Sun Hints At Open-Source Database Offering 167
An anonymous reader points out a ZDNet story which begins "Sun Microsystems has raised the possibility that it might offer customers its own database, a move that could trigger displeasure at Oracle but curry favor with open-source advocates," writing "Last week, during a meeting with financial analysts, Chief Executive Scott McNealy showed a slide that placed the words 'Sun DB' next to a list of existing database products. McNealy offered no details besides 'stay tuned.'"
Uhm... (Score:4, Interesting)
Oh really? Um, yay.. (Score:5, Interesting)
Re:Uhm... (Score:1, Interesting)
Build or Buy ? (Score:5, Interesting)
Given that a reasonably useful database system would be several hundred thousand lines of code, and, that Oracle & IBM have a 25 year head start not to mention MicroSofts 10 year head start. I don't think it would make sense for SUN to roll thier own database software.
So the question is who are they gonna buy? IBM has already snapped up Informix. CA has "given" Ingres to the Open Source community. SAP has donated SAP/DB to MySql. MicroSoft is unlikely to sell Access or SQLServer. Which leaves -- Sybase?
Could be intersting.
I doubt they can unseat MySQL... (Score:3, Interesting)
This begs the question; exactly what role would high-end Open Source DB software be able to fill today? Oracle is well entrenched with both DBAs and businesses -- Unless there are serious flaws in it that I am unaware of, I don't see the SunDB going very far.
Say Ingres (Score:5, Interesting)
a) It is Open source
b) CA is a non-competitor (no application server)
c) CA has been harmonizing their open source license with Sun's (I wonder why?)
d) CA hopes to make some buck from Ingress and even if they split it even, they're going to make a shitload more than by cooperating with Oracle.
e) Ingres has parallel features like Oracle RAC so it's more suitable for Sun's vision and for enterprise customers than PostgreSQL or other open source databases.
f) Oracle is competing with Sun (Oracle's application servers compete with Sun's J2EE servers/apps); there's no reason for Sun to help Oracle.
I'd really really enjoy see Oracle on their own. I've really had enough of their sales people...
The time for them to pause and think real hard how they're going to compete in the future.
Did they really think their competitors were going to stand idly and watch them take all the money (Oracle + Linux).... Hahahaha....
Clustra anyone? (Score:5, Interesting)
Used it a lot myself [leyton.org], and felt that - like many other companies Sun have bought - the pointy haired bosses there just didn't realise what they'd acquired.
Maybe a much-needed clue has finally hit home at Sun, and they're going to give Clustra the lease of life it sorely needs and deserves.
No matter what DB it is, Sun needs its "stack" (Score:4, Interesting)
With MySQL being dual-licensed, and questionable for Enterprise-level DB use, it's not really an option to sell incorporated into the stack. PostgreSQL would be an option, since they could fork it (and the PostgreSQL team not having heard anything is irrelevant to an extent, since it's BSD-licensed). I think we can sit back and see what happens pretty safely. They're certainly not going to make things incompatible with Oracle for a back-end, but I'm sure they'd like to offer a cheap solution since they're obviously trying to lower-cost solutions in order to stay alive.
What's the status of compatibility with native Java bits with Ingres? Oracle has obviously bitten on the Java-compatibility of everything, but I think that anything Sun would want to do DB-wise would keep Java squarely in the mix.
Re:Ahem (Score:3, Interesting)
Keep using your GPL software.
It really goes on my nerves to see people claim that GPL is the only "free" license?
CDDL? Big deal - download the software and use it - you never have to pay a single cent to Sun.
The point here is that
a) As the fucking article said, Sun's fed up with giving Oracle 50% of every DB deal they close while Oracle constantly competes with their application server platform
and
b) Lack of their "own" database is deterimental to Sun's utility strategy (presumably even IBM can take their lame DB2 and make money by renting it to their utility or Websphere customers), while Sun is stuck with Oracle.
I think GPL databases like mySQL and PostgreSQL are too "Linux-biased" for Sun's liking, so they might be looking for a non-GPL open source DB with enterprise features.
bad for Open Source (Score:3, Interesting)
Very few people in the enterprise world trust MySQL or PostgreSQL for anything other then web apps so this isn't going to win Sun any new business.
Oracle is an amazingly vindictive company, they will put the screws to Sun if they feel even slightly threatened. This is bad for Open Source because it just gives IT managers one more reason to replace Unix based systems with WinNT. Convincing your boss to move from Unix based commercial OS to Linux or BSD is a lot easier then trying to get Linux or BSD into a Windows shop. So in the end this will be bad for Open Source.
Makes sense - here's why. (Score:4, Interesting)
Here is why: Oracle is now officially pushing linux on its customer base (they are slowing moving Oracle Hosting Services (OHS) over to a Linux based service. IBM is removing support for Solaris (Domino, Websphere, DB2). And Checkpoint is pushing Linux appliance servers. And so Sun is seeing an assult from all quarters.
In fact most people buy Oracle per CPU (typically $50K per CPU). Those running a machine with AMD Opterons running 64 bit Suse Linux and Oracle can expect to see a 4x improvement in performance per dollar of Oracle licensing fees. PowerPC also outperform Sun machines - and so many Banks are switching to AIX to reduce Oracle licensing fees.
What does that leave for Sun? To move up the value chain and start selling a system with a database integrated right into the OS. Sun will want a database that they can control though - so I bet the relationship with CA Ingris will sour (joint ventures almost never work) and they will switch to supporting Postgresql or another database they can dominate and buy up most of the developers.
This Is Only Relevant (Score:3, Interesting)
Sun is in no position to beat Oracle, MS SQL Server, Sybase, or (in the OSS community) MySQL, FireBird, and PostgreSQL with something new in that space. No community for one thing, no rep for another.
If it's just a "warm fuzzy" for their locked-in customers nervous about open-sourcing Solaris, then it's irrelevant to the rest of us.
The database in question is... (Score:4, Interesting)
The database in question is probably a database originally created by a Norwegien company called Clustra. This company was acquired by Sun 2 or 3 years ago. Clustra built a distributed database system that was seen by Sun to be a good fit for Sun's J2EE platform.
If it's true that this database is being offered as an open source product, it could be very interesting because it's a very good database from what I hear.
Re:Ahem (Score:1, Interesting)
Sadly, you are referring to pre-dotcom Sun. I started here at the end of dotcom era... and let me tell you, geeks do not run the company in any way or form. It's process ("business") drones, with their beloved Sigma bullshit, assisted by grossly incompetent mid-level (and, lately, high-level...) management. Geeks exists, and in some limited cases, thrive, in their niches (Solaris, maybe [parts of] Javasoft)... but they are an endangered species. On top of process folks, there's also the second annoying group of folks: buzzword believers. You'll find plenty of these jokers as "architects", and "staff engineers", at least outside truly competent parts of organization. Oh and finally, we have the geniuses who started during dotcom years. Take a sysadmin, or DBA, add some inflated expectations, and voila: here you have a new Sun "developer". I hope you never have the "pleasure" to work with these self-taught "experts".
But salaries are nice -- I have no complaints on that part. Just make sure you start with nice salary; for past few years increases have been minimal (consistent with financial results, yes). And if you want more money, take the management route and learn your buzzwords (and Sigma!).