What makes you think CentOS is not willing to be commercially sponsored? (Or only work developing CentOS?)
I would LOVE to be able to do CentOS as my only job.
No one that we know of is willing to pay a full time salary for 1 or 2 or 3 people to develop CentOS. If they would pay for it, we would likely do it.
They might be willing for us to let their current employees do some CentOS things ... but not willing to pay for CentOS development.
Sorry, that was just my impression from previous posts. I guess I have that wrong. Maybe I am confusing the reluctance to take donations at the moment with commercial sponsorship. Thanks for correcting me.
Couple of questions, then....
What is the average current time commitment per week, i.e. man hours that is currently volunteered by the core developers?
What would that need to increase to, to significantly reduce release times (which I think was the overall goal)?
What would the *market rate* be for the skills required? Just to give a rough figure to work with and shouldn't be related to any particular person's current day job.
Thanks in advance,
Ian.
At Sun, 27 Mar 2011 13:36:02 +0100 (BST) CentOS mailing list centos@centos.org wrote:
What makes you think CentOS is not willing to be commercially sponsored? (Or only work developing CentOS?)
I would LOVE to be able to do CentOS as my only job.
No one that we know of is willing to pay a full time salary for 1 or 2 or 3 people to develop CentOS. If they would pay for it, we would likely do it.
They might be willing for us to let their current employees do some CentOS things ... but not willing to pay for CentOS development.
Sorry, that was just my impression from previous posts. I guess I have that wrong. Maybe I am confusing the reluctance to take donations at the moment with commercial sponsorship. Thanks for correcting me.
Couple of questions, then....
What is the average current time commitment per week, i.e. man hours that is currently volunteered by the core developers?
What would that need to increase to, to significantly reduce release times (which I think was the overall goal)?
What would the *market rate* be for the skills required? Just to give a rough figure to work with and shouldn't be related to any particular person's current day job.
Thanks in advance,
Ian.
I expect that from a *corporate* POV the CentOS 'team' would work for maybe a man-day (a few update RPMs) to a man-week or three (point release, major update, etc.), and the rest of the time have little to do *with respect to CentOS* (not worth being on a full time payroll). Unlike Red Hat's staff who are working on fixing bugs, writing and testing back ports, etc. between updates and releases. And fielding support calls from paying customers, etc. And I expect Oracale and Novell have a similar work flow, except that they are piggybacking on Red Hat *for free*.
I belive SL is maintained by a *research* organization, where the maintainers are like researchers or support staff, who are paid to do research or to administer research machines most of the time and then work the few hours (minor updates) or days/weeks (point release / major update, etc.) and the research organization gives them 'leave' to concentrate on the SL updates on an as needed basis (eg the SL maintanence is a *part* of their job description, but not all of it).
The CentOS developers have full time 'day jobs' and can't work on CentOS while at their day jobs. It *might* make sense if the 'day jobs' the CentOS developers work for *also* were corporate sponsors of CentOS, but I suspect that is not going to happen for all sorts of reasons.
CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
On 03/27/2011 07:36 AM, Ian Murray wrote:
What makes you think CentOS is not willing to be commercially sponsored? (Or only work developing CentOS?)
I would LOVE to be able to do CentOS as my only job.
No one that we know of is willing to pay a full time salary for 1 or 2 or 3 people to develop CentOS. If they would pay for it, we would likely do it.
They might be willing for us to let their current employees do some CentOS things ... but not willing to pay for CentOS development.
Sorry, that was just my impression from previous posts. I guess I have that wrong. Maybe I am confusing the reluctance to take donations at the moment with commercial sponsorship. Thanks for correcting me.
Couple of questions, then....
What is the average current time commitment per week, i.e. man hours that is currently volunteered by the core developers?
What would that need to increase to, to significantly reduce release times (which I think was the overall goal)?
What would the *market rate* be for the skills required? Just to give a rough figure to work with and shouldn't be related to any particular person's current day job.
What the CentOS project would be interested in (from a corporate provider) would be to hire people and allow them to do CentOS related things.
We are not interested in being paid in addition to our current work, but making taking care of CentOS our only work.
There are many things other than building packages that have to be maintained for making CentOS go. These include:
1. We have dozens (more than 100) servers that need to be maintained in tens of countries all world. These machines need to be updated and managed, including monitoring and taking corrective action for any services that go down.
2. We have to maintain lists of update mirrors, rsync mirrors, DVD mirrors and verify that the "Dynamic DNS list" for all these machines stay in sync when mirrors drop out or can be added back.
3. Manage the CentOS DNS services, the CentOS mail services, the mailing lists, the IRC Channels, and the main website.
4. We have to write/configure/change software to ensure our mirrors are up-to-date and control the release of updates. Our update system gives out GEO-IP relevant targets for download of ISOs and updates.
