August 20, 2009 at 7:54 am
Maybe MYSQL was getting too much functionality. Oracle's charging model - massive upfront cost and then only renting the licence - will never compete with the outright purchase model of SQL Server. Their model makes the TCO for SMB's prohibitive, so as long as MS keeps improving SQL Server they will not have anything to fear from Oracle. MYSQL could give Oracle a vehicle for a low-cost solution, or as a previous post suggested, a cut-down starter version of oracle.
DaveT
BR DaveT
August 20, 2009 at 8:38 am
Actually, they bought RDB to get the Clustering management and RDB's MVCC stuff. RDB was already on the chopping block and VMS already in sharp decline and no longer strategic for Oracle
Oracle was MVCC from version 3 (some sources say 4), in 1983 (1984). What might they have gotten from DEC in 1994?
August 20, 2009 at 8:48 am
davtt (8/20/2009)
so as long as MS keeps improving SQL Server they will not have anything to fear from Oracle.DaveT
No version of windoze will scale as well as *nix. That *nix can do what z machines do is a (the) reason Larry wants Sun. If you need Oracle horsepower, SQLServer won't do. If you need real MVCC and don't have the $$$, then Postgres is where you'll go. At least in the short term. Whether Snapshot Isolation can be made to behave enough like real MVCC is also an open question. The real issue is the OS. SQLServer has that albatross round its neck.
August 21, 2009 at 9:16 am
No version of windoze will scale as well as *nix. That *nix can do what z machines do is a (the) reason Larry wants Sun. If you need Oracle horsepower, SQLServer won't do. If you need real MVCC and don't have the $$$, then Postgres is where you'll go. At least in the short term. Whether Snapshot Isolation can be made to behave enough like real MVCC is also an open question. The real issue is the OS. SQLServer has that albatross round its neck.
I am not sure about transaction related limitation in SQL Server because Oracle implemented and always improve the .NET based transaction model where different implementations are propagated to the RDBMS to be either consumed as Unit of Work ANSI SQL transaction or ADO.NET transaction.
A side note I think Postgres had a DARPA grant.
And the US Government have approved the sale so MySQL users mostly ISP that charges a few dollars a month for unlimited use may have to pay more because Oracle is not known for nickle and dime business model.
http://tech.yahoo.com/news/ap/20090821/ap_on_hi_te/us_oracle_sun
Kind regards,
Gift Peddie
August 21, 2009 at 10:02 am
A side note I think Postgres had a DARPA grant.
And the US Government have approved the sale so MySQL users mostly ISP that charges a few dollars a month for unlimited use may have to pay more because Oracle is not known for nickle and dime business model.
http://tech.yahoo.com/news/ap/20090821/ap_on_hi_te/us_oracle_sun
The DARPA $$ were for Stonebraker at Berkley in 1986. That relationship is long gone.
Not sure of the timing any more (it was just before or just after the announcement of the buyout by Sun), but there are, at least, two forks of MySql; allegedly to return it to its roots, which is to say, a mere SQL parser in front of a file system. Those folks will not be Oracle customers, those folks don't do relational data. I'd wager that most moved out since Sun bought MySql. The first release under Sun is widely reported to have been less than useful.
In any case, MySql is GPL, so for those who don't want to deal with Larry, alternatives already exist.
August 21, 2009 at 10:48 am
I think Godaddy and many ISP run the version that includes INNO DB because I have helped both US and UK users deploy MySQL version of the Asp.net 2.0 Club starter kit based on code uploaded by a user in Belgium. Now that was pre Sun MySQL. The UK user and I was offline email conversation the main reason to use MySQL was unlimited storage for a few Euros.
Kind regards,
Gift Peddie
August 21, 2009 at 1:01 pm
gnuoytr (8/20/2009)
If you need Oracle horsepower, SQLServer won't do.
This rash and unsubstantiated (and always disproven) claim has been made so many times over the decades by Oracle wonks and industry wags, that at this point it illustrates little more than their ignorance of the actual facts.
If there was even a shred of truth to this ridiculous claim then SQL Server (and Windows) wouldn't even show up in the TPC benchmarks (nor would anything else if you swallowed the Oracle hype). But the fact of the matter is that NO vendor has EVER dominated the TPC benchmarks, nor even any single category. Everybody wins some and loses others and over the decades it has always been very dynamic. And Oracle has never been the performance-dominant player that it counts on weak-minded pundits believing.
