CentOS Project Joins Forces With Red Hat

Home » CentOS-Announce » CentOS Project Joins Forces With Red Hat
CentOS-Announce 111 Comments

With great excitement I’d like to announce that we are joining the Red Hat family. The CentOS Project ( http://www.CentOS.org ) is joining forces with Red Hat. Working as part of the Open Source and Standards team ( http://community.redhat.com/ ) to foster rapid innovation beyond the platform into the next generation of emerging technologies. Working alongside the Fedora and RHEL ecosystems, we hope to further expand on the community offerings by providing a platform that is easily consumed, by other projects to promote their code while we maintain the established base.

We are also launching the new CentOS.org website (
http://www.CentOS.org ).

– ———–

111 thoughts on - CentOS Project Joins Forces With Red Hat

  • Fantastic news!

    CentOS and RHEL have been mutually beneficial projects for years. As a user of both, I am extremely happy to see the ties grow between the communities.

    digimer

  • effort.

    Does this have any impact on security metadata so that the yum security plugin works without using things like the CEFS script for spacewalk?

  • Dne 7.1.2014 22:09, Karanbir Singh napsal(a):

    Something I have dreamt of. Good news. Congratulations.

    Regards, David Hrbáč

  • That is amazing news, I hope this proves to be a great relationship.

    Congratulations, looking forward to the future.

    Aly

    Sent from my BlackBerry device on the Rogers Wireless Network

    —–Original Message—

  • Essentially Red Hat is slowly taking over and developing/assisting CentOS to be a more regular and structure organisation. The fact that Red Hat now owns the CentOS brand worries me but that’s life. Absolutely nothing remains static.

    CentOS was created by many ordinary people to whom I am very grateful. Its a really great operating system. I genuinely like it, hence I
    abandoned M$ completely about 5 years ago and have never regretted it. I
    just wish I had migrated to CentOS many years earlier.

    Red Hat will gain commercially from their de facto take-over. One of the general beneficial effects will be bringing Linux into the vast mainstream of everyday computing and attracting (or should that be enticing?) M$ business users. Working as a single team with Red Hat will inevitably mean speedier updates for CentOS users.

    I love Linux and want this merger to succeed. I’m patiently waiting for CentOS on my Cube (Android) tablet.

    Happy New Year.

  • But there is more to Red Hat’s de facto “take-over” including the imposition of USA’s domestic law on citizens all around the world.

    The compulsory imposition of USA law on all CentOS downloaders creates the possibility of being arrested in one’s home country and sent to the USA for a criminal trial. A few people in Britain have been extradited to the USA for criminal trials for matters which are not criminal in Britain.

    Can anyone remember seeing this on the old CentOS …. ?

    http://www.CentOS.org/legal/

    Export Regulations

    By downloading CentOS software, you acknowledge that you understand all of the following: CentOS software and technical information may be subject to the U.S. Export Administration Regulations (the “EAR”) and other U.S. and foreign laws and may not be exported, re-exported or transferred (a) to any country listed in Country Group E:1 in Supplement No. 1 to part 740 of the EAR (currently, Cuba, Iran, North Korea, Sudan
    & Syria); (b) to any prohibited destination or to any end user who has been prohibited from participating in U.S. export transactions by any federal agency of the U.S. government; or (c) for use in connection with the design, development or production of nuclear, chemical or biological weapons, or rocket systems, space launch vehicles, or sounding rockets, or unmanned air vehicle systems. You may not download CentOS software or technical information if you are located in one of these countries or otherwise subject to these restrictions. You may not provide CentOS
    software or technical information to individuals or entities located in one of these countries or otherwise subject to these restrictions. You are also responsible for compliance with foreign law requirements applicable to the import, export and use of CentOS software and technical information.

    This is a Community mantained site. Red Hat, Inc is not responsible for its content.

    ————————————–


    Paul. England, EU.

    Our systems are exclusively Linux. No Micro$oft Windoze here.

  • […]

    Whether this was there, before, is irrelevant. If the software was subject to EAR then it was subject to it regardless of a web page stating it.

  • [EAR = USA’s Export Administration Regulations]

    How would a mere downloader from a mirror, or a purchaser of a CentOS
    disk or even a beneficiary of a free CentOS disk at a CentOS event beware of USA law restrictions and understand the full legal implications of USA law ?

    Its reminiscent of the PGP farce from nearly 20? years ago.

    With Google slowly removing, or not updating, the open source bits of Android and replacing them by closed sources, will the same commercial strategy emerge from Red Hat into CentOS ?

  • RH has a long history of being a benevolent supporter of 3rd party projects under their umbrella. Look at Fedora, Gluster, KVM, etc.

    What RH did was guarantee the long term health and sustainability of the CentOS community. They’ve provided that same community access to tremendous resources, both technical and human, to grow and maintain the project.

    Red Hat’s benefit is that they help grow the community of EL users. A
    very many paid RHEL users got their start with CentOS. They’ve given CentOS instant corporate credibility that will help grow that
    “incubator” user base even further, increasing the pool of users who might one day grow into needing commercial support. They’re building the foundation for their future customer base.

    I am very confident that this will prove to be a very good thing for both CentOS and RH.

  • You’re missing the point.

    This is not RedHat causing “[t]he compulsory imposition of USA law on all CentOS downloaders” (your words); that imposition _already existed_
    regardless of a web page telling you. The difference, now is that you’re told about it (presumably standard RedHat legal boiler template
    ‘cos RH lawyers believe it adds some protection to _them_ – and thus the CentOS board – by having it there).

