SYMBOL-TO-JS has been renamed to SYMBOL-TO-JS-STRING (patch attached).
Also, the current cl-json web page leads users to the obsolete darcs repo of Parenscript. Guys have switched to git.
Thanks,
Leslie
On 6 Dec 2008, at 13:16, Leslie P. Polzer wrote:
SYMBOL-TO-JS has been renamed to SYMBOL-TO-JS-STRING (patch attached).
Also, the current cl-json web page leads users to the obsolete darcs repo of Parenscript. Guys have switched to git.
There was some talk on this list [1, 2] about altogether removing the dependency on Parenscript. (Indeed, SYMBOL-TO-JS[-STRING] is the only feature that CL-JSON makes recourse to.) I'm in the middle of preparing a set of patches which do that, among other (more substantial) things. But anyway, many thanks for pointing out the change.
Sincerely, - B. Sm.
[1] http://common-lisp.net/pipermail/cl-json-devel/2008-August/ 000168.html [2] http://common-lisp.net/pipermail/cl-json-devel/2008-August/ 000172.html
There was some talk on this list [1, 2] about altogether removing the dependency on Parenscript. (Indeed, SYMBOL-TO-JS[-STRING] is the only feature that CL-JSON makes recourse to.) I'm in the middle of preparing a set of patches which do that, among other (more substantial) things. But anyway, many thanks for pointing out the change.
From what I gather here this will take you at least a few days.
Can you apply my fix? This problem prevents clbuild users from running Weblocks.
Leslie
On Sun, Dec 7, 2008 at 11:04 PM, Leslie P. Polzer sky@viridian-project.de wrote:
There was some talk on this list [1, 2] about altogether removing the dependency on Parenscript. (Indeed, SYMBOL-TO-JS[-STRING] is the only feature that CL-JSON makes recourse to.) I'm in the middle of preparing a set of patches which do that, among other (more substantial) things. But anyway, many thanks for pointing out the change.
From what I gather here this will take you at least a few days.
Can you apply my fix? This problem prevents clbuild users from running Weblocks.
I have now removed the dependency of parenscript, so it should work. I copied some code from an old branch of parenscript, it does not seem to be very optimized so even better code from Boris is welcome.
By the way Boris, I get three errors when I run the tests of the style: SB-MOP:CLASS-SLOTS called on #<STANDARD-CLASS GOO>, which is not yet finalized. Maybe you can take a look at this, then I guess it should be time for a new release soon...
Thanks, Henrik
On 8 Dec 2008, at 02:28, Henrik Hjelte wrote:
By the way Boris, I get three errors when I run the tests of the style: SB-MOP:CLASS-SLOTS called on #<STANDARD-CLASS GOO>, which is not yet finalized. Maybe you can take a look at this, then I guess it should be time for a new release soon...
That's some SBCL. For one reason or another they won't finalize a properly defined class, probably until an instance is initialized. Explicitly calling FINALIZE-INHERITANCE before the relevant tests fixed that. (Please find the dpatch attached.)
Sincerely, - B. Sm.
P.S. It's a pleasure to see you around. I was worrying you had stepped away from the process.
On Mon, Dec 8, 2008 at 1:13 PM, Boris Smilga boris.smilga@gmail.com wrote:
. Explicitly calling FINALIZE-INHERITANCE before the relevant tests fixed that. (Please find the dpatch attached.
That did the trick, applied.
P.S. It's a pleasure to see you around. I was worrying you had stepped away from the process.
I had to do other things for a while, but now I can fire up slime again. Feels great :-)
/Henrik
-
Boris Smilga -
Henrik Hjelte -
Leslie P. Polzer