[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [Elist Home]
Subject: Re: Proposed agenda for meeting 2000-09-06 09:00-12:00EDT
Greetings: With apologies for the lateness, I've only just managed to use the holiday weekend to get back to duties for this committee (thanks to all for understanding), and so have some input here. 1. Sun has a suite of tests from the development of the XSLT engine for their Forte Fusion project which Ron Ten-Hove developed. They've forwarded them to me for submission to the committee on Sun's behalf. 2. I also append Ron's comments, for any edification it might add, on his strategy for referencing between the tests and the specification which seemed worthy of mention in connection with Tony, Eduardo, and Ken's "Re: Which file should be the basis for pointing into XSLT?" thread. I don't know to what degree the submission of Sun's tests are to be included on the agenda, but please consider making appropriate room for them to mentioned and described. respectfully submitted, Rohn Robert <test-to-spec-linkage> ------------- Begin Forwarded Message ------------- From: "Ron Ten-Hove" <ronald.ten-hove@ebay.sun.com> To: "John Robert Gardner" <John.Robert.Gardner@east.sun.com> Subject: RE: XSLT Conformance Committee Input/Meeting . . . This sounds similar to the approach I took. I actually printed out a copy of the spec with page and line numbers, and named each test after the piece of the spec it was to test: xx.yy.z xx = page number yy = line number z = 0, 1, 2... for multiple tests from a single line (or multiple files for include/import stuff). > It stands to reason then, that we should have a W3C specification > version for Conformance, yes? With links therefrom to respective > test stylesheets? Would this be a fair and viable extrapolation > on my part from what you've written below? It seems a very > useful way to go about things, and could be coupled with a per-component > > >directory structure > arrangment as well, allowing tests to be accessed either via-spec > or via-topic. Agreed. Please, feel free to share these comments with the committee, and whomever else you deem appropriate. I hope my experience can be of help. Other thoughts: a) it would be nice to divide the tests up according to the various sections of the PR. The same should be done for XPath. Testing a single feature at a time makes debugging the processor easier. b) it should be required that the XML parser being used passes the current OASIS test suites. . . . . </test-to-spec-linkage>
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [Elist Home]
Powered by eList eXpress LLC