[asdf-devel] clisp, logical pathnames, and ASDF

[Faré]

On Mon, Feb 3, 2014 at 1:40 PM, Pascal J. Bourguignon
<pjb at informatimago.com> wrote:
> Faré <fahree at gmail.com> writes:
>
>> On Sun, Feb 2, 2014 at 7:38 PM, Robert P. Goldman <rpgoldman at sift.info> wrote:
>>> I have been grappling with my inability to upgrade clisp, and find
>>> myself confused by the following events in my transcript:
>>>
>>> [17]> (uiop/pathname:ensure-pathname "home:lisp;asdf;")
>>> #P"home:lisp;asdf;"
>>>
>> BEWARE!
>> ENSURE-PATHNAME uses PARSE-UNIX-NAMESTRING, not PARSE-NAMESTRING.
>> In this case, you want PARSE-NAMESTRING (possibly through PATHNAME).
>>
>> (describe (uiop/pathname:ensure-pathname "home:lisp;asdf;"))
>>
>> #P"home:lisp;asdf;" is a pathname, with the following components:
>> NAME = "home:lisp;asdf;"
>> VERSION = :NEWEST
>>
>> (Also, most of the time that you use MERGE-PATHNAMES, you should probably
>> instead be using UIOP:MERGE-PATHNAMES* or UIOP:SUBPATHNAME.)
>>
>>> I am not at all experienced with clisp, and find myself on strange and
>>> shaky ground here.
>>>
>> CLHS pathnames are strange and shaky. UIOP tries to paper over the issues,
>> making them notably less shaky and the price of being a bit stranger.
>
> there's also a difference between clisp -ansi and clisp without -ansi.
>
> You can also test by setting or resetting custom:*ansi* (a symbol macro,
> use ext:letf if you want to change it temporarily).
>
> I always use clisp with -ansi,
>
Indeed, I find that my Ubuntu-provided CLISP binary has custom:*ansi*
nil by default, whereas my self-compiled one has it being T by
default, and that both fail to parse the logical namestring in
-traditional mode and succeed in -ansi mode.

I also note that test/run-tests.sh does explicitly specify -ansi as a
flag when running tests.

—♯ƒ • François-René ÐVB Rideau •Reflection&Cybernethics• http://fare.tunes.org
You don't have to like everything about me, but if you don't love me
the way I am, it's not me you love, only some fantasy of yours.