[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: :accessor-prefix bug
- To: MBM@CAF.MIT.EDU
- Subject: Re: :accessor-prefix bug
- From: Gregor.pa@Xerox.COM
- Date: Sat, 3 Oct 87 13:08 PDT
- Cc: CommonLoops.pa@Xerox.COM
- In-reply-to: <2769195635-16374562@CHAPLIN>
- Line-fold: no
Date: Fri, 2 Oct 87 17:00:35 EDT
From: MBM@CAF.MIT.EDU
Date: 30 Sep 87 11:10 EDT
From: Rick.Busdiecker@H.GP.CS.CMU.EDU
Subject: Re: :accessor-prefix bug
I would think that the correct Common Lisp way to deal with the
question would be to handle it the same way that defstruct handles
conc-name. This first proposal seems like the right way to go. Are
the certain circumstances that you mention also problematic for
defstruct, or are the problems unique to defclass?
Has this question been answered? If so, I missed it.
--mike
After OOPSLA, I will send a more detailed reply to this. The short form
of the answer is that the current mechanism used by defstruct has the
package problems described in an earlier message. defstruct interns its
constructed symbols in the package which is the value of the dynamic
variable *package* at <<some not very well specified time>>. This can
lead to surprising results under certain not very well defined
circumstances.
-------