    The legal situation for downloaders _has not changed_ by the presence of that section on the web site (and the page has even less importance considering you can download the DVDs without even having to see that page; it’s not an agreement you sign or click through).

    It’s the same farce.

  • What ARM ver? Is there Fedora for it? This takes lots of time, but f20
    works on a lot of ARMv7 units. Now what ver of RH will that map into? :)

  • of course, tablets also require touch pad support, and a touch-oriented UI… trying to use a mouse interface with a touch pad is an exercise in frustration

  • No. That was ITAR, and no farce. ITAR is very proscriptive, and only the loophole on printed algorithms allowed PGPv3 to be shipped out legally. It took us some time, but we finally weakened ITAR. I
    remember well, as I was running the IPsec international interoperablity work back then and had a major hand in showing the non-enforceablity of ITAR wrt cryptography as munitions.

  • My first thought as well. Redhat already has Fedora as a testing ground. So for Redhat acquiring another free distribution makes me wary, unnecessarily so maybe…

    I hope CentOS will continue to be The “free” stable enterprise solution.

  • I wouldn’t worry so much. RedHat has every incentive to keep CentOS
    very much alive and well. Consider that if RedHat were to try to kill CentOS then they would have Oracle chomping at the bit to take over with Oracle Linux and that is a scenario that RedHat almost certainly does not want.

    Peter

  • Think about this. RDO, GlusterFS, oVirt, and OpenShift Origin are all Red Hat community offerings that need to have a long lived community base OS to speed their usage and growth.

    All of those also have a paid equivalent (Open Stack Platform, Storage, RHEV, and Open Shift) where Red Hat gets paying customers if the community projects thrive. It is absolutely in Red Hat’s best interest for all of the community software listed above to do well.

    Red Hat wants their paid platforms to continue to be successful, they therefore want their community projects to be successful.

    CentOS and Red Hat are joining forces to make those (and other)
    community projects more successful. It is a simple as that and it is in both the CentOS Project’s and Red Hat’s best interest for both of us to thrive and grow.

    Fedora, a Linux distribution to deliver “state of the art” features, is also always going to be “Red Hat Enterprise Linux … Next”. Fedora is also a great Linux distribution in its own right. It is obviously still very much in Red Hat’s best interest for Fedora to continue to grow.

    Is Red Hat in business to make money … of course they are. Does Red Hat make more money or less money if their community projects do well?
    Of course they make more money if more people use their community projects. Red Hat wants CentOS, Fedora, RDO, GlusterFS, oVirt, OpenShift Origin, and every other project where they provide support to do thrive and grow.

    Is it in the CentOS Project’s best interest for RHEL and Fedora to continue to grow … of course it is.

    Karanbir Singh (the Chair of the CentOS Project board) and Robyn Bergeron (the Fedora Project Leader) have both posted blog entries that discuss these items in further detail:

    http://www.karan.org/blog/2014/01/07/as-a-community-for-the-community/

    http://wordshack.wordpress.com/2014/01/07/CentOS-welcome/

    This is not rocket science folks. We all want all of these open source projects to do well.

    I am very excited about this arrangement and I think we all win.

    Thanks, Johnny Hughes

  • Thanks for linking to the blog posts, it made me appreciate why the move was made, incl some of the “softer” values. I think I understand it clearer now.

  • Hello Alain,

    The more stability to have from Open Source, the better product and happy customers you have for RHEL.

    So Red Hat has huge support for Open Source, including CentOS. But you are right, Red Hat is also stright on having paying customers stay with RHEL and they do not give away their base RHEL product for free.

    This is for sure, Red Hat has taken over. It is not a cooperation on infrastructure or similar, but kind of the “community of CentOS” to move into Red Hat proper…

    best regards,

    Florian La Roche

  • They do that right now.

    – CentOS Plus-Kernel
    – CentOS Extras
    – Xen4CentOS

    Are all NOT upstream but CentOS original projects. Read the part about SIG
    in the FAQ and you know what this will eventually become.

    —–Urspr

  • Is there Xen Support in Fedora?

    You’re completely missing the point here. Redhat ist not going to influence the direction CentOS is going, but merely saying it will advocate
    ‘variations’ of CentOS, i.e. additional package sets much like Xen4CentOS or the much abused OpenStack (at least in that FAQ) ;-)

    —–Urspr

  • I have talked a bit with Redhat people at the ETSI NFV meetings. We will see how NFV is met and what part OpenStack plays. Of course the customers just want the reality of NFV with the economic drivers that are pushing us in that direction. If OpenStack is part of the solution, and is secure (my role in the process), then it will get deployed.

  • Sorin Srbu wrote:

    wary,

    One thing that struck me in Karanbir’s message was the marketing mumbo-jumbo such as “the next generation of emerging technologies” and
    “a platform that is easily consumed”.

    Karanbir usually writes better than that, so I suppose that someone else had a important role in drafting the message.

    Yves Bellefeuille

  • I hope the ‘CentOS minimal’ iso continues to be supported. That has become my favorite install approach, especially where someone in a remote office has to bring a box up to where I can SSH to it and add our applications. I do wish it included openssh-clients and rsync to make the next steps easier, though…

  • That won’t go away. This is purely an additive process. Nothing changes for what we already do. If you want to use the new stuff we do or a variant that someone else comes up with, great. If not, the base/updates will always be there, minimal included.

