July 28, 2004 | Comments: (0)
The news broke today, first on InfoWorld, then perhaps on other sites, that next week at the LinuxWorld conference in San Francisco, Computer Associates is planning to fulfill a promise it made this spring and deliver its Ingres r3 database to the open source community.
CA made that public at its CA World conference back in May, with the stated goal of open sourcing Ingres within 90 days. But the more interesting piece of CA's plan is that the company is hoping to compete with proprietary databases and has its sites set on surpassing Oracle 10g in terms of performance.
Tony Gaughan, senior vice president of CA, told me last week that although Ingres r3 admittedly lags behind Oracle 10g for now, Ingres blows SQL Server away on the same hardware, and is comparable to Oracle 9i. Furthermore, he said CA plans to catch up to and perform better than Oracle 10g.
Now those are some bold claims.
Although it may have appeared to most people that CA was riding out the Ingres wave until it crashed upon the shore, in fact the company has spent the last 12-18 months developing the database.
Naturally, naysayers maintain that CA is embarking on the nearly impossible: trying to fight its way into an already mature market and elbow entrenched database soldiers aside. While I don't expect any enterprises to replace existing databases with Ingres r3 or even to add Ingres into, say, an Oracle shop where they otherwise would choose to add another Oracle database, CA and at least one industry analyst claim that there are opportunities, namely emerging companies and SMBs that need either a first or even a new database.
Here is my challenge to CA: put Ingres r3 through the TPC (Transaction Processing Performance Conuncil) benchmarks.
I'd like to see how Ingres stacks up against Oracle, SQL Server and DB2 in the benchmark tests, even if those tests do not always reflect real-world scenarios.
The TPC benchmark, at the very least, would bring CA's claims out of the abstract and into concrete, quantifiable terms.
And for the record, I'd hope that the race is closer than doubters might anticipate it to be.
Posted by Tom Sullivan on July 28, 2004 07:20 AM
July 23, 2004 | Comments: (0)
Is Salesforce.com just a one product wonder?
Salesforce.com continues to ramp up its software as a service capabilities as exhibited this week with the introduction of their Summer O4 release.
The company's business model rests on its ability to have many customers reside on one instance of the database, called multi-tenancy, while creating virtual secure environments for each.
Up until now, Salesforce had hit a brick wall in terms of the number of
customers, individual users and data sets it could support with the multi-tenant architecture. This week's announcement is all about overcoming that hurdle.
Now, according to Salesforce.com execs they are able to target much larger
organizaions, announcing recent wins of Corporate Express with 2,000 subscribers and ADP with 3,000 subscribers as proof.
The question for me is, is Salesforce.com a one-off phenomenon or just the tip of the iceberg of what's to come? Is software as we know it really going to go away big time?
Interestingly enough, Salesforce has one big ally in its just say no to software campaign.
Despite the fact that IBM recently partnered up with rival Siebel for Siebel On-Demand, CRM as a service, it appears to me IBM is betting that Salesforce.com is not a one product wonder.
Big Blue's message, if you listen closely, is that software is irrelevant. It's the middleware, WebSphere Whatever, that connects anything to anything, seamlessly, transparently, easily, etc., etc.
Is Siebel sowing the seeds of its own demise by supporting an on-demand
service? Is Siebel OnDemand killing off the branding the company spent so much time and money on creating?
The company's entire business model is built around a suite of packaged applications that of course can be customized. Now, along comes Siebel OnDemand, and users don't even see the package, or the CD. Don't discount the psychological effect of that as well.
The software market is definitely changing but it is still a bit too early to tell where it will come to rest.
Posted by Ephraim. Schwartz on July 23, 2004 10:00 AM
July 16, 2004 | Comments: (0)
When has an IT project failed?
I'm certain most of you have read, or heard, someone somewhere quoting statistics about the failure rate of IT projects.
"Eighty percent of all IT projects fail," they may have explained.
For clarity's sake, that number 80 is an arbitrary one, perhaps you've heard 70 percent, or 60 percent, or another percentage.
The point is not the precise percentage of IT projects that fail, but the definition of failure itself. Is going over budget on an application rollout, for instance, really failure? I don't think so.
Is delivering a month late on a project that ultimately saves your company mil-lions while enhancing productivity the same thing as a failed effort? Well, no, it's not.
