[OpenIndiana-discuss] Update info?

[Richard L. Hamilton]

Presumably there's a balance with this, as with most things.

But giving in is not without costs.  Making the quarterly look good often leaves one out of position for the long haul.

Maybe one could do both.  Use familiarity tools, but _only_ as a temporary measure, while concurrently encouraging native familiarity.  Encourage standards (and the isolation of code that needs to use non-portable features).  Clean code costs less to maintain, gets a better reputation, and is an insurance policy when the hot platform changes and porting is needed.

The problem I have with expediencies is not recognizing that they're sometimes needed.  It's that they become the norm even when _not_ needed, with little or no commitment to clean up after; it's always move on to the next thing, and curse the old as "legacy", which usually flies since the management has changed since then anyway.  It's like reorgs; it takes a couple years to really make one work, but by then the managers have padded their resumes with "led X through reorg" and moved on, so they're almost always a symbolic exercise that leaves chaos and _more_ bureaucracy and turf battles in its wake.  Changing organizational structures, or OS's, does not confer magic results.  It takes changing _people_, or at any rate their applied skills, to approach that.  It takes both techies and managers that know the difference between white papers and _judgement_.

Doing something right means _both_ being responsive now _and_ maintainable for the long run.

On May 24, 2011, at 6:11 PM, Blake wrote:

> I certainly understand your point of view, Richard.  Unfortunately, fewer
> developers nowadays understand systems, and that's not likely to improve.
> 
> My point is more that tool choice in business is driven by opportunity cost
> and the bottom line.  If I have to spend an extra two weeks training devs to
> use a Solaris platform, management will often prefer to use the faster,
> albeit sloppier, Linux platform instead.  Hence the meteoric rise of Ubuntu,
> Ruby, RVM and other easy-but-not-that-stable tools.  Stability ends up
> getting handled with horizontal scale across many systems, rather than
> making a smaller set of monolithic systems highly reliable.
> 
> I predict that if we as a Solaris community don't adapt to this changing
> corporate landscape we will be relegated to a corner along with AIX and
> HP-UX.  In *every* development environment I've worked in, developer
> response to my suggestion to try Solaris has always started with "why?  it's
> too hard to use/compile on/install".  I've been able to change these
> opinions in some cases, but it's a hellish uphill battle when the Solaris
> community has refused (until fairly recently), to compromise in the
> userspace.
> 
> All I want is to get as many people as possible using and contributing to
> this project :)
> 
> 
> Blake
> 
> On Tue, May 24, 2011 at 2:12 PM, Richard L. Hamilton <rlhamil at smart.net>wrote:
> 
>> 
>> On May 24, 2011, at 12:13 PM, Blake Irvin wrote:
>> 
>>> As a longtime Solaris user I don't have to be convinced that it's better
>> in many ways than Linux.  But there is a reality that I've been forced to
>> embrace - developers don't care.
>>> 
>>> I'm in the process right now of helping a large development team make the
>> transition to a Solaris hosting solution.  I've spent hundreds of hours with
>> my operations team customizing our environments so that they 'look and feel'
>> like Ubuntu.
>>> 
>>> Why?  Because that's what the current crop of developers know, and it's
>> too expensive to train them otherwise.
>> 
>> Being both a programmer and a sysadmin, I wonder how good a developer is
>> that needs to be trained to learn something new.  There's got to be a
>> "Solaris for Linux users" doc around somewhere.  If someone can write code,
>> they _should_ be able to read English, too.
>> 
>> If they're too stumped to _use_ more than one environment, they're
>> certainly not going to be the brightest about writing portable code.
>> 
>> People who have trouble learning something new are just dead men walking.
>> Flip the switch, enjoy the crackly sound, and get new ones.
>> 
>> 
>> _______________________________________________
>> OpenIndiana-discuss mailing list
>> OpenIndiana-discuss at openindiana.org
>> http://openindiana.org/mailman/listinfo/openindiana-discuss
>> 
> _______________________________________________
> OpenIndiana-discuss mailing list
> OpenIndiana-discuss at openindiana.org
> http://openindiana.org/mailman/listinfo/openindiana-discuss
> 

-- 
eMail:				mailto:rlhamil at smart.net
Home page:			http://www.smart.net/~rlhamil/
Facebook, MySpace,
AIM, Yahoo, etc:		ask