  • I’ve always thought Red Hat was at its best before they started restricting access to the finished product, even to the community that contributed most of the code and bug reports that made it possible and usable. That is, when they just sold support and the released code was the same for everyone, including the binaries. Without that, I
    don’t think they would exist today. While I greatly appreciate the CentOS project and the way it has continued this access in a practical sense, I still don’t understand why Red Hat thinks it is a good idea to dilute their brand name or make it less visible and well known.
    (Well, I can understand it with Fedora as the never-finished work in progress, but not for the equivalent of the base CentOS as an exact clone.).

  • True, but Fedora is a bleeding-edge Linux, while CentOS is a stable Linux. Both have their place.

    Red Hat knows there are pieces of the Linux market it will never be able to grab significant share in. Low-end web hosting, for example.

    Think about it: is it better for Red Hat to spike CentOS’ wheels on the hope that people will go running to RHEL, or is it better for Red Hat to make sure the CentOS project runs smoothly, so that it can keep some kind of fingerhold on these sections of the market?

    I know people like conspiracy theories, but do you think the pain of getting CentOS 6 out the door did Red Hat any real good? No. All that did was make Red Hat look bad.

    I think that’s the real reason Red Hat is doing this: they want to make sure CentOS 7 launches smoothly, and are helping out the best way they can.

    Another good reason for Red Hat to do this is that they now have a serious answer to Ubuntu Server and Debian. Before, they were saying,
    “Well, if you want no-cost Red Hattish Linux, you can go to *those*
    people over *there*.” Now they can point to an official Red Hat sponsored offering. When/if those people want commercial support and such, they can use CentOS as an on-ramp to RHEL.

    This is a good thing.

  • Warren Young wrote:

    1++ (and boy, do I *hate* bleeding edge)

    Yup. And since RH’s big thing, like many old computer companies, is service… and with this, the shops that are CentOS only will be more likely to buy a RHEL license or two, to get guaranteed response to issues… which makes it a lot more palatable to upper management, who often only knows WinDoze.

    If that’s it, I agree. We’ll just have to see how it all plays out.

    mark

  • Some years back I REALLY tried living with CentOS on my notebook. I now put up with the Fedora eol joys. Just jumped from 17 to 20. You just got to love, ahem, what the Gnome team is doing….

  • I can think of 101 different reasons as to why this move is good for RedHat, and no reason why them picking up CentOS only to kill it would be good for them. If RedHat were to kill CentOS it would just leave the door open to other clones to step in and fill the gap, other ones such as SL or puias which are not as strict in cloning RHEL, so could give RedHat a bad name if the clone has problems, or worse yet, Oracle.

    The worst thing I can see happening here is that RedHat may decide at some point in the future that the relationship isn’t working out and will simply release CentOS back as a fully community-driven project, they’ve done this before on other projects and it’s not a malicious move that kills the project it’s just going separate ways. If this were to happen then CentOS would be no worse off than it was before the move to RedHat in the first place.

    Peter

  • Robert Moskowitz wrote:

    No. (And if you *had* to do fedora, why not 19?)

    A thought: have you considered trying to install dual boot with the current CentOS? I’ve been considering redoing my netbook, with the thought of getting rid of the Ubuntu netbook remix….

    mark

  • Because it will be eol before you blink. So even if f20 is brand new, you at least live with it for a while before the next update. Plus i got to find a bug with NVRAM update with my Lenovo x120e which is now an RFE
    for f21. If I had been on f19, might not have caught this, or had a bigger struggle to get them to realize that the NVRAM update should be the LAST step of the installation, not so early, so if it fails you still have a bootable install. I got some good help to get things working.

    Well since I am now running on an SSD drive, I have my old HD to test with. Maybe. Or maybe I will wait until CentOS 7 comes out. ;)

  • This is great news!! Congrats and thanks to everyone that has put in all the hard work to make this a great environment.

    Michael Weiner

  • hi,

    i am sorry about that – some of those things came through in my attempt to reduce the post from just over 2400 words to under 1000. I didnt make it, it was still over a 1k words. But i do mean that, lots of really cool stuff was bootstrapped on CentOS, but not here – and in many cases, so far away from the project that people had to do it again and again.

    the voip setup for home/small users is the best example. if asterisk@home were done as a part of the CentOS community, how cool would that have been?

    the next generation of cool stuff is all also happening out there, and I
    really do want to bring as much of that into the CentOS community as possible – after all, we are a user and problem lead community, not a developer led one where someone is just churning out new code to see what works and what does not.

    its been hard to do in the past, mostly down to the constraints – Red Hat TM issues, protect the buildservice, handle 100+ sponsors, work on community issues, get that update out in the 45 min between dinner time and kid’s bedtime etc.

    And remember, a critical artifact of this group : we are a user led community, not a developer led one. Massive win, in my opinion. But the lack of developer density has been a problem. And I dont know how much of that we will get access to, but at >0 we are already winning. right ?

    Now ‘easily consumed’… becuase we can start opening up the buildsystem, publish all the scripts we write, post instance and image specs – and anyone/everyone is welcome to join the effort since the needs of privacy and secrecy are dramatically reduced ( i assure you, this is one of the top wins in my books ).

    Come join me in an officehours meetup (
    http://wiki.CentOS.org/OfficeHours ) – lets talk about these things :)

    – KB

  • I’d throw SMEserver, ClearOS, and the old (up to CentOS5) version of K12LTSP in that bucket too. Maybe someone will roll a new K12LTSP
    that comes up working as installed again now.

    Is this likely to result in Scientific Linux converging with the base version?

