Collins Richey wrote:
endeavors justify the decsion - ethics, community vs. vendor support, etc., etc. I've heard many comments to the effect that we CentOS users are just leaches, since RedHat does the major work. I would like to know how those of you who use CentOS in commercial
Collins,
I'm sorry if I am unable to follow the question of ethics here. At what point did Red Hat's efforts in the open source community become magical and change the ethical environment Red Hat lives in for people downstream of Red Hat? Your question pre-supposes that because CentOS is downstream of Red Hat then somehow it is different than Red Hat being downstream of the Linux kernel developers & others (fill in other names as you see fit). Red Hat is downstream of well over 99.99% of the 'major work'.
Red Hat has built a business around using freely available software. As part of the licence provided by upstream authors who created, enhanced, and often support the software Red Hat is required to make the software source others created and they use for their business available to others to use as they see fit (within terms of licence). Exactly the same as Red Hat taking the source from upstream and using as they see fit (WTOL).
Commercial vs. community support is not an ethical choice. Neither is Red Hat's leveraging the efforts of the community to build a commercial enterprise. The community knowingly provided their efforts to others to do exactly what we, and Red Hat are doing.
regards Dave HornfordAssociates.com
On Apr 4, 2005 10:14 PM, Dave Hornford OSD@hornfordassociates.com wrote:
Collins Richey wrote:
endeavors justify the decsion - ethics, community vs. vendor support, etc., etc. I've heard many comments to the effect that we CentOS users are just leaches, since RedHat does the major work. I would like to know how those of you who use CentOS in commercial
[ snips ]
I'm sorry if I am unable to follow the question of ethics here. At what point did Red Hat's efforts in the open source community become magical and change the ethical environment Red Hat lives in for people downstream of Red Hat? Your question pre-supposes that because CentOS is downstream of Red Hat then somehow it is different than Red Hat being downstream of the Linux kernel developers & others (fill in other names as you see fit). Red Hat is downstream of well over 99.99% of the 'major work'.
Red Hat has built a business around using freely available software.
Actually, as I stated earlier in this thread, my question about ethics pre-supposes and implies nothing of the sort. I asked the question because it was brought up by others. I agree with you 100%. Linux is a GPL product whether distributed by a for-hire vendor or a community maintainer.
I'm interested in every aspect of the CentOS project and its relationship to RedHat and its policies. I'm still hoping to convince my employer that CentOS would be a good choice for at least some of the servers and desktops I maintain. Most of these are still at the RH9 level. As a frugal person, I question what the company really gets much out of a maintenance contract other than expense, most especially so if you read the companion thread about experiences with paid support from RedHat.
Thanks for your insightful comments.
Collins Richey wrote:
I'm interested in every aspect of the CentOS project and its relationship to RedHat and its policies. I'm still hoping to convince my employer that CentOS would be a good choice for at least some of the servers and desktops I maintain. Most of these are still at the RH9 level. As a frugal person, I question what the company really gets much out of a maintenance contract other than expense, most especially so if you read the companion thread about experiences with paid support from RedHat.
Maintenance contracts are most valuable when they are used. Far to often support vendors are only called after an 'insoluble problem' is at a crisis and time & tempers are short. If you have contracts that are unused do two things - first ask what support issues were worked without calling, gently explain that no further support issues can be worked without an open ticket with the support vendor (if necessary enforce within your own trouble tracking system). Second, after the support is regularly used for 6 months review value received for cost - you will find a need to have a difficult conversation with some staff & some vendors. I've seen far to many known-fixes independently rediscovered by in-house staff who wouldn't call a support desk - when I've got support I never want one of my people to reinvent the wheel. It is a waste of their time. (I've also noticed that using support process clarifies thinking - it is way more fun to complain about the support tech who asked if it was powered on than to discover it wasn't when asked)
Personally I can't see buying the low-level Red Hat support, if you can live with 9x5 next day don't pay for support you will never get value. However, 24/7 speak to me NOW! clarifies what is valuable in the environment. No skipping change control & testing, documentation &they had better be properly configured (in HW & SW terms). If not, well see above.
I see a place for RH, CentOS & Fedora in many environments. Put the ERP & Mailserver on RH, when there is a problem & the boss' boss is 'providing leadership' in the data centre being able to say you have a ticket open with RH & Oracle is very wise. Saving money on 2nd tier, infrastructure & most dev systems with a fully compatible OS is also very wise. Then someone will need wizbang app #43 that requires some library just out of the cradle, you'll find that library on FCx about 4 years before RHEL. When you are worrying about the untested, unstable nature of FCx remember wizbang #43 is still in the cradle.
For you boss it may be a question of scale & risk management. Spending money on RH means never having to explain why she thought 'some kid in a basement could support our business' in an outage review.
regards Dave www.HornfordAssociates.com
On Apr 4, 2005 11:50 PM, Dave Hornford OSD@hornfordassociates.com wrote:
Collins Richey wrote:
For you boss it may be a question of scale & risk management. Spending money on RH means never having to explain why she thought 'some kid in a basement could support our business' in an outage review.
Yep, most of the responses I've gotten support the conclusion(s):
1. A support contract is good for CYA (and perhaps not more?). If it feels good, do it. Most ingrained business risk analysis types do not understand community support in the first place. Having a contract in place gives them a (not necessarily justified) warm fuzzy.
2. The kid in a basement may be able to provide more responsive support (especially for environments that the vendor chooses not to support!) than the vendor. In actuality, community support is a world-wide, effective team, and certainly not just a kid in the basement.
3. Multi-tier structuring of the OS choice can certainly make sense. There's probably not a good reason to pay for support in a (possibly) bleeding-edge test environment.