More things to do

[hadrielk]

This is more of a TODO list (and mostly to myself), than an issue for github:

• Clean up the naming scheme for fields in terms of the "shb_userappl", "if_MACaddr", "ns_dnsIP6addr" type names. Either use lowerCamelCase, or underscore_name_scheme. Don't have all combinations.
• We use inconsistent number bases: for block types we use hex numbers, for others we use decimal. Pick one.
• Add a section for "Recommendations for Applications", or some such - for providing guidance on what apps should do when merging, saving only some blocks, anonymizing, etc.
• If we decide to continue on the IETF path, then add an appendix of a (very small) example file in hex-ascii.
• If we decide to continue on the IETF path, then write up the Security Considerations section text. (it's mandatory for IETF RFCs)
• For the XML file itself, the anchor names, spacing between sections/entries, etc., could use a cleanup.

[packetfoo]

• naming scheme: I'm bad with this, using both inconsistently myself, so it's hard for me to decide :-) I'd probably prefer lowerCamelCase as it allows for more Variantions
• number bases: this probably needs to be HEX then, as the block type for the SHB has a meaning in HEX that is lost when translating to decimal. But I'm open for discussion.
• Good idea about application recommendations, maybe even twofold: MUST and SHOULD, because some things like ordering of multiple identical options (e.g. comments) need to be consistent as they do not have timestamps or other ordering information. So an application reading a file has no chance of knowing the correct order otherwise.
• good idea about the example file, let's do that
• I'll have to check the security considerations section requirements, but we should do that anyway, and especially if it's mandatory.

[guyharris]

• number bases: this probably needs to be HEX then, as the block type for the SHB has a meaning in HEX that is lost when translating to decimal. But I'm open for discussion.

I'd vote for hex.

Any objections to hex? If not, I'll update the spec.

[guyharris]

• Clean up the naming scheme for fields in terms of the "shb_userappl", "if_MACaddr", "ns_dnsIP6addr" type names. Either use lowerCamelCase, or underscore_name_scheme. Don't have all combinations.

if_MACaddr and if_EUIaddr aren't CamelCase, they're "initialism/acronym case" - MAC is an initialism (and acronym) for Medi{um,a} Access Control (yes, the IEEE uses both "Medium" and "Media", sometimes even using both in the same document) and Extended Unique Identifier.

Similarly, in ns_dnsIP6addr and the like, "IP" is an initialism for Internet Protocol. The same applies to if_IPv4addr and if_IPv6addr. For "IP", however:

1. we also have nrb_record_ipv4 and nrb_record_ipv6, which don't capitalize "ip";
2. neither ns_dnsIP4addr nor ns_dnsIP6addr have "v" between "IP" and the number;

so we don't have a consistent convention for referring to the two main versions of the Internet Protocol.