objections.


Rumble, Nick wrote:

> I'm not sure the transaction isolation is really an issue if the
> RowSetWriter ensures the original values in the rowset match the current
> values of underlying database when it comes to flush the changes out. What
> do you think ?
>
> -----Original Message-----
> From: Richard Monson-Haefel [mailto:[EMAIL PROTECTED]]
> Sent: 11 October 1999 15:12
> To: Rumble, Nick
> Subject: Re: Recommendation: CachedRowSets
>
> No. I know all about CachedRowSets and the fact that they are updatable. I
> was saying that they should be set so that they are not updatable when
used
> in EJB because EJB has very specific transaction
> considerations that necessitate that updates be done through bean methods
> and not the CachedRowSet.  One of the reasons I didn't include
CachedRowSets
> was (1) smart people like you would figure it out
> anyway (2) they are not well supported yet.
>
> Rumble, Nick wrote:
>
> > Sorry - I don't understand you - are you saying CachedRowSets aren't
> > updateable ? .... because they are !
> >
> > (please ignore the rest of this if you are already aware of this )
> >
> > The implementation of CachedRowSets allows the developer to plug in a
> > RowSetWriter class to flush rowset changes back to an underlying data
> source
> > or to supply their own, home grown solution if the default functionality
> > doesn't suffice. Sun's default implementation uses an optimistic locking
> > algorithm to check if the underlie database data has changed since the
> > CachedRowSet was created - if not it flushes the changes back to the
data
> > source.
> >
> > Regards
> >
> > Nick
> >
> > -----Original Message-----
> > From: Richard Monson-Haefel [mailto:[EMAIL PROTECTED]]
> > Sent: 11 October 1999 13:13
> > To: Rumble, Nick
> > Subject: Re: Recommendation: CachedRowSets
> >
> > Thanks.  I've thought of doing that since, as you observed, its a nice
> fit.
> > One thing make sure that the CachedRowSet is
> > not updatable (read only).
> >
> > Richard
> >
> > Rumble, Nick wrote:
> >
> > > Hi - just thought I'd let you know that you should update URL
> > > http://www.ejbnow.com/ejbtips/ejbtip_4.html to include information on
> > SUN's
> > > new CachedRowSet implementation.
> > >
> > > Regards
> > >
> > >  <<Nick Rumble.vcf>>
> > >
> > >
> ------------------------------------------------------------------------
> > >
> > >   Nick Rumble <[EMAIL PROTECTED]>
> > >   Developer
> > >   SageTetra Ltd
> > >
> > >   Nick Rumble
> > >   Developer
> > <[EMAIL PROTECTED]>
> > >   SageTetra Ltd
> > >   Cynosure Centre Vanwall Business Park;Maidenhead;Berkshire;SL6
> 4UB;Unted
> > Kingdom  Work Voice: +44 (0) 1628 677167
> > >
> > Home Voice: Dream On!
> > >
> > Voice Cellular: +44 (0) 411 764627
> > >
> > Work Fax: +44 (0) 1628 771634
> > >   Additional Information:
> > >   Version   2.1
> > >   Last Name Rumble
> > >   First NameNick
> > >   Label WorkCynosure Centre Vanwall Business Park Maidenhead,
Berkshire
> > SL6 4U United Kingdom
> > >   Revision  19991007T123443Z
> >
> > --
> > Richard Monson-Haefel
> > EJB Expert for jGuru.com
> > ( http://www.jguru.com  )
> >
> > Author of Enterprise JavaBeans
> > Published by O'Reilly & Associates
> > ( http://www.ejbnow.com )
> >
> > The statements and opinions expressed here are my own and may not
> represent those of the company.
> >
> > This e-mail is subject to copyright and the information in it is
> confidential. It is intended only for the named recipient. You are advised
> not to disclose the contents of this e-mail to another pers
> > on or take copies of it.
>
> --
> Richard Monson-Haefel
> EJB Expert for jGuru.com
> ( http://www.jguru.com  )
>
> Author of Enterprise JavaBeans
> Published by O'Reilly & Associates
> ( http://www.ejbnow.com )
>
> The statements and opinions expressed here are my own and may not
represent those of the company.
>
> This e-mail is subject to copyright and the information in it is
confidential. It is intended only for the named recipient. You are advised
not to disclose the contents of this e-mail to another pers
> on or take copies of it.



--
Richard Monson-Haefel
EJB Expert for jGuru.com
( http://www.jguru.com  )

Author of Enterprise JavaBeans
Published by O'Reilly & Associates
( http://www.ejbnow.com )








The statements and opinions expressed here are my own and may not represent those of 
the company.

This e-mail is subject to copyright and the information in it is confidential. It is 
intended only for the named recipient. You are advised not to disclose the contents of 
this e-mail to another pers
on or take copies of it.

===========================================================================
To unsubscribe, send email to [EMAIL PROTECTED] and include in the body
of the message "signoff EJB-INTEREST".  For general help, send email to
[EMAIL PROTECTED] and include in the body of the message "help".

Reply via email to