4 ms·

Sure, sure. But while it's definitely not in the standard, I do see it all over, and it's something I picked up from reading a lot of other people's lisp. SBCL

by vzcx 9y ago

Sure, sure. But while it's definitely not in the standard, I do see it all over, and it's something I picked up from reading a lot of other people's lisp. SBCL internals, for example. You might not like the convention and chose not to follow along with it, but I would be surprised if, after 18 years of lisp, you had never seen it before, and would choose to misspell an accessor to prevent a clash instead of naming it differently. Edit: ah, I think I understand. I don't mean to imply that all accessors should be named in this way. That would be gross.

kazinator 9y ago
I've never seen this outside of internals. In application code, this is just a smell. To prevent clashes, there are packages. What will you do if three or more modules want the same function name? Tack on %%, %%%, ...
- vzcx 9y ago
  Only in implementation internals? I'm fairly certain I've encountered it in other unexported package symbols in library code, things like struct constructors where you want a "smarter" "make-my-struct" function. But perhaps I misremember. > What will you do if three or more modules want the same function name? Tack on %%, %%%, ... Certainly not! I think your points are all very fair: it would be a smell, and is evidence of a need for improvements at a more structural level. It just smells a little less than a seemingly purposeful misspelling to my nose.
jjwiseman 9y ago
To be explicit, the old Lisp convention I know of, which is not a super common one but one I've definitely used quite a bit, is that you might name a slightly lower-level/more primitive version of a function with a %-prefix. So if we have a user-email function that is just an object slot accessor, We could have %user-email that actually does a database query. I can't remember for sure, but I wouldn't be surprised if I'd seen a double %%-prefix used, too. You don't always want to put those functions in separate packages (traditionally in Lisp, packages are relatively heavyweight--e.g. an entire web server framework might have only one or two packages). Another similar example might be the convention of defining a macro with the name my-macro, and a function that actually implements much of the macro, or is an equivalent of the macro that takes a thnk, that is named my-macro*. (24 years of lisp here--I wouldn't be surprised if the conventions of mainstream lisp during the earliest six years were somewhat different than the following decades.) Edited to add: Now that I think about it, this convention may have been heavily used in the Macintosh Common Lisp community.
- vzcx 9y ago
  Thank you, and thank you to the grandparent. It's quite fantastic to be schooled by old-time lispers (only 6 years of lisp here).
- kazinator 9y ago
  > Macintosh Common Lisp community. That's why I linked to some CCL code (descendant of MCL). I use foo-impl and such for the implementation of foo. If I have a my-macro, which is implemented by an expander function, then I call that function expand-my-macro.

by vzcx 9y ago

kazinator 9y ago

I've never seen this outside of internals. In application code, this is just a smell. To prevent clashes, there are packages. What will you do if three or more modules want the same function name? Tack on %%, %%%, ...

vzcx 9y ago
Only in implementation internals? I'm fairly certain I've encountered it in other unexported package symbols in library code, things like struct constructors where you want a "smarter" "make-my-struct" function. But perhaps I misremember. > What will you do if three or more modules want the same function name? Tack on %%, %%%, ... Certainly not! I think your points are all very fair: it would be a smell, and is evidence of a need for improvements at a more structural level. It just smells a little less than a seemingly purposeful misspelling to my nose.

jjwiseman 9y ago

To be explicit, the old Lisp convention I know of, which is not a super common one but one I've definitely used quite a bit, is that you might name a slightly lower-level/more primitive version of a function with a %-prefix. So if we have a user-email function that is just an object slot accessor, We could have %user-email that actually does a database query. I can't remember for sure, but I wouldn't be surprised if I'd seen a double %%-prefix used, too. You don't always want to put those functions in separate packages (traditionally in Lisp, packages are relatively heavyweight--e.g. an entire web server framework might have only one or two packages). Another similar example might be the convention of defining a macro with the name my-macro, and a function that actually implements much of the macro, or is an equivalent of the macro that takes a thnk, that is named my-macro*. (24 years of lisp here--I wouldn't be surprised if the conventions of mainstream lisp during the earliest six years were somewhat different than the following decades.) Edited to add: Now that I think about it, this convention may have been heavily used in the Macintosh Common Lisp community.

vzcx 9y ago
Thank you, and thank you to the grandparent. It's quite fantastic to be schooled by old-time lispers (only 6 years of lisp here).
kazinator 9y ago
> Macintosh Common Lisp community. That's why I linked to some CCL code (descendant of MCL). I use foo-impl and such for the implementation of foo. If I have a my-macro, which is implemented by an expander function, then I call that function expand-my-macro.