[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Russ Rew: Re: 961127: NetCDF vs HDF
- Subject: Russ Rew: Re: 961127: NetCDF vs HDF
- Date: Wed, 27 Nov 1996 14:30:27 -0700
Oops, forgot to CC: support-netcdf ...
------- Forwarded Message
Date: Wed, 27 Nov 1996 14:29:28 -0700
From: Russ Rew <address@hidden>
To: Roy Mendelssohn <address@hidden>
Subject: Re: 961127: NetCDF vs HDF
>To: address@hidden
>From: Roy Mendelssohn <address@hidden>
>Subject: NetCDF vs HDF
>Organization: Pacific Fisheries Environmental Group
>Keywords: 199611272016.AA13458
Hi Roy,
> I know awhile ago there was a brief document comparing NetCDF and HDF,
> but now that at the minimum there is a similar multi-file API for
> scientifc datasets, and not really understanding the internal workings
> of either format, I was wondering if there is anything written about, or
> perhaps generate discussion about, the advantages and disadvantages of
> each format given the recent changes.
>
> Specifically we are interested in issues such as size of files created,
> speed of read and writes, speed of access to the data, flexibility etc.
No, sorry, we don't have anything like that that's up-to-date. I wrote
the original comparison, but deleted it because it was difficult to
maintain with both HDF and netCDF adding features and improving in
performance. I downloaded HDF 4.0r2, but haven't got around to running
benchmarks comparing it with netCDF-3.
If you do such a comparison, we would be interested in the results.
Other Possibilities would be to ask NCSA or post a query to the
netcdfgroup mailing list.
In a slightly related development, we're currently working with the NCSA
HDF group to see if it would be practical to share a common format (not
just a common interface) for the next version of HDF and netCDF. We
both agree it would be very desirable if such a convergence was
possible. NetCDF-4 will need a new format anyway to support
enhancements we're committed to provide, and HDF also needs to change
their format to support larger files, so we decided it made sense to try
to design one common format that would meet both of our requirements.
An announcement about this will probably be made sometime in the next
month.
- --Russ
_____________________________________________________________________
Russ Rew UCAR Unidata Program
address@hidden http://www.unidata.ucar.edu
------- End of Forwarded Message