On Fri, Mar 16, 2018 at 10:13 AM, Scott Kostyshak <skost...@lyx.org> wrote:

> On Fri, Mar 16, 2018 at 08:56:15AM +0000, Pavel Sanda wrote:
>
> > I agree that the language of the dialog can be a problem. But there is
> no rush
> > and if you can't stand for the version with english dialog, we can
> advertise
> > the installer once you get the translations you need and are happy with
> the
> > state of the installer.
>
> +1
>

+1


> Indeed. I'm very happy that we are continuing this discussion. We are
> still using logic, combined with examples (e.g. Uwe's example of his
> Mom), and trying to understand each other's arguments. This is not a fun
> debate and we are all tired, but we must continue. Although it might
> take time, I really hope that we can come to a solution. It would be a
> sad day for users and for developers if this escalated more and we
> abandoned the goal of providing Windows binaries on the website. I will
> not abandon that goal. From what I see, we are still working towards a
> solution. Thank you to all for participating in this issue that is
> frustrating for everyone.
>

+1; I too appreciate this discussion to ensure that the best possible
software is provided to all end users on all OSs.  I have little ability to
contribute to the technical aspects of the debate, but a couple personal
(and perhaps unpopular) thoughts and an anecdote:

   1. I would rather see proper time taken to release 2.3.0 coherently for
   all supported platforms in as good of shape as possible (as determined by
   developer consensus, and perhaps with some compromise) rather than
   suspension of cross-platform support (temporary or permanent).  I work and
   play with others in heavily cross-platform environments.  A major selling
   point for both LaTeX and LyX (c.f., MS Word) is that they are
   cross-platform on the three major OSs.  I would have a hard time fairly
   explaining to Windows colleagues why I've received a new version of LyX
   ahead of them.
   2. If taking proper time means delaying to await fixes to MikTeX,
   translations for dialogs, etc. (as determined by the development team) then
   so be it.  Ultimately, if someone installs something from lyx.org and it
   causes a problem, the blame and burden falls on the LyX team regardless of
   the underlying cause.  Using an analogy in the spirit of this thread: if my
   car breaks down I don't care which sub-supplier made the part that failed.
   The car's manufacturer is who takes the blame.  However, new car models get
   delayed and recalls happen.  That's fine as long as the problem is dealt
   with as timely and ethically as possible.
   3. Unfortunately, and unfairly, the damage caused by creating user
   frustration may never be observed by the LyX team and thus be unable to be
   addressed ending up with an unaware development team and a marooned user.
   The development team is aware of a potential Windows installer issue, so
   I'm happy to see everyone discussing it to get the best possible solution
   despite the frustration involved.  Imagining things as a
   (new/inexperienced/etc.) user: if I saw a (temporary or permanent)
   unavailability or failing of a Windows installer I would see it as a
   failing of LyX and not some underlying dependency.  The omission of the
   Windows installer (or it failing during installation), would have me
   question the quality of the software as a whole and the development team's
   practices (regardless of which OS I use).  I have personal experience with
   LyX failing me (turning into a complete abandonment of it between
   2004-2014).  Was I an inexperienced user in 2004?  Yes.  Should I have
   pursued asking about the problems I was having?  Yes.  Did I?  No.  I took
   the "easy" path and returned to what worked for me when it came to
   LaTeXing.  I don't like to imagine another user choosing to do that.

- Joel

P.S. One reason I see no need to push the release before it's ready (other
than potentially reducing Scott's stress as release manager) is because the
RCs are out there.  A user wanting/needing 2.3.0 (in as good of form as
possible when released) and its features can find and use them.  Do I like
a delay in 2.3.0?  No.  But I'd much rather see the right product go out
later than a known incomplete or flawed product go out sooner.

Reply via email to