preserving history with breakpoints [Was Re: [squeak-dev] Squeak 4.6 release candidate]

Eliot Miranda eliot.miranda at gmail.com
Sat Jul 4 15:10:35 UTC 2015


Hi David,

On Sat, Jul 4, 2015 at 7:25 AM, Eliot Miranda <eliot.miranda at gmail.com>
wrote:

> Hi Dave,
>
> On Jul 4, 2015, at 7:15 AM, "David T. Lewis" <lewis at mail.msen.com> wrote:
>
> > On Sat, Jul 04, 2015 at 12:10:41AM -0400, David T. Lewis wrote:
> >> On Fri, Jul 03, 2015 at 07:23:27PM -0700, Eliot Miranda wrote:
> >>> Hi Chris, Hi All,
> >>>
> >>> On Thu, Jul 2, 2015 at 12:26 PM, Chris Muller <ma.chris.m at gmail.com>
> wrote:
> >>>
> >>>> We have a release candidate image.
> >>>>
> >>>>  http://ftp.squeak.org/4.6/
> >>>>
> >>>> The new sources file is required.
> >>>>
> >>>> Please test your apps.  This could be the final image unless major
> >>>> issues are uncovered.
> >>>
> >>> One issue that I'm very tempted to address before the final release is
> the
> >>> bug to do with inadvertent editing of a breakpointed method.
> >>
> >> Squeak 4.6 needs to be frozen. This sounds like something that can be
> >> addressed in a next release after 4.6, so lets do that please.
> >
> > I guess that is entirely the wrong thing for me to say, since I don't
> > even know what would be involved in the fix.
> >
> > What I should have said is that it is great that we have the release
> > candidate prepared in time for ECOOP (thanks!), so I think it would be
> > good to avoid any late changes to that release package unless really
> > necessary.
>
> I was about to say that I agree with your first reply even though, having
> slept on it, I woke up realizing the #2 fix above is easy.  So given yiur
> second message I'll implement #2 and leave #1 for after the release.
>
> To get the override in place, create a config for the status quo, commit a
> version of Kernel containing the default CompiledMethod>>#hasBreakpoint,
> create a config, then commit a version of System which restores the
> override.  I guess I was tired last night cuz this is obvious.  Ok time to
> get up, have a cup of tea and do this :)
>

OK, #2 is implemented and committed.  #1, the warning on defining a method
containing a break, can wait until after the release.  I think it's
important to fix #2, though, because destroying method history is a bad
bug.

Cheers!

> Dave
> >
> >

-- 
best,
Eliot
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.squeakfoundation.org/pipermail/squeak-dev/attachments/20150704/9fec186f/attachment.htm


More information about the Squeak-dev mailing list