W3C home > Mailing lists > Public > public-qa-dev@w3.org > September 2004

Re: [wmvs] Prerequisites, Minimum Versions, Platforms?

From: Bjoern Hoehