Follow-up Comment #2, bug #99175 (project xrootd):
Well, needless to say there is no definitive standard for "query strings" as
they are called. The string may or may not start with an ampersand.
Typically, they do not but xroot (and some HTTP servers) do allow for it.
There is no discussion of what to do when you get a null field in the query
string (i.e. &&). In xroot's case, it's not that the server looses the values
after the double ampersand but it assigns the wrong field name to that value.
All items with a single ampersand do get properly recorded. The question here
is what should be the appropriate fix (or bypass). Do we fix the clients to
avoid including an extra ampersand? Do we tell developers to never echo back
a leading ampersand? Do we simply change the server to ignore null fields?
Perhaps some combination?
As for the 3.2.6 change, yes, we finally changed he client to include the
opaque information on all requests not just open (which was a historical
omission).
Anyway, please do propose the preferred solution. I suspect the answer will
be to ignore null fields.
_______________________________________________________
Reply to this item at:
<http://savannah.cern.ch/bugs/?99175>
_______________________________________________
Message sent via/by LCG Savannah
http://savannah.cern.ch/
########################################################################
Use REPLY-ALL to reply to list
To unsubscribe from the XROOTD-DEV list, click the following link:
https://listserv.slac.stanford.edu/cgi-bin/wa?SUBED1=XROOTD-DEV&A=1
|