[Cuis] A few proposed is: additions

Phil (list) pbpublist at gmail.com
Wed Mar 11 21:30:27 CDT 2015


On Tue, 2015-03-10 at 14:39 -0400, Phil (list) wrote:
> On Tue, 2015-03-10 at 10:49 -0300, Juan Vuletich wrote:
> > Hi Phil,
> > 
> > On 3/10/2015 2:00 AM, Phil (list) wrote:
> > > Juan,
> > >
> > > I just wanted to touch base on some changes we discussed a while back
> > > related to #is: / protocols.  I was wondering if this got lost in the
> > > shuffle or if you decided against doing it?
> > >
> > > Thanks,
> > > Phil
> > 
> > This was loaded as #1313 in June 2012. But the idea didn't really catch 
> > on, and nobody worked on unifying the concept of 'protocols' all over 
> > the system. This was needed, as the aim for Cuis is to try to be 
> > consistent, and use few concepts system wide.
> > 
> > I removed it in #1629 on March 2013, as the advantages were not really 
> > exercised, and it added complexity that didn't seem worth it.
> > 
> > But we can reopen discussion, if you want.
> > 
> 
> Well that's a bummer and unfortunately I've been running with it in 4.0
> since we discussed it so obviously I wasn't active in submitting
> changes.  It's something I really want but if it's not something others
> see the value of I won't push for it in core.  It seemed like a logical
> fit for use with #is: but I can always just implement it using a
> different method for my code.
> 

One thing that would be helpful: can you recommend a way to
programmatically add a class variable to an existing class?  One thing
that I'm recalling is that if a changeset with class overrides is saved
to add them, over time this will become a maintenance chore as other
subsequent upstream changes will be lost unless I'm constantly
re-merging these changes in. (i.e. if my changeset adds a class variable
foo and the class later adds bar, my changeset will override effectively
removing bar)

> > Cheers,
> > Juan Vuletich
> 






More information about the Cuis mailing list