Loading ...
Sorry, an error occurred while loading the content.

Interpreting the Profile

Expand Messages
  • Sam Ruby
    Randy kindly provides some answers to some of the questions I raised: http://www.therssweblog.com/?guid=20071111122717 I m copying the questions and answers
    Message 1 of 12 , Nov 11, 2007
    • 0 Attachment
      Randy kindly provides some answers to some of the questions I raised:

      http://www.therssweblog.com/?guid=20071111122717

      I'm copying the questions and answers here, as I'm mostly interested in
      finding out if these answers are something the board can come to a
      consensus on.

      Here goes:

      * 4.1 - Should warnings be issued when version="0.91" or
      version="0.92" is encountered?

      "Yes, in fact, it should issue a warning whenever the version
      is not 2.0, this includes any remaining 0.93 and 0.94s."

      Follow up: is this something the board can come to consensus
      on? In particular, are there any known problems with 0.91?
      0.92? 0.93? 0.94?

      * 4.1.1.4 - If read one way, this statement implies that all
      category values should have a slash in it. If read
      any other way, this statement is meaningless.

      "I don't understand the confusion."

      Follow up: in Randy's feed, I see an item with a category of
      "rss". Is this a slash delimited value? Put another way,
      can anybody come up with an example of a category which
      is *NOT* a slash delimited value? If not, what does this
      recommendation mean?

      * 4.1.1.7 - Should the feed validator recommend this?

      "Sure, why not?"

      Again, is this something that the board can come to consensus
      on? Are there any known problems with other values?

      * 4.1.1.20.2 - Same issue as in section 4.1.1.4

      "Same response."

      Fair enough.

      * 5.1.1 - To help seed this discussion, is it the intent of this
      requirement to make the following suggested replacements?

      "Both dc:publisher, dc:source can be used for unique purpose and
      can be funky. Warnings here may or may not be helpful."

      Hmmm. Not it is my turn to not understand. Let's make it simple:

      Should a warning be issued for admin:generatorAgent ?
      Should a warning be issued for dc:date ?
      Should a warning be issued for dc:language ?
      Should a warning be issued for dc:publisher ?
      Should a warning be issued for dc:rights ?
      Should a warning be issued for dc:source ?
      Should a warning be issued for dc:subject ?
      Should a warning be issued for dcterms:modified ?

      - Sam Ruby
    • Randy Morin
      Sam, You ask several times is this something that the board can come to consensus on? We already did. Thanks, Randy ... interested in ... this ...
      Message 2 of 12 , Nov 11, 2007
      • 0 Attachment
        Sam,
        You ask several times "is this something that the board can come to
        consensus on?" We already did.
        Thanks,

        Randy

        --- In rss-public@yahoogroups.com, Sam Ruby <rubys@...> wrote:
        >
        > Randy kindly provides some answers to some of the questions I
        raised:
        >
        > http://www.therssweblog.com/?guid=20071111122717
        >
        > I'm copying the questions and answers here, as I'm mostly
        interested in
        > finding out if these answers are something the board can come to a
        > consensus on.
        >
        > Here goes:
        >
        > * 4.1 - Should warnings be issued when version="0.91" or
        > version="0.92" is encountered?
        >
        > "Yes, in fact, it should issue a warning whenever the version
        > is not 2.0, this includes any remaining 0.93 and 0.94s."
        >
        > Follow up: is this something the board can come to consensus
        > on? In particular, are there any known problems with 0.91?
        > 0.92? 0.93? 0.94?
        >
        > * 4.1.1.4 - If read one way, this statement implies that all
        > category values should have a slash in it. If read
        > any other way, this statement is meaningless.
        >
        > "I don't understand the confusion."
        >
        > Follow up: in Randy's feed, I see an item with a category of
        > "rss". Is this a slash delimited value? Put another way,
        > can anybody come up with an example of a category which
        > is *NOT* a slash delimited value? If not, what does this
        > recommendation mean?
        >
        > * 4.1.1.7 - Should the feed validator recommend this?
        >
        > "Sure, why not?"
        >
        > Again, is this something that the board can come to consensus
        > on? Are there any known problems with other values?
        >
        > * 4.1.1.20.2 - Same issue as in section 4.1.1.4
        >
        > "Same response."
        >
        > Fair enough.
        >
        > * 5.1.1 - To help seed this discussion, is it the intent of
        this
        > requirement to make the following suggested
        replacements?
        >
        > "Both dc:publisher, dc:source can be used for unique purpose
        and
        > can be funky. Warnings here may or may not be helpful."
        >
        > Hmmm. Not it is my turn to not understand. Let's make it
        simple:
        >
        > Should a warning be issued for admin:generatorAgent ?
        > Should a warning be issued for dc:date ?
        > Should a warning be issued for dc:language ?
        > Should a warning be issued for dc:publisher ?
        > Should a warning be issued for dc:rights ?
        > Should a warning be issued for dc:source ?
        > Should a warning be issued for dc:subject ?
        > Should a warning be issued for dcterms:modified ?
        >
        > - Sam Ruby
        >
      • James Holderness
        ... If you re referring to the vote on the profile, I must have misunderstood what we were voting for, because I didn t think it implied we would be
        Message 3 of 12 , Nov 11, 2007
        • 0 Attachment
          Randy Morin wrote:
          >> "Yes, in fact, it should issue a warning whenever the version
          >> is not 2.0, this includes any remaining 0.93 and 0.94s."
          > You ask several times "is this something that the board can come to
          > consensus on?" We already did.

