It sounds like a gating factor to switch to cmake then...

On Wed, Jan 16, 2019 at 10:40 AM Randy Abernethy <randy.aberne...@rx-m.com>
wrote:

> People have maturity expectations associated with a 1.0 release. So
> yes, my vote to do a release might be +1 while my vote for doing the
> same release as 1.0 might be -1 because I do not want to mislead the
> public.
>
> On Tue, Jan 15, 2019 at 7:32 PM James E. King III <jk...@apache.org>
> wrote:
> >
> > To do this we need to retire the autoconf build and make the cmake
> > environment as prolific as autoconf is, and be able to run cross tests on
> > Windows.  That's a lot to ask, and we need to release at least twice in
> the
> > upcoming year, and three times in the next.  No more once-per-year or
> more
> > releases,  We have folks interested and engaged and we need to help them
> > contribute as much as possible.
> >
> > Votes aren't supposed to have conditions - do they? :)
> >
> > - Jim
> >
> > On Tue, Jan 15, 2019 at 9:48 AM Randy Abernethy <
> randy.aberne...@rx-m.com>
> > wrote:
> >
> > > I am very pro the 1.0 moniker on the next release. However I would put
> > > a few key criteria on it. Without these things I would be a strong -1.
> > > Here's my list:
> > >
> > > 1. A single build system and no trace of a duplicate/confusing second
> > > system (e.g. cmake everywhere)
> > > 2. No claims of support for anything that does not have a passing cross
> > > test
> > > 3. TBinaryProtocol support everywhere
> > > 4. A published specification for the RPC protocol
> > > 5. 0 or close to 0 open bug jira issues (there are over 300 right now)
> > >
> > > Each of these is tied to this statement at the top of the Thrift home
> page:
> > >
> > > "The Apache Thrift software framework, for scalable cross-language
> > > services development, combines a software stack with a code generation
> > > engine to build services that work efficiently and seamlessly between
> > > C++, Java, Python, PHP, Ruby, Erlang, Perl, Haskell, C#, Cocoa,
> > > JavaScript, Node.js, Smalltalk, OCaml and Delphi and other languages."
> > >
> > > I would expect a 1.0 project to have few if any known bugs, to be
> > > fairly simple to build, to be specified and to do what is says (cross
> > > platform rpc), which must be born out in tests.
> > >
> > > A 1.0 release is a great goal.
> > >
> > > --Randy
> > >
> > > On Tue, Jan 15, 2019 at 6:27 AM James E. King III <jk...@apache.org>
> > > wrote:
> > > >
> > > > I'd like us to consider the next version number to be 1.0.  The
> project
> > > is
> > > > mature enough, and some folks won't want a version 0.13.  There are
> > > already
> > > > a number of accumulated breaking changes in interfaces of the C++,
> > > > JavaScript, and Java libraries.  C++ especially, with the break away
> from
> > > > C++03 and boost as a link-time dependency has allowed us to change
> our
> > > code
> > > > significantly interface-wise.  In js/node.js we not have correct
> 64-bit
> > > > integer handling.  Of course, the wire protocol is still backwards
> > > > compatible.  None of that has changed (not to my awareness).  These
> > > changes
> > > > are documented in the top level CHANGES.md and in each language's
> > > README.md
> > > > file.
> > > >
> > > > Let's vote.
> > > >
> > > > [ ] +1 Next version number is 1.0.
> > > > [ ] 0 Don't care
> > > > [ ] -1 Next version number is 0.13.
> > > >
> > > > Voting ends in 72 hours, Friday January 18 at 13:00 UTC.
> > > >
> > > > Thanks,
> > > >
> > > > Jim
> > >
> > >
> > >
> > > --
> > >
> > > --
> > > Randy Abernethy
> > > Managing Partner
> > > RX-M, LLC
> > > randy.aberne...@rx-m.com
> > > o 415-800-2922
> > > c 415-624-6447
> > >
>
>
>
> --
>
> --
> Randy Abernethy
> Managing Partner
> RX-M, LLC
> randy.aberne...@rx-m.com
> o 415-800-2922
> c 415-624-6447
>

Reply via email to