Doc: Added note to CG about disable-optimizing

Doc: Added note to CG about disable-optimizing

This is for Tracker 1905

Added sentence to add --disable-optimising for ./autogen.sh or ../configure
when doing regression testing

[Trevor Daniels, 2011-09-21 08:43:09]

LGTM, with one comment

http://codereview.appspot.com/5081048/diff/1/Documentation/contributor/regres...
File Documentation/contributor/regressions.itexi (right):

http://codereview.appspot.com/5081048/diff/1/Documentation/contributor/regres...
Documentation/contributor/regressions.itexi:143: be run with the
@code{--disable-optimising} option.  Then you will need
I think I'd mention just running ./autogen, just in 
case the reader is doing the build for the first time.
But if you leave the configure option in it should be
./configure.

[Trevor Daniels, 2011-09-21 09:01:28]

On 2011/09/21 08:43:09, Trevor Daniels wrote:
> But if you leave the configure option in it should be
> ./configure.

Ah, that was a bit simplistic.  In-tree and out-of-tree
builds are different wrt configure, I think.

[pkx166h, 2011-09-21 09:26:56]

On Wed, Sep 21, 2011 at 10:01 AM,  <tdanielsmusic@googlemail.com> wrote:
> On 2011/09/21 08:43:09, Trevor Daniels wrote:
>>
>> But if you leave the configure option in it should be
>> ./configure.
>
> Ah, that was a bit simplistic.  In-tree and out-of-tree
> builds are different wrt configure, I think.
>
> http://codereview.appspot.com/5081048/
>

Well I wasn't sure even this was the right place to put this
information, we could for instance put this in the section(s) 4.4.1
and 4.4.2 instead where we explicitly talk about ./autogen.sh or
../configure (which as far as I can tell is only referred to when we
use out of tree builds - i.e. I don't see any ./configure, only
../configure.

James


-- 
--
James

[Colin Campbell, 2011-09-21 12:55:55]

http://codereview.appspot.com/5081048/diff/1/Documentation/contributor/regres...
File Documentation/contributor/regressions.itexi (right):

http://codereview.appspot.com/5081048/diff/1/Documentation/contributor/regres...
Documentation/contributor/regressions.itexi:143: be run with the
@code{--disable-optimising} option.  Then you will need
On 2011/09/21 08:43:09, Trevor Daniels wrote:
> I think I'd mention just running ./autogen, just in 
> case the reader is doing the build for the first time.
> But if you leave the configure option in it should be
> ./configure.

I believe the ../configure is appropriate for an out of tree build: it would be
run from a /build directory, which we strongly recommend, no?

[Graham Percival (old account), 2011-09-21 19:26:53]

http://codereview.appspot.com/5081048/diff/1/Documentation/contributor/regres...
File Documentation/contributor/regressions.itexi (right):

http://codereview.appspot.com/5081048/diff/1/Documentation/contributor/regres...
Documentation/contributor/regressions.itexi:143: be run with the
@code{--disable-optimising} option.  Then you will need
On 2011/09/21 12:55:56, Colin Campbell wrote:
> On 2011/09/21 08:43:09, Trevor Daniels wrote:
> > I think I'd mention just running ./autogen, just in 
> > case the reader is doing the build for the first time.
> > But if you leave the configure option in it should be
> > ./configure.
> 
> I believe the ../configure is appropriate for an out of tree build: it would
be
> run from a /build directory, which we strongly recommend, no?

Yes, we very strongly recommend ../configure.  And as far as I'm concerned,
there is no such thing as a 
./autogen.sh
command; there is merely
./autogetn.sh --noconfigure
(followed by mkdir -p build/ ; cd build/ ; ../configure --options)

At this point in the CG, I don't think it's worth worrying about somebody
starting from a totally untouched git tree -- or rather, if anybody _is_ in that
state, they'll know to run autogen.sh first.

[Graham Percival (old account), 2011-09-25 10:18:02]

LGTM, go ahead and push

[pkx166h, 2011-09-25 11:35:08]

Pushed as commit	0290695173284ccf68b08c0e92734d76df7120fd

Graham ok'd this.

James