[LLVMdev] [Frustration] API breakage

Villmow, Micah Micah.Villmow at amd.com
Wed Jul 13 16:38:49 PDT 2011



> -----Original Message-----
> From: llvmdev-bounces at cs.uiuc.edu [mailto:llvmdev-bounces at cs.uiuc.edu]
> On Behalf Of Chris Lattner
> Sent: Wednesday, July 13, 2011 10:16 AM
> To: Jay Foad
> Cc: llvmdev at cs.uiuc.edu
> Subject: Re: [LLVMdev] [Frustration] API breakage
> 
> 
> On Jul 13, 2011, at 8:09 AM, Jay Foad wrote:
> 
> > On 13 July 2011 15:47, fly language <flylanguage at gmail.com> wrote:
> >> I don't really buy the manpower argument. Updating the release doc
> when
> >> breaking the frigging API is the Right Thing To Do and shouldn't
> take that
> >> long, when done when the change is fresh in memory.
> >
> > I regularly make small API-breaking changes in the name of cleaning
> > things up. Sorry! I'd be happy to update the release notes if folks
> > reckon this is the right thing to do. Would it just mean adding a
> <ul>
> > to the (currently empty) list in docs/ReleaseNotes.html#api_changes ?
> 
> Sure, that makes perfect sense.
> 
> -Chris
[Villmow, Micah] If possible to add my own two cents. I would recommend 
to add not just api changes, but default setting changes. Took us awhile
to figure out that with 2.9, NoAA was set to default instead of what was
in 2.8 and it caused some performance regressions on our side.
> _______________________________________________
> LLVM Developers mailing list
> LLVMdev at cs.uiuc.edu         http://llvm.cs.uiuc.edu
> http://lists.cs.uiuc.edu/mailman/listinfo/llvmdev






More information about the llvm-dev mailing list