As noted in prior forum postings, we've been working on adding support
for MVCC into Ingres. (See 'MVCC - Ingres Community Wiki'
(http://community.ingres.com/wiki/MVCC) for details on the feature.)

The feature will be available in the next few weeks in a community
edition version of Ingres 10. If you have an interest in testing the
feature, please send me email directly at zelaine.fong (AT) ingres (DOT) com, and I
can then email you back when the product is available for download.

Thanks in advance for your interest.

-- Zelaine


--
zelaine

Ingres Forums wrote:

First the VectorWise tech preview program, and now first sight of MVCC!?
What a week. I'm like a dog with two tails.

--
Roy

UK Ingres User Association Conference 2010 will be on Tuesday June 8 2010
Go to http://www.iua.org.uk/join to get on the mailing list.

That's sounds so much better than being a one legged Chinese Elvis
impersonator.

But the effect is much the same.

Marty

Martin Bowes wrote:

Tell me *you're* not impressed.

--
Roy

UK Ingres User Association Conference 2010 will be on Tuesday June 8 2010
Go to http://www.iua.org.uk/join to get on the mailing list.

I feel like Roy to be honest.


--
dejan

Hi All,

I've been working on a little Opensource project that would allow me to express integer constants in kbytes.
This would allow things like 'with page_size=4k'.

I've made the necessary changes to back!psf!psl!pslsscan.c and it appears to work OK. The problem occurs when
I try to get it to raise an error if the number would exceed what could be held by an int8 item.

I've defined an error message in common!hdr!hdr!erusf.msg and throw it from the pslsscan.c with the following code.
switch (byte_scale)
{
case KIBIBYTE:
TRdisplay("Have found kibibyte %ld\n", int_number);
if (int_number >= MINI8K && int_number <= MAXI8K)
{
int_number = int_number << 10;
}
else
{
TRdisplay("Want to Generate error E_US0AD0_2768,\n");
_VOID_ psf_error(2768L, 0L, PSF_USERERR,
&err_code, &psq_cb->psq_error, 0);
return (-1);
};
break;

/* case MEBIBYTE: etc... */
}; /* switch(byte_scale) */


Trouble is that my error message is not printed, instead I get: E_US0A94 line 1, Unexpected character ''.

I've traced back!psf!psf!psferror.c and it appears that the problem is that my message comes with a
setting for pss_retry = 1 in the session CB. This gets anded to PSS_REPORT_MSGS = 2, and hence my message
is not printed. The E_US0A94 message comes with a pss_retry=10 and is hence printed.

Anyone know ehat incantation I need when defining the error message to get the pss_retry = 10?

Martin Bowes

On Feb 1, 2010, at 4:05 AM, Martin Bowes wrote:

A query that fails is parsed twice. (!)
The first time has messages suppressed. The second time runs with REPORT_MSGS
and a flag that tells RDF to refresh anything it's asked for. The idea behind this
craziness is that RDF is unreliable and might be giving the parser bad
information. So, your message should be seen the second time around.
The question is why you are getting US0A94 instead, and when.
I'd try to determine whether the US0A94 is happening on the first
parse attempt or the second, that might help you locate the problem.

Karl

Hi Karl,

Thanks for that. My message is generated on the first pass but not the second. On the second pass the E_US0A94 is generated.
No idea why just yet!

Marty

-----Original Message-----
From: info-ingres-bounces (AT) kettleriver...ting (DOT) com on behalf of Karl Schendel
Sent: Mon 01/02/2010 11:46
To: Ingres and related product discussion forum
Subject: Re: [Info-Ingres] Raising PSF_USERERR from pslsscan.c

On Feb 1, 2010, at 4:05 AM, Martin Bowes wrote:

A query that fails is parsed twice. (!)
The first time has messages suppressed. The second time runs with REPORT_MSGS
and a flag that tells RDF to refresh anything it's asked for. The idea behind this
craziness is that RDF is unreliable and might be giving the parser bad
information. So, your message should be seen the second time around.
The question is why you are getting US0A94 instead, and when.
I'd try to determine whether the US0A94 is happening on the first
parse attempt or the second, that might help you locate the problem.

Karl



_______________________________________________
Info-Ingres mailing list
Info-Ingres (AT) kettleriverconsulting (DOT) com
http://ext-cando.kettleriverconsulti...fo/info-ingres