I noticed that the current implementation for web2py uses pickles. That is a design choice. There are pros and cons. Right off of my head, the biggest cons may be retricting cache-use to python, and performance penalties. When I think of all that redis can do, I can not help imagining a better solution - especially for caching query results. All result-sets are flat and simple in nature - before the dal steps in and converts them to row objects. This makes it an ideal candidate for redis. Has anyone thought of this already? A simplistic (naive) solution aould be to store every result in a hash, and stlre all the ids in a sorted set. This way, the result-sef in the cache may be queried by redis, and not necessarily be pulled in an all-or-nothing fasion, improving read-performance and resources dramatically, while opening the possibilities for external non-python processes to access the cache talking to redis directly. It may not be desierable for all use-cases, as there are obviouse security concearns, but for ipc stuff and/or intranet applications (which are a common use-case in the web2py world), this can be most beneficial. What do you say?
-- Resources: - http://web2py.com - http://web2py.com/book (Documentation) - http://github.com/web2py/web2py (Source code) - https://code.google.com/p/web2py/issues/list (Report Issues) --- You received this message because you are subscribed to the Google Groups "web2py-users" group. To unsubscribe from this group and stop receiving emails from it, send an email to web2py+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/groups/opt_out.