  • The focus from the project side is going to be creating the infra and resources that allows arbitary stuff like that to come in and be successful at doing what they are doing on CentOS.

    But keep in mind that opportunities to get involved will come up on both sides – ie. help the CentOS project do the buildout as well as on the SIG’s side to do the work that people like k12ltsp folks need to consume those resources.

    That would be nice. And they are certainly welcome.

  • I really dont know how oracle’s linux rebuild effort works – but as far as I -do- know, the sources are available at the same time to everyone right ? its a case of what you do with them and how you do it.

    Also, i think people are reading too far into the delays for C6 were caused by redhat – it was also down to limited resources, machines, time and almost no QA infra at .CentOS.org :: that contributed quite a lot. Things that we have overcome and built up in the last few years.

    Might also be worth noting that we have a CentOS7beta up internally already.

    Look at it another way – we are not working with the RHEL teams, we are working with the RH open source and standards team ( that has no real input into RHEL ) – to expand what we do with the platform, rather than carry on with the single focus of the platform. And I think being more open and more community driven, we -can- improve across the board.

  • At this point I really don’t see why RedHat doesn’t just offer RHEL + updates + extra channels for free and then only charge for support. This would put them on a real equal ground with Canonical. It would save money and time freeing up all the duplicated effort of ripping out all the redhat logos and rebuilding the core OS and then rebuilding all the updates. The core CentOS team and volunteers working on CentOS would be freed up and could focus their effort on extending third party open source projects mentioned earlier in the thread to work better with RHEL. Just my 2 cents. In any case, this is interesting news. Ever since the CentOS team got everything going smoothly for CentOS 6.x version, my biggest concern was Redhat

  • I suspect doing so would cut heavily into their revenue stream, as many business IT operations types who are told they have to run RHEL because
    ___ requires it would just install it and never pay for support. by keeping the free version separately branded, however slight the actual difference, discourages this except by those in the know.

  • Probably has something to do with being able to require paid support for _all_ instances of RHEL you are running to get any. It then takes at least a little effort on the user’s end to install CentOS on the less critical hosts instead of just cloning everything and paying for support on one copy.

  • I see 2 problems with this. First is that it would create a conflict between free RHEL and CentOS, and would be seen as attempt in destroying CentOS.

    Second is “supported on RHEL” for many products. Today, when they want security/peace of mind/business insurance, you buy both RHEL and app to have support for app. With free RHEL they could only buy app and have that support without paying for RHEL support.

    It is same with SuSE and OpenSuSE, right?

  • David Miller wrote:
    >

    That’s an *easy* one to answer: try selling “we can use it for free, we just download it from the net and install it….

    Right. You want to see 66.6% of CTOs, much less 95% of CEOs, go with that as a business plan? They almost comprehensively want Someone To Get On The Phone (and I do *not* mean someone in India, with a heavy accent, asking if they’re rebooted their computer) to resolve this within an SLA.

    Tell them you can try it out, and if they like the results, they can pay for a license and support for RHEL, the “real” thing, and that’s a *lot*
    easier sell.

    mark

  • Especially if there’s a migration script to convert existing CentOS images to point to RHEL repos and refresh packages :-)

  • Fedora 20 allows for installation of the latest hadoop with yum. I don’t know if this is in CentOS 7 but i can’t see it in the SotB.

    I am gearing up for a project based on using a hadoop cluster (along with HiveQL) and my current plan was going to use fedora 20 for the nodes with CentOS used for everything else. I wasn’t overly happy about the prospect of the inevitable churn but this good news has made me reconsider.

    I am willing to bet my time that the trade off between helping maintain hadoop/hbase/hive/avro/mahout/zookeeper as part of a CentOS SIG in the “new era” against the reduced change control required to avoid fedora breakage will be worth it.

    I’ll drop by #CentOS-devel

    regards

    mike

  • Apparently nto all is well with the take-over. Here is an example. Should I stop mirroring CentOS in the fear of being arrested next time a I visit the US on vacation?


    Hi,

    We are mirroring CentOS in Iran.

    http://CentOS.iranmirror.ir

    IP: 94.182.146.125

    Protocols: http

    Location: Asia / Iran / Tehran

    Bandwidth: 1 Gbps

    Version: All

    Architecture: All

    Direct DVD Download: Yes

    Organisation: http://iransamaneh.com (Web application development and web hosting)

    Email: admin@iranmirror.ir

    Thanks

    XX XXXX

  • I dont understand your question or statement, what are you saying here ?
    Can you say the same thing, but a bit in a more verbose manner ?

  • Am 17.01.2014 um 01:18 schrieb Karanbir Singh :

    I think he refers to:

    “You may not provide CentOS software or technical information to individuals or entities located in one of these countries or otherwise subject to these restrictions.

  • These restrictions were always inherited. Theoretically if you use cryptographic software developed in USA you are “bound” to these rules. In many cases if you use for example OpenSSL in Windows, Ubuntu, Android etc etc you are still affected (I think), it’s just that now it’s written somewhere. In practice this is not very relevant and also pretty unenforceable;
    not to mention that – to my understanding – it contradicts the GPL. RH needs to specify this legal bit so uncle Sam is happy. Just do whatever everyone else does, ignore it.

  • There are no *new* export restrictions. You’re just now aware of them. It’s the US gubmint that puts those restrictions, not RedHat, and they’ve always applied to CentOS.

