Quick reminder that the coloring in the previous version had semantic value we don't want to lose: it indicated whether a completion came from the local text scope or from the object's namespace.

One problem with the previous one, though, is that it would repeat the same completion if it was in both. In that case, it should remove the duplicates, leaving say only the namespace one. Because otherwise it refuses to 'finish' a completion that appears both locally and in the namespace even when it's the only one left, since both copies are kept making the logic mistakenly think it's non-unique.

Agreed with Fernando,
I'll be back from Japan in 24 hour and a little more. I see There is a
lot to catch up with. Give me time to get better from jet lag and I'll soon
be back in. I'll appreciate if you can point me to important point I'll
miss when skipping through 2 weeks of mails.

Le mardi 30 avril 2013, Fernando Perez a écrit :

Quick reminder that the coloring in the previous version had semantic
value we don't want to lose: it indicated whether a completion came from
the local text scope or from the object's namespace.

One problem with the previous one, though, is that it would repeat the
same completion if it was in both. In that case, it should remove the
duplicates, leaving say only the namespace one. Because otherwise it
refuses to 'finish' a completion that appears both locally and in the
namespace even when it's the only one left, since both copies are kept
making the logic mistakenly think it's non-unique.

—
Reply to this email directly or view it on GitHubhttps://github.com//pull/3238#issuecomment-17205888
.

OK I have styled things in a way that takes into account the different completion types. Styling select options is a horrible mess - most things simply don't work. I think what I have now is a nice balance between being clean, but a good visual difference.