          If you're referring to the vote on the profile, I must have misunderstood
          what we were voting for, because I didn't think it implied we would be
          deprecating older versions of RSS (which I believe is what you're
          suggesting). Or had they already been deprecated at some point in the past
          and I just didn't notice?

          Regards
          James
        • James Holderness
          ... c/c - a common abbreviation for credit card . I could quite easily imagine someone using that as a category in a financial feed of some sort. While
          Message 4 of 12 , Nov 11, 2007
          • 0 Attachment
            Sam Ruby wrote:
            > Follow up: in Randy's feed, I see an item with a category of
            > "rss". Is this a slash delimited value? Put another way,
            > can anybody come up with an example of a category which
            > is *NOT* a slash delimited value?

            "c/c" - a common abbreviation for "credit card". I could quite easily
            imagine someone using that as a category in a financial feed of some sort.
            While there's no way to validate it, that's technically not a valid
            category, since it's not identifying an hierarchical location.

            Regards
            James
          • Sam Ruby
            ... if category == c/c : warn(...) The feed validator generally won t implement a check because somebody could quite easily imagine something, but if there
            Message 5 of 12 , Nov 12, 2007
            • 0 Attachment
              James Holderness wrote:
              >
              > Sam Ruby wrote:
              > > Follow up: in Randy's feed, I see an item with a category of
              > > "rss". Is this a slash delimited value? Put another way,
              > > can anybody come up with an example of a category which
              > > is *NOT* a slash delimited value?
              >
              > "c/c" - a common abbreviation for "credit card". I could quite easily
              > imagine someone using that as a category in a financial feed of some sort.
              > While there's no way to validate it, that's technically not a valid
              > category, since it's not identifying an hierarchical location.

              if category == 'c/c': warn(...)

              The feed validator generally won't implement a check because somebody
              "could quite easily imagine" something, but if there are real feeds that
              use categories which clearly aren't following the spec and/or cause real
              problems, then I would be quite willing to accept patches and/or code
              myself checks for specific patterns or even specific values for which
              warnings will be issued. Example:

              http://feedvalidator.org/docs/warning/NotSufficientlyUnique.html

              - Sam Ruby
            • rcade
              ... The reference to the RSS specification in the rss section ... http://www.rssboard.org/rss-profile#element-rss ... refers to the RSS 2.0 spec described in
              Message 6 of 12 , Nov 12, 2007
              • 0 Attachment
                --- In rss-public@yahoogroups.com, Sam Ruby <rubys@...> wrote:
                >Follow up: is this something the board can come to consensus
                >on? In particular, are there any known problems with 0.91?
                >0.92? 0.93? 0.94?