  • I say “new” because the original email referred to what I believe was about an existing CentOS mirror in Iran. This prompted me to look at the CentOS website and I found the export restrictions to which the email was referring. Then I looked at the Fedora project website and found the same restrictions. I don’t know much about open source export restrictions. I know they must exist for proprietary software developed in the United States. I was merely pointing out what the the email stating and what it was referring to. Please retract the word new.

    Joe

  • That’s the point though. If “you” (for generic values of “you”) export code under US legal restriction from the US then you’re in breach of US regulations. Whether you know about it or not.

    Fun, huh?

    If “you” run a mirror then you get to determine your legal risk and whether you should keep the mirror. The CentOS team are not lawyers;
    they can’t tell you.

    It’s a fun legal question as to who does the export; the person making available for export on a web site or the person downloading from that website. As far as I know it’s not really settled. In my opinion the RedHat wording is a prayer hoping that’ll cover them :-)
    But I’m not a lawyer, either!

    If you’re really concerned then consult a lawyer.

    (This actually applies to _any_ downloader, not just people who mirror).

  • I have found at times the community to CentOS-leadership relations to be quite poor.

    I have witnessed the summary judgements against people like Dag Wieers driving them away. Useful members driven out.

    I have seen release dates slip for months at a time with no word from the people in control of CentOS. The project has come into jeopardy many times. CentOS 5.4 was a fiasco.

    I have always suspected that after each release the exact build environment
    / script to create the RPMs is not made available to bring about this end –
    whereby the “secret sauce” of how to build the SRPMs on http://ftp.redhat.com are still kept hidden by the CentOS leaders. This is not a community project. Its a free rebuild with all the mock magic hidden by those who just got a huge payout.

    I am also wondering if the serially rude and dismissive behavior by some of the folks in control of CentOS will continue now that they cash massive checks from Redhat. I guess when you sell out one needs to be more polite.

    Now we need to possibly find a new rebuild. I think that release dates will still be something that the leaders here do whatever and whenever they want. I think that there will be significant differences in RHEL and CentOS now. I
    think the secret build sauce will remain hidden from view and the people receiving big pay for Redhat will serve their new masters well.

    I’ve been a user since the WBEL/cAos days. I worry about this state of affairs. Deeply.

  • ITAR is a 1947 treaty the binds all signatures to treat cryptographic
    ‘artifacts’ as munitions and abide by the export restrictions that exist for all munitions. Period. Full stop.

    This includes Crackerjacks ™ encoder rings that I played with as a kid! Really! Someone in the US State department figured this out.

    The only exception in the treaty is cryptographic academic papers (how we got pgpv3 exported, in book form); but even this got challenged because of the pgp export.

    And like all treaty provisions regarding munitions export, they are open to interpretaton and enforcement. I leave the rest of the logic, or lack thereof to you.

    (I lived this very closely back in the late ’90s. I could, and have, tell you stories of the conversations back then)

  • At one point a major unix manufacturer tried to get around this by having the crypto code written in another country by citizens of that country. They got shut down as re-exporting. In the end, they had to ship broken software that required customers to optain the critical code from this other country. This was part of our action to show how unenforceable ITAR was wrt cryptography as munitions. Some likened it to shipping guns without firing pins or ammo; which were readily available from other sources.

    But at any point, someone in State can decide someone’s actions violate the law and go after them. Ask Phil Zimmerman…

  • I see you haven’t read announcements and explanations, or you haven’t understood them.

    You ARE aware that RepoForge is forzen solid because Dag Wieers does not want to release control to others but has no time to build packages ready for build? I wonder how is that different of what you accuse CentOS devs did.

    FYI, I am not on either side, I do not accuse anyone, but I think every comment should be balanced.

    I can understand that someone is not willing to explain “secret sauce”
    they spent 100’s of hours poured into to make it work in their free time, just so others can jump in and create a competitor to their
    “product” thus invalidating their work with lesser gratification. I am first who would not do it. Not without monetary reward. Weather I
    personally liked it or not is irrelevant.

    Red Hat wants RHEV and their other products to have rebuilt versions. They need it so their products get bigger user base. It would be stupid to create entire community from scratch when CentOS only needs little help to open up and producing Variants, and then compete with CentOS.

    So Red Hat will get opensource rebuilds for RHEV and other products and CentOS gets second wind and opens up entire process.

    Every one of “you”, unhappy ones, could have created your own rebuild, you could have also teamed up and found sponsors from all those unhappy community members you say exist. So, where is the product of your open collaboration?

  • I lack knowledge of how the community inspired CentOS project started. I
    remember squabbles over the domain name which was satisfactorily resolved.

    Not many people have the time and mental ability (both are needed) to acquire the knowledge to create a rebuilding of RHEL. Using CentOS
    requires less intellectual effort than literally starting from the absolute beginning with RHEL sources.

    Thinking positively about CentOS, we share as users/installers/administrators and problem solvers a really great and very practical alternative to the world of M$.

    CentOS is used for millions, if not trillions, of operating systems. Many use it but very few contribute technical assistance or money to the continuing CentOS project. Without CentOS what would we do ? SL or the Debian family or the BSDs or Solaris ?

    Despite negative, unhappy and wrong things that have occurred, the CentOS project has continued to our personal advantage. It would be nice if the unhappy things of the past could be amicably resolved and we all become one big, happy and very satisfied world-wide family.

    Lots of people have contributed directly in CentOS or as package re-builders for CentOS suitable repositories. To all those people, I
    would like to say “Thank You”.

    Great to see you are still in love – she must be very special :-)

  • “Every one of “you”, unhappy ones, could have created your own rebuild, “

    A lot of Redhat rebuild projects gave up their very existence to support a single CentOS.

