development release newLISP 9.2.15

[Lutz]

More bugfixes for the upcoming 9.3 release. Source only.

files and changes notes: http://newlisp.org/downloads/development/

Lutz

[cormullion]

assoc, set-assoc (old replace-assoc) and new assoc-set all take mutiple keys for multilevel association lists

That's an interesting bug fix! :) I'd love to try it out, but it's source only ...

[newdep]

Hi lutz,

Scripts that use replace-assoc dont give a warning they just behave strange..

..Are you stripping replace-assoc from 9.3.0...?

Norman.


PS: actualy... set-assoc & replace-assoc both dont work...

[Lutz]

Scripts that use replace-assoc dont give a warning they just behave strange..

can you be more specific? 'replace-assoc' runs the usual tests fine. 'set-assoc' is the renamed 'replace-assoc', but the old naming will stay for a while.

Can you give a specific example where 'replace-assoc' now behaves differently? I want to release 9.3 in a few days, so this would be important to know.

Lutz

[newdep]

See the pinballs.lsp demo... there is replace-assoc in there ;-)

Run it and it wont work..change replace-assoc to set-assoc it wont work eighter? I cant see whats happening ;-)

[Lutz]

Yes, I am seeing this, it seems not to do the increment of the variables, thanks for catching this.

Lutz

[Lutz]

v.9.2.15 has been retracted, I will post a 9.2.16 later

Lutz

[newdep]

Hi Lutz..

Why "string index out of bounds" ? ->

> (set 'a 10)
10
> (set 'b '( "one" "two" "three"))
("one" "two" "three")
> (set 'c '( "high" "low" "vertical"))
("high" "low" "vertical")
> ( (b 2) a (c 0))

string index out of bounds

[Lutz]

Because (c 2) evaluates to a string, the next expression a -> 10 is taken as an index into the string "vertical" and overflows, because vertical only has 8 characters.

Lutz