Ignore:
Timestamp:
Jun 6, 2009, 8:36:40 PM (11 years ago)
Author:
mikel
Message:

lots more errors get caught now. repeatedly using the emacs utility to request a swank load and connection correctly creates multiple swank connections. It's possible to disconnect from and reconnect to a CCl using a swank server loaded by the load request. It's possible to kill CCL and relaunch it, and reconnect to the new CCL using the emacs request.

the handshake is hairy enough that additional uncaught errors probably remain.

File:
1 edited

Legend:

Unmodified
Added
Removed
  • trunk/source/cocoa-ide/swank-ccl-ide.el

    r12164 r12214  
    2626
    2727(defvar *ccl-swank-output* nil)
    28 ;;; TODO: make this filter function start up a connection to
    29 ;;;       the CCL swank server if it reads a success message,
    30 ;;;       or display an informative error if it reads a
    31 ;;;       failure message
     28
    3229(defun slime-ccl-swank-filter (process string)
    3330  (let* ((status (read string))
    3431         (active? (plist-get status :active)))
     32    (setq *ccl-swank-output* status)
    3533    (if active?
    3634        (let ((port (plist-get status :port)))
    37           (slime-connect *ccl-swank-listener-host* port)))))
     35          (slime-connect *ccl-swank-listener-host* port))
     36        (error "CCL failed to start the swank server. The reason it gave was: '%s'"
     37               (plist-get status :message)))))
    3838
    3939(defvar $emacs-ccl-swank-request-marker "[emacs-ccl-swank-request]")
Note: See TracChangeset for help on using the changeset viewer.