Well, the sad thing is that there has been so many changes that my
attempts to merge my existing distribution and the cvs checkout of today
produced much too many differences, most of which are due to
re-indenting... one file had 121 differences...
I think I can only give up.
As far as I understand, the security exception issue with applets is
fixed as you're having a (Apache Licensed, thanks) Base64 class in there.
So my only concerns are now:
-> UTF-8 aware exchange of string values, is this done ?
(this included modifying XmlRpc.java and MinML.java)
Wishes:
-> "saxon-aelfred" as "com.icl.saxon.aelfred.SAXDriver" in the parser
list.
-> still make a string element when writing a string in the methodCall
or methodResult so that it is most portable even though the specs don't
require it. I.e.:
startElement("string");
write(obj.toString());
endElement("string");
-> Have a dynamic check of character values to decide automatically
between UTF-8 and ISO-8859-1 (this is what I had made, I think you're
sticking to XmlRpc.setEncoding() which raises the problem of using
XmlRpc with, on one side, non-UTF-8 aware server and UTF-8 aware servers
at the same time.
An alternative would be to set the encoding in the XmlRpcClient(Lite).
-> and of course know that MinML support UTF-8 at least (I had submitted
the patches for this, was a bit dirty indeed.
Paul
On Vendredi, mars 15, 2002, at 06:02 , Daniel Rall wrote:
> Daniel Rall <[EMAIL PROTECTED]> writes:
>
>> It didn't come out quite right. It looks like you took your modified
>> version, checked out a new copy from CVS, and did a file system diff
>> between the two. This makes your changes impossible to decipher, as
>> it's impossible to separate the regressions in the diff from any
>> forward progress (CVS has changed quite a bit since you got your
>> copy). What would be useful to us is if you made a back up of your
>> working copy, then did a `cvs update` on it, and merged in the changes
>> that you're missing from the repository (for instance, the Base64
>> class has been replaced entirely).
>
> And then created a unidiff from your new working copy via
> `cvs diff -ub`.
>
> Thanks, Dan
>
