![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
| an angel dressed in the blackest lace (lithiana) wrote,
on May 13th, 2011 at 04:49 am... |
So, apparently The SCO Group (nee Caldera) has been acquired by UnXis, a new company set up for this purpose.
I've always thought SCO (both the old and the new one) had fairly decent products; OpenServer was the standard x86 Unix for a long time, and the company understood how to sell to the SMB market, something other Unix vendors (except Dell, perhaps) never really managed. Unfortunately, under Darl McBride, the company rather lost its way, culminating in the ridiculous Novell/Linux litigation which ended up bankrupting the company.
A side effect of this was a notable lack of investment in their software technology; there have been no real updates to the SCO's Unix products since the 90s, and while OpenServer 6 (the last major release, in 2005) added several important new features, this was mostly the result of porting technology from UnixWare (their other operating system) rather than new development. Today, SCO's products are nearly obsolete (not even supporting 64-bit x86 hardware, for example) and it's hard to imagine choosing that platform for a new deployment.
Despite that, I think UnXis has a good opportunity here. Other than Linux, there are only two serious Unix platforms for x86: Solaris and OpenServer. Sun was never able to sell to the SMB market effectively, and Oracle has made it clear that they don't care about this market either (and even Oracle's enterprise customers are grumbling about changes since Sun was acquired). As a small Oracle customer, I am actively looking for alternatives, and an updated Unix product from SCO would certainly be interesting.
Bringing a product (or company) back from the brink of death is not easy, but if I was UnXis, this is how I'd do it:
Firstly, UnXis needs to retain its existing large customers. This means making the ownership transfer simple and non-disruptive to customers (something Oracle failed at with the Sun purchase), and making a clear statement about the future of OpenServer. Retaining existing customers will provide a revenue stream to fund future development.
Next, they need to make a large investment in updating their product line. The most obvious update needed is support for 64-bit hardware, and support for large systems (multi-core systems with 64+ cores and 256GB+ RAM). This might involve an overhaul of the VM system. The userland needs to be updated to add common utilities and options that users expect. The packaging system should be improved to make finding and installation software easier, and there needs to be a simple mechanism for installing OS updates. It should be simple to install common open source software.
OpenServer has always had decent support from hardware vendors, and UnXis needs to build on this to ensure it continues to have good driver support. It should be possible to install OpenServer on a modern system and be sure the hardware will work well without fiddling. This includes X drivers (NVIDIA, KMS, etc.) for the desktop.
A new desktop should be provided, probably based on GNOME, and common software (such as Firefox) should be available. Linux emulation might not hurt. Enterprise customers don't care about this, but enabling end-users to use OpenServer in the same way they use Linux now is important in attracting new customers; if people use it at home, they're more likely to use it at work and pay for it. Some version of OpenServer, suitable for home users, should be available for free.
Veritas Volume Manager should be updated to the current version and available in the base system. With the prevalence of SAN storage now, not having a decent volume manager available as part of the OS is inexcusable.
SCO always had a good relationship with Microsoft, and this should be built on, especially since Windows is the standard platform in the SMB market. This means good support for Windows technologies like CIFS and Active Directory, both as a client and a server.
Then, OpenServer needs to be promoted, and available everywhere. This means if you buy a new system from Dell or HP, it should be possible to buy an OpenServer license with the system, and the system should be certified for use with SCO products. Since OpenServer has a reputation for being old and outdated, a marketing campaign is needed to promote the new features that make it interesting again.
"Available everywhere" includes virtualisation. It should be available pre-installed on virtualisation platforms like EC2, and include full support for common virtualisation technologies (VMware, Hyper-V, Xen, KVM). Some kind of lightweight virtualisation platform should be added, comparable to HP-UX Containers or Solaris Zones.
There's a few other features, I wouldn't mind seeing, but this is a reasonable starting point. I don't know if this is enough to fix SCO, and if UnXis were a public company, I don't think I'd be buying shares in them at the moment. However, I do think they have a chance to do something interesting.
