“Oracle on Friday announced a surprise $5.1 billion takeover bid for enterprise software maker PeopleSoft, only a few days after PeopleSoft said it was acquiring rival J.D. Edwards for $1.7 billion. Oracle is offering $16 cash per share for each share of PeopleSoft, a roughly 6 percent premium over PeopleSoft’s closing price Thursday of $15.11.” Read more at ZDNet.
Oracle knows they need a second franchise for longterm success; this has been clear for a while.
However, for Oracle to go after PeopleSoft is just too funny. Oracle is well known for being a very unfriendly and anti-customer company.
The better cultural match would be Siebel having been founded by an ex-Oracle guy. However, “Evil Siebel” is a dead company walking and Oracle knows this.
So if Oracle succeeds in buying PeopleSoft, it will kill the pro-customer culture at PeopleSoft. It would be pure hell for the fine folks at PeopleSoft.
Perhaps Oracle should consider a merger with Monsanto, Exxon, Lockheed, or some other fuck-the-world company that has a compatible corporate culture.
“anti-customer” ?
Simple.
If you sell the best product in your market, people will buy it, no matter how “anti-customer” you are.
Until another enterprise database in Oracle’s class (don’t even talk to me about SQL Server!) can best Oracle in price, performance AND scalability, they can remain the way they are.
At least the Peoplesoft software that I’ve used has been crap. The people that complain about the horrible user friendliness of oss have obviously not used any proprietary, intra-company software. There’s some stuff floating around the halls of the phone company I work at that would make some of you want to remove your eyes.
But it is easy to see how a company can be anti-customer. After, just because you don’t like people dosen’t mean you don’t like money…
Despite Oracle’s comments to the contrary, they are much better at building sales brochures and relationships with the decision makers than they are at building databases. Informix, for one, has always had superior technology (and reliability and price) compared to Oracle. The “best” products don’t always win in the marketplace (OS/2, Beta, etc.). The marketplace does not operate strictly on logic, which is why free enterprise is so interesting!
Oracle’s applications apparently suck and are losing revenue (one article I saw said they were down 26%) so they are trying to acquire the expertise to make them better. Note that they don’t plan to continue selling PeopleSoft. They want to migrate everyone to Oracle apps. PeopleSoft runs on more databases than just Oracle, so this would be an interesting tangle to watch.
That’s what I have never understood about Oracle; they work so hard to compete with their partners (the application builders who can choose to run on the Oracle database platform).
Oracle is losing ground to IBM’s DB2. IBM is a lot more frendly and DB2 is in the same class. MS-SQL is not enterprise ready (it’s still only ANSI SQL-89B compliant; they don’t even support the SQL-92 or SQL-98 standards).
We use Infinium instead of PeopleSoft. A lot more frendly and they give us the source code with the application so we can freely modify it for our needs (makes upgrades a little more fun since we need to roll our modifications into each new release, but thats the price of customizing software).
At least I think so.
Yes, they have one excellent product, but the gap is getting smaller every day. Just look in the progress that DB2, MS SQL, PostgreSQL, MySQL and Firebird had experienced.
And with that, how long will take to people start to think seriously if all that price tag is worth of be paid?
I mean, if they want to survive to all the changes that are happening in the software industry, they have to have a closer presence on real market. In other words, to make software not only for depelopers but for all business chain.
On paper Oracle would be a much bigger enterprise player, more of a threat to both IBM and Microsoft. I noticed Borland’s shares were up almost 8% today – I could see Oracle going after them next, and IBM and Microsoft entering the bidding as a defensive measure.
How about a Oracle + SUN + Borland + Peoplesoft + Corel merge and create the ultimate uber company?
Vision this, centralised Starfire server with 128 Ultra Sparc IV CPU’s running Oracle and the Corel software range. At the end user, the employees are logging on using SUN Ray appliances with Java cards, and get to work on their daily tasks. Developers in the IS department are able to now deploy custom written applications on Solaris with minimum fuss thanks to the ease of use Borland brings.
Then, coupled with GNOME 2.6 (or 2.8, which ever is out by then), SUN/Oracle/Corel/Peoplesoft/Borland will have an end to end solution for every customers need. Essentially they’ll be a mini-IBM, most likely larger than HP.
White elephant. Solaris and Sparc are dying, and WordPerfect is dead (except at a few law firms). Besides, Oracle doesn’t want to field workstation products – that’s why Larry keeps talking up ultrathin clients.
