MySQL veteran drifts clear of Oracle Borg Oracle has lost a MySQL veteran who helped the plucky database start-up sink permanent roots in the developer and services communities. Kaj Arnö has left Oracle quietly, having submitted his resignation in June two days before Sun Microsystems' legal entity in Germany ceased to operate. Arnö was based in Munich. Arnö had …
Oracle certainly has more than enough people who know how to build a real RDBMS for real enterprises who make millions of revenue every day and account for that on the Oracle database system.
Losing all that MySQL "talent" is just the process of removing the second-rate guys. Those who were so smart to "invent" MyISAM and similar stuff. The real threat to Oracle is Postgres, but they also lack many enterprise features.
And more important, Mr Larry's personality (and consequential action) is more and more a major threat to Oracle's prospects. MySQL is a sideshow which matters to bulletin board systems and second-rate websites.
I've actually worked with Oracle software - a larger piece of tat has never been created (and I'm including Windows here).
Overly expensive and complex (need to keep the cash flowing in with training for the arcane rituals that involve chicken slaughter!)
It's a lumbering giant, and look what happend to the last lot of those 65 million years ago...
For the full list see http://www.mysql.com/customers/?id=287
5 years ago, i might have agreed with you. However, 5 years is a long time in computing.
I expect the core talent who wrote the original version of Oracle are long gone now as well.
Its much easier to improve a product than to initially write it.
The only advantage i see these days for Oracle is that its a much more mature product, however, anyone who has ever done any work with the oracle tools and that *really* awful universal installer they insist you use will tell you any of this "Oracle Talent" hasnt been let near them, although i suspect the "Oracle Interns" may have been.
And i doubt EXP and IMP have had an update since the 80's. Certainly MysqlDump is far superior. (And gives you human readable output!)
'And i doubt EXP and IMP have had an update since the 80's. Certainly MysqlDump is far superior. (And gives you human readable output!)'
Utter twaddle. Mysqldump is crap by comparison. Export / Import have been updated with each release, though the basic command set remains the same. But if you had anything more than half an idea about using Oracle why wouldn't you use DataPump?
A rose by any other name is still a rose. Look how quick XFree86 died as an example of what can happen if users get dissastisfied with a project.
I expect that people using MySQL will jump ship if they perceive that Oracle is not properly maintaining / developing the official branch. I expect migration would be relatively simple, and possibly made simpler if dists like Ubuntu, Debian or whomever dump the mainline packages for the branch.
I also expect PostgreSQL picking up a lot of business precisely because it isn't suffering from the same development uncertainty. In some ways it probably offers a smoother migration path for people considering Oracle anyway. For example it uses a dialect of PL/SQL (called PL/pgSQL) and would be considered a more "correct" database with none of the confusion or mess that goes with MySQL and choosing a backend.
But I've been enormously happy with the improvements in MySQL 5.5 (and 5.6 preview). I've started using 5.5 over the last few months and finally put it into production now that it's RC stage. The number of improvements in each point release are staggering, and add up to a much faster, more reliable, and less disk-gobbling database.
My big downer is that I never know whether Oracle is going to keep the mainline open source or kill it like OpenSolaris. If it turns into a no more previews, no more betas, just closed updates here on out, I'll probably look elsewhere.
Is El Reg really going to chronicle every person that resigns from Oracle? Last I checked they had around 80,000 employees, most from acquisitions so this is going to keep you busy. Is this really newsworthy? I've not worked on a single engineering team that couldn't absord attrition and still execute, its a sign of a well managed operation. Any particular reason El Reg doesn't detail every IBM, HP, and Cisco resignation with equal fervor?
Thing is, companies gamble - to get maximum benefit out of employing you, they have to keep you just happy enough.
For high-flyers that are a) rich enough to work for other reasons than paying the bills and/or b) employable enough elsewhere that another job is merely a phonecall away, that required level of happiness is significantly higher than for the mass of generic employees.
And what help is it if you have some primadonna who may do the work of 30 "generics" if he not only costs 5x as much but also requires a level of happiness that, for you to provide him with, increases your "costs" (as in: your time and/or your money spent on non-wages costs) another 5x ?
Also, given that these primadonna-type engineers don't reflect in "performance numbers" because they blow the scale - you can't do more than exceed targets, can you - what help is it to you as a manager if you can't show you improved your department's performance simply because your department's performance already is super-optimal ?
Just hire another 10 cheap minions that don't ask questions, don't make demands, and don't have ambitions. People that are happy to appear at 9am, shovel whichever stuff is thrown at them, and leave 5pm saying "glad to get home". Your work as senior level manager becomes so much easier, as you now have a mass of bodies to work with, and numbers average out better, and you (as a manager) get some freedom to "rule".
Performance suffers, somewhat. Not necessarily catastrophically though. Or, at least, not catastrophically till you've been promoted out of the mess. Creativity and ambition suffer a lot, but then you don't really need that in your mature core organization - you can always purchase inventions elsewhere !
That's a bit of the sad side of reality.
If you can afford to work for other reasons than money, why should you stay with a position that doesn't make you happy ? If you can get a position elsewhere for sufficient money but higher levels of happiness for yourself, why should you not take it ?
And if you're a manager for a person like that, why would you even want to retain them ?
When these people leave Oracle, a lot of faith goes with them.
Success of open-source products is quite often simply based of faith and persistence. If your community looses faith in the product, they will simply migrate to an other open source product. Oracle doesn't understand how this works in this segment. They only understand enterprise behaviour.
I expect that a lot of smaller MySQL users/products will migrate to PostgreSQL and MySQl forks in the next 5 years.
Not only Mysqldump, PostgreSQL dump is also FAR superior to Oracle's exp/imp. I used exp/imp for years and I was quite surprised when I saw pgdump in action for the first time. I just couldn't imagine that a database export/import can actually be fast and efficient.
PostgreSQL is catching up with features (features that matters) to Oracle. The new streaming replication and warm standby features are specifically created for building large, high-availability clusters, typically found in enterprise and/or cloud environments.
They also appear to have lost people from their storage group where you thought they would be keen to make money.
Jeff Bonwick recently left http://blogs.sun.com/bonwick/entry/and_now_page_2 as did Bryan Cantrill http://dtrace.org/blogs/bmc/2010/07/25/good-bye-sun/
Never was sure how they would make money from MySQL but the ZFS and the Fishwork groups should have been a nice entry point to bringing in money.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
