Re: Food for thought...

On Dec 14, 1:29 pm, InDeep <ind...@indeep.com> wrote: > Ian Michael Gumby wrote: [SNIP] > > Sun is a bit more interesting. In purchasing MySQL for a billion, > > they've essentially bought the rights to the software that millions of > > site use in LAMP and have put unlicensed in to production > > environments. (Sun won't chase them for license fees because of the > > negative publicity.) > > The license allows basically unfettered production use Gumpy.  They > don't have much if anything to chase. Uh no. > > > But by buying MySQL, they have a "framework" or > > "core" to a database engine that they can release on a commercial > > scale. The thought is to capitalize on the "brand" and potentially > > convert growing LAMP installations to their commercial version which > > would offer features like data/table partitioning, shared nothing > > clusters, and they could code the engine to run more efficiently on > > their Rock hardware. > > Partitioning is available today.  MySQL = SE + partitioning TODAY. > Uh not really. If you read the blog of one of MySQL's founders, he admits that the partitioning is not ready for prime time. I'm not saying that they are not going to get it right, but that its not there just yet. > > Rock is multi-core, however Sun's concept is unlike Intels where the > > cores are more lightweight and have a different I/O structure to > > memory and the other cores. Sun claims that their processors are more > > efficient for multi-threaded applications. They acknowledge that for > > the majority of applications out there, they would have to rewrite the > > apps to allow for better parallelism.  Informix is probably the best > > commercial engine that can take advantage of the Rock because of how > > it has been engineered. > > Why think in the theoretical when you can think reality.  IBM has a > bigger problem of shrinking mind share about IBM as a company.  IBM isn't > on the top of company lists of hardware choices anymore, at least not from > where I see.  Everywhere I turn I see Dell and HP, not IBM, and not much > Sun either.  And what apps do we get to pick from IBM?  __________________. > Crapplications.  Lotus? yawn.  DB2? yawn. Domino?  Websphere?  Tivoli? > Absolutely no mind share in mainstream computing for IBM's stack.  Just a > bunch of old, ugly, 1980s style apps.  Nothing innovative about any of it. > Most of these apps are geared toward big companies, they have no reach into > the larger market. > You need to look at the bigger picture, IBM's SWG is still making money. I agree that certain products are getting long in the tooth. I also agree with the attitude of why purchase RAD (Rational Application Developer) when you get get Eclipse and NetBeans for free? (And yeah, I prefer NetBeans). As to thinking theoretical, if you don't start with theory, you're just a code monkey who's hoping to get it right. An example? All these 20 something code monkey's who rave about GoF. (GoF == Gang of Four which is a reference to a pattern's book. Of course what these kiddies don't really recognize is that these patterns existed before GoF and have their roots to guys from the 60's and 70's.) If you didn't learn the theory, then you'd look at GoF as a holy bible. If you did learn theory, you'd understand where GoF is coming from and find it to be a good reference book. If you don't like that example, then you're looking at it from the standpoint of an MIS degree vs a Software Engineering degree. Or if you want to put this in terms you may appreciate... Looking at the Ramones vs classically trained musicians. > You can bet that as IBM shrinks it will have to dance and start to innovate > to stay alive, besides just outsourcing and offshoring their projects to India. > Offshoring is a separate issue. IBM is growing because they're acquiring more infrastructure software pieces. (Datastage anyone? ;-) > IBM may even be forced to unload Informix or other products simply because they > can't sell or support it, in realigning their core products.  Don't think it > wouldn't happen.  The best time to sell Informix is on its way up as it emerges, > not after it flattens out. > No, they won't. Even though the IBM's project arrowhead didn't work, there's still to much shared technology for IBM to spin off IDS. Either you see IBM kill it off or relegate it to the same fate as IMS, or you'll see IDS take over the distributed (LUW) and DB2 reside on the AS/400 I Series and the mainframe. In the end it will be an either or situation. > Reality is this:  Name brand big ticket items are at the bottom of the > list now.  Without going into details I am working with a media company that > is moving aggressively towards a total open source stack, and may even be > looking at the desktop as part of this strategy. They have a strong Oracle and > Java product but the periphery is rich with open source tools.  Instead of > buying expensive monitoring software they have Nagios, and Cacti.  Instead of > BEA they have JBOSS. Instead of a payware ticket system they have an open > source ticket system.  Agility is the key. > LOL... There are some things that you can use that are open source. There are other things that you can't find an open source equivalent product. Databases are one of those. Sun's new and improved MySQL most likely won't be open source, however there will be an open source version out there. Greenplum's latest and greatest isn't open sourced. I've seen a lot of companies push for the LAMP stack. But when things go wrong. Who do you think they're going to kill? Not you. You're the consultant who's long gone. Scalability is a big issue. What happens when a retailer's website melts under the pressure of black friday? Who get's tossed under the bus. I'm not saying that Open source doesn't have its place. I have 1 windows box and 4 servers running OpenSuSE. (ok 2 are up, 2 are in pieces as I'm doing something interesting with one of the boxes.) I run Postfix instead of Exchange or Lotus because well, I don't have to support 100's of users on my mail server. I chose Postfix over Sendmail because it was easier. But I will rather run IDS over any other db because of what it offers. > > Clearly Sun's long term goal is to enhance MySQL, however Rock is > > coming out soon and MySQL will take more time (years) before it will > > be able to take advantage of Rock.  This puts Sun in an interesting > > position. > > I wouldn't link MySQL's success or failure to Rock.  MySQL is already > out there, entrenched.  Download the latest version and really take a > look at it.  Compile it from source, and look at its options. > Uhm reverse your thinking. Sun is still a hardware company, in spite of what they want you to believe. Just look at their qtrly report and annual report. Sun made a billion dollar gamble on converting MySQL freeloaders in to MySQL licensees and also to give them their own database as part of their young software stack. Again, you have to take a step back and look at the bigger picture. You're myopic when it comes to LAMP and don't see the bigger picture. Remember the 80/20 rule when it comes to revenue.@@NL