    Not giving up the secret sauce is about control and power in the hand of a few that have now financially benefited and retain a dictatorship on roadmaps, release information and code.

    Community here is a consumer of a built OS, but there is no community in how it gets built. And with this centralized power comes the takeover and payouts.

    If Redhat wasnt trying to block OEL or SL or trying to control CentOS and make it different, they would simply offer RHEL for free on their own. This allows them to wean the world off of CentOS at what is likely to be a glacial pace at first then by Redhat we will have all given up.

  • “You ARE aware that RepoForge is forzen solid because Dag Wieers does not want to release control to others but has no time to build packages ready for build”

    There is quite a bit of open-source surrounding rpmforge and rpmforge doesn’t have the work “Community” it its very name.

  • I view this as a takeover. I view this as a few who kept how to rebuild RHEL
    a state secret benefiting financially. I don’t see how a community benefits when we cannot recreate for ourselves what is being done here. I don’t see how we benefit when a large company comes in and buys their way into the board and pays off all members. Where is the Community’s say in this? This is a payoff. Will we get releases sooner? Will we know how to rebuild the build environment for ourselves? What if Redhat slowly makes using CentOS
    painful to incentivize using RHEL? If Redhat had good intentions why don’t they give unsupported RHEL for free themselves. Granted the probably want to keep OEL and the like from being able to freely rebuild and plagiarize and charge money for their stuff, but we , the Community, the masses of users, are stuck now between behemoths and their lackeys taking payouts throwing us whatever table scraps they want and we are powerless to change this.

    There is no makeworld or emerge world here, just binaries that magically get produced and peppered on an ftp whenever someone gets around to it.

  • “Essentially Red Hat is slowly taking over and developing/assisting CentOS to be a more regular and structure organisation. The fact that Red Hat now owns the CentOS brand worries me but that’s life. Absolutely nothing remains static.”

    Interesting how a _community_ “Brand” can be bought.

    Seems that we get magical binaries for free but no insight into the build process or timelines to said creation.

    Surely this was done to keep OEL at bay, but we are still caught in the crossfire and the holders of the build secrets are getting $paid$ to keep the secret.

    This is opensource without useful makefiles. Something Sony and Cisco do.

  • Actually, “Ljubo” in both my first and last name means closely to someone who loves, kisses someone. Ljubomir means “one who loves/kisses peace (peace = mir). “Ljuba” for example means “one you love”, designates mostly females.

    So if you look and Internet as “cloud” in the “air”, signature means I
    am still around :)

  • I am on repoforge mailing list from 2008, and I know times when no package was built for several months, and guy working with Dag saying why got no responses from him. And when he has responded with “I do not have time”, they where denied any option to build packages without him.

    If you are ignorant of this, then you need to dig into mailing list and learn true status, this being one-man show with helpers. If anything changed, I somehow missed it.

  • IonPacepa wrote:

    Most projects have specially authorized people. This is a Good Thing… unless you really enjoy having a distro larded with malware and bugs that crackers, crooks, other organizations and governments have deliberately, or when IMSOHOT updates code with bugs galore.

    I’d prefer not to have any of that (or I’d be on, say, another distro that shall remain nameless but is also a style of hat….

    mark

  • Path to CentOS core member is simple. You join CentOS Q&A team, and after some time proving you are reliable, you might join them.

    Unless you prove your self, you can not even get job of supervisor to a bunch of clerks in supermarket, right? It is dangerous to allow unproven persons messing with such trusted OS like CentOS.

    So you just skipped everything else I said and just reiterated what you said in first e-mail?

    Ok, what ever, I am done wasting time on you.

  • I really didn’t want to get dragged into this, and this will probably be my only post on the matter. But I feel the need to address some ‘facts’
    that have been laid out.

    Let’s clear a few points up here:

    The benefit we gained is time. We are able to work on this fulltime now instead of after hours following a job doing something else.

    As to not giving up the secret sauce, we publish the changelog and packages we’ve had to modify to deal with TM compliance. It’s in the wiki for every release. The build scripts for isos were for the early releases were on the mirrors and are still published on the vault.

    What we didn’t do was create a support mechanism to fracture the community every time someone got an idea. That seeks only to tear away at the community rather than to build it up.

    Several groups took the distribution we put out and changed it to suit their own needs just fine. ClarkConnect as an example.

    Please stop the FUD here. The centralized power you’re talking about is the origin of the source. It was never ours. We, SL, Puias/SpringDale and the rest all had to go through the same motions.

    Hugely incorrect and outright FUD. The point of this is to *build*
    community. Offering free RHEL would fracture and destroy several communities, as well as damaging likely damaging Red Hat’s reputation in the eyes of everyone inside those communities and anyone outside who wanted to throw stones.

  • One of the beautiful things about open source is the ability to fork, create a new project, etc.

    CentOS was never under any requirement to release their build methods. Whether that was a good or bad choice is not very relevant now.

    If you (and others) feel that the build process needed to create a binary compatible is a worthy goal, you can start a project to do just that.

  • Is that how you describe every other open source project? Ones where the tools to rebuild are easily available? Are they all really that bad?

    I think you are missing a bit of history in that project and its clearos successor. Notably the issues around the delay of a 6.x release. Not to revisit those issues, but still everyone _must_ stay away of the dependency chain and the potential of upstream problems when that dependency is forced.

