This is the mail archive of the cygwin-developers@cygwin.com mailing list for the Cygwin project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: Cygdaemon - planning


Corinna Vinschen wrote:

On Thu, Jul 03, 2003 at 09:59:04AM -0400, Christopher Faylor wrote:

On Thu, Jul 03, 2003 at 11:17:54AM +0200, Corinna Vinschen wrote:

Do we want this?  Wouldn't it be better to have a separate export library
for the functions only available with cygserver?

In this scenario, we have always gone with "what does UNIX do?" I think it is pretty likely that the main users of this code will be existing programs which expect not to have to look in other libraries to find Sys V shm and semaphores. So, putting the exports in some other library will result in configury changes for existing programs which is something that cygwin tries to avoid.


But if you want such a package to work w/o cygserver, you would have to
change the configury, too, with the additional disadvantage that the
package would use cygserver automatically unintentionally as long as the
configurey hasn't been changed.

The question is, what is worse?


Can't this be solved by packaging?  If the package was configured
assuming the existence of cygserver, seems to me that making the package
depend on a cygserver package is enough to make sure that it's clear to
the person installing what's required.  If necessary, maintainers could
create a cygserver and a non-cygserver version, assuming they're into
configure hacking. ;-)  I'm not sure this is ideal either but it would
address the issue in some form.  Or am I missing something?

Larry


-- Larry Hall http://www.rfk.com RFK Partners, Inc. (508) 893-9779 - RFK Office 838 Washington Street (508) 893-9889 - FAX Holliston, MA 01746


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]