Hi Daniel,

        Collating both your mails:

On Thu, 2008-11-27 at 09:06 +0100, Daniel Rentz wrote:
> Michael Meeks schrieb:
> > and splits out all of the calc filters (except for xml) into a
> > separate library: 'scfilt' which is demand loaded via an
> > abstract interface.
> 
> Looks much better, thanks.

        Good stuff.

> > hopefully that will avoid symbol clash problems in future.
> 
> for consistency with SingleRefData, you may also change
> ComplexRefData.

        Certainly; changed ComplRefData -> ScComplexRefData & committed.

On Thu, 2008-11-27 at 09:30 +0100, Daniel Rentz wrote:
> 1) I noticed that you changed filter names and file type names in the 
> configuration (module filter). This is not a good idea as these names 
> can be considered public API.

        Good catch; since we don't want to enable the filter in the default
builds as yet, anyway - I've just reverted all the filter/ changes. I
read through the whole cws diff more carefully this time, and pruned
another couple of bits that had drifted in too.

> 2) What is the new STREAM_FOR_OUTPUT for, that is added besides the 
> OUTPUT_STREAM in the media descriptor? Additionally all XOutputStream's 
> in oox/core have been changed to XStream.

        That looks somewhat inelegant but it seems that it is just there to
compensate for the package/ code, which can't cope with an
XOutputStream, only either an XStream or an XInputStream - which is a
shame: seemingly the XInputStream is used for copying and re-reading the
written data - perhaps in use-cases that we don't care about. This is
Radek's work-around for that; it'd be good to do something more elegant;
what do you recommend:

        a) update package/* to handle XOutputStream
        b) create a simple XStream wrapper in oox for the XOutputStream,
           that proxies XSeekable with no real XInputStream
                + test to ensure we don't hit any re-reading cases.
        c) some other option ;-)

        b) looks easy enough in the short term, I can knock that up if it seems
reasonable; and we can drop the sfx2 changes - then again, perhaps there
is some much better way.

        Thanks,

                Michael.

-- 
 [EMAIL PROTECTED]  <><, Pseudo Engineer, itinerant idiot


---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to