Are you really living up to your name? what the hell do you think is going to run on the server which these applications will need to interact with? think about that sunshine before making such stupid remarks.
Solaris and Sparc aren’t dying, Scott just needs a right royal kick up the ass and realise he is there to serve the shareholder through creating more wealth. He isn’t there to “bash Microsoft”, have a temper trantrum over IBM or blame the voodoo man for a plumet in sales.
A thin client is absolutely bloody useless if there is nothing running on the server. I might as well try and hook up a thin client to a toaster if I went with your way of thinking.
it will be pretty slow for sure, not to mention an ugly UI creeping up into view for every repaint event.
Hence the reason why I said 2.4/2.6 or some future version, depending on whether STSF is up and mature by then plus other factors that need to be taken into account.
SUN isn’t too far into the hole, so they can actually pull themselves out IF they’re willing to get over their Microsoft-abuse-addiction and start getting customers on board and workstations/servers starting to sell at a reasonable volume.
Regarding running X over a network, having done it over a VERY busy University network running at 10Mbps without any issues of slowness, I would say that you comment regarding the speed is more BS than reality.
these days, the benchmark should be a remote GUI session over a 28.8 kbps dialup network.
VNC is barely usable over a 128 kbps connection
If something looks to you like BS, that’s because u are filled with BS 😎
We’re talking about a frigging corporate network. Who gives a crap what Joe luser has. The money is made selling big friggin servers, software and maintainance to businesses. They the hell should SUN even CARE about people like you and your 28.8K connection when that is not their focus.
What we’re talking about are frigging corporate networks. Who gives a crap what Joe luser has. The money is made selling big friggin servers, software and maintainance to businesses. Why the hell should SUN even CARE about people like you and your 28.8K connection when that is not their focus.
If you took atleast 5 second to read what I wrote, you can clearly see that I was talking about the BUSINESS WORLD not the Joe Luser world which you consider the be-all and end-all of “profit making”.
—
Btw, could the site owner please make a system with accounts so that people can EDIT what they post if they have made a mistake.
G’day mate! Oracle and PeopleSoft are both becoming Linux stalwarts, as opposed to Solaris or proprietary Unix. That’s what pushes down TCO for their enterprise solutions compared to Microsoft’s. As for Sparc, it will always suffer economies to scale problems relative to Intel. Nobody wants to buy Sun except to get Java and J2EE. Solaris is what you’re stuck with after you’ve bought Sun.
Well here are the plain facts:
1. Oracle has licensing charges based on the following:
– Product
– Annual Transactions (Difficult to define.. even for Oracle)
– Class server (Enterprise, Workgroup, etc.)
– Number of CPU’s
2. Linux is cheap and runs on cheap hardware that can’t address large amounts of memory (over 4GB’s).
3. Sun has large-scale hardware (Lots of CPU’s and Memory.. look at the SF15K).
4. Even with RAC clustering you can’t go beyond 8 nodes because of latency for Cache Fusion.
5. By supporting Linux, Oracle insures larger revenues because companies have to buy more PC’s to place into RAC clusters due to the memory and SMP scaling issues. Which means that you’ll have to have a RAC cluster for each of your major databases. This means more DBA’s, more SA’s, more servers, more storage, and more consulting business for Oracle.
6. It makes more sense to have a large system like a SF6800 or a SF15K to consolidate your Oracle products onto than a bunch of PC’s that will not only take up more space.. but will consume more management costs, not to mention the licensing costs from Oracle.
7. Oracle and Sun have a long relationship, even with development of new products. Sun and Oracle are working on getting RAC to work with Sun’s blade servers. Oracle gets new Sun equipment all the time, even before it’s available to the market.
8. Oracle, Sun, and Veritas has been partnered for a while to provide out of the box solutions for RAC configurations for a reason.
9. By pushing Linux on the low-end, it locks people into using Oracle products.. but opens them to needing bigger systems from people like Sun when their companies get larger.
10. By using Linux, Oracle and Sun will have a perfect life cycle to bring in new customers and lock them into hardware and software solutions. That is the real business of Oracle.
So do you still think Solaris and Sparc are going away? Grow up people, Linux is just bate to bring in new customers. If Oracle and Sun had their way.. everyone would be using Sun Rays with java web browsers to access Oracle products running on large Sun servers.
