PostgreSQL TGuid fields - broken again

[ertank]

Hello,

I confirm that between SVN 4052 and 4087 something broke working TGuid fields. I am using same SVN link
http://svn.code.sf.net/p/zeoslib/code-0 ... esting-7.2

I do not have time to identify exact revision number at the moment. I shall see to it as soon as I can.

Currently, I am reverting back to SVN 4052.

Thanks & regards,
Ertan

[marsupilami]

Hello Ertan, in which way did TGuid for PostgreSQL break? What doesn't work anymore?

[ertank]

Hello Jan,

I get same GUID displayed in all rows of a table. GUID's last 10-15 numbers being a zero. GUID seems to change each time I run the application. I would say, GUID is not initialized and displaying its variable's memory footprint as I once has my own bug in my application and GUID values look familiar.

If I revert back to earlier SVN all good and back to normal. Problem I am observing is not on Raspberry Pi 3 this time. This is a Debian 9.3 64bit linux OS running on a regular PC I am having mentioned problem.

[EgonHugeist]

Hello ertrank,

is it possiple to pinpoint the revision the issue reintroduces?
If so we quickly can fix it..

[Fr0sT]

Try r4100. I did bugfix there regarding GUIDs

[ertank]

Hello,

Sorry, I am quite busy. Will check as soon as possible and post back in here.
Both the broken SVN version and SVN 4100 test result.

Thanks.