    I strongly disagree with that. Red Hat’s community and reputation were just fine back in the day when they did not restrict access to binaries. In fact, if it were not for those days, we’d probably all be using debian. Their problem would be in how to enforce the requirement that all copies of RHEL in an organization have to be under paid support to have any if not for the distinction between the rebuilds and their own.

  • John R. Dennison wrote:
    Can’t resist: I think he’s trying to get our goat… and everyone knows what happens when a troll runs into a goat….

    mark “your folks *did* tell you that story, right?”

  • What the heck are you talking about … rpmbuild -ba .src.rpm

    It builds if you install the proper packages from the CentOS repos.

    Using mock and a CentOS Tree can reproduce CentOS just as easily.

    We are creating git.CentOS.org so that everyone can look at and build any of the packages.

    We are creating a variants program so that projects can take CentOS as a base and create (on our servers) respins of the ISOs and/or repositories that get branded as CentOS. They can collaborate, ON OUR SYSTEMS, to build things for the community to use.

    I have no earthly idea what you are talking about … although, you are certainly free to use (or not use) CentOS however you choose.

    I just wish you would research your fact before you post garbage on the list.

  • https://en.wikipedia.org/wiki/List_of_Linux_distributions

    I count 115 Debian/Ubuntu variants. (Could be off by a few, since my eyes started to cross there near the end.) 15 of those are directly under the Ubuntu umbrella; apparently they feel the need to capture at least a handful of these forks, to prevent their “community” from going all to pieces. That leaves a hundred non-official forks.

    I count only 10 RHEL derivatives, plus RHEL itself.

    If you throw in Fedora and its derivatives, then the total goes to 32, which only goes to prove my (and Jim’s) point: the more “open” Fedora branch gets forked more often.

    Anyway, if you want a wide-open Linux, Les, you know where to get it.

  • Sigh…, It’s complicated. I want stability and reliable security updates. But I don’t like being dependent on any single entity to provide that. Maybe that goes back to relying on some AT&T unix systems in what seems like another life. Even though semi-compatible alternatives were available, being forced to change was somewhat painful. So I don’t necessarily want wide-open, just a little more open than being married.

    I don’t really think the CentOS team has an evil plan here, but they should take it as a compliment that I think they are smart enough to fool me if they did want to do something like inject a hidden backdoor with their builds. But, the bigger question is where it leaves us if they just decide to quit after assimilating most of the related systems under a build ecosystem that no one else can reproduce easily.

  • And I view you as unbelievably dense … how about you actually see something tangible actually CHANGE for the worse before you make your proclamations of the end of CentOS. When something happens that actually takes something away that is important, you can then come back and post about it. If a frog had wings it would not bump its ass on the ground when it jumped. That statement is as relevant as your proclamations of doom before anything has changed in any way.

    This has absolutely NOTHING to do with CentOS the base OS or any restrictions to or for anything … it has to do with adding the ability for the community (Like Xen4, like RDO, like GlusterFS, like OpenStack Origin, like OpenNebula, like Ceph, like RackSpace, like , being able to start a community project, on OUR HARDWARE, and build things to use with CentOS by the community.

    If the source code is available, any one can build it … both Red Hat and CentOS already all the source code available. Every package that is changed in CentOS and every srpm (changed or not) is published. All it takes is time to build and compare and build again in the proper order.

    Some things, like a dot zero (ie, 5.0, 6.0, 7.0) release take a lot of time to figure out the proper build order and mechanics .. which is why there is http://seven.CentOS.org/ . We (The CentOS Project)
    specifically went out and got permission to get this site up and discuss, from the beginning of the first beta release of rhel7b1, the ability to build this software. How to get it to build, what is required (rhel7b1, f19, other packages from rawhide, etc.). We are doing it in the public, publishing mock configs and everything else on git.CentOS.org:

    https://git.CentOS.org/summary/sig-core!bld-seven.git

    We will, it the coming weeks, publish our beanstalk client (nazar) and build system (reimzul) on that git site as you can see in:

    https://git.CentOS.org/summary/CentOS-git-common.git

    We could not possibly be more open than this.

    In summary, opinions are like … well, you know the rest. Opinions are a dime a dozen. Actions are relevant. Our actions show we want to continue to provide the best OS in the world to the community AND we want to also bring in many more members to make CentOS better than ever.

    Take a look at the CentOS-devel mailing list at all the groups that want to start a new Special Interest Group:

    http://lists.CentOS.org/pipermail/CentOS-devel/2014-January/thread.html

  • The CentOS-devel mailing list has had more traffic in the last 10 days than it had in the previous 9 months … I’d say that some people have figured it out :)

    It is also a link on the front page to variants/SIGs:

    http://www.CentOS.org/variants/

    We are not ready to actually create these yet, as we need to get more infrastructure and processes in place … but we are discussing how we are going to do it and getting ready now.

  • In summary, opinions are like … well, you know the rest. Opinions are a dime a dozen. Actions are relevant. Our actions show we want to continue to provide the best OS in the world to the community AND we want to also bring in many more members to make CentOS better than ever.

    –Snip

    +1 from me I like the direction and the additional openness, I believe it is headed in a better direction, thanks for all the hard work!

    :)

  • I don’t expect that it would ever be necessary, but it wouldn’t be terribly difficult to reproduce the distro from source packages. It would require a lot of work and a lot of build time, but it’s not really very difficult. The most challenging component would be the initial bootstrap build, we could produce altered trademarks packages in less than an hour. I wouldn’t lose any sleep over it.

  • You want your Linux to be under control, but not controlled. Is that it? :)

    Someone has to have their hand on the tiller.

