Well, I posted this obviously into the wrong slot. It's more an open question and not an open discussion where I wrongly posted it. Stupid me, next time I'll take more care.
Besides, the problem is solved - at least until it turns up again which can very well happen.
For those who are stumbling upon this because they are like me following the 'mobile first' approach:
The mysterious h..p://1.2.3.13/bmi/ addition to an image link represents a server of my mobile network provider, which attempts to compress all images to reduce general bandwidth use. It's done with a bmi.js and this practice is very controversial. Can it e.g. break your image link as in my case and who knows what malicious else is happening along this way. I am very careful with my images and compression and I hate the idea that third party elements unknown to me are tampering with my creative work.
However - I sorted it by adding the [[++site_url]] tag in front of the relative URL and so forced it to output an absolute link to the image in question. The bmi.crap is still there but at least the link isn't broken.
If you want to know more about this practice here is an article about it:
http://skippy.org.uk/bmi-js-is-not-my-baby/
[ed. note: kdmeloper last edited this post 13 years, 5 months ago.]
☆ A M B ☆