Re: cors testing build

2012-12-01 Thread Noah Slater
On 13 November 2012 22:44, Randall Leeds randall.le...@gmail.com wrote: https://issues.apache.org/jira/browse/COUCHDB-1602 Is this a blocker for 1.3? -- NS

Re: cors testing build

2012-12-01 Thread Robert Newson
No, but problem when we support 1.8.5 Sent from the ocean floor On 1 Dec 2012, at 19:49, Noah Slater nsla...@apache.org wrote: On 13 November 2012 22:44, Randall Leeds randall.le...@gmail.com wrote: https://issues.apache.org/jira/browse/COUCHDB-1602 Is this a blocker for 1.3? -- NS

Re: cors testing build

2012-12-01 Thread Randall Leeds
It presents a real problem for packaging on debian unstable. On Sat, Dec 1, 2012 at 11:58 AM, Robert Newson robert.new...@gmail.comwrote: No, but problem when we support 1.8.5 Sent from the ocean floor On 1 Dec 2012, at 19:49, Noah Slater nsla...@apache.org wrote: On 13 November 2012

Re: cors testing build

2012-12-01 Thread Noah Slater
If we get a fix, Debian can include a patch. On 1 December 2012 21:52, Randall Leeds randall.le...@gmail.com wrote: It presents a real problem for packaging on debian unstable. On Sat, Dec 1, 2012 at 11:58 AM, Robert Newson robert.new...@gmail.com wrote: No, but problem when we support

Re: cors testing build

2012-11-13 Thread Randall Leeds
Issue-ified: https://issues.apache.org/jira/browse/COUCHDB-1602 On Tue, Nov 13, 2012 at 2:41 PM, Randall Leeds randall.le...@gmail.comwrote: I bisected mozilla-central and believe I found the commit that removed that operator overload[0]. Looks like we have a little more work to do before

Re: cors testing build

2012-11-12 Thread Dave Cottlehuber
On 12 November 2012 00:06, john.tiger john.tigernas...@gmail.com wrote: removed the xulrunner stuff - ran configure with libmozjs (as per wiki) - still same newer version error - this is on Debian Testing amd64? i386? I'll try with 7.0 b3 amd64 here. I see that there was a lot of discussion

Re: cors testing build

2012-11-12 Thread john.tiger
On 11/12/2012 10:59 AM, Dave Cottlehuber wrote: On 12 November 2012 00:06, john.tiger john.tigernas...@gmail.com wrote: removed the xulrunner stuff - ran configure with libmozjs (as per wiki) - still same newer version error - this is on Debian Testing amd64? i386? i386 but hopefully w/

Re: cors testing build

2012-11-12 Thread Randall Leeds
./configure --enable-js-trunk The rationale is that in the newest builds of spidermonkey (actually post-1.8.5, which Debian ships in testing), a bug was corrected which was previously tolerated by the interpreter. The result is that our typical map/reduce examples break. If I recall correctly,

Re: cors testing build

2012-11-12 Thread john.tiger
On 11/12/2012 03:36 PM, Randall Leeds wrote: ./configure --enable-js-trunk thks on debian testing i386 configure runs okay but make breaks couch_js/util.c: In function ‘couch_print’: couch_js/util.c:216:33: error: invalid operands to binary == (have ‘jsval’ and ‘jsval’) make[4]: ***

Re: cors testing build

2012-11-11 Thread Dave Cottlehuber
On 11 November 2012 21:36, john.tiger john.tigernas...@gmail.com wrote: just went to compile on another debian testing machine I'm getting that spider monkey version is too new ? xulrunner-devel-10.0 that same version formerly compiled the stable couchdb package. maybe something else is at

Re: cors testing build

2012-11-11 Thread john.tiger
On 11/11/2012 02:33 PM, Dave Cottlehuber wrote: On 11 November 2012 21:36, john.tiger john.tigernas...@gmail.com wrote: just went to compile on another debian testing machine I'm getting that spider monkey version is too new ? xulrunner-devel-10.0 that same version formerly compiled the