Re: release scheduling, cygwin 1.7 et al (was Re: X11R7.5 & fontcache..)

[Linda Walsh <cygwin@tlinx.org>]

* Yaakov (Cygwin/X) wrote: "http://cygwin.com/acronyms/#RSN :-)"
** LW wrote: "--I don't suppose you could express that in ISO format? :-)"

Larry Hall wrote:
> Since you've asked this on multiple lists, I'm going to assume this is
> more than just a humorous comment that you don't expect an answer to.

   Posting this as a reply on cygwin would have been more appropriate, as
there, it was obvious, I was *repeating* the question, (I mentioned it came
up on this list), and it was there that I was *probing*, to see if a more
firm day had been given that I had missed.

> Cygwin 1.7 will be released as soon as it's ready.
   So I've heard.

> There is no specific date at this time. 
---

   Same as before.  So I haven't missed any news and there is still no
timeframe (could still be a year away) for release...  firm dates are one
thing, but "we are hoping before Xmas, or Beltane (May Day)" would give an
idea.  That said, I very well know  the unpredictable nature of SW
development.  How can one give a "real schedule" about something that has
never been done before?  

   Managers have pushed the moniker 'software engineering' onto the field --
as though software is something that can be engineered and manufactured like
off-the-shell nuts and bolts.  That does the field a great disservice.  As
writing software is all about creativity.  It's not about creating the nuts
and bolts, but how they go together -- just like painting is not about the
paint, it's about how it goes onto the medium.

   Sure you can give estimates, but reality is they are only best guesses
and any number of things can come up to alter them.  This was known as far
back as 1985.  See good paper at:

   http://dspace.mit.edu/handle/1721.1/48174?show=full

   Says basically, even the "act" of coming up with a schedule can
[adversely] affect the outcome of not creating one.  The compare it to the
'Heisenberg Uncertainty Principle' But even with such knowledge at your
disposal, you'll still run into Dilbertian manager types who live in a
fantasy world, where, by force of will, they believe they can force a
product into existence on their schedule.  :^)  However, I am not the
manager type -- I'm not ruthless enough.

> That's not a reason not to use it however, if you prefer.  And if you need
> something that's only on 1.7, this is a good time to try it out.  And
> since you can install 1.7 beside 1.5 if you like, your risk of borking
> your Cygwin installation is pretty minimal.
---

   I saw support for dual installations in a recent announcement on Cygwin's
main list.  I've also seen quite a few problems reported against 1.7 in the
compatibility department, but I realize that gives no indication of the
number of users who don't have problems.  

The announced wasn't clear if the dual installation support was backwards
compatible to 1.5.  My history with cygwin goes back a bit, to, when having
more than one cygwin install on your disk at the same time was grounds for
being drawn and quartered if you reported any problem in such a setup.  You
know how the cygwin-standard team can be...any excuse for flaying a user
provides them joy (in their 'we take pride in our meaness' way).  :-)

   Maybe when I'm not swamped w/other probs...

   Thanks for the ~informative~ response.  ;^)

-linda


--
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple
Problem reports:       http://cygwin.com/problems.html
Documentation:         http://x.cygwin.com/docs/
FAQ:                   http://x.cygwin.com/docs/faq/