Summary: | JJ Incorrect error codes in error() calls in KIO slaves lead to error message garbage | ||
---|---|---|---|
Product: | [Frameworks and Libraries] kdelibs | Reporter: | George Kiagiadakis <mail> |
Component: | general | Assignee: | kdelibs bugs <kdelibs-bugs> |
Status: | RESOLVED FIXED | ||
Severity: | normal | CC: | cullmann, egorov |
Priority: | NOR | Keywords: | junior-jobs |
Version: | 4.1 | ||
Target Milestone: | --- | ||
Platform: | Compiled Sources | ||
OS: | Linux | ||
Latest Commit: | Version Fixed In: |
Description
George Kiagiadakis
2008-12-03 19:07:37 UTC
This is a general KIO problem, and not slave-specific (although many slaves have their own incorrect error messages. In particular, the host not found message comes from TCPSlaveBase, I think. The problem is that the slaves (and SlaveBase and TCPSlaveBase) call error() with the incorrect error ID, which is not compatible with KIO::buildErrorString(). See the documentation of SlaveBase::error(). This needs somebody with a bit of time on his hands to wade through all error() calls in the slaves and in SlaveBase/TCPSlaveBase (there are many) and fix them. Is this bug still here? It is not reproducible anymore. It looks like TCPSlaveBase uses correct error IDs with "host: socketError" as error string. Calls to error() from other slaves looks correct too. Then let's close it. |