                The reference to "the RSS specification" in the rss section ...

                http://www.rssboard.org/rss-profile#element-rss

                ... refers to the RSS 2.0 spec described in the introduction:

                http://www.rssboard.org/rss-profile#element-rss

                We never discussed making the requirements in the profile apply to
                versions of RSS prior to 2.0.

                >Follow up: in Randy's feed, I see an item with a category of
                >"rss". Is this a slash delimited value? Put another way,
                >can anybody come up with an example of a category which
                >is *NOT* a slash delimited value? If not, what does this
                >recommendation mean?

                Logically, a slash-delimited record with only one field could have a
                trailing slash or omit it. This recommendation is a pick up directly
                from the spec. It means that categories in a hierarchy should use
                slashes to separate levels of the taxonomy.

                >Again, is this something that the board can come to consensus
                >on? Are there any known problems with other values?

                We've already reached consensus on it -- if you rely on the spec
                published by the board, use http://www.rssboard.org/rss-specification.
                I am not aware of any problems with using another docs value or
                omitting it entirely. I don't see how the validator could determine
                which copy of the spec was relied upon, so a warning seems impossible.

                >Should a warning be issued for admin:generatorAgent ?
                >Should a warning be issued for dc:date ?
                >Should a warning be issued for dc:language ?
                >Should a warning be issued for dc:publisher ?
                >Should a warning be issued for dc:rights ?
                >Should a warning be issued for dc:source ?
                >Should a warning be issued for dc:subject ?
                >Should a warning be issued for dcterms:modified ?

                No to all questions. We were conservative in our inclusion of
                namespace elements, adding the ones with significant interop issues
                with core elements.
              • Randy Morin
                We didn t deprecate, but we definately recommended the version attribute be 2.0. The language is pretty clear IMHO. Thanks, Rany ... misunderstood ... be ...
                Message 7 of 12 , Nov 12, 2007
                • 0 Attachment
                  We didn't deprecate, but we definately recommended the version
                  attribute be 2.0. The language is pretty clear IMHO.
                  Thanks,

                  Rany

                  --- In rss-public@yahoogroups.com, "James Holderness" <j4_james@...>
                  wrote:
                  >
                  > If you're referring to the vote on the profile, I must have
                  misunderstood
                  > what we were voting for, because I didn't think it implied we would
                  be
                  > deprecating older versions of RSS (which I believe is what you're
                  > suggesting). Or had they already been deprecated at some point in
                  the past
                  > and I just didn't notice?
                  >
                  > Regards
                  > James
                  >
                • Randy Morin
                  A vote to deprecate is a good idea. I think we should do that for all 0.9x. Thanks, Randy ... misunderstood ... be ... the past
                  Message 8 of 12 , Nov 12, 2007
                  • 0 Attachment
                    A vote to deprecate is a good idea. I think we should do that for all
                    0.9x.
                    Thanks,

                    Randy

                    --- In rss-public@yahoogroups.com, "James Holderness" <j4_james@...>
                    wrote:
                    >
                    > If you're referring to the vote on the profile, I must have
                    misunderstood
                    > what we were voting for, because I didn't think it implied we would
                    be
                    > deprecating older versions of RSS (which I believe is what you're
                    > suggesting). Or had they already been deprecated at some point in
                    the past
                    > and I just didn't notice?
                    >
                    > Regards
                    > James
                    >
                  • James Holderness
                    ... The text in the profile says: A feed that conforms to the RSS specification must contain a version attribute with the value 2.0 . I read that as a simple
                    Message 9 of 12 , Nov 12, 2007
                    • 0 Attachment
                      Randy Morin wrote:
                      > We didn't deprecate, but we definately recommended the version
                      > attribute be 2.0. The language is pretty clear IMHO.

                      The text in the profile says:

                      A feed that conforms to the RSS specification must contain
                      a version attribute with the value "2.0".

                      I read that as a simple restatement of the RSS 2.0 spec text which says:

                      At the top level, a RSS document is a <rss> element, with a
                      mandatory attribute called version, that specifies the version
                      of RSS that the document conforms to. If it conforms to this
                      specification, the version attribute must be 2.0.

                      In other word, if you're creating an RSS 2.0 feed, the value of version
                      attribute should be "2.0".

