On Tue, 18 Oct 2016 21:42:13 -0500, Chris Bennett wrote:

> On Wed, Oct 19, 2016 at 02:06:51AM +0000, Ralph Siegler wrote:
>> 
>> Linux on Power8 provides a way to run certain  closed source softwares
>> that are certified to run on  Linux on PowerPC. Of course, those
>> softwares generally run even faster on AIX with less "loose ends" and
>> bugs because they were specifically developed and tuned for a couple
>> decades using bespoke tools before recent porting to Linux.
>> 
>> 
> Hmm. OpenBSD's license does not in any way prohibit using closed source
> software with it. Would it be better than AIX or Linux?

None of that closed source software will ever run on OpenBSD.  There will 
never be IBM Websphere for OpenBSD, nor Lawson Financials for OpenBSD, 
nor JDEdwards EnterpriseOne for OpenBSD. 


> Maybe. Or maybe not. Speed has not been the primary goal of OpenBSD.
> Which is OK for me. But, for such a high priced box, I think speed might
> be way up there or would reliability be more important?
> 
In most cases the reliability of the OS and the kind of  app I'm talking 
about is a given, uptime will be indefinite for most customers.  It's a 
OS plus app plus libraries plus tools solution groomed over a couple 
decades for that architecture.  Security will be external, right or wrong 
that's how it is.



> For example, scientific software would definitely place more emphasis on
> precise and accurate calculations rather than speed.
>
There are specs for that and standard libraries, certainly for HPC where 
power architecture plays speed is huge consideration but then they have 
nvidia gpu doing the bulk of the work along side of the power8, different 
beast than business power8.


> 
>> 
>> Developers are interested in that architecture you say...yes I believe
>> that.  I'm a photographer and am interested in a Hasselblad H5D-200c
>> which the body alone goes for $45,000.  I can't afford one, let alone
>> say  put a lens on it, have no earthly use that would  require
>> one.....but dang if it isn't cool.
>> > 
>> > Below you suggest getting going with Power6 and 7, which are much
>> > cheaper to purchase.
>> > 
>> > Would it be reasonable to look at this the other way around:
>> > 
>> > Develop the Power8 architecture now so that when prices fall later,
>> > companies can then afford to buy them and immediately use a developed
>> > and tested OpenBSD on them?
>> 
>> Develop for architecture none of the userbase has or will have for five
>> plus years?   Of course the BSD licensed open source drivers that IBM
>> will provide for all that poop that flies by in the five minute POST
>> time will make the job easier, and the megabytes of spec docs they've
>> written...cause they wouldn't have the gall to  hand over BLOBs for
>> hardware without full specs....
>> 
>> 
> Supposedly, IBM is energetically supporting Open Source for this.
> 
> 
>> We'll have to make the BSD foundation thermometer taller for N devs
>> times $170 month or more extra electric bill.
>> 
>> 
> If IBM is really interested in Open Source, they might just decide to
> donate to the OpenBSD Foundation. Would pay the electric bill
> (hopefully).

OpenBSD foots electric bill for devs? 


Oh, and when they start power8 box will put a HP blade chassis  to shame 
with the volume level for a few minutes, I couldn't imagine having one in 
a home and repeatedly booting to port an OS.  
> 
> But talk is not action.

Action is too expense, too pointless.  It's out of league just as a Cray 
vector supercomputer  OpenBSD port would be

> 
> 
> General question:
> Would Power8 lead to using Power9, Power10, etc?


 too expensive to have for development, too expensive to run, to 
expensive for a userbase while businesses waited for a mature version, no 
compelling use case in the open source world that couldn't be done with 
Xeon drawing half to a third the power.  

> 
> Chris Bennett

Reply via email to