OSEC

Neohapsis is currently accepting applications for employment. For more information, please visit our website www.neohapsis.com or email hr@neohapsis.com
 
From: ArunKumar S (akumarOMNESYSINDIA.COM)
Date: Thu Feb 07 2002 - 23:11:06 CST

  • Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

    Another emthod again breaks the compatibilty rules - and as for clients
    which need the 'other' interface Qi will do just fine -

    Well if u could just make it mandatory that the new dll (with new
    interface)
    is used everywhere it should be fine - scripting clients should have no
    problem with interface ID change - if compile clients are there atleast they
    wouldnt
    bomb coz u have kept interface constant and changed the vtable - they
    would get a chance to report something and get out

    ----- Original Message -----
    From: Madhusudanan R <madhuRSI.RAMCO.COM>
    To: <DCOMDISCUSS.MICROSOFT.COM>
    Sent: Thursday, February 07, 2002 3:28 PM
    Subject: Re: Interfaces and new methods

    > For scripting clients having to access the 'non-default' interfaces, why
    not
    > use a method in the default interface which returns an interface pointer
    to
    > the 'non-default' interface. Another form of QI.
    >
    > > -----Original Message-----
    > > From: Steve Johnson [SMTP:sjohnson3TSYSTEMS.COM]
    > > Sent: Friday, January 25, 2002 10:08 PM
    > > To: DCOMDISCUSS.MICROSOFT.COM
    > > Subject: Re: Interfaces and new methods
    > >
    > > CG> What would the approach be if an existing method was changed?
    > >
    > > CG> Would you rename it to MethodName2 in the new interface?
    > > CG> Is there no other option?
    > >
    > > Theoretically, you could add a new method with a different ID to the
    > > existing interface. This would provide compatibility for existing
    > > clients and allow new clients to take advantage of new functionality.
    > > The problem is that if clients depending upon the new method
    > > encounter a component implementing the old interface, they will fail.
    > > A better solution would be to create a new interface. However, if
    > > you are servicing scripting clients, they have no means of QI'ing
    > > for an interface, so this won't work. There has been discussion this
    > > problem on this
    > > list in the past, but I don't remember the solutions. If you are in
    > > this situation with scripting clients, I'm afraid I can't be of
    > > further assistance as I've never had to solve this problem myself.
    > >
    > >
    > > Steve Johnson
    > > 3t Systems
    > >
    > > ----------------------------------------------------------------
    > > Users Guide http://discuss.microsoft.com/archives/mailfaq.asp
    > > contains important info. Save time, search the archives at
    > > http://discuss.microsoft.com/archives/index.html .
    > > To unsubscribe, mailto:DCOM-signoff-requestDISCUSS.MICROSOFT.COM
    >
    > ----------------------------------------------------------------
    > Users Guide http://discuss.microsoft.com/archives/mailfaq.asp
    > contains important info. Save time, search the archives at
    > http://discuss.microsoft.com/archives/index.html .
    > To unsubscribe, mailto:DCOM-signoff-requestDISCUSS.MICROSOFT.COM

    ----------------------------------------------------------------
    Users Guide http://discuss.microsoft.com/archives/mailfaq.asp
    contains important info. Save time, search the archives at
    http://discuss.microsoft.com/archives/index.html .
    To unsubscribe, mailto:DCOM-signoff-requestDISCUSS.MICROSOFT.COM