[Info-vax] Layered products, the HP view !?!

[David Froble]

johnwallace4 at yahoo.co.uk wrote:
> On Saturday, 20 June 2015 02:31:12 UTC+1, David Froble  wrote:
>> JF Mezei wrote:
>>> On 15-06-19 09:56, clairgrant71 at gmail.com wrote:
>>>
>>>> That was the intent. The big thing we did was fix a long-standing bug that preventing booting on a particular i4 blade memory configuration.
>>> When you guys announced you managed to do a build of 8.4 from the
>>> sources HP gave you, would it be correct to state that this included not
>>> only VMS but also all the layered products that you inherited that are
>>> still "alive" ?
>>>
>>> And and you did that first build from the sources you got from HP, I
>>> take it that that build would have included all those small fixes HP did
>>> but never released ?
>>>
>>>
>> I believe that was already mentioned by Robert Brooks.
>>
>> Bugs get fixed, in the sources, but never sent out as patches and such, 
>> other than to whomever filed the support request.  So, when you rebuild 
>> the sources, all the fixes are there.
>>
>> As for knowing you have a problem, but not distributing the fix to users 
>> other than the one who filed the support request, well, I find that 
>> rather disgusting.  Why?  Because other users could spend significant 
>> time trying to find out what's going wrong before filing a support request.
> 
> Mr Froble sir, you need to get out more (or maybe I misunderstand).
> 
> Standard operating practice in the traditional part of the industry
> is to fix bugs and keep silent about it till the next release.
> 
> An improvement on that is to fix bugs and (as you go along)
> document what's been fixed but not necessarily release the fix
> till the next version comes out. 
> 
> A variation on that is to fix bugs as you go along and make the
> fixes available to customers on support contracts.
> 
> It'd be a "brave" company that fixed things as they went along
> and made the fixes available to all, for free. It does happen.

I'm old enough to remember the DEC software organization(s) in the past. 
  I'm talking 1970s and 1980s.  While I was not familiar with the inner 
workings, my impression was that they "did it right".  I also remember 
the 11/750 floating point problem, and DEC's response, which was rather 
different than Intel's "so what" attitude with their 386 FP problem.

So I've been conditioned to expect excellence, because that is what we 
had back then.  You'll just have to live with my folly.