[font="Times New Roman"]-- RBarryYoung[/font], [font="Times New Roman"] (302)375-0451[/font] blog: MovingSQL.com, Twitter: @RBarryYoung[font="Arial Black"]
Proactive Performance Solutions, Inc. [/font][font="Verdana"] "Performance is our middle name."[/font]
August 21, 2009 at 1:29 pm
RBarryYoung (8/21/2009)
gnuoytr (8/20/2009)
If you need Oracle horsepower, SQLServer won't do.This rash and unsubstantiaed (and always disproven) claim has been made so many times over the decades by Oracle wonks and industray wags, that at this point it illustrates little more than their ignorance of the actual facts.
If there was even a shred of truth to this ridiculous claim then SQL Server (and Windows) wouldn't even show up in the TPC benchmarks (nor would anything else if you swallowed the Oracle hype). But the fact of the matter is that NO vendor has EVER dominated the TPC benchmarks, nor even any single category. Everybody wins some and loses others and over the decades it has always been very dynamic. And Oracle has never been the performance-dominant player that it counts on weak-minded pundits believing.
You've quoted me out of context. Oh well. The horsepower referred to is the MVCC semantics and *nix scalability. SQLServer (modulo snapshot) has neither. I'm not asserting that one *must have* these things, only that SQLServer currently (and so far as *nix support is concerned, always will) has nothing comparable. It is clear at this point that MVCC semantics and multi-tier architecture are a very good match. Thus snapshot in SQLServer (and "Oracle compatibility" in DB2/LUW 9.7); but not quite the same. I'm not a particular fan of Oracle, by the way. I've spent most of the last decade with DB2 on *nix and z. But the fact is that Oracle's semantics simply make more sense for many types of applications. M$ and IBM are playing catchup; revealed by their actions with their databases. To argue otherwise is, at best, disingenuous.
August 21, 2009 at 3:01 pm
Perhaps explaining yourself a little would help. I don't know what you mean by "z", and alluding to "*nix" is possibly even worse. I'm not even sure that you can call "*nix" a "thing" in operating systems. Which of the Six million versions and names of Eunichs do you mean? Some *are* good at this kind of thing, most suck worse than late-version windows (suffereing as they do from the same kind GUI IPL over-load as most modern OS's). If you have insights on which ones are good for this and why, I would like to hear it, but the same statements I made about Oracle's mthycal performance dominance apply geerally to Unix also. Why doesn't nix with Oracle, or Sybase to mke it even more pointed, dominate the TPC benchmarks then? They don't they're just in that horse race along with everyone else.
gnuoytr ... You've quoted me out of context. ... The horsepower referred to is the MVCC semantics ...
We may have a miscommunication here, but its not because I am quoted you out of context. In my 35 years in the computer field, "horsepower" has only ever meant one thing: performance. I am not aware of any context in your statement that magically gives it a defintion it never had before. And as for Nix scalability, TPC measures that too: it's still not running away with it.
.. SQLServer (modulo snapshot) has neither. I'm not asserting that one *must have* these things, only that SQLServer currently (and so far as *nix support is concerned, always will) has nothing comparable. It is clear at this point that MVCC semantics and multi-tier architecture are a very good match. ...
Lots of interesting statements here tha I would like to hear more about, but no real explanations or substantiation. I am especially intrigue about how the Nix communities opinion that MS "will never have" something, could be regarded as anything other than prejudice (given the imposssibilty predicting the future with that kind of accuracy).
M$ and IBM are playing catchup; revealed by their actions with their databases. To argue otherwise is, at best, disingenuous.
The only thing disingenuous here is acting like you should be able to foist a bunch MS/SQL Server hating hand-waving on a site dedicated to SQL Server and not get called on it.
[font="Times New Roman"]-- RBarryYoung[/font], [font="Times New Roman"] (302)375-0451[/font] blog: MovingSQL.com, Twitter: @RBarryYoung[font="Arial Black"]
Proactive Performance Solutions, Inc. [/font][font="Verdana"] "Performance is our middle name."[/font]
August 21, 2009 at 3:54 pm
Umm. This thread is about Oracle and MySql. Whether such a thread should be on SQLServer dedicated site is not my call.
August 21, 2009 at 4:46 pm
gnuoytr (8/21/2009)
Umm. This thread is about Oracle and MySql. Whether such a thread should be on SQLServer dedicated site is not my call.
Actually, this thread is a discussion based on an editorial regarding Oracles purchase of MySQL. It is totally relevant on this site. We may be MS SQL Server centered, but we realize that there are other systems out there and SQL Server does need to interact with these systems at times. So it makes sense to pay attention to things going on outside our normal world of SQL Server.
Viewing 11 posts - 16 through 25 (of 25 total)
You must be logged in to reply to this topic. Login to reply