[kwlug-disc] Microsoft Patches Linux; Linus Responds | Linux Magazine

[unsolicited]

Bob Jonkman wrote, On 07/27/2009 9:23 PM:
> unsolicited wrote:
>>  I hear you, but ... how is the situation you describe different than
>>  today's situation (pre-MS code release).
> 
> The difference is that with the code release Microsoft will get some 
> traction in Linux shops, then siphon that off to their own VM platform.

How? (Assuming I understand the code contribution correctly.)

If you're a Linux shop, you're not running Hyper V.

If you're a Windows shop running Linux VM's ...
(Which might keep a Windows customer a little longer.)


> By slowly Extending and Embracing this code release they'll get some 
> lock-in on HyperV, and then possibly pull out the rug.

How? If you're running a Linux vm under windows and they pull they 
plug ... people (presumably) are going to find another way to run 
Linux under Windows, or Linux another way. Not switch their vm's to 
Windows.

[What am I missing?]

> Now, my entire argument is based on speculation, but each step has 
> precedent and none of it is impossible.

That was then. (And, <sigh>, probably now.)

But ... the Windows marketing model / revenue stream has changed since 
then. And is probably still changing. If that was then, and we're all 
smarter now ...

> 
>>  [...] It still baffles me why similar success on the desktop still 
>> evades
>>  us. Especially at MS Office costs.
> 
> I've worked for large industry, a major bank, and two levels of 
> government. The larger the organization, the more the thinking is that 
> it's important to get vendor support.  Not that the vendor support is 
> actually used to solve problems, but it's a way to divert blame.  If you 
> buy applications from a vendor and the wheels fall off then you can 
> always sue the guy that made them.  If you get FLOSS applications you 
> can get third-party support, but the only one to blame for lost wheels 
> is yourself.

This is only practically true if you have the cash to finance the 
attempt, think you can succeed, and the fortitude to carry through. 
Which is doubtful. But I take your point. And, obviously, this logic 
doesn't convince them, or you would have mentioned that.

> Unsolicted wrote:
>>  I hear you, but ... how is the situation you describe different than
>>  today's situation (pre-MS code release). i.e. Consider the MS
>>  solution as non-viable / unacceptable, today, or then, and migrate to
>>  the other solution. (As noted, e.g. Xen, vmware, etc.)
> 
>>  Is vendor lock in the same issue today as it has been in the past?
>>
>>  No, at least in some ways - e.g. Xen, Open Office.
>>
>>  Yes, at least in some ways - e.g. MS Project and Visio still seem to
>>  be the industry wide standard, without similar, competing, universal
>>  industry wide FOSS alternative.
>>
>>  If Linux servers are finally industry standards, look how long it
>>  took for that to occur. It still baffles me why similar success on
>>  the desktop still evades us. Especially at MS Office costs.
>>
>>  Bob Jonkman wrote, On 07/27/2009 4:08 PM:
>> > Khalid says:
>> >
>> >> But as far as the embrace, extend, extinguish thing goes, they
>> >> can do the first 2 all they like. The GPL prevents them from
>> >> doing the 3rd!
>> >
>> >
>> > The GPL can ensure that the newly released driver code lives on in
>> > its current incarnation.
>> >
>> > But no such restrictions exist for HyperV.  Nothing prevents
>> > Microsoft from changing HyperV so that it becomes incompatible with
>> > the open source drivers just released.  And open source programmers
>> > are not going to invest much time in reverse-engineering HyperV to
>> > keep these drivers compatible, when other open source alternatives
>> > to HyperV already exist. A business that has invested in HyperV is
>> > not going to continue to use an old version of HyperV to maintain
>> > compatibility with some open-source drivers. For most businesses it
>> > makes more business sense (ie. it's cheaper in the short term) to
>> > buy the compatible Microsoft product.
>>  . . .