Information is not knowledge.
Knowledge is not wisdom.
Wisdom is not truth.
Truth is not beauty.
Beauty is not love.
(Frank Zappa from the album Joe's Garage, Track: Packard Goose)
Feeling Blue - Cobalt, Onyx and Quartz, a Forking disaster?
Submitted by Martyn on 7 May 2009 - 10:38am
First there was Cobalt - in the form of a nice little 1U server appliance (or a desktop box aptly called "The Cube"). Basically this was a Linux based "Internet in a box" solution - handling web, mail and DNS wrapped in a nice understandable web based user interface.
The Cobalt became popular with hosting companies (we bought lots of them) as we could (reasonably) safely let the customer loose to manage their own VISP (Virtual ISP), with each box capable of handling a huge number of websites, domains and users.
From the RaQ2 and Cube the Cobalt range grew through RaQ3, XTR and finally into the RaQ550 at which point Sun Microsystems bought the company and immediately ceased all production, such was the perceived threat. The Sausalito code base was effectively released into the open world and morphed into a new Open Source release known as Blue Quartz - keeping the nice user interface and functionality, but ostensibly built upon CentOS. Blue Quarts was, and still is, maintained by Hisao Shibuya
As the RaQs reached EOL many migrated to Blue Quartz and the platform became respected and very well supported. Many easy quick installers were developed, as were additional components and upgrades (special thanks to Brian N Smith at Nuonce Networks).
The project however stalled and development stagnated. YUM updates, bug fixes, software upgrades and UI functionality did not appear in core releases and were essentially left to third party providers (Nuonce Networks PHP5, MySQL5 packages are two examples).
Toward the end of 2008 the lack of pro-activity saw a collaboration between Michael Stauber (Solarspeed) and Brian N Smith (Nuonce Networks) leading to the development and release of a new fork; Blue Onyx.
Blue Onyx claims to have everything that should have been developed into Blue Quartz, including PHP5, MySQL5, Centos 5.2 base, improved account management and site creation all there straight out of the box, and yes, it looks good!
A few weeks after the release, Brian N Smith closed down Nuonce Networks and disappeared into the ether (gone, but not forgotten Brian).
So we now have a nice, fresh, up-to-date new release of what was once the Cobalt classic, but is this really a good development?
I think 'no', it isn't - and this is for one fundamentally flawed reason;
The project is still pretty much dependent upon one person; Michael Stauber.
Don't misinterpret this - Michael is an excellent coder and still offers a great amount of free support and advice on both Blue Quartz and Blue Onyx mailing lists - but he is one person, and as we have seen with Hisao, and Brian - people can have other things going on in their life, or lose interest for whatever reason - or simply do not have the time they'd like to dedicate, and this, therefore, is the main vulnerability facing Blue Onyx.
This isn't new. The development of Etomite (a very capable CMS) forked to ModX very successfully and various versions of Asterisk (VoIP PBX) have been developed, including Elastix and TrixBox - however in each example of a fork here, there is a development team collaborating on the project, which means it isn't so important if people get 'other things' in their life distracting them from the project.
So, what's the future for Blue Onyx?
I have every faith in Michael Stauber to do what's needed in the immediate term, however unless the project develops to encompass a dedicated group of developers with shared motivations, the long term could be fraught with the same issues facing Blue Quartz.
ADDENDA - 2009-05-14:
To clarify a little, the above was originally written for a Board Meeting when looking at our Corporate Exposure. We use both Blue Quartz and Blue Onyx widely within both customer facing and operational units - and will continue to do so. I have every confidence in Michael Stauber and cannot fault Blue Onyx - however the longer term vulnerability must remain a concern and will be continually reviewed as a part of normal business continuity strategy.
We all must work together, especially with the lack of adequate
We all must work together,
but what is your advice centos Bluequartz is a great server software
Alex Bajan
RAQport.com
You get what you pay for.. and a lot more..
