Ludvig Ericson
Ludvig Ericson
That seems to be the case! There really shouldn’t be two code paths for cas and non-cas sets, so if you want to fix this, do it right!- LudvigOn 4...
Interesting. I'll look at this and get back to you in a second.
The way pylibmc does it at the moment is just pass the value on to `libmemcached`, but since it's unclear what `memcached` even _should_ be doing with negative expiration times,...
I agree, negative timeouts should raise an exception.
Sorry been super busy, I’ll get back to you in the next 48 hours. > On 1 Sep 2016, at 15:44, Ed Morley [email protected] wrote: > > @lericson https://github.com/lericson what's...
Ah yes I think it would make sense to check for signedness at the Python C API level; alternatively fixing libmemcached. Whichever you prefer. Sorry, those 48 hours got drawn...
Should be real simple to implement, I don't really have that particular itch to scratch but if you want to take a dive into C Python extensions, feel free!
+1. I fiddled with this for an hour, only to realize that you cannot do this using just rewrites. You should be able to with path-to-regexp, but it seems that...
It seems even if you specify `/base/:path+`, which does capture the subpath correctly, the rewritten URL will have its slashes escaped and so it will fail.
Another interesting note is that if I collapse the "inside" summation into `sum()` over a generator, it is much slower (2x).