Dear all,
thanks for your voices! As it seems that the client mode of BaseXHTTP is used by none of you who reacted on my call -- and it's not used in any of the commercial projects we are/have been involved in -- we will most probably remove it in the next official release. Instead, we can already offer a first version of the announced Client Module [1], which you can find in our latest snapshot [2].
Regarding the ZIP proposal, I got 5 reactions; 2 of you said they could easily switch to a new version, and 2 of you asked me to keep the existing proposal alive. As a result, we will probably offer the new proposal (which will first have to be standardized anyway) as an optional alternative, and assign a different prefix to it.
@Pascal: thanks for your sharing your concerns, which I can well understand. There is indeed quite a number of existing features/limitations in BaseX which we would like to revise, but which are still kept alive due to legacy reasons and backward compatibility. Beside that, now and then, we are removing features that (as far as we can find out) are rarely applied in practice. The main reason for that is actually not the file size of the core library, but our limited resources: the more features we provide in parallel, the more time we need to maintain the code base. Some of the feaures on our GitHub issue list have now been listed for nearly one year, which feels frightening... and the removal of obsolete features and a more flexible core (..and some more potential sponsors?) will increase chances that those will be implemented in a foreseeable future.
After all, we'll do our best to take our customer, and all of you participating in the community, as serious as possible; the more feedback we get, the better we'll know how BaseX is used out in the wild.
Cheers, Christian
[1] http://docs.basex.org/wiki/Client_Module [2] http://files.basex.org/releases/latest/