fixing the test samples (was Re: initial review of sc1.2.1_l1_002?)

Hi Tim, Christophe, all,

I think Tim pointed out several things that need to be fixed in this 
specific test sample. However, based upon his other review I assume that 
these problems will be found in several test samples as we go along. To 
save valuable time and effort I propose that we hold off from doing any 
additional reviews and get the test samples fixed by the BenToWeb folks 
who submitted them first.

Christophe, if you agree then I would ask you as the facilitator to 
contact the BenToWeb submitters and check if they could fix the samples.

Regards,
   Shadi


Christophe Strobbe wrote:
> 
> Hi Time,
> 
> Thanks for the review.
> 
> 
> At 14:35 29/08/2007, Tim Boland wrote:
> 
>> Initial "manual"  review of sc1.2.1_l1_002:
>>
>>
>> Name and Email Address of the Submitter (File) Are Available -
>>
>> Not sure if files pass -
>> I could not name and email of submitter in the metadata file itself; the
>> submitter name is in the test sample list, but the email associated 
>> with that submitter does not
>> appear to me to be the personal address of the submitter (rather a 
>> generic email
>> address). (QUESTION: Should the name and email address of the 
>> submitter be
>> included in the metadata file directly? - if this has already been 
>> discussed
>> and I've forgotten, please accept my apologies). Similarly, I could 
>> not find submitter
>> information in the actual file).
> 
> Name and e-mail of the submitter are not specified in the metadata file 
> but in the Test Sample Status List. When the test sample was submitted 
> by or on behalf of BenToWeb, we use a generic and stable e-mail address, 
> so that it is still possible to contact BenToWeb folks after the project 
> ends. (See <http://www.w3.org/2007/05/15-tsdtf-minutes.html#action01>.)
> 
> 
>> Organization on whose behalf the test sample (file) was submitted -
>>
>>
>> Not sure if files pass - this organization name is not given 
>> explicitly in the
>> metadata file itself (although it is given in the list of test 
>> samples); the
>> organization may be derived by implication ("bentoweb") in "xmlns"? 
>> (QUESTION: Should this
>> organization be included in the metadata directly? - if this has 
>> already been
>> discussed and I've forgotten, please accept my apologies). Similarly, 
>> I did not find the
>> submitting organization mentioned explicitly in the actual file, but 
>> it may be derived by
>> implication from the "copyright" notification contained?
> 
> The "organization" is BenToWeb; this is specified in parentheses in the 
> "Submitter" column.
> 
> The BenToWeb copyright notice in the test file needs to be removed 
> because the materials are meant to be subject to W3C license after they 
> are submitted.
> 
> 
>> Test Files
>>
>>
>> All the files that are necessary to execute the test procedure have been
>> submitted -
>>
>>
>> not sure? In the actual file, I could not locate (or get to play? the
>> file "sc1.2.1_l1_002.wmv", but maybe that's just me) Also, for link in 
>> "files" containing a file in
>> subdirectory "testfiles", "sc1.2.1_l1_002.html", shouldn't the path 
>> include subdirectory "video"
>>  before the actual filename?
> 
> It appears that the video file ended up in the wrong location: the 
> 'video' folder is supposed to be a subfolder of the 'resources' folder; 
> we will need to fix that.
> 
> 
>>  All the submitted files follow the naming convention and directory 
>> structure -
>>
>>
>>
>> not sure? (see note just previous) Other naming conventions and directory
>> structure issues seem to be satisfied by these two submitted files, 
>> except that the
>> required "techniques" element (from the metadata document) does not 
>> seem to be included
>>  in the metadata file?
> 
> Indeed, the technique mapping needs to be added.
> 
> 
>> (NOTE: The lower-
>> case "l" and the number "1" may appear almost identical in certain 
>> renderings - should
>> it be considered to use upper-case "L" which might be less confusing? 
>> Just a
>> thought..)
>> (NOTE: There are additional elements (for example, "functionalOutcome" 
>> and "TestElement" - twice -)
>> included in the metadata file which were not mentioned (that I could 
>> see) in the metadata
>> document.. Should these be mentioned in case they have inadvertent 
>> "side-effects"?)
> 
> testElements is OK (optional), but functionalOutcome needs to be removed.
> This also implies an action item for me to check and update the XSLT 
> that converts TCDL 1.1 (from BenToWeb) to TSDTF metadata.
> 
> 
>> (QUESTION: "file" is "required" in the Test Samples Metadata Document, 
>> but the four
>> choices afterwards "http: GetRequest", "http: PostRequest", 
>> "http:PutRequest", and "http: HeadRequest"
>> are all "optional" - is it possible to have none of these four choices 
>> included since they're all "optional"?
>> Perhaps the document can be reworded more precisely?  In any case, 
>> none of these strings is explicitly
>> included after the "file" element..)
> 
> The xlink:href attribute is something I am a bit reluctant to remove 
> from TCDL 2.0 until 'HTTP Vocabulary in RDF' is finalized. That said, 
> for a simple GET request, which is what we have in this test, the file 
> element should not use the xlink:href attribute but http:GetRequest.
> The intention is to get exactly one of http:GetRequest, 
> http:PostRequest, http:PutRequest or http:HeadRequest. Should I tighten 
