Talk:OPERAT
raphael 20:33, 27 October 2009 (UTC): Ideas for a better name than OPERAT are welcome. The rationale for this variable and the reasons for not using "function" are at Talk:UKCVAR.
I suggest the following addition to the definition:
Remarks:
OPERAT is intended to be used in conjunction with other sub-attributes of a complex attribute to indicate how their values must be combined in order to describe a condition. Example: The combination
UKCLRN (complex attribute)
UKCFIX=2.5, UKCVAR=10.00, OPERAT=1 (sub-attributes)
indicates that the under-keel clearance required is the greater of 2.5 metres or 10% of the ship's draught.
jens 06:16, 28 October 2009 (UTC) I agree. The remarks are useful.
Draft extension for logical connectives
raphael 00:15, 16 February 2010 (UTC): Here is a draft of the extension discussed on Talk:CHALIM.
Definition: A process or action, such as addition, substitution, transposition, or differentiation, logical conjunction or disjunction, performed in a specified sequence and in accordance with specific rules.
Attribute values: add
3 logical conjunction
4 logical disjunction
logical conjunction
all the conditions described by the other attributes of the object, or sub-attributes of the same complex attribute, are true
logical disjunction
at least one of the conditions described by the other attributes of the object, or sub-attributes of the same complex attributes, is true
Remarks: Add:
When bound to the object CHALIM, OPERAT can take only the values 3 or 4.
CHALIM/MAXLOA = 50.0, CHALIM/MAXDRF=10.0, CHALIM/OPERAT=3
implies the limitation applies only when LOA > 50.0 and draught > 10.0
I have drafted it as an extension to OPERAT so we can discuss it, but am a little reluctant to make the concept of "logical connection" an extension of OPERAT instead of a different attribute. As drafted, the combination looks a little awkward to me, even if it is not mixing apples and oranges, it looks like to different varieties of apples...
jens 11:29, 16 February 2010 (UTC) Thought more on that. Can we say that logical conjunction/disjunction are few of the bitwise logic operations (AND (logical conjunction), NOT (logical negation), OR (logical disjuntion), XOR (exclusive disjunction))? Does S100 support that? I am in doubt; see e.g. 1-4.3 (page 18), 3-5.2.9 (page 73) and 3-6.4.2 (page 79). But further down in the document they talk too much about attributes and operations. Can one give me an advice which assumption is correct now? Does S100 support the operation idea in conjunction with attribute values or not? If yes. than I would propose to move the logic operations to a separate attribute.
raphael 18:13, 16 February 2010 (UTC): The reason I wanted to make a new attribute is to separate logical from non-logical combination. To answer your queations: Yes, logical conjunction/disjunction are two of the logic operations AND, NOT, OR, XOR. ("Bitwise logic" has a slightly different meaning.) S100 does not support operations, but only in the sense that one cannot say CHALIM has operations sizeOfText, logicalAND:
int sizeOfText(): returns the size of the file named by TXTDSC, 0 if the TXTDSC attribute is absent or null, or -1 if the file is not found
boolean logicalAND(): returns TRUE if the OPERAT attribute exists and has value=3, FALSE otherwise
But this is not what we are doing. The attribute OPERAT just says how the values of other attributes are expected to be used. It is for the application to code a function that implements the logical AND/OR operation and produces the result for use by the application.
raphael 22:40, 16 February 2010 (UTC): Coordinate Operations in S100 Part 6: The concepts described in this Part are like what we are doing with OPERAT, but a lot more complex. Section 6-4.6 (page 173) says "The mathematical formulas for an operation are specified in text form or by referencing a source document." We are specifying our formula in text form, in the specification of OPERAT and its allowed values.
jens 09:35, 18 February 2010 (UTC) o.k. I understand that now. A new attribute LOGCON is drafted.
DavidAcland 23:35, 25 March 2010 (UTC) I have messed with the format to conform with the layout requirements of enumerations. In particular we do not have the option of a top level Definition. So without really understanding why it is so complicated, I have adapted your earlier definition for each ID. As it stands I would expect to substitute the word "subtraction" for "addition" in both meanings but have held back because I am obviously missing a lot here. I hope you can straighten these out.
raphael 02:56, 26 March 2010 (UTC): I thought S-100 had changed the format. Anyway, I think the following may be OK even without a top level Definition:
ID Meaning
1 : largest value
2 : smallest value
Definitions:
largest value:
The largest value computed from the applicable attributes or sub-attributes must be used
smallest value:
The smallest value computed from the applicable attributes or sub-attributes must be used
Remarks:
OPERAT is intended to be used in conjunction with other attributes (or sub-attributes of a complex attribute) to indicate how their values must be combined in order to describe a condition. Null attributes are ignored.
Example:
Complex attribute UKCLRN with sub-attributes UKCFIX=2.5, UKCVAR=10.00, OPERAT=1
indicates that the under-keel clearance required is the greater of 2.5 metres or 10% of the ship's draught.
DavidAcland 08:08, 26 March 2010 (UTC) Great. Overnight I had also come to similar simpler definitions. I have further simplified by removing the final phrases "...must be used"
I expect you are right about S-100. Unfortunately the Registry expects the format used here.
jens 11:24, 26 March 2010 (UTC) ok. I remove the bold text
raphael 14:30, 26 March 2010 (UTC): Agreed.
DavidAcland 09:20, 23 September 2010 (UTC)
I have ammended the definitions i.a.w. the discussion at Tokyo to include the term "numerically largest" or smallest.
raphael (talk) 03:37, 6 February 2017 (CET):Suggested definition: Indicates whether the minimum or maximum value should be used to describe a condition or in application processing.
The registry currently has this as the definition: "OPERAT is intended to be used in conjunction with other attributes (or sub-attributes of a complex attribute) to indicate how their values must be combined in order to describe a condition." but that seems more of a remark and I think it should be moved to the Remarks section, which would consist of two bullet points:
- Attribute operation is intended to be used in conjunction with other attributes (or sub-attributes of a complex attribute) to indicate how their values must be combined in order to describe a condition.
- Null attributes are ignored. Example use: Complex attribute underkeelAllowance with UKCFIX=2.5, UKCVAR=10.00, operation=1 inicates that the under-keel allowance required is the greater of 2.5 metres or 10% of the ship's draught.