RFC INTERNET MESSAGE ACCESS PROTOCOL – VERSION 4rev1, March Canonical URL: ; File formats . [RFC ] IMAP/POP AUTHorize Extension for Simple [RFC ] INTERNET MESSAGE ACCESS PROTOCOL – VERSION 4rev1. RFC List. RFC. Problem Solved. Plan for new Protocol. RFC Internet Message Access Protocol – Version 4. Replaced by See
|Country:||Turks & Caicos Islands|
|Published (Last):||16 January 2011|
|PDF File Size:||11.7 Mb|
|ePub File Size:||16.58 Mb|
|Price:||Free* [*Free Regsitration Required]|
RFC List Every RFC is the work of someone who felt enough pain from the lack of something that they would make the effort to write the document. These problems need to be solved somehow – maybe inherently in a different protocol, maybe explicitly. There’s not even an attempt to include address book, calendaring, etc in this list yet!
Make sure protocol supports simple access to emails.
Ideally not too different – reuse as much as possible. Mixed responses a la Rf would render many more of them moot. Heirarchy separator will either be forced or not exist tree data structure on the wire. An Email Filtering Language and all the extentions Support Sieve – possibly even sieve fragments as “run this filter on this mailbox”.
Specify quota and ACL handling currently different for keywords and annotations. Specify shared vs private “same keyword”, e. Except making LIST extensible and giving tiny bandwidth improvements and tiny server performance improvements There are a few cases covering list andling – it’s a lot more complex than the small amount of actual data involved justifies. In particular, replacing pairs of mutually exclusive options with a “tristate” makes sense.
But this one has a problem, it doesn’t let you sort across folders, which makes it rdc half the solution. Need to fix the sort as well [ http: Less necessary with higher resources available today. Will need an Access Control mechanism. May need rrc consider individual folders vs “entire user” in the context of folder-level ACLs. Will need some 35011 of Quota.
Expand scope to include other objects annotations, keywords, num-folders, etc consistently.
RFC’s – hMailServer – Free open source email server for Microsoft Windows
Expand scope considerably – IDLE is insufficient for many users, because it only monitors one mailbox.
Not using sequence numbers would render many of these moot. Also need to handle the case of “partial IO error” – where data exists but can not be returned immediately for some reason.
Specify one mailbox layout only. Fix insufficient information being returned from commands to synchronise state.
This will need to be addressed in individual sections. Ideally part of a more generic “batch operations” – more discussion required. Ideally composible actions – first characters of decoded value of part “1. See comment – will be mooted.
Avoid having such a complex ABNF – use a structured substrate of some sort. It will be less human-typing-friendly, but more machine friendly. Definitely want token based access so you can authenticate once and then re-connect with the same token, at least during the same “session”. BURL is too complex from a “interacting systems, firewalls and authentication methods” perspective.
RFC – INTERNET MESSAGE ACCESS PROTOCOL – VERSION 4rev1
This is almost a replacement protocol in itself – hopefully it will map almost untouched. The hard bit will be extra things like metadata, condstore, which it doesn’t seem to consider – so it’s incomplete. Explicity allow space for extention.
Ideally, part of a more general “compose commands” – but the ability to store a search result for later manipulation will be needed. An Email Filtering Language.
Support Sieve – possibly even sieve fragments as “run this filter on this mailbox”. Common use case is “add a rule to filter messages matching X into a folder, and also apply that to the messages in INBOX now so existing ones get copied”. Definitely, need to keep this – language negotiation needs to be part of the initial connection. Doesn’t go far enough – particularly across folders.
Also fuzzy searching and partial response is an interesting area. Yet another axis of data – need to merge with keywords and regularise. Was there really a point? Except making LIST extensible and giving tiny bandwidth improvements and tiny server performance improvements.
There are a few cases covering list andling – it’s a lot more complex than the small amount of actual data involved justifies. Massive deadwood cleaning required here. Set a baseline of “required supported features” to get a better experience. Exactly what we’re planning here – everything you can do with Lemonade should be supported – not necessarily in the same way. Either use them as-is, or map into annotation-space.
Internet Message Access Protocol
Make it easy to not only get status data, but to get a list of which folders have changed since last request. Would love to make this more general – sort by a somewhat arbitrary function on the message. Extend more – allow both private and shared “special uses”, because there are contexts where both make sense. Folder listing is one of the major pain points for client authors. Need to fix the sort as well.