[Openmcl-devel] COMPILE-FILE maybe-bug

[james anderson]

good afternoon;

On 2010-09-30, at 00:32 , Robert Goldman wrote:

> On 9/29/10 Sep 29 -5:02 PM, Gary Byers wrote:
>
>> The fact that this is the first time that I remember user code
>> apparently trying to modify *FEATURES* at load time doesn't
>> mean much of anything, but coupled with the fact that doing
>> so seems like a bad idea my first reaction is to not want
>> to encourage user-level code to do that sort of thing.
>
> I'm actually having a hard time seeing a case where one would want the
> changes /not/ to persist, which shows how opinions can vary!

a cursory search for changes to *features* in the source code for  
several hundred open source lisp systems yields numerous cases in  
which it appears that the author expects the change to be pervasive.

   find .  -name '*.lisp' -exec egrep -i -H -e ".*push.*features" \ 
{\} \; | fgrep -v setf | wc

yields 723 lines in source files which are not mine. in my memory  
these are all cases which run at load time.
without looking further i cannot say whether any expect to be _loaded  
at compile time_, but i would doubt it. a cursory scan of the systems  
yields

  babel
  bordeaux-threads
  cffi
  cl-http
  clsql
  clx
  elephant
  lift
  lisa
  portableaserve
  trivial-features
  trivial-garbage

in each case, the intended pervasive effect is implied by file names  
such as 'depdefs" or "package" and/or the absence of respective  
conditional in the library code itself.