RE: attrsetvaluenomodificationallowederr.xml

Joseph Kesselman wrote:

> That's  Level 1 behavior if the node is read-only. See 
http://www.w3.org/TR/2000/WD-DOM-Level-1-20000929/level-one-core.html#ID-F68D080

>(I seem to remember that there may have been an erratum in earlier versions of the >DOM which failed to define this behavior, whcih might explain the
>confusion...?)

Attr.value is not declared to raise any exceptions in the original DOM Level 1 Recommendation, http://www.w3.org/TR/REC-DOM-Level-1/level-one-core.html#ID-221662474  But was declared to raise
NO_MODIFICATION_ALLOWED_ERR in the DOM Level 1 errata (http://www.w3.org/DOM/updates/REC-DOM-Level-1-19981001-errata.html) and
DOM Level 1 Second Edition, http://www.w3.org/TR/2000/WD-DOM-Level-1-20000929/level-one-core.html#ID-221662474.

The DTD's and Schemas were generated from the original DOM Level 1 spec, so it did not recognize raising NO_MODIFICATION_ALLOWED_ERR as something to be expected from setting Attr.value.

I will move the test back to the DOM Level 1 set.  The test (and only this test) will not validate against the DTD and Schema's generated from the DOM Level 1 Recommendation due to an erratum.  The
test generation transform does not require that the test validate and so should properly convert this test using original recommendation sources.

There are a couple of options:

1) attrsetvaluenomodificationallowederr.xml could be explicitly excluded from the dom1-core-validate-tests task.

2) The DTD and schema building process could use the Second Edition working draft instead of the original recommendation.

My preference would be 1 since DOM Level 1 is still a working draft and other than failing to validate this one test there seem to be no negative consequences of continuing to use the original
recommendation.

However, I will attempt to build tasks for generating resources from the SE, at least as a check for the SE.

Received on Monday, 20 August 2001 12:35:14 UTC