|Re: [eigen] About Complex numbers Todo|
[ Thread Index |
| More lists.tuxfamily.org/eigen Archives
- To: eigen@xxxxxxxxxxxxxxxxxxx
- Subject: Re: [eigen] About Complex numbers Todo
- From: Gael Guennebaud <gael.guennebaud@xxxxxxxxx>
- Date: Fri, 20 Aug 2010 12:33:37 +0200
- Dkim-signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=gamma; h=domainkey-signature:received:mime-version:received:in-reply-to :references:from:date:message-id:subject:to:content-type :content-transfer-encoding; bh=1liaG7lts6nFz7KewCizWCeVgnnGSjFz6Y1BGftzZkM=; b=KHmG/4v/idLzJVjH8oZTOGIOFVyxQPocSgKHKS+pmwB6G6846UN5gbFhhKD72F7fTb G60Zchqv0tKVE7MKGF5nDB7tWtisYg2V1/zwdeD3fY9gq/4plWffXycb9a3Yimz4W7MM wgW62mhoz6W22BFVDszym9li4ZM2Rzp59OmnQ=
- Domainkey-signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=mime-version:in-reply-to:references:from:date:message-id:subject:to :content-type:content-transfer-encoding; b=QTUODmK+jUTBRkqxZBDM6qabLyCFlnbQtdM2O+IxjaraPQpPu/rjQ7EPVWxEMPq098 2FkfphgFdgKwq8uJz4lnJi79QzmzqpX/rEkhmZzc9pES2h4Vf59jXVLXjWyT3+BLDXq/ YQ6myO6JNz6iG3ESyer6WdcDB8sn6tXQYbUZM=
actually we already rely on that, and this todo item have been
unofficially dropped but we forgot to remove it from the todo page.
I'll try to do a cleaning pass on this page. thanks.
On Fri, Aug 20, 2010 at 11:09 AM, Bastien ROUCARIES
> On Fri, Aug 20, 2010 at 10:41 AM, Carlos Becker <carlosbecker@xxxxxxxxx> wrote:
>> Hi Bastien. If I get things right then there might be two issues with this:
>> first it is only for C++0x, so older C++ versions would not be supported..
> In theory yes, but in pratice no. All the major compiler team are part
> of the standard board. And if g++ does not support this we could open
> soon a bug report.
>> Furthermore, and if I remember correctly, complex numbers are stored in
>> memory as an array, even though it might be hidden by a struct or class.
> That is the point. At least it "array storage layout which makes
> many optimizations unsafe" is not true because standard board has
> agree since 2004 that the next standard will use implicitlty this
> layout. So the layout is guaranted by pratice (see the remark in the
> standard rationnale) in the current state of standard and officially
> for the next standard.
> And BTW c++ user should use recent compiler see for instance the bug :