Ludvig Ericson

There is an earlier discussion about this and how the API would look. Do you have any suggestions? How do we keep the API from becoming too crowded?

Did it cause errors? - Ludvig > On 12 Nov 2018, at 10:29, Chow Loong Jin wrote: > > While comparing python-memcached and pylibmc behaviours, I found that python-memcached was...

I see. Maybe we should reintroduce bool storage but as int-with-extra flag. Going through the pickle machinery for something so simple seems... counterproductive, if you see my reasoning. - Ludvig...

I made a test case for this in the test suite; however, as it stands, it is sadly _not_ something we can fix on our side of things. It seems...

Hello @henning! To my knowledge no, there is no communication with upstream going on about this. It might prove instructive to try to look at what the intended usage of...

Hm, yes that is very confusing. I did know about this behavior, but libmemcached doesn't actually document this behavior either. IMO this is pretty weird design, I agree that we...

Ping @BrianAker as this is relevant to libmemcached as well.

We could probably check if the object given to incr_multi is a mapping or a sequence. However, I dislike the idea of having a self-contradictory API -- what does mc.incr_multi({'key':...

No, that's not a good idea. A mapping would be better for such a purpose, I think. ``` python mc.incr_multi(['a', 'b'], delta=2) mc.incr_multi({'a': 2, 'b': 3}) ``` These two are...

This is a very interesting feature! Perhaps we could make a more intuitive API than to extend get/set and friends, as this breaks new ground. ``` py with mc.master_key("somekey"): mc.set_multi(dict.fromkeys("abcdef",...