This issue tracker has been migrated to GitHub, and is currently read-only.
For more information, see the GitHub FAQs in the Python's Developer Guide.

Author loewis
Recipients
Date 2006-11-19.10:23:59
SpamBayes Score
Marked as misclassified
Message-id
In-reply-to
Content
The patch looks good, however, I have a few questions:

- what is the purpose of abstracting fault code constants? For interoperability, I think they should always have the specified values. Furthermore, if register_fault_interop_spec isn't called, it will give 1 as the fault code if some of these faults occurs, which is bad.

- why is system.getCapabilities registered in register_fault_interop_spec? That sounds backwards: there should be a register_get_capabilities function, or better yet, this should get registered with register_introspection_functions.

- if you agree to these changes so far, I think register_fault_interop_spec should go away, or invoked by default. What is the purpose of making it optional (especially since raising the faults is not optional)?

- Please provide documentation changes (i.e. a patch to Doc/lib/libsimplexmlrpc.tex, and Doc/lib/libxmlrpc.tex)

- I think ParseError should support "nested" exceptions: the original exception should be included into the ParseError object, in case people want to know the details of the underlying parser error.
History
Date User Action Args
2007-08-23 15:44:38adminlinkissue1360243 messages
2007-08-23 15:44:38admincreate