                      If you're creating something else (like say an RSS 1.0 feed, or an RSS 0.92
                      feed), then the profile doesn't apply. IMHO.

                      Now an argument could be made that the board should be deprecating certain
                      older version of RSS, but that's something that should be decided
                      separately. Personally I don't think it's a good idea, but it's certainly
                      worth discussing.

                      Regards
                      James
                    • Randy Morin
                      I just want to make it clear that none of this discussion relates to version 1.0 or 1.1. Those versions are clearly not in our domain and in fact are a
                      Message 10 of 12 , Nov 12, 2007
                      • 0 Attachment
                        I just want to make it clear that none of this discussion relates to
                        version 1.0 or 1.1. Those versions are clearly not in our domain and in
                        fact are a separate XML that happens to share the same name as RSS.
                        Confusing, but true.

                        All other versions of RSS are previous instances of RSS 2.0 and thus
                        fall within the existing spec and profile instances.
                        Thanks,

                        Randy

                        --- In rss-public@yahoogroups.com, "James Holderness" <j4_james@...>
                        wrote:
                        >
                        > If you're creating something else (like say an RSS 1.0 feed, or an
                        RSS 0.92
                        > feed), then the profile doesn't apply. IMHO.
                        >
                      • James Holderness
                        ... I don t think the feed validator should be trying to warn about such things. I was just trying to give you an example of what an invalid category might
                        Message 11 of 12 , Nov 12, 2007
                        • 0 Attachment
                          Sam Ruby wrote:
                          >> "c/c" - a common abbreviation for "credit card". I could quite easily
                          >> imagine someone using that as a category in a financial feed of some
                          >> sort.
                          >> While there's no way to validate it, that's technically not a valid
                          >> category, since it's not identifying an hierarchical location.
                          >
                          > if category == 'c/c': warn(...)
                          >
                          > The feed validator generally won't implement a check because somebody
                          > "could quite easily imagine" something,

                          I don't think the feed validator should be trying to warn about such things.
                          I was just trying to give you an example of what an invalid category might
                          look like since you didn't seem to get what the profile recommendation
                          meant.

                          FWIW I have a category database with hundreds of slash separated items. Some
                          are clearly hierarchical, many are hard to say either way, but some are
                          definitely *not* hierarchical. Some examples: "I/O", "C/C++", "H/P",
                          "8/11/2006". Unfortunately I can't tell the source - they could be from
                          atom:category elements, dc:subject, or some other category-like extension
                          (so not really much better than my hypothetical "c/c" example).

                          Either way I don't think this is an issue worth worrying about.

                          Regards
                          James
                        • Sam Ruby
                          ... It looks to me that you are successfully making the case that in strings that contain a slash in position two are likely to be mistakes. - Sam Ruby
                          Message 12 of 12 , Nov 12, 2007
                          • 0 Attachment
                            James Holderness wrote:
                            >
                            > Sam Ruby wrote:
                            > >> "c/c" - a common abbreviation for "credit card". I could quite easily
                            > >> imagine someone using that as a category in a financial feed of some
                            > >> sort.
                            > >> While there's no way to validate it, that's technically not a valid
                            > >> category, since it's not identifying an hierarchical location.
                            > >
                            > > if category == 'c/c': warn(...)
                            > >
                            > > The feed validator generally won't implement a check because somebody
                            > > "could quite easily imagine" something,
                            >
                            > I don't think the feed validator should be trying to warn about such
                            > things.
                            > I was just trying to give you an example of what an invalid category might
                            > look like since you didn't seem to get what the profile recommendation
                            > meant.
                            >
                            > FWIW I have a category database with hundreds of slash separated items.
                            > Some
                            > are clearly hierarchical, many are hard to say either way, but some are
                            > definitely *not* hierarchical. Some examples: "I/O", "C/C++", "H/P",
                            > "8/11/2006". Unfortunately I can't tell the source - they could be from
                            > atom:category elements, dc:subject, or some other category-like extension
                            > (so not really much better than my hypothetical "c/c" example).

                            It looks to me that you are successfully making the case that in strings
                            that contain a slash in position two are likely to be mistakes.

                            - Sam Ruby
                          Your message has been successfully submitted and would be delivered to recipients shortly.