  • That is reasonable fear, but unless you are going to build everything yourself, you can never be sure in anyone else. And even if you have an accessible build system, there is a question if it was compromised in a way that others can not notice, but producing backdoor.

    So it all comes down to trust vs convenience.

  • Controlled as in having a currently authoritative version, but not secret or restricted beyond not calling something different the same name.

    Yes, but if the boat sinks it would be nice if the blueprints didn’t go down with the ship. (Or even if it goes off in a wildly wrong direction…). Anyway, per Johnny’s comment that it is all going to be published – that’s all anyone could ask.

  • Complicated is a good word…

    There is no secret sauce to the buildsystem. It’s not like we’re converting TRS-80 Model II TRSDOS files to LS-DOS or anything here, where things aren’t well-documented or completely undocumented.

    I have reproduced to an extent the buildsystem for CentOS 5 on IA64, and with a little nudging in the right direction by some folks I was able to figure it out. The hard part, as as been said I don’t know how many times, is getting the x.0’s first tree’s dependency tree and build sequence correct with the buildroot populated with a good starter set of packages (and this is very well documented in the Fedora documentation).

    Anyway, the current work with seven.CentOS.org is a really good start.

    But we really shouldn’t feed the trolls.

  • Hey, I could have done that with my eyes closed and in z80 code. But those were simpler times.

    I hope someone manages an ARM build eventually. It would be fun to play with cheap hardware and reliable code.

  • [TRS-80 in-joke and pointer to OS source code with Les’s name in the comments taken off-list…… suffice to say that Les actually has done
    (or worked on the code, at least for) my referenced conversion in Z80
    assembler for one of the TRS-80 operating systems]

    Is the F19 ARM build workable? (I know there are graphics module issues). I might have to try that myself on my GuruPlug or one of our Pi’s. If F19 on ARM is stable enough, once the build chain is proven for x86_64 it shouldn’t be too difficult for you or someone else to build from the source, taking F19 as the base for the buildsystem and initial buildroots. I would expect the build would take a long time, though.

  • My understanding is that RHEL/CentOS 7 is based on Fedora 19. F19 has been ported to arm, but each arm is different and needs personal attention. So PERHAPS a couple arm boards will be worked on. I will probably be in a position to do this in the spring. I am looking at a cubie2 or truck to build a pbx. Regardless if it is CentOS 7 or Fedora
    20, someone will have to build the FreePBX modules for arm. But if CentOS 7 is running on arm, it MIGHT be easier and more stable over the long run.

  • Unfortunately forget any armv5s and pretty much through v7s. Start with armv8s or armv9s. Are you on the fedora-arm list? Remixes on ‘older’ arm archs are left up to parties that want to do it.

  • Home page is static from back in ’12, but I see mailing list has some activity. I have a pogoplug that has f18 port on it, I will see if I can get redsleeve on it. But won’t be for a couple weeks.

  • Maybe it might be a good idea to do some research on Debian systems?…and using them for file and system servers?……I’m just sayin’ LoL!

    EGO II

  • When there is discernible evidence of a deterioration of service, maybe. But until then it’s all just FUD.

    If anything, the evidence currently points to a vastly improved picture since the delays of a few releases back. Back then there was cause for concern. At present I see far less cause for concern. Of course things can change, but at present I see no reason to be concerned. I’ve never been very good at predicting the future so I will stick to looking at what the present is telling me, and currently the CentOS team are doing a good job on delivering the core product in a timely fashion. That is a metric I can measure today and it tells me something meaningful. IF that changes and things observably deteriorate then there are alternatives but I’d rather make decisions based on what I observe today rather than predictions about what might happen in the future.

  • Well I for one will not be “jumping ship” anytime in the foreseeable future. CentOS (wish they would change the way it appears to the world…the “e” should be capitalized…as the “OS” is….its the start of a real word!….but I digress!) CentOS has been good to me….and has never given me problems since installing it at 6.0’s release. If anything this should solidify the fact that CentOS is TRULY an
    “Enterprise Class” OS available to the masses from a Community that has the (strength?….clout?….resources?) of Red Hat Enterprise Linux…(this might make my taking the RHCSA a bit easier too!…….(wonder if there are any CentOS certification exams?…..or would that be an “over-saturation” of the market?….like…if you’re not RHCSA approved…then you go for “second string” CentOS?……maybe its better to NOT have one then!…)

    EGO II

  • Here is my take (just a CentOS user).

    The communication from Red Hat/CentOS during this change has been somewhat poor. By reading various blog posts, etc.. A lot of people are confused about what this change actually means. When people read things like “CentOS
    will allow Red Hat to innovate and test new things” or however they word it, people read that to mean RHEL != CentOS.

    I know to a lot of the developers CentOS is a “community” or something, a collection of repositories and whatnot, but to the average person, CentOS
    is a product, a clone of RHEL.

    The average person wants to know this: if I download CentOS 7, and choose
    “Basic Server” in the installation, will I get the same packages (sans trademark) that RHEL 7 has? Will it have the same version of gcc and httpd, etc?

    This hasn’t been clear. If I understand the plan properly, CentOS will remain a RHEL clone, but there will be modified versions (variants?) of CentOS with added functionality, and maybe some repositories with extra goodies. If the communication was clearer, people wouldn’t be as worried about Red Hat making CentOS some sort of unstable testing grounds, and you’d receive better press.

    Logan

    On Sun, Jan 19, 2014 at 6:25 AM, Eddie G. O’Connor Jr.

LEAVE A COMMENT