[media-types] Re: [IANA #1444997] application/vnd.hd fgroup.hdf5 registration request
Darrel Miller <[email protected]>
| Newsgroups | gmane.ietf.types |
|---|---|
| Message-ID | <PH7PR01MB80998A30EFFD09AD4705742AA343A@PH7PR01MB8099.prod.exchangelabs.com> |
Hi Amanda, Looks good to me. One minor item: 1. Under "Additional information", Object Identifiers is listed as "non" - should be "None". With that correction, this is approved. Darrel ________________________________ From: Amanda Baber via RT <[email protected]> Sent: Tuesday, March 10, 2026 1:41 AM Cc: Darrel Miller <[email protected]>; [email protected] <[email protected]> Subject: [IANA #1444997] application/vnd.hdfgroup.hdf5 registration request Hi Darrel, Sending a reminder for this request from February 26th. #1 of 2, but I didn't manage to copy this first one to the list. thanks, Amanda On Thu Feb 26 23:46:18 2026, amanda.baber wrote: > Hi Darrel, > > Could you review this one for us by March 12th? This is #1 of 2. > > thanks, > Amanda > > ===== > > Name: Gerd Heber > > Email: [email protected] > > Media type name: application > > Media subtype name: vnd.hdfgroup.hdf5 > > Required parameters: N/A > > Optional parameters: N/A > > Encoding considerations: binary > > Security considerations: HDF5 is a general-purpose binary container > for arbitrary datasets and metadata. Implementations should treat > files as untrusted input and defend against: > > - vulnerabilities in the HDF5 library or dependent codecs/filters, > - resource exhaustion (very large datasets, extreme chunking, > many objects, pathological metadata), > - decompression bombs (high compression ratio data), > - unintended external file access if external links or external > storage > are honored, > - execution risks associated with dynamically loaded filter plugins > (if enabled by the application environment). > > HDF5 does not provide built-in content confidentiality or integrity > (no inherent signing/encryption); such services must be layered > externally. > > Interoperability considerations: Filename extensions are not a > reliable indicator of the underlying format: the extension “.hdf” is > commonly used for both HDF4 and HDF5 content (and may also be used for > other HDF-family files). Implementations SHOULD identify content by > inspecting the on-disk format signature rather than relying on > extensions. HDF4 files begin with the 4‑byte header signature > 0x0E031301 at offset 0. HDF5 files contain the 8‑byte signature > 0x894844460D0A1A0A at offset 0, or at offset 512 and successive > powers-of-two multiples (to allow an optional user block). If the > expected signature is not present, the content SHOULD be treated as > unknown or in a different format rather than inferred from the > filename. > > Published specification: HDF5 File Format Specification (The HDF > Group). > > https://support.hdfgroup.org/documentation/hdf5/latest/_f_m_t4.html > > Applications which use this media: Scientific/engineering data tools > using the HDF5 library, including HDFView and many language bindings. > > Fragment identifier considerations: none > > Restrictions on usage: none > > Provisional registration? (standards tree only): No > > Additional information: > > 1. Deprecated alias names for this type: application/x-hdf5 (historic, > unregistered) > 2. Magic number(s): HDF5 format signature (8 bytes): Hex: 89 48 44 46 > 0D 0A 1A 0A ASCII C: \211HDF\r\n\032\n Location: The HDF5 superblock > signature may appear at offset 0, or at offset 512 and successive > powers-of-two multiples (0, 512, 1024, 2048, ...), to allow a user > block/wrapper. > 3. File extension(s): .h5 (common), .hdf5 (also common) > 4. Macintosh file type code: none known > 5. Object Identifiers: non > > Person to contact for further information: > > 1. Name: The HDF Group > 2. Email: [email protected] > > Intended usage: COMMON > > Author/Change controller: The HDF Group _______________________________________________ media-types mailing list -- [email protected] To unsubscribe send an email to [email protected]