2.1.61?

classic Classic list List threaded Threaded
2 messages Options
Reply | Threaded
Open this post in threaded view
|

2.1.61?

John W. Eaton-6
Unless anyone has objections, I'm planning to make the 2.1.61 snapshot
today.  There are a couple of things I'd still like to add, but they
may take some time and I think a new snapshot with all the recent bug
fixes is more important.  Does anyone know of any bugs that remain and
are critical to fix before I make a new snapshot?

The two things I know about are:

  * Comparison operations for 64-bit integers will not work correctly
    for large values (> 2^53) because we are converting to double to
    do the conversions.

  * Things like [int32(1), int16(1)] will fail.  Concatenation
    operations like this should return an object of the smaller type
    (int16 in this case).

Comments?

Thanks,

jwe


Reply | Threaded
Open this post in threaded view
|

Re: 2.1.61?

David Bateman-3
Dapr├Ęs John W. Eaton <[hidden email]> (le 05/11/2004):

> Unless anyone has objections, I'm planning to make the 2.1.61 snapshot
> today.  There are a couple of things I'd still like to add, but they
> may take some time and I think a new snapshot with all the recent bug
> fixes is more important.  Does anyone know of any bugs that remain and
> are critical to fix before I make a new snapshot?
>
> The two things I know about are:
>
>   * Comparison operations for 64-bit integers will not work correctly
>     for large values (> 2^53) because we are converting to double to
>     do the conversions.
>
>   * Things like [int32(1), int16(1)] will fail.  Concatenation
>     operations like this should return an object of the smaller type
>     (int16 in this case).

There was the issue of passing the first arg to the concat function as
non-const so that it could be used direct, like

NDArray
concat (NDArray& ra, const NDArray& rb, const Array<int>& ra_idx)
{
  if (rb.numel () > 0)
    ra.insert (rb, ra_idx);
  return ra;
}

rather than what is current done that makes an unecessary copy. This is
a pretty trival change, which you asked me to do but I haven't had the
time, mainly due to the fact that it implies a complete recompile. Its
probably worth having it for 2.1.61 though, especially if you are going
to call 2.1.61 the testing version.

Regards
David

--
David Bateman                                [hidden email]
Motorola CRM                                 +33 1 69 35 48 04 (Ph)
Parc Les Algorithmes, Commune de St Aubin    +33 1 69 35 77 01 (Fax)
91193 Gif-Sur-Yvette FRANCE

The information contained in this communication has been classified as:

[x] General Business Information
[ ] Motorola Internal Use Only
[ ] Motorola Confidential Proprietary