Re: PSPP should remove the scary, misleading message about "UNRELEASED TEST SOFTWARE! NOT FOR PRODUCTION USE!"

[John Darrington]

On Wed, Feb 24, 2016 at 11:57:27AM -0600, Alan Mead wrote:
     
     John,
     
     I don't want to be augmentative but I cannot ever agree (for several
     reasons, outlined below) that "scary" is good.  And I know you're a
     "FLOSS guy" so I'm surprised that we disagree about those points; so
     I've tried to clarify.

No offence taken.  We are all entitled to a difference of opinion.
     
     > I do not follow your reasoning why it could possibly be construed as a 
GPL
     > violation (but if we choose to exercise Option 7b of the licence, then
     > removing this text would be a violation).
     
     I made the point that the message is a kludge added to fix a problem
     already addressed by the GPL. So there is sense in which our kludge
     supersedes or overshadows the GPL and we are ignoring the GPL guidance. 


     I don't know that we are in "violation" but we should adhere more
     closely to the GNU guidance.

The GNU guidelines are here:  
http://www.gnu.org/prep/maintain/maintain.html
I don't see anything that we are not following.
     
     And we are "scaring" users; especially Windows users who don't know
     "git" from "production."  Isn't scaring users and explicitly telling
     them not to use a snapshot ("for production") directly contradictory
     with FLOSS practices like "release early and release often" or the
     concept that "given enough eyeballs, all bugs are shallow?" 

I don't think so.  If a user does not know what git is, they shouldn't be
using it.

     What other FLOSS software includes such an intentionally scary message?

What other FLOSS software publishes binaries of git snapshots?  I don't know
of any.  This is the crux of the matter - we have a somewhat unusual situation
which was causing problems for us.  It requires a somewhat unusual solution.
     
     You related a story about a bad review. I think those concerns are
     perhaps natural but fairly contrary to FLOSS. But assuming The Message
     precludes another bad review, is that worth it to "scare" users to
     preclude that bad review?  Does the message actually preclude (or even
     reduce the chances of) another bad review?
     
     Is the message even sensible?  Aren't there likely to be bugs and
     limitations in official releases?  And so isn't the "not for production
     use" part of the message _approximately_ as true for the release as for
     newer snapshots?  Are Windows users _really_ better off using 0.8.5,
     rather than the most recent snapshot that Harry has provided? Obviously
     not, right? So why would we tell them not to use the best, most recent
     version from Harry rather than 0.8.5?

You are right that we can never guarantee that a release will be free of
bugs.  But for released versions we have taken extra care that they are 
as solid as possible.  That is not true for a git snapshot.

As for your question "are users better off using 0.8.5 than the newer
snapshots" the answer in my opinion is an unequivocal YES.  The current git head
has bugs which are NOT present in 0.8.5

So.  My advice to users (on whatever OS) is:

Use the most recent release (currently 0.8.5) unless:

1.  You want to help the development process by "beta testing" the upcoming 
release
    and reporting any bugs you found.

2.  You need a feature or a bug fix which is not in the most recent release.  
There
    are not many of them.  They are listed at 
http://git.savannah.gnu.org/cgit/pspp.git/plain/NEWS


I think "The Message" helps users avoid any unwanted and possibly embarrasing 
suprises.

     
... just my opinion.

J'



-- 
Avoid eavesdropping.  Send strong encryted email.
PGP Public key ID: 1024D/2DE827B3 
fingerprint = 8797 A26D 0854 2EAB 0285  A290 8A67 719C 2DE8 27B3
See http://sks-keyservers.net or any PGP keyserver for public key.

signature.asc
Description: Digital signature