Use Advanced Search to search the entire archive.
[jsr363-experts] Profiling and optionality in the TCK
- From: Werner Keil <
>
- To: "
" <
>, "
" <
>
- Cc: Heather VanCura <
>
- Subject: [jsr363-experts] Profiling and optionality in the TCK
- Date: Mon, 27 Jul 2015 15:20:38 +0200
Dear All,
Creating the TCK for JSR 363 is on its way.
As we plan some level of optionality, the TCK will have to take it into
consideration and some kind of profile an implementation is made for
("Full", "Minimal" and some other options or combinations)
Especially for the "quantity" package, which we now scaled to support a
"reasonable subset of SI", see
https://docs.google.com/document/d/12KhosAFriGCczBs6gwtJJDfg_QlANT92_lhxUWO2gCY/edit#heading=h.w9qmpfuv3rcg
A related question was raised by Martin a while ago:
https://java.net/jira/browse/UNITSOFMEASUREMENT-96
After we heard, the idea of a "Stripped Implementation" was dropped, we
dismissed the idea. The number of quantities and units based on them is now
smaller (leaving room for extension modules like "Full SI"
https://github.com/unitsofmeasurement/si-units) but would it be legitimate
to have the TCK assert a group of them, while leaving others optional, too?
An example from JPA would be
http://docs.oracle.com/javaee/7/api/javax/persistence/Persistence.html
Running the TCK (I have not personally requested to see it, though I am in
the JPA EG and others did, it's not Open Source AFAIK;-) there must be a
choice between "SE/Standalone" and "Java EE Container" or similar,
otherwise the TCK could not decide how to test the Persistence class either.
Would it be OK if we group the quantities like the Spec mentions into
-
Base units
<
http://www.bipm.org/en/publications/si-brochure/section2-1.html>
-
Coherent derived units in the SI with special names and symbols
<
http://www.bipm.org/en/publications/si-brochure/table3.html>
-
Derived units expressed in terms of base units
<
http://www.bipm.org/en/publications/si-brochure/section2-2-1.html> (a
reasonable subset based on broad usage)
and therefore allow implementations to stick to the 7 base quantities only
as minimum requirement, while leaving the others to one or more additional
profiles?
Thanks and Regards,
Werner
[jsr363-experts] Profiling and optionality in the TCK
|
Werner Keil |
07/27/2015 |