And if you think IBM is going to help out, you’re wrong there too. IBM see’s Linux as the next “Consulting Oil Field”. By getting you to switch to Linux, means you have to port your Windows and Mainframe apps, which doesn’t happen magicly with “Pixie Dust”. It happens with IBM consulting services, hardware, software, and project management. And that is the real reason for IBM to support Linux. It has nothing to do with them loving Linux. It has to do with them squeezing Linux into submission. It’s not like they don’t use new Linux clients as lab animals to build commercial products out of.
So too all of you who think Linux will remain pure and able to beat M$.. you’re wrong. Linux will be manipulated into defeating M$, and be transformed into the vechile for the few remaining large companies to take it’s place. It’s not about whose OS is better, it’s about whose hardware, software, and consulting services are *better*.
The dot-com implosion has made for some interesting times and weeded out the weaker companies regardless of how good their products were. So keep your ears to the ground and learn from experience.. the mass media marketing battle is designed to win the minds of the weak.
Soap box mode off..
Get your facts straight:
2. Linux is cheap and runs on cheap hardware that can’t address large amounts of memory (over 4GB’s).
Linux IS cheap but it also runs on 64 CPU NUMA systems with up to 64 GB of physical RAM. If you don’t believe me ask SGI and IBM about their NUMA support project. Linux kernel version 2.5 and above will have a new threading system that will rival performance of even the best of the traditional database platforms like AIX and Solaris. Linux allows for farming processes out to other nodes both with and w/o storage IO capabilities and I’ve seen clusters using Linux many times bigger than your “8 node” limit. Linux is also open source which allows for in-house support indefinitely on current or future hardware investments. Kernel 2.6 will most definitely be ready for large scale enterprise class deployments.
Next time you take out your soap box, be sure you have current information or someone else will burn it again and you again.
Oracle is feeling the heat from IBM which has 2 database systems that are between them, suitable for everything from workgroup type database servers up to full data mining systems with hundreds or even thousands of terabytes of storage with your choice of Informix on the low end, or DB2 for the middle to highest ends. BOTH systems run on Linux and the new 2000 TB database servers from IBM were designed to run a SUSE Linux-derivative from the start.
I’d hand you a fire-extinguisher from the flames, but I’m a pyro.
I’m well aware of the HPC clustering capabilities of Linux. I’ve done consulting work for several vendors of Linux HPC solutions.
The problem is that not many business have use for such things. It’s something that you don’t call up Oracle to sell software for.
The only way that Linux gets around the inherent I/O limitations in such situation is with low latency transports like memory channel. Otherwise, you are dividing up your workload with a batch system. Not really that new, people have been doing this in labs for years.
As for NUMA systems, SGI is not doing well at all. They have some very impressive products, but most of the technology that went into building those systems has been commercial viable by systems from Fujitsu, HP/Compaq (Tandem and GS Alpha series servers), and Sequent (Which IBM bought and killed off mind you).
While having the source code is essential for labs and companies that do research.. “every-day” companies don’t care about such things. They are more concerned about who will be there to support the system regardless of if someone leaves the company. You’re not going to get that with Beowulf Linux systems. You get that with large support contracts and big name software products.
Of course Oracle is behind on large scale clustering.. so is DB2. That is why the government uses Informix a lot:)
Also, when it comes to databases like Oracle, you need to have shared storage or a really large I/O proxing backend. From reading your comment you obviously have not used Oracle in RAC configurations. It has to maintain the same shared memory, cache, and disk locking across all nodes in order to maintain it’s “active/active” clustering. It’s not something as silly as a batching system like on Beowulf. It’s real-time clustering of your database and applications. Something that you can do on Linux, Solaris, AIX, and HPUX, but only with commercial software to provide the framework around it ( things like Sun Cluster, Veritas DBED/AC, etc.).The difference has more to do with your OS and hardware. With Solaris, AIX, or HPUX you have high quality hardware to run them on and large companies to support them. With Linux, the problem is the hardware, it’s not that high in quality. Also, when you buy a Sun system it’s not something you are going to throw out the window within a year because it’s completely obsolete. Having a system you can use for 3-5 years makes more sense than having a disposable asset. But you probably don’t know that since you obviously aren’t a decision maker involved with both the technical and financial aspect of things.
Let me know when you wake up and realize the difference between commercial products that real companies depend on and lab projects that companies can’t rely on. People don’t want something they have to build from scratch and support forever.. it’s too expensive.. that’s why people like Oracle and Peoplesoft are around:)