
As many of you may remember
I did a review of Windows Services for UNIX 3.0 (SFU) a few months ago. I remember being frustrated with that release because it seemed to me that all Microsoft did was throw something together just to be able to say "Hey look, we have this". I thought, since Microsoft released version 3.5, I would revisit and see what changes were done with it. I downloaded the beta version a while back and from the beta I was very impressed with the improvements that Microsoft made. Being a beta version it was buggy and some things just didnt quite work. I finally got the final version of the OpenBSD-based
SFU 3.5 and this release makes dynamic leaps and bounds over previous releases of this software package. I am glad to see a lot more work was put into this release.
Great article. I had been looking for a comparison between SFU and Cygwin so very good timing...
The idea of MS Linux begs a certain question. I agree with the following article:
http://www.eweek.com/article2/0,4149,1400161,00.asp
It doesn't look like MS would have their own Linux distro any time soon (the only way that happens is if the cost of maintaining their proprietary code for OS/Office becomes unfavorable when compared to the revenue they take in) but it does look like they are trying to find ways to "embrace and extend" Linux/Open Source or at least coexist. Their moves with Virtual PC and SFU demonstrate their desire to not get locked out of environments with a substantial and/or increasing Unix/OSS/Linux investment (and hopefully win some converts to Windows in the process). For example, both products require you have AT LEAST 1 Windows license - on SFU this is obvious whereas on Virtual PC it doesn't allow you to use a non-Windows server as the host OS (but you can use anything as a guest).
As to the author's closing statement:
My only hope is that people do not snub SFU just because it is an Microsoft product but rather embrace it and use it and support Microsoft for their efforts.
My response is that it's good to see MS working toward interoperability (with a solid product too) but potential customers are definitely going to question their long-term commitment. For example, Cygwin is likely to be supported indefinitely whereas MS could pull the plug at any time on SFU. Too many customers believe MS does what is good for MS and MS's platforms - backward compatibility and a customer's prior investment notwithstanding. Anyone seen MS Internet Explorer for Unix recently? I rest my case.