Forum OpenACS Q&A: ACS 3.4, bind parameters and Postgres

The new ACS db_* API looks great (http://acs-
staging.arsdigita.com/doc/core-arch-guide/database-access-api).

Is anyone looking at upgrading the Postgres driver to support bind
parameters?

Collapse
Posted by Don Baccus on
I've not looked at it, but I've thought about it.

Since our plan is still to skip 3.4 and tackle things in earnest for 4.0, I'm planning to wait to do driver work.  Of course, I'll probably  want to look into this earlier than 4.0 if possible (I'll be gone most of September, though).  There's another driver bug I want to look into, as well (my recover-from-hosed-backend code doesn't work with AOLserver3 though it worked fine in AOLserver2).

Collapse
Posted by Peter Vessenes on
Something tells me 3.4 is going to be around for a while before 4.0 comes out. The amount of effort they put into poritng 3.4 seems pretty high, from the ASJ bboards. Combined with 4.0's vaporous status, my bet is that we're going to be waiting a while.
Collapse
Posted by Don Baccus on
If it stretches out too far, then we'll revisit the issue, most certainly.  They plan to release the ACS 4.0 core followed by the individual modules on their own schedule.  On the  one hand this means the core and early modules should arrive fairly early (fall), on the other it means that the entire toolkit won't arrive for months.

Unfortunately, so much has been redone for ACS 3.3 that it might boil down to essentially a report of the toolkit.  Followed by what we know will be a large effort to port over the new ACS 4.0 core and components.

Limited resources have driven this thinking.

Of course, it appears as though aD is willing to pay Sebastian to work  on the InterBase port, so perhaps that will be out earlier ...

Collapse
Posted by Michael Feldstein on
Does anyone have any insight into the reason for the aD Interbase decision?
Collapse
Posted by Michael A. Cleverly on
Sebastian ported an early version of the ACS (1.3) to InterBase 4.0/Linux.  Though I'm not 100% positive, I believe it was before he became employed by aD (which, I'm sure only helped him get a job).  His ACS 1.3 port has not been maintained through ACS 2.x & 3.x.  (Porting a 1.x version of the ACS would be lots simpler than 2.x or 3.x.  In fact, had Solid at that time supported triggers & sequences directly in inserts, I would have attempted a 1.x port to Solid myself.)

Now that InterBase 6.0 has been open sourced, I imagine he's     (personally) interested in ACS/InterBase again.  Now, as to why aD (as opposed to Sebastian personally) would want to do an InterBase port.....? I don't understand why they'd suddenly want to be in the porting business (after having the all-mighty Oracle outlook-on-life, they've had) and when OpenACS is already working on porting to open-source databases.  (It would seem more effective, and also a better PR move, for aD to allow Sebastian to spend time working as an OpenACS volunteer.)  Who knows.

Collapse
Posted by Michael A. Cleverly on
Apparently Sebastian isn't the only aD employee with prior InterBase involvement. Things are becoming a bit clearer over at http://www.arsdigita.com/bboard/q-and-a-fetch-msg?msg_id=0008dx&topic_id=21&topic=web%2fdb, especially after Michael's eloquent post. I loved the closing line: "If aD wants to run a soup kitchen, that's admirable enough. But a lot of us came here to garden. And now that we're here, we're armed with pitchforks."

😊