Remove wrong filter order assumption in the README.

Works great! A few core filters also have the weight set to "-10", so making it even lower could make sense?

Please review!

I'm wondering about the README description a bit, as I would have expected the SpamSpan filter to run AFTER "Convert URLs into links" because I thought that filter may also transform email addresses into clickable email addresses.

Could you please test that without having the spamspan filter enabled? Just to be sure what that core filter does and if it only handles URLs?

If it doesn't touch email addresses, I think this is fine! If core uses -10, should we use -11 then to always put it on top?

BTW other modules validate the order to ensure the filtering works as expected, we could add that as minor feature request. Not sure which module does that, I think it's even in a core filter. You may try to drag the filters into crazy order to get the message...

"Convert URLs into links" Does actually convert a simple mail adress like "aaa@bbb.com" into a mailto link <a href="mailto:aaa@bbb.com">aaa@bbb.com</a>!!

So I guess you are right, and the README is wrong? Really weird...

Should I also check the "Convert line breaks into HTML" filter?

"Convert URLs into links" Does actually convert a simple mail adress like "aaa@bbb.com" into a mailto link aaa@bbb.com!!

So I guess you are right, and the README is wrong? Really weird...

Nice, good to know, so we should always put SpamSpan behind that filters weight and correct the README here.

Should I also check the "Convert line breaks into HTML" filter?

No, SpamSpan should be able to work before and after it. I'd tend to remove that from the README, as it's just confusing. For performance reasons it might be a good idea to place it before that one by default, but if the weights are in conflicts with the paragraph above, ignore that.

All done, please review now!

We should emphasize that, in the README, the Spamspan filter should be always configured to run as last.

We could also still initiate the formatter with a very low weight, using the doctrine annotation provided in the class.

I'd be fine with both... will be only for new installs anyway, I don't think we should touch existing sites.

I think we have an agreement on #16, what do you think about #17 @joevagyok?

I agree with both.