That said, it is true that some IT projects fail, are abandoned and eventually begun again, typically with a different vendor.
How do you determine when an IT project fails? And, what percentage of IT projects that you have worked on qualifies as a failure? Write to me at: tom_sullivan@infoworld.com.
Posted by Tom Sullivan on July 16, 2004 07:07 AM
July 12, 2004 | Comments: (0)
Although it happened last May, news of what looked like a classic hacker exploit at Microsoft's TechEd conference is just coming out now.
Using AirMagnet, a WiFi LAN management application, Microsoft network security people detected something interesting happening on the show floor. There was a roving outage on the WLAN.
As best they could determine someone was running a wireless DHCP [Dynamic Host Configuration Protocol] server on a laptop in a backpack walking up and down the aisles. It had the effect of sending user traffic to his laptop.
In the parlance of the day it's called a honey pot, trying to snag traffic going someplace else, and the hacker is giving you a bad address which is his.
Microsoft dispatched people using the AirMagnet software on handhelds and laptops finally cornering the perpetrator who had innocuously secreted himself in the corner of a large room off the show floor.
Sounds like a dangerous character but actually it turned out to be fairly harmless, according to the Microsoft official I spoke with who was on the scene.
It seems the fellow did indeed have a laptop but he also had a booth at the show and his corporate IT people set him up with a wireless AP and the wireless DHCP server. According to the Microsoft official, who I spoke to without his or her ubiquitous PR handler in tow and so it is best not to name him or her, his laptop had two protocols installed. When he closed it it became the wirless DHCP server.
He set his laptop up at the booth but every time he closed the lid to talk to someone he would lose his IP address. Trying to figure out what was wrong with his system, he stowed the laptop away in his backpack and went looking for a quiet spot to investigate.
Meanwhiel Microsoft people determined, by the strength of the signal, the area it was operating out of, narrowing down the location of the rogue server to one corner of a large lunch room with hundreds of people sitting at tables, most of them with laptops.
Once narrowed to one small area, they went round and politely asked individuals if they were having any trouble with their network. When the alleged perpetrator said yes, they inquired if they might look at his laptop.
Problem solved.
See, it's like my mother always told me, you catch more flies with honey, or something like that.
Posted by Ephraim. Schwartz on July 12, 2004 09:32 AM
July 08, 2004 | Comments: (0)
No one gets fired for buying IBM?
Back in the days when IBM was the only King of all it surveyed in the Land of IT, there was an axiom that many professionals in larger shops lived by when undecided about which systems to buy: No One Ever Gets Fired For Buying IBM.
And while this saved and/or accelerated more than a few IT careers, it may have ended or seriously limited some. Anyone out there know of a case that proved the axiom wrong, or is willing to talk about a personal experience?
Or even, perhaps, talk about a case that proves an updated version of the axiom wrong: No one ever got fired for buying Microsoft or Linux.
(I posted for Ed Scannell.)
Posted by Tom Sullivan on July 8, 2004 08:55 AM
July 07, 2004 | Comments: (0)
Close the door on open source Java
Lately, there has been a lot of talk about how Sun Microsystems should give up its role as steward of the Java programming language and offer up Java to an open source format, enabling developers to do what they want with the code.
Proponents argue that new possibilities might arise for Java through open source. Sun counters that Java already is malleable enough and that compatability needs to be maintained with the Java platform.
Funny how two of the major proponents of open sourcing Java are two Sun competitors - IBM and BEA Systems. How many of us are open to business advice from major competitors, particularly on matters related to our company's crown jewel technology?
While some may have gripes with the Java Community Process for amending Java, it seems to be a good enough system for expanding the realm of Java. An open source format could lead to variants of Java that do not work with each other.
Open source is not the solution for all ills, if there is even an ill here to be considered. Let's keep the status quo for Java.
Posted by Paul Krill on July 7, 2004 01:25 PM
TOP STORIES
ADDITIONAL RESOURCES
- Remote Access: Maintain Security and Decrease the Burden on IT
- Beyond AntiVirus: Symantec Endpoint Protection
- What Every Enterprise Needs to Know About VDI
- Disaster Recovery in Minutes
- Protecting Microsoft(R) Applications
- Reduce Recovery Times and Tape Costs
