Subject | Re: History of Interbase's failure to make it to the big time. |
---|---|
Author | plinehan |
Post date | 2005-10-19T14:47:15Z |
In Firebird-general@yahoogroups.com,
Dmitri Kouzmenko <kdv@i...> wrote:
call it MVCC, MGA or record-shadowing or
record-versioning or Row Versioning. 8-)
I've just thought of a nice definition -
record-piggybacking. Any takers?
the error if your undo log gets too big. Their
architecture does allow for PITR though.
I think they had a read version of MGA since
version 4 but the whole system only became
MGA since 7 - but I can't remember where I
read this - have been reading a lot of
different sites about this.
was there from day 1 though - but can't
find site.
Here
http://pgsqld.active-venture.com/history.html
It seems to imply that there used to be
a locking mechanism but that it was
replaced in and around 1995, but it's
not that clear.
Many went there, most died poor and cold.
versioning engine supplier.
Microsoft motto - Embrace and Extend (meaning
smother)
Oracle motto - Buy, buy, buy (to competitors),
you may or may not get the pun in English,
"Bye, bye, bye (to competitors)".
Check out the URL in the thread I posted to
http://dev.mysql.com/doc/refman/5.0/
en/innodb-multi-versioning.html
And then check out Roman Rottsky's article
and you will see that they're virtually
identical.
since - ditto for MGA 1980.
Paul...
Dmitri Kouzmenko <kdv@i...> wrote:
> p> I was engaged in a debate about IB/FB'sAw, come on! That's just semantics, whether we
> p> MVCC model and various other issues about
> Need some correction. InterBase's multiversion
> architecture is named as Multi-Generational Architecture
> (MGA). So, most of documents about Interbase refer to
> MGA, not MVCC.
call it MVCC, MGA or record-shadowing or
record-versioning or Row Versioning. 8-)
I've just thought of a nice definition -
record-piggybacking. Any takers?
> - Oracle have versioning (named as semi-verisioningI think it's ORA-1555 (or something like that) is
> because it have some problems with very long snapshot
> transactions) since version 8.0 (am I right?).
the error if your undo log gets too big. Their
architecture does allow for PITR though.
I think they had a read version of MGA since
version 4 but the whole system only became
MGA since 7 - but I can't remember where I
read this - have been reading a lot of
different sites about this.
> - PostgreSQL have MVCC since ... (don't know).unable to find this. I don't think it
was there from day 1 though - but can't
find site.
Here
http://pgsqld.active-venture.com/history.html
It seems to imply that there used to be
a locking mechanism but that it was
replaced in and around 1995, but it's
not that clear.
> - MS SQL 2005 will have versioning same as inWill be similar (if ever realeased) as Yukon.
> InterBase/Firebird.
Many went there, most died poor and cold.
> - MySQL will (? or already) have versioning ... don't knowIf Oracle let them!!! Oracle bought their
> about it's implementation or limitations.
versioning engine supplier.
Microsoft motto - Embrace and Extend (meaning
smother)
Oracle motto - Buy, buy, buy (to competitors),
you may or may not get the pun in English,
"Bye, bye, bye (to competitors)".
Check out the URL in the thread I posted to
http://dev.mysql.com/doc/refman/5.0/
en/innodb-multi-versioning.html
And then check out Roman Rottsky's article
and you will see that they're virtually
identical.
> These all show that commercial RDBMS' are moving toRelational model - 1970 - no significant improvements
> versioning systems. And this also shows that
> there is no new (extraordinal) thoughts in that area
> since 1983-1985.
since - ditto for MGA 1980.
Paul...
> Dmitri Kouzmenko, www.ibase.ru, (095) 953-13-34