The original poster, if I understand well, want a free lunch.. delivered to his desktop. We've been using Cobalts (Raq3/4/550) since mid 90's. Was the Cobalt and Sun support good?? No, not at all.. questions took weeks to get answered via the official way, updates.. well they arrived.. sometimes and after a long time.
Not even talking about re-installing with the weird x-over cable procedure and a specific ethernetcard. And then again... 4 hours loading all the packages.
SOOO.... BQ came, install a snap, but a BETA quality product with no follow up, except from Brian, Michael and a few others. We bought a lot of Brian's packages because they actually worked and the support was great. But the core product was outdated and we had sometimes strange issues.
Then BlueOnyx came.. after 1 week of testing we migrated 80% of our servers to it.. a risk? maybe.. but we had confidence in BO.. and we were not disappointed at all. Support from the list and from Michael was great, we bought some packages from Solarspeed and will do so for all our other servers as well. And if he comes up with a paid support plan, we will subscribe; Not because we need it, but because we want to support the company in development. I would not mind if Michael charges 500 EUR for the BO software.. remember what you paid for your Cobalts? Remember what you pay NOW for WIndows (with a big development team, but unable to create a stable product let alone a clear path to the future).
I agree that MAYBE a more formal development team would be nice, but with so many BO addicts arounf (see mailinglist) help is not far away and future development is almost guaranteed in my opinion.
But to have a dedicated development team means someone has to pay.. So either we all pay $$ or €€ for this great product and then we can 'complain' about issues and development; or we don't pay and just be happy.
BTW: IF you are so worried about the future.. just get the source code and hire a couple of coders to do whatever you want with it in terms of support and development.
Just my opinion.. we are happy with BO, the development going on, the speed of fixes and the support from the whole usergroup.
THANKS ALL!
jan
Great post
We use BQ since 5 years and really would be happy to migrate von BX. But our problem really is that we cannot relay our hosting business on a one man show. So Michael, please think about it...
All the eggs in one man's basket...
While having one person responsible for everything might be good in some respects, if you want another example of what happens when that one person has other priorities pop up, check out Whitebox Linux (http://www.whiteboxlinux.org). What once started as one man's endeavor to provide something for a small public library, developed quite the following. Then the unthinkable happened. That man quietly vanished amid other more important things. Then the forums filled with people making suggestions to take over the project, offers of assistance, plans to fork, and more and more people arguing over the status of WBEL, and why people should move to CentOS.
While I do use BlueQuartz on some of our production mail servers, we don't use it to it's potential. Furthermore, now that I know more about Linux and setting up LAMP servers, when these boxes finally give up, and I have to wonder what I replace it with... I know I won't have to rely on one man other than myself. I'll have to take responsibility for my own servers and, yeah, I might just use CentOS, but it will be on my terms, and it will be a larger community supported environment.
Some of you will agree with me, some will not, but in the end we each have to be responsible for our own.
I agree, but...
BQ stalled and much debate ensued and it still continues.
But with no-one taking onus/direction, it quickly fell into the mire.
Perhaps with one person directing themself, they cannot get as lost as BQ did!
Has anyone found another free out-of-the-box hosting software company that was transparent, helpful and devoted?
If, like me, you loved Cobalt's simple but effective boxes, we stayed with them as long as we could.
But I was having to look at buggy expensive solutions to replace - with forums having users fix problems and post them back to the company - despite all paying for support!
Simply SUPPORT Michael, like it, use it, then buy upgrades from Solarspeed - we have and plan to do so in the future.
Long live BO!
While I agree with the
While I agree with the overall premise, having the code in the hands of one individual can reduce the number of issues that crop up in collaborative development environments. Having worked in both, I can definitely see benefits to each. So we just need to ensure that Michael has a backup and remains sufficiently motivated (either financially or by finding purpose and meaning in what he's doing--or both) to ensure project momentum.
I've been extremely impressed with Michael's improvements and upgrades to date and will continue purchasing from him to ensure the longevity of BO.