Just from what I have seen. People care if we break a SConstruct from working 
because we have to change the file to make it work.  Having it work better ( ie 
we can remove extra Depends calls which are now redundant) is not an issue and 
are welcome. updating SCons and having it rebuild stuff is not an issue for 
anyone I know of. Most of the time people update SCons and they rebuild from 
scratch, and when they don’t know no one I have seen ever noticed or cared if 
it rebuilt some code one time.

The only case that I would see happening with the scanner fixes is some rare 
case in which SCons rebuild some file it now sees correctly that get modified 
every build for some reason causing the build to never be “up-to-date”. I think 
given what SCons claims we want this to be seen if it was missed before. 

Jason

From: Gary Oberbrunner 
Sent: Tuesday, July 28, 2015 11:57 AM
To: Bill Deegan 
Cc: SCons developer list 
Subject: Re: [Scons-dev] Cross-language support

Hi Bill!  I don't think it's "compatibility breaking" in that existing 
SConscripts will continue to work without change, but it _will_ require (cause) 
a rebuild in many cases, and we do usually pre-announce those changes and call 
them out in the release notes so people with huge projects don't get surprised. 
 (Just to be clear, I'm not as averse to changes that cause rebuilds as Steven 
used to be -- it's sensible to avoid them when possible, but I don't think we 
need to avoid otherwise-sensible changes to avoid rebuilds every now & then.)

-- Gary



--------------------------------------------------------------------------------

  From: "Bill Deegan" <[email protected]>
  To: "SCons developer list" <[email protected]>
  Sent: Tuesday, July 28, 2015 11:56:00 AM
  Subject: Re: [Scons-dev] Cross-language support

  Gary & Dirk,


  Thoughts on whether this change introduces compatibility breaking change?


  -Bill


  On Tue, Jul 28, 2015 at 7:24 AM, Bill Deegan <[email protected]> 
wrote:

    William,


    It seems likely that since the change to scanning behavior will likely 
change many builds (as it's more accurate in tracing dependencies).


    As such I think we should pre-announce it.


    Is it safe to say this change "breaks compatibility"?

    (If you ran a build to completion without change, and reran it you'd get a 
new build, switch to this change and it may rebuild some files)


    -Bill



    On Tue, Jul 28, 2015 at 12:16 AM, William Blevins <[email protected]> 
wrote:

      Once we have finalized the patch, so that the behavioral changes can be 
concretely defined, I will update those two files or should we do a pre-release 
announcement like with the slots changes?


      V/R,

      William


      On Mon, Jul 27, 2015 at 8:34 PM, Bill Deegan <[email protected]> 
wrote:

        William,


        I just got around to doing a thorough read of your pull request and 
added a couple comments.


        Notably c++ doe (in the standard) support and require usage of header 
files with no extension:
        http://en.cppreference.com/w/cpp/header


        Another item is that since this is a change in functionality, 
documentation will need updates.

        And we should probably put a section in the src/CHANGES.txt and 
src/RELEASE.txt


        -Bill


        On Thu, Jul 9, 2015 at 7:18 AM, William Blevins <[email protected]> 
wrote:



          On Thu, Jul 9, 2015 at 5:56 AM, Russel Winder <[email protected]> 
wrote:

            <Likely going off-topic…>

            On Thu, 2015-07-09 at 00:20 -0400, William Blevins wrote:
            > Thanks for responding everyone.  I just wanted a "heart beat" so 
to
            > speak,

            You could always play the start of Dark Side of the Moon ;-)

            > since I wasn't sure how many members were watching the devs list.
            >  I'm not
            > asking anyone to stop what they are doing, but a lot of what I 
have
            > left is
            > requirements related questions.

            Whilst I note every email, I mostly delete and move on due to not
            having enough time to properly contribute.

            > I will hopefully still be able to work on SCons after early
            > September, but
            > I am going to be a little disorganized during the move and culture
            > adjustment.  I will be overseas for a year getting my MSc in Great
            > Britain.

            Just to note that Great Britain is a geographic but not political
            entity, something the ISO committees handing out country codes 
chose to
            forget when trying to solve the UK/Ukraine problem.

            Where will you be studying and living when here?

          University of Sussex in Brighton; approximately Sept 2015 - Sept 2016.



            > Also, I may not have my high-end workstation. I'm still debating
            > whether or
            > not I want to break it down and ship it.

            I guess this depends on cost. It always seems that countries 
shipping
            to UK pay about 0.5 or 0.3 the cost of shipping the same from the 
UK.
            Basically all companies (especially USA ones) charge far more in 
the UK
            for everything than they charge anywhere else in the world.

          Cost plus risk of it getting damaged.  I generally build my own 
workstations, so it's not like shipping X-U server form-factored machines.  I 
will have to dismantle it prior to shipping.  I'm tempted to ship it case less 
and buy another one in Britain because it'll be cheaper than shipping 
(probably).



            --
            Russel.
            
=============================================================================
            Dr Russel Winder      t: +44 20 7585 2200   voip: 
mailto:sip%[email protected]
            41 Buckmaster Road    m: +44 7770 465 077   xmpp: 
[email protected]
            London SW11 1EN, UK   w: www.russel.org.uk  skype: russel_winder

            _______________________________________________
            Scons-dev mailing list
            [email protected]
            https://pairlist2.pair.net/mailman/listinfo/scons-dev




          _______________________________________________
          Scons-dev mailing list
          [email protected]
          https://pairlist2.pair.net/mailman/listinfo/scons-dev




        _______________________________________________
        Scons-dev mailing list
        [email protected]
        https://pairlist2.pair.net/mailman/listinfo/scons-dev




      _______________________________________________
      Scons-dev mailing list
      [email protected]
      https://pairlist2.pair.net/mailman/listinfo/scons-dev






-- 

Gary


--------------------------------------------------------------------------------
_______________________________________________
Scons-dev mailing list
[email protected]
https://pairlist2.pair.net/mailman/listinfo/scons-dev
_______________________________________________
Scons-dev mailing list
[email protected]
https://pairlist2.pair.net/mailman/listinfo/scons-dev

Reply via email to