5. We have to research/answer bugs and maintain the bugs.centos.org website.
There are many things that we could do if CentOS was our only responsibility.
The cost to a corporate entity would be to hire one or more developers full time and designate them to working only on the project. If someone were willing to do that, we would be willing to listen.
Dne 27.3.2011 17:33, Johnny Hughes napsal(a):
What the CentOS project would be interested in (from a corporate provider) would be to hire people and allow them to do CentOS related things.
We are not interested in being paid in addition to our current work, but making taking care of CentOS our only work.
Well, Financial donations to project are suppressed by CentOS for a few years now. DH
on 3/27/2011 5:36 AM Ian Murray spake the following:
What makes you think CentOS is not willing to be commercially sponsored? (Or only work developing CentOS?)
I would LOVE to be able to do CentOS as my only job.
No one that we know of is willing to pay a full time salary for 1 or 2 or 3 people to develop CentOS. If they would pay for it, we would likely do it.
They might be willing for us to let their current employees do some CentOS things ... but not willing to pay for CentOS development.
Sorry, that was just my impression from previous posts. I guess I have that wrong. Maybe I am confusing the reluctance to take donations at the moment with commercial sponsorship. Thanks for correcting me.
Couple of questions, then....
What is the average current time commitment per week, i.e. man hours that is currently volunteered by the core developers?
What would that need to increase to, to significantly reduce release times (which I think was the overall goal)?
What would the *market rate* be for the skills required? Just to give a rough figure to work with and shouldn't be related to any particular person's current day job.
Thanks in advance,
Ian.
A good linux sysadmin in the US makes from 60K to 80K USD a year... High level programmers a bit more... So with benefits, and other support costs... How about a half million to three quarters of a million a year to commercialize CentOS... In US dollars... Get your checkbook out...
Anyone know someone who can front at least 2 years working capital to get started and productive?
Anyone know someone who can front at least 2 years working capital to get started and productive?
From a pure business standpoint, it would be near impossible to pull off. No one is going to pony up $2,000,000 to start CentOS up as a for-profit company. Aside from the small point that you would be competing with RedHat using its own product, think about it in simple financial terms...let's say we charged each licensee $200/year /machine (and yes at that price we would be competing with RH directly). We would have to sell 5,000 licenses a year just to break even. Bring the price point down to take into account the fact that most users of CentOS can't afford $200/month (or they would probably be using RHEL now..), and your number of conversions goes up exponentially to make up for the low sticker price. Aside from that inconvenient truth, also consider that CentOS does not produce unique products that are protected in any way/shape/form, there is no way to protect the investment or the business from disappearing overnight.
On Mon, Mar 28, 2011 at 7:47 PM, David Brian Chait dchait@invenda.com wrote:
Anyone know someone who can front at least 2 years working capital to get started and productive?
From a pure business standpoint, it would be near impossible to pull off. No one is going to pony up $2,000,000 to start CentOS up as a for-profit company. Aside from the small point that you would be competing with RedHat using its own product, think about it in simple financial terms...let's say we charged each licensee $200/year /machine (and yes at that price we would be competing with RH directly). We would have to sell 5,000 licenses a year just to break even. Bring the price point down to take into account the fact that most users of CentOS can't afford $200/month (or they would probably be using RHEL now..), and your number of conversions goes up exponentially to make up for the low sticker price. Aside from that inconvenient truth, also consider that CentOS does not produce unique products that are protected in any way/shape/form, there is no way to protect the investment or the business from disappearing overnight.
You're just not getting it. The economics of most OSS projects have absolutely nothing to do with "fronting capital", forming a company to sell licenses, or scraping together enough donations to hire someone to quit their day job to work on the project. This has happened maybe once or twice in history. I'm talking about existing OSS projects, not something that was always intended to use the freemium model.
When one says "corporate sponsorship", they are talking about a company with employees able to devote some of their paid time to working on the project. Almost always this paid development also benefits the company, but they also release the work to the project. This is the exact structure that companies like Redhat, IBM, Oracle, Google, Novell, etc... use for their "corporate sponsorship" of Linux. The .info registrar supports PostgreSQL this way.
Discussion of any other type of structure, especially when related to CentOS, is just absurd. Anyone looking to pay someone is going to buy RHEL. Ideally, the big companies using CentOS should be devoting some employee time to CentOS builds, QA, etc... This is the only viable option for a project like CentOS, and is exactly the type of structure Johnny was talking about in an earlier post.
-
Brian Mathis -
David Brian Chait -
David Hrbáč -
Ian Murray -
Johnny Hughes -
Robert Heller -
Scott Silva