AS THE PLUCKY developers announced Open Indiana, a fork of the Open Solaris operating system, it was hard not to think 'here we go again'.
Since Oracle pulled the plug on Open Solaris, the race was on for a community led effort to build upon the small but dedicated installed base that it had attracted. As lead programmer Alasdair Lumsden took to the stage, it seemed that Open Solaris in the guise of Open Indiana, now free from the shackles of Oracle, might finally be allowed to stretch its legs. But alas, first impressions suggest otherwise.
Open Indiana's goal is very clear - to become the default free Solaris distribution out there. It aims to be binary and package compatible with Solaris 11 and Solaris 11 Express and act as a "drop-in replacement" for Open Solaris.
While the Open Solaris project concentrated on the desktop, Lumsden is saying that Open Indiana will focus on servers. It's a good idea, as frankly users are overwhelmed with choices for desktop oriented operating systems. The only problem is that server distributions need stability, a pedigree and perhaps even a sense of conservatism, things that are sorely lacking in Open Indiana at this stage.
Lumsden has lofty goals for Open Indiana, saying that the project will be "100 percent free and open source". He went on to qualify that statement by saying that while packages meant that it would be hard to attain the 100 per cent open source aspect, it is a good goal to have. Lumsden continued by saying that not only did he want Open Indiana to be the open Solaris distribution but to try to attract users of Ubuntu and Centos.
There's nothing wrong with setting the bar high, but the statements made when releasing the first beta of Open Indiana, oi_147, left a lot to be desired. It might sound harsh given that it's a community driven effort, but with Open Indiana targetting Centos and Ubuntu users, coming out with statements such as "it has barely been tested" is simply not going to instill the level of confidence needed to convince users to change.
Many examples of this haphazard attitude were on show throughout the talk. The fact that the current installer is a graphical one, which Lumsden admitted isn't right for servers, is one. He went on to say that it's a simple task to create a text based installer and that one will be incorporated in a matter of days. If it is so trivial, then why wasn't it done for the first release?
Lumsden wouldn't publicly predict a stable release date for Open Indiana or be drawn to a roadmap, though a stable version is expected sometime within the next six months.
The Linux community ditched the Heath Robinson attitude almost a decade ago and because of that distributions such as Ubuntu, Redhat and Suse have started to make serious inroads within industry. The fact is, clear, reliable releases, schedules and support are the factors that help sell any operating system to management.
With no public release date penciled in, it's hard to get too excited for Open Indiana. Even one of Solaris' most highly regarded features, the ZFS file system, is soon going to be available as a native port on Linux. ZFS has been extensively updated in FreeBSD 8.1 and while Lumsden proudly said that he wouldn't trust his data on any other file system, the truth is, one of Open Indiana's biggest selling points is about to vanish.
It might sound harsh to beat up on the Open Indiana developers like this, but it's not an indictment of their technical ability, rather their attention to the details that will help the project get the recognition it deserves. Lumsden said that the faults in its io_147 release were due to the project sticking to a tight release date. The problem is, publicising such an experimental release to a wider audience is likely to do more harm than good.
Server oriented Linux distributions such as CentOS not only have years of experience on their side but something more valuable in the server space, reputation. Open Indiana needs to adopt a more professional attitude from the beginning if it is to compete with Linux and BSD.
If the Open Indiana project wants to provide a viable alternative to Linux and BSD in servers, then the project needs to adopt a low profile, quietly work on producing a stable and complete release, formulate release schedules and support structures, and then come back to beat the drum. Otherwise it will be seen as a rabble of hobbyists playing around with the long since discarded scraps of an industry behemoth, which won't do justice to the talents of the coders involved. µ