minutes of ESC call ...

[David Ostrovsky]

On Thu Mar 10 16:40:50 UTC 2016, Michael Meeks wrote:

>* embedded database / firebird mess (Lionel/David O)

[...]

>+ or bundle their binaries (urgh) (Lionel)
>             + deep non-enthusiasm (Norbert, Michael, Bjoern)

But isn't this is exactly what we did for years with Seamonkey based
Mozab driver ... and TBH if /me wouldn't remove it in: [2], we would
probably still ship the prebuilt binaries of this mess for years. So
what is wrong to ship parts of our suite that we can't build? Why it
was the right thing to do, to ship prebuilt Mozab mess for years, but
is wrong thing to do to ship prebuilt FB?

>    + decision needed: (Norbert)
>        + effort to support VS 2015
>        + not eager to turn it off.
>AI:        => regret; disable for now for CI (Norbert)

I have strong impression that there is still some misunderstanding
left: we don't "regret; disable (FB) for now for CI", but we can't
build it any more in next 10 years with our current toolchain at all,
after this patch is merged: [1]. After [1] is merged, say in a couple
of weeks, after we have enough TB and gerrit slaves settled with MSVC
2015, we have exactly two options:

* a) drop FB
* b) ship the prebuilt binaries on windows, built with old and removed
toolchain (and on Mac, because it cannot be built there as well?)

And no, we don't have the option:

* c) fix FB to build on MSVC 2015

because the FB devs rejected support for the new toolchain
(C++14/C++17) for the next 10 years (at least) because they need the
compatibility with WinXP (see the thread from their ML I mentioned in
my previous mail).

In case someone is in doubt, why should we merge [1]: we urgently need
MSVC 2015 to support recent Python 3.5, because Python project dropped
support for MSVC 2013 last year, and we are shipping outdated Python
3.3 in LO 5.1 on Windows because of the outdated compiler toolchain on
this platform. Not to mention better C++ standard C++14/C++17 support.

So, what option should we choose to move forward with merging [1]: a)
or b)?

[1] https://gerrit.libreoffice.org/22588
[2] https://gerrit.libreoffice.org/19560