Re: [lmi] Empty xml child text nodes [Was: product editor patch]

[Greg Chicares]

On 2012-02-27 14:23Z, Václav Slavík wrote:
> On 24 Feb 2012, at 14:22, Greg Chicares wrote:
>> And then I rethought it. The real issue isn't that we add element
>> nodes whose content is empty; it's that we create such nodes in
>> the first place. So I'm thinking of applying the 'xmlwrapp' patch
>> below, which I believe will do exactly what we want; can you see
>> any reason why that would be a bad idea?
> 
> No, such behavior is desirable. It's not backward compatible (some
> xmlwrapp-using code may rely on the fact that a child text node is
> always created), but I think the incompatibility is extremely minor
> and will apply this patch on xmlwrapp proper.

I doubt that lmi relies on the existence of empty child nodes, but
it's not trivial to prove; if that incompatibility arises, will an
exception be thrown? (I glanced at enough of the code to see that
it wouldn't be trivial for me to answer that latter question either,
but I thought you might just know.)

> BTW, if you're looking into xmlwrapp-related things, it's probably
> worth upgrading to 0.6.2. Or maybe this would be a good time for me
> to release 0.7.0 (with your patch included) so that you can use
> unpatched xmlwrapp once again?

Thanks, but don't release anything just for my sake. Despite the
drawbacks you might perceive with my handwritten makefiles, they
do accommodate patches, and I've already prepared a patch for this;
it would be more psychologically satisfying to me to commit it than
to erase it.

> Here are xmlwrapp changes since 0.6.0 for reference:
> https://github.com/vslavik/xmlwrapp/blob/master/NEWS

Is that xmlwrapp's primary residence now? I ask because a large
corporation that uses lmi has started blocking sourceforge.net,
and we're asking them to unblock it; should we include github.com,
too, or do you plan to mirror new releases on sf?