james-server-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Danny Angus" <danny.an...@gmail.com>
Subject [VOTE] Non compliance disclaimer (was: Change in policy to for JAMES to violate RFCs)
Date Thu, 05 Oct 2006 12:33:44 GMT
All,
Please indicate your vote in the usual way (+1,0,-1) for the following:


The standards compliance policy statement on
http://james.apache.org/server/design_objectives.html is a pragmatic
one.
However the idea that the documentation be hidden is flawed, as it
highlights neither the feature nor the issue.

I believe that it should be highly visible, and contain a simple disclaimer.

I therfore propose that the statement be extended to read as follows,
and linked to the code standards at
http://james.apache.org/code-standards.html

the proposed statement can be found in full at:
http://wiki.apache.org/james/StandardsComplianceStatement

and reads as follows:

standards compliance

This statement occasionally uses terms that appear in capital letters.
When the terms "MUST", "SHOULD", "RECOMMENDED", "MUST NOT", "SHOULD
NOT", and "MAY" appear capitalized, they are being used to indicate
particular requirements of this specification. A discussion of the
meanings of these terms appears in [RFC2119].

It is the existence of published "open" standards which allows
independant teams to develop interoperable software.

James attempts to support a number of these standards (most of which are
IETF RFC's) and in the areas covered by these standards the published
standard is our requirements document.

This sometimes leads to confusion where behaviour is not the subject of
a relevant standard, or conflict where common (de-facto) behaviour is
actually at odds with a supported standard.

We believe that it is our responsibility to adhere to the published
standard. If we allow our implementation to deviate it means that we are
tacitly encouraging the situation whereby interoperability is no longer
guarenteed by standards compliance alone, but also requires access to
undocumented and possibly even commercially licenced technology. There
is no easy route for a newcomer to aquire these secrets, and
interoperabilty becomes something only available to the elite.

The James policy for issues of non-compliance tries to tread the fine
line between a pragmatic acceptance of differing interpretation and
defects in implmenentations of the RFC's and an evangelical defence of
open standards as the key to freedom of interoperation.

In practice this policy is that certain well argued of cases of
non-compliance which can be *safely* worked around, will be tolerated by
James.

In cases (like jira issue JAMES-344) where variance from a published
standard is required this functionality SHOULD be disabled in James by
default, it MUST be prominently and clearly documented that this causes
James to violate the relevant standard and SHOULD require to be enabled
by explicit configuration, making its use a conscious decision of the
user rather than a decision taken by the James team.

The feature MUST be clearly documented and the documentation MUST
contain the following statement which MUST describe every violation of
each standard with which James claims to comply which is enabled by the
feature:

    "Certain features allow Apache James to handle mail which has been
    constructed or sent in a manner not in compliance with the standards
    which James implements.

    The feature documented here permits James to handle messages which do
    not comply with the following standards which James claims to implement:

    RFC XXXX para Y.Y

    This feature has been disabled by default because James developers
    intend that James itself fully complies with the relevant published
    standards in the form in which it is distributed.

    The James project's policy is to encourage the developers of other email
    software to comply with published standards. It is only by all parties
    conforming to published standards that interoperability can be
    guaranteed, this is a fundamental charateristic of the internet.

    You are free to enable this feature which has been developed for your
    benefit as carefully as any of James' other features."

In cases where the required behaviour is not within the scope of any
standard which James claims to support (such as behaviour which is a
de-facto standard or an internet draft RFC but not yet subject of a
standards track RFC) it is acceptable to implement the behaviour but it
SHOULD be adequately documented (for instance by refrence to an internet
draft or other public document) so that users can be clear about its
status and what to expect from James.


d.

---------------------------------------------------------------------
To unsubscribe, e-mail: server-dev-unsubscribe@james.apache.org
For additional commands, e-mail: server-dev-help@james.apache.org


Mime
View raw message