Hi @vitamind. I guess that people around here wouldn't objected even if you left your question - Reference in a comment of a deleted question - on meta (without deleting it).

It's probably ok either way - but since it is now deleted, you won't get additional response there. (Some 10k+ users could be able to say whether there were also some additional comments other than what I found in SEDE.)

Just in case somebody notices this here in chat, it was about a deleted question: mathoverflow.net/questions/406605/…

> However I remember reading a comment about a paper where $li(xα+iβ)$ is defined. I don't have enough reputation to see it. I would very much appreciate a comment with the reference of a user who is able to see it.

I won't copy my answer in full, but I basically just looked at the comments based on a SEDE query: data.stackexchange.com/mathoverflow/query/1340978/…

The most relevant comment seems to be this one:

> Ultimately ${\rm li}(x^\rho)$ is just notation. That it is not literally a composition is similar to a nonvanishing holomorphic function $f(z)$ on a disc (say) having a logarithm that is denoted $\log f(z)$ but is not defined as a composition of ""$\log$"" and $f$; the definition of $\log f(z)$ is more technical. In Narkiewicz's ""Development of Prime Number Theory from Euclid to Hardy and Littlewood"", ...

> at the top of p. 146 the notation ${\rm li}(x^{\alpha+\beta i})$ is defined for $x > 1$ and real $\alpha$ and $\beta$. (The explicit formula you ask about is on pp. 145, 193-194, and 287.)

I will add also Google Books link to the page 146 of Narkiewicz's book.