1: \begin{abstract}

2: Broadly speaking the present is a homotopy

3: complement to the book of Giraud, \cite{giraud}, 

4: albeit in a couple of different ways. In the

5: first place there is a representability theorem

6: for maps to a topological champ

7: (the translation stack will

8: be  eschewed

9: \ref{faq5.1})

10: and whence an extremely convenient global atlas,

11: {\it i.e.} the path space, which permits an

12: immediate importation of the familiar definitions

13: of homotopy groups and covering spaces as

14: encountered in elementary text books, \ref{faqR}.

15: In the second place,

16: it provides the adjoint to Giraud's co-homology,

17: {\it i.e.} the homotopy 2-group $\Pi_2$,

18: by way of the 2-Galois theory of

19: covering champs. 

20: In the sufficiently path connected case this is achieved by

21: much the same construction employed in constructing

22: 1-covers, {\it i.e.} quotients of the path space by

23: a groupoid, \ref{faqII}. In the general case, 

24: so inter alia the pro-finite theory appropriate for

25: algebraic geometry, 

26: the development parallels the axiomatic

27: Galois theory of \cite[expose\'e V]{sga1},

28: \ref{faqIII}.

29: The resulting

30: explicit description of the homotopy 2-type

31: can be applied to prove theorems in algebraic

32: geometry: 

33: optimal generalisations to $\Pi_2$

34: (by a very different

35: method, which even gives improvements

36: to the original case) of the Lefschetz theorems (over a

37: locally Noetherian base) of \cite{sga2}, \ref{faq4.5}-\ref{faq4.7},

38: and a counterexample to the extension from 

39: co-homology to homotopy of the smooth base

40: change theorem, \ref{faq4.3}. These limited goals 

41: are achieved, 

42: albeit arguably at the price of obscuring the

43: higher categorical structure,

44: without leaving the 2-category of groupoids.

45: \end{abstract}