Re: [Firebird-devel] Firebird Maintenance Policy

2012-08-23 Thread Dimitry Sibiryakov
23.08.2012 15:29, Tony Whyman wrote: This issue needs to be resolved long before the next end of life decision is made. I would propose that time should not be the only criteria for deciding on end of life. This is not an issue. Yes, free support of old Firebird version is ended, but

Re: [Firebird-devel] Firebird Maintenance Policy

2012-08-23 Thread Adriano dos Santos Fernandes
On 23/08/2012 10:29, Tony Whyman wrote: However, Redhat Enterprise/Centos 5 uses an even older kernel - 2.6.18 - and that is in support until 2020. Anyone running Firebird on this distro is going to be left without an upgrade path if the Firebird 2.1 series gets the chop in 2013. It is in

Re: [Firebird-devel] (CORE-3291) New pseudocolumn to get number of transaction that created this record version

2012-08-23 Thread Dimitry Sibiryakov
23.08.2012 17:52, Adriano dos Santos Fernandes wrote: But maybe we can return NULL for views of more than one table and maintain it as a number? At first, I'm more for that. No, it must return version from real table record. Like this: create view v (a,b) as select t1.rdb$record_version,

Re: [Firebird-devel] (CORE-3291) New pseudocolumn to get number of transaction that created this record version

2012-08-23 Thread Adriano dos Santos Fernandes
On 23/08/2012 12:59, Dimitry Sibiryakov wrote: 23.08.2012 17:52, Adriano dos Santos Fernandes wrote: But maybe we can return NULL for views of more than one table and maintain it as a number? At first, I'm more for that. No, it must return version from real table record. Like this: create

Re: [Firebird-devel] (CORE-3291) New pseudocolumn to get number of transaction that created this record version

2012-08-23 Thread Leyne, Sean
23.08.2012 18:04, Adriano dos Santos Fernandes wrote: I'm talking about: create view v as select t1.*, t2.* from t1, t2; select v.rdb$record_version from v; IMHO, it should give Column unknown error. I.e. this pseudocolumn must be defined for real tables only. I tend to agree,