I've always thought SCO (both the old and the new one) had fairly decent products; OpenServer was the standard x86 Unix for a long time, and the company understood how to sell to the SMB market, something other Unix vendors (except Dell, perhaps) never really managed. Unfortunately, under Darl McBride, the company rather lost its way, culminating in the ridiculous Novell/Linux litigation which ended up bankrupting the company.
A side effect of this was a notable lack of investment in their software technology; there have been no real updates to the SCO's Unix products since the 90s, and while OpenServer 6 (the last major release, in 2005) added several important new features, this was mostly the result of porting technology from UnixWare (their other operating system) rather than new development. Today, SCO's products are nearly obsolete (not even supporting 64-bit x86 hardware, for example) and it's hard to imagine choosing that platform for a new deployment.
Despite that, I think UnXis has a good opportunity here. Other than Linux, there are only two serious Unix platforms for x86: Solaris and OpenServer. Sun was never able to sell to the SMB market effectively, and Oracle has made it clear that they don't care about this market either (and even Oracle's enterprise customers are grumbling about changes since Sun was acquired). As a small Oracle customer, I am actively looking for alternatives, and an updated Unix product from SCO would certainly be interesting.
Bringing a product (or company) back from the brink of death is not easy, but if I was UnXis, this is how I'd do it:
Firstly, UnXis needs to retain its existing large customers. This means making the ownership transfer simple and non-disruptive to customers (something Oracle failed at with the Sun purchase), and making a clear statement about the future of OpenServer. Retaining existing customers will provide a revenue stream to fund future development.
Next, they need to make a large investment in updating their product line. The most obvious update needed is support for 64-bit hardware, and support for large systems (multi-core systems with 64+ cores and 256GB+ RAM). This might involve an overhaul of the VM system. The userland needs to be updated to add common utilities and options that users expect. The packaging system should be improved to make finding and installation software easier, and there needs to be a simple mechanism for installing OS updates. It should be simple to install common open source software.
OpenServer has always had decent support from hardware vendors, and UnXis needs to build on this to ensure it continues to have good driver support. It should be possible to install OpenServer on a modern system and be sure the hardware will work well without fiddling. This includes X drivers (NVIDIA, KMS, etc.) for the desktop.
A new desktop should be provided, probably based on GNOME, and common software (such as Firefox) should be available. Linux emulation might not hurt. Enterprise customers don't care about this, but enabling end-users to use OpenServer in the same way they use Linux now is important in attracting new customers; if people use it at home, they're more likely to use it at work and pay for it. Some version of OpenServer, suitable for home users, should be available for free.
Veritas Volume Manager should be updated to the current version and available in the base system. With the prevalence of SAN storage now, not having a decent volume manager available as part of the OS is inexcusable.
SCO always had a good relationship with Microsoft, and this should be built on, especially since Windows is the standard platform in the SMB market. This means good support for Windows technologies like CIFS and Active Directory, both as a client and a server.
Then, OpenServer needs to be promoted, and available everywhere. This means if you buy a new system from Dell or HP, it should be possible to buy an OpenServer license with the system, and the system should be certified for use with SCO products. Since OpenServer has a reputation for being old and outdated, a marketing campaign is needed to promote the new features that make it interesting again.
"Available everywhere" includes virtualisation. It should be available pre-installed on virtualisation platforms like EC2, and include full support for common virtualisation technologies (VMware, Hyper-V, Xen, KVM). Some kind of lightweight virtualisation platform should be added, comparable to HP-UX Containers or Solaris Zones.
There's a few other features, I wouldn't mind seeing, but this is a reasonable starting point. I don't know if this is enough to fix SCO, and if UnXis were a public company, I don't think I'd be buying shares in them at the moment. However, I do think they have a chance to do something interesting.
Post a comment