Great Confusion Regarding Legal and Illegal URL Characters

Question

I am researching what characters are illegal in URLs and what to do about them and I am finding little help from my Google searches.  There are two context questions which don't seem to be addressed.

1) Some characters seem to be OK when invoking a browser but not OK when including the URL in an e-mail.

2) Some hits are saying that, e.g., & and ? are illegal but I am pretty sure that that is wrong as ?x= and &y= are used to specify parameters, aren't they? 

Is there one encoding which works when invoking a browser and which works for links in emails? 

In my research I came across HttpUlility.URLEncode and URLPathEncode.  So I got all excited and tried them.  URLEncode did not seem to do anything!  URLPathEncode was at the other extreme, it translated, e.g., http:// to http%3a%2f%2f - yielding a result which is not so human friendly.  

Any guidance?

Thanks,  Bob

 

 

 

Answer 1

I don't know where you looked, but this ain't the place.  All answers to questions regarding the world wide web can be found at the W3 Consortium.

http://www.w3.org/Addressing/  

The link just below is found on the above page.

http://labs.apache.org/webarch/uri/rfc/rfc3986.html

 

Rudy   =8^D

---

Mark the best replies as answers. "Fooling computers since 1971."

http://thesharpercoder.blogspot.com/

Answer 2

Thanks Rudedog2, but I had seen that web page and didn't find it very digestible.  Maybe if I were a lot smarter.  Also it didn't seem to address the problem of what can be pasted (as a link) into an email.  And, since it doesn't have a Windows orientation, doesn't mention Windows services (e.g. HttpUlility.URLEncode and URLPathEncode). 

What I have done is to use HttpUtility.URLPathEncode but only for the parameter values.  For my purposes that seems to provide an acceptable result.

Thanks for your assistance.

Bob