SquareBracketAssociates / Booklet-uFFI

about the new Unified FFI frameworks
14 stars 9 forks source link

Revise the intro of uFFI booklet Chapter 1 #4

Closed tbrunz closed 5 years ago

tbrunz commented 5 years ago

I edited (and in places, revised or extended) Chapter 1 of Guillermo's "version2" branch, as it appeared to be the latest (and it matched the description provided to me in the Users ML). I originally decided to edit only the intro and submit a PR, mainly to make sure I'm on the right track, and allow you provide feedback. Then I decided to continue (and forgot I was going to make a branch for each section; probably not needed anyway). Any/all comments/guidance is appreciated...

guillep commented 5 years ago

Wow this is cool! Thanks :)

I've read through it and I like it very much. I will merge it!

However, the only thing that popped in my mind while looking at it was the comment on the spaces in the function signatures. AFAIK the extra spaces should not be needed, the parser gracefully manages it, have you found a problem with that?

tbrunz commented 5 years ago

I wasn't 100% sure about the spacing. I thought #'funcname( args )' is parsed differently from #'funcname ( args )' in Pharo; i.e., the space would be needed to set off the embedded string array as a separate array element..??

This is a good example of one of those "finer points" that are essential to spell out thoroughly in the guide, because someone new to this will still be forming a model of how it all works in their mind. This is an interesting (and clever) way that "Pharo does it a little different". The key (and unexpected) differences need particular highlighting -- and rationale, too.

I have a few other issues to discuss with you, but no direct way to message you. (Other than the Dev ML.. and who wants to see sausages being made? :^) What do you suggest?? (You have my direct email addr...)

guillep commented 5 years ago

Sure, private email is ok. You can also drop me a message in discord (see http://pharo.org/community), my handle is Guille#6069)