Re: Mulberry - the future?

--On Wednesday, February 12, 2014 12:16:22 PM -0700 Eric S Pulley
<[email protected]> wrote:

> I have looked and looked for a native replacement for Mulberry since
> switching to FreeBSD for my main OS a few years ago. Mulberry works there
> under Linux emulation but not as well as I'd like...
[...]
> I'd dearly love to see a version of Mulberry come out
> with the GUI API switched to... well pretty much anything open source GTK
> or QT at the top of the list. But since I can't do the work for free I
> don't expect anyone else too either.

Oddly enough, I've been thinking on and off (as recently as last night)
about what it would take to resurrect things.

I'm personally interested in having a modern Mulberry run natively in
CentOS (Linux) and OpenBSD.  My concerns are:
  1. things just don't build nicely right now
  2. dependencies on older libraries that don't appear to be well
     supported, if at all
  3. the SSL problem
  4. certain features that are currently broken under CentOS6 (such as
     deleting accounts)
  5. the problem with building under 64 bit natively
  6. various annoying UI artifacts that exhibit under Mulberry 4 but not
     under Mulberry 3 (yes, I still use a commercially licensed copy
     on one laptop because it works better).

I have to admit, I've been waffling about what would be better: 

A. Do the absolute minimum to address things like (1), (3), and (4), or
   do a major code rework.  

B. A new implementation based on Qt4 and steal code shamelessly from
   Mulberry as much as I can (and perhaps doing things like using a
   currently supported 3rd party IMAP library).  The intent would to
   keep the same general layout and workflow as Mulberry, but bring
   the code base up to date.

I would much prefer to do (B), but I have serious concerns about the
workload.  I also program only on UNIX platforms, and while I have 
OS-X boxes kicking around, I've not had a Windows box around for years.
Therefore I wouldn't be able to deal with the cross-platform aspects
without concerns.

(BTW, I say Qt4 because that is what is currently available natively
for OpenBSD and CentOS, however I suspect that dealing with Qt4/Qt5
conditional compilation should not be infeasible.)

So out of curiosity, do we have any Qt programmers in the crowd?
How about Qt programmers with cross-platform experience? Do we have
any Windows or Mac developers in the crowd?

Of the *developers* out there who could reasonably expect to commit
some amount of effort, is there any interest in this?

I am *not* commiting to this at this point; I'm merely bouncing ideas.

Devin