>> As Bruce wrote the release notes for 9.0, I read it to check if we still >> have a lot of things to do to fully support 9.0. We're not that far but >> we still have two major missing features. >> >> Here is the list of the work *I think* we still have to do before a 1.12 >> beta 1 release: >> >> * Support for 9.0: >> * exclusion constraint (#118) >> * default ACL (#99) >> * deferrable unique constraint >> * typed table (#135) ... patch available, awaiting comments >> * autonamed index (#158) ... patch available, awaiting comments >> > > Is there any way, I can help you on this? >
Thanks, I would love to. #135 and #158 are already commited. I sent the patch for frmOptions yesterday. So you can grab anything of the rest of the list.
Here is what I did on the other tickets
* Ticket 118: exclusion constraint See ticket118_NOTCOMPLETE.patch. IIRC, I copied the dlgIndex and pgIndex to make it easy building the dlgExclude. But it didn't work out that great.
* Ticket 99: default ACL See ticket99_NOTCOMPLETE.patch. My idea was to create a copy of dlgSecurityProperty named dlgDefaultSecurityProperty. It kind of works. This new tab is available on the database, but I have issues with it and the old security one.
* Ticket 142: SQL/Med objects See ticket142_NOTCOMPLETE.patch. The Foreign Data Wrapper object is already available. I wanted to have the Foreign Server object as a child of the first one, but all I'm able to have is a big crash. If you're able to fix this crash, I think I'll be able to handle the rest of the patch. But this crash stopped me. BTW, to anwser Dave's question, dblink can make use of SQL/Med objects (but AFAIK, it's the only one... which is too bad).
So, the easier one should be ticket 142. It would help me to continue on this patch. The more important one, as Dave already noticed are exclusion constraint and default ACL.
BTW, I'm not sure you should start from my patches. I sent them if it can help you, but I probably was wrong from the beginning of my patches (for example, copying dlgIndex instead of dlgCheck could be a serious mistake).