- From: ¼ÛÁ¤±â <jungkee.song@samsung.com>
- Date: Tue, 17 Jul 2012 07:57:15 +0000 (GMT)
- To: Marcos Caceres <w3c@marcosc.com>
- Cc: "public-device-apis@w3.org" <public-device-apis@w3.org>, "public-web-intents@w3.org" <public-web-intents@w3.org>
Hi Marcos, > ------- Original Message ------- > Sender : Marcos Caceres<w3c@marcosc.com> > Date : 2012-07-16 22:38 (GMT+09:00) > Title : Re: About testable assertion > Hi Jungkee, > On Monday, 16 July 2012 at 14:12, ¼ÛÁ¤±â wrote: >> Hi all, >> >> Thanks everyone for good discussion in F2F last week. >> >> I have one question about testable assertions we discussed at the second day of the meeting. > A specification's primary purpose is to list the "conformance requirements" (i.e., the MUSTs, SHOULDs, etc.). Each conformance requirement will then have a number of testable assertions. For example, given the following conformance requirement: > "If the document is missing a foo, then user agent MUST add a bar at the end of the document." > You could create the following testable assertions: > 1. test to see if the bar is added to end of the document when foo is missing. > 2. test to see if the bar is not added when there is a foo. > 3. test to make sure that two bar are not added when one one bar already. > and so on¡¦ then each testable assertion becomes a real test in whatever you are testing. I got it. We use "conformance requirements" in spec writing and create relevant testable assertions and the matching test cases in the test suite. Thanks for the clarification. -- Marcos Caceres http://datadriven.com.au ---------------------------------- Jungkee Song SAMSUNG ELECTRONICS CO., LTD.
Received on Tuesday, 17 July 2012 07:57:53 UTC