[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [h5md-user] Dataset layouts
From: |
Pierre de Buyl |
Subject: |
Re: [h5md-user] Dataset layouts |
Date: |
Wed, 1 Jul 2015 16:18:40 +0200 |
User-agent: |
Mutt/1.5.23 (2014-03-12) |
Hi,
On Wed, Jul 01, 2015 at 10:43:37AM +0200, Felix Höfling wrote:
> Am 24.06.2015, 22:55 Uhr, schrieb Peter Colberg <address@hidden>:
>
> If I understand your examples correctly, the issue of parallel
> reading/writing is only an issue within the same application using MPI. If
> the programme (not the file format) is implemented consistently, there is no
> problem. For me, there is no need to put more restrictions in the H5MD
> specification. We should keep it as simple/minimal as possible, also to
> avoid conflicting specifications.
>
> However, it is a good idea to share your experience. I think that the
> "implementation" section on the H5MD web page is the perfect place for this.
> Would this meet your concerns?
> http://nongnu.org/h5md/implementation.html#compact-datasets
Even though I see the benefits to the storage change, I doesn't change in any
way the code one has to write to read the data and as such, I am also in favor
of putting these aspects in "implementation".
Pierre