Is part of gcc3 missing?
Larry Hall (Cygwin)
Wed Nov 3 19:53:00 GMT 2010
On 11/3/2010 3:25 PM, Andy Koppe wrote:
> On 3 November 2010 14:59, Larry Hall (Cygwin) wrote:
>> On 11/3/2010 10:10 AM, Lee Maschmeyer wrote:
>>> Hmm. Is that really the best approach unless absolutely necessary? "That
>>> doesn't work so do something else" has always struck me as a less than
>>> approach to debugging. :-) Is it possible that caml could be repaired so
>>> doesn't depend on GCC4?
>> I'm not sure "repaired" is the right word for this
> I don't think it is. Gcc-4 is Cygwin 1.7's system compiler, so there's
> nothing wrong with (parts of) the ocaml package depending on it.
> But I guess the ability to switch the default compiler back to gcc-3
> should come with a health warning: it may break stuff. Time to get rid
> of the gcc alternatives setup perhaps, and require users to specify
> gcc-3 explicitly if they still want it?
I don't believe this is an issue with which compiler is used. The
issue is simply that the OP is trying to build brltty with gcc-3 while
not rebuilding (or using) dependencies (ocaml) built with gcc-3. I'd
wager that just installing the old Cygwin ocaml packages that were
built with gcc-3 would fix the problem, though like I said in my
reply, I didn't actually test that theory. Of course, that's not a
general recipe for success, since older packages built on gcc-3 may
not always be available. So anyone that wants to build with gcc-3
must be prepared to build all dependencies. Like most software, the
compiler is backward-compatible but isn't forward-compatible. But
you're right. In a nutshell, if you don't consider and address these
issues when using gcc-3 to rebuild packages, "it may break stuff". :-)
Larry Hall http://www.rfk.com
RFK Partners, Inc. (508) 893-9779 - RFK Office
216 Dalton Rd. (508) 893-9889 - FAX
Holliston, MA 01746
> Q: Are you sure?
>> A: Because it reverses the logical flow of conversation.
>>> Q: Why is top posting annoying in email?
Problem reports: http://cygwin.com/problems.html
Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple
More information about the Cygwin