Product cycles too short for small businesses (esp. Red Hat)
Jay Smith <jay-BB6z0fKkvBxWk0Htik3J/[email protected]>
| Newsgroups | gmane.linux.usability.annoyances |
|---|---|
| Organization | Jay Smith and Associates |
| Message-ID | <[email protected]> |
Red Hat's recent change in their product release cycle strategy has major negative implications for a small business such as mine. A couple years ago I was able to assume (excuse me!) that when a major Red Hat linux version was released that I would be able to install it and keep using it for 2-4 years. That I would be able to update it, install security patches, and even upgrade it for 2-4 years. I understood that after a couple years they would no longer backport new applications or major upgrades, but I had reason to assume that security releases would continue for a reasonable and practical amount of time. Red Hat has turned that on its head. They split their product line into two major directions: 1) Inexpensive/free product but with short product cycles. This is where all the new stuff would be available, etc. However, within a year or two they would no longer be supported, including security fixes would not be offered. The intended user is the casual desktop user, hobbiest, geek, guru, etc., etc.; either those who don't use it hard or use it very hard and want the latest. 2) Costly product with very long product cycles. Little or no enhancement; it is pretty much locked down. However, security updates and the like would be available for a long time. The intended user is the corporate enterprise running big databases (i.e. Oracle and the like). Well.... I am stuck in the middle. I am a SMALL company, but I have to run a big database (Oracle). I feel that Linux is not yet ready for true 100% desktop use (for us) because we can't yet run all the apps that my company needs to run on it. (Thus we run Windows workstation PCs that runs Linux apps in XWindows provided by Cygwin.) I desperately need the new features that are appearing in the new linux releases so that I can move away from Windows toward 100% linux. However, I also need the long-term stability of knowing that I can still get essential security fixes. The combination of new linux versions always seems to coincide with the need to upgrade equipment. If doing one, it only makes sense to do both -- but we usually would not do only one and not the other. However, this is a big deal in a small company and usually takes one person 6-8 weeks to fully build, install, configure, test, and debug the new environment. That is 6-8 weeks full time at $60/hour -- it absolutely kills us. Part of the problem is that linux seems to lag on drivers for the new equipment. Because we can only afford to replace equipment every few years, when we do, we try to get at the front of the curve. However, that means that we end up with weeks fighting SCSI RAID drivers that don't work; video drivers that conflict with this or that. This costs real time and real money; more than my small organization can justify. Yet if we buy 1-2 year old [new] equipment, by the time we replace it in five years, it is ancient technology and the last couple years of its life are spent hoping that it has enough horsepower to do the job. Jay