Minor change to clarify that 'custom' is an object #125
Conversation
joshuagl
force-pushed
the
joshuagl/issue50-custom-object
branch
2 times, most recently
from
46fb9d2 to
b9a066e
Compare
tuf-spec.md
Outdated
| If defined, the elements and values of the CUSTOM object will be made | ||
| available to the client application. The format of, and information in, |
There was a problem hiding this comment.
Minor nit: Do you think that it is clear that the CUSTOM object, if supplied, must be a json object (i.e. deserialize to a Python dictionary), if we say the "format ... is opaque"?
There was a problem hiding this comment.
That's a good question. I wanted to try and avoid mentioning JSON in the description, because we are trying to make the actual format of the metadata looser in the specification.
I felt like the change here was as clear, from the prose alone, as other places we talk about objects. Though admittedly those other places in the specification where we talk about objects also have an explicit example of the object format and don't refer to it as opaque.
I will try and clarify this some more tomorrow. Either by clarifying the language here, by pointing out in section 4.1 that all objects in the file formats are JSON objects, or perhaps a combination.
There was a problem hiding this comment.
I think it is not a very big deal. I guess my association with the word "format" here is biased through securesystemslib.formats, where the format also specifies the datatype. It's quite likely that others don't have that association.
There was a problem hiding this comment.
I've pushed a fixup commit with some further clarification that the framework only needs to know that the value mapped to the "custom" attribute is an object.
joshuagl
force-pushed
the
joshuagl/issue50-custom-object
branch
from
d55587e to
821942b
Compare
|
Force pushed an update to merge the fixup and increase the patch number to 11, assuming that this will land after #127 |
joshuagl
force-pushed
the
joshuagl/issue50-custom-object
branch
from
821942b to
f7ec856
Compare
|
In the spirit of #132, could I get a second review/approval @trishankatdatadog or @mnm678 ? |
Clarify that the custom field of TARGETPATH objects is itself an opaque object with format and contents defined by the application. Fixes: theupdateframework#50 Signed-off-by: Joshua Lock <[email protected]>
joshuagl
force-pushed
the
joshuagl/issue50-custom-object
branch
from
f7ec856 to
48df293
Compare
Minor tweak of the description of the custom field of TARGETPATH objects to explicitly state that the value is an object which is opaque to the framework.
Fixes #50