In terms of readability, a problem is that many aspects of current RDF standards are defined in external documents. In general, this is a great idea, as it eliminates redundancies and creates synergies, but it is problematic for the occasional user. Taking the SPARQL function REPLACE as an example, this is currently defined with reference to XPath fn:replace, but it order for it to work, users need to
understand the SPARQL/RDF specifications,
understand the XPath/XML specifications,
retrieve the relevant bits from the XPath spec,
integrate it conceptually with their interpretation of the SPARQL/RDF specification, and
figure out common implementation details that are not documented in the XPath spec (e.g., double escaping of "\")
The second, third and fourth step in this process are completely unnecessary (other than for formal reasons), and with decreasing relevance of XML standards they even become a major obstacle, so having a compact (maybe even non-normative) description of these aspects in a single, coherent document (https://en.wikibooks.org/wiki/SPARQL/Expressions_and_Functions is quite good at this, actually) will improve readability a lot.
In terms of readability, a problem is that many aspects of current RDF standards are defined in external documents. In general, this is a great idea, as it eliminates redundancies and creates synergies, but it is problematic for the occasional user. Taking the SPARQL function REPLACE as an example, this is currently defined with reference to XPath fn:replace, but it order for it to work, users need to
The second, third and fourth step in this process are completely unnecessary (other than for formal reasons), and with decreasing relevance of XML standards they even become a major obstacle, so having a compact (maybe even non-normative) description of these aspects in a single, coherent document (https://en.wikibooks.org/wiki/SPARQL/Expressions_and_Functions is quite good at this, actually) will improve readability a lot.