> the wording in <http://www.w3.org/WAI/ER/tests/usingTCDL>?
> 
> 
> 
>> All the files include valid markup unless otherwise required by the 
>> test -
>>
>>
>> not sure?
>> I ran the metadata file through Oxygen validator, and result indicated 
>> file was
>> valid, but I'm not sure whether I validated against schematron or just 
>> XML schema?,
>> since I don't know Oxygen that well yet..).
> 
> I don't know it either. I hope I will find some time to check in the 
> next few weeks.
> 
>> For the actual file, a check of the URL
>> against the new W3C Validator revealed four errors (was not valid 
>> XHTML1.0 Strict - 1) line 11,
>> column 13: there is no attribute "src", 2) line 11, column 55: there 
>> is no attribute "loop", 3)
>> line 11, column 72: there is no attribute "autostart", 4) line 11, 
>> column 79:
>> element "embed" undefined.
> 
> Hmm, I think the test file uses the embed as a fallback for browsers 
> that don't support the object element. Either the valid markup 
> requirement is a bit too tight or the purpose should mention why embed 
> is used and that it should not be treated as invalidating the test file. 
> Any thoughts?
> 
> 
> 
>> All the files include correct links unless otherwise required by the test
>>
>> - not sure? It depends on what the definition of "correct" is in this 
>> context.. The
>> links presented in the two files all seem reasonable, but might need 
>> to check further..
> 
> The links are correct in theory, but the video file is not where it is 
> supposed to be...
> 
> 
> 
>> All the files include correct spelling unless otherwise required by 
>> the test -
>>
>>
>> pass? A cursory examination reveals spelling seems to be correct. I 
>> ran a spell
>> checkers on the metadata file and "HyperText" was an unrecognized word 
>> (suggested
>> replacing with "hypertext")?
> 
> I assume you mean "HyperText" in the specName element? That is the 
> spelling used by the XHTML spec, so that's OK. Do we need to refine the 
> spelling requirement?
> 
> 
> 
> 
>> Metadata
>>
>>
>>
>> All the dates and other integer or literal values have the correct 
>> format -
>>
>>
>> not sure?
>> It depends on what the definition of "correct" is in this context. A 
>> cursory
>> examination of the formats of such values indicate that they seem 
>> reasonable, but bears
>> further examination..
>>
>>
>> All static values (especially copyright notices) are included and 
>> accurate -
>>
>>
>> pass? Copyright notices are included in both files, but for the 
>> metadata file, the
>> copyright notice seems to be W3C-related, whereas in the actual file, 
>> the copyright
>> notice seems to be "BentoWeb" related.. I can't speak fully to the 
>> accuracy of these
>> notices..
> 
> The copyright notices are currently only available in the example file at
> <http://www.w3.org/WAI/ER/tests/scx.x.x_lx_xxx.xml> (which is referenced 
> from
> <http://www.w3.org/WAI/ER/tests/usingTCDL#resources>).
> The ISO Schematron file contains checks for these data.
> 
>> (QUESTION: I notice "Microsoft" is mentioned explicitly in the 
>> metadata file -
>> should there be some sort of attribution or other qualification as to 
>> the use of a
>> company name in this way?  There is a trademark on the XHTML reference 
>> - maybe
>> something similar for Microsoft as well?)
> 
> Your attention to detail is impressive!
> As far as I can make out, Microsoft uses a registered sign (R in a 
> circle) after its name. I guess we need to fix this.
> 
> 
> 
>> All titles, descriptions, and other required fields are included and 
>> accurate -
>>
>>
>> pass for metadata file? - For the actual file, see notes previous.. 
>> Need more
>> information on definition of "accurate"..
> 
> I assume we use the normal dictionary meaning ;-)
> In my opinion the description is incomplete, because it does not say 
> anything about the embed element.
> 
> Best regards,
> 
> Christophe
> 
> 
>> All identifiers (especially ID for techniques and rules) are used 
>> correctly -
>>
>>
>> pass? The id attribute value "sc1.2.1_l1_002" seems reasonable and 
>> consistent with
>> naming conventions, technique "G93" seems to be referenced correctly, 
>> and the primary
>> rule seems to be consistent with the BentoWeb ruleset in terms of naming
>> (I assume the "WCAG2_20070517_1.2" and "media-equiv-captions" are 
>> concatenated"
>> in the reference?) but didn't have time to check it further?
>>
>>
>> All structures such as rules, techniques, and pointers are used 
>> correctly -
>>
>>
>> not sure?  It depends upon the definition of "correct" in these 
>> contexts - the approach
>> seems reasonable, but without further definitions/specification it's 
>> hard to tell?
>>
>> (COMMENT: It would be good to define precisely what "correct" means, 
>> so that reviewers and
>> submitters have more information to "get it right")
> 

-- 
Shadi Abou-Zahra     Web Accessibility Specialist for Europe |
Chair & Staff Contact for the Evaluation and Repair Tools WG |
World Wide Web Consortium (W3C)           http://www.w3.org/ |
Web Accessibility Initiative (WAI),   http://www.w3.org/WAI/ |
WAI-TIES Project,                http://www.w3.org/WAI/TIES/ |
Evaluation and Repair Tools WG,    http://www.w3.org/WAI/ER/ |
2004, Route des Lucioles - 06560,  Sophia-Antipolis - France |
Voice: +33(0)4 92 38 50 64          Fax: +33(0)4 92 38 78 22 |

Received on Thursday, 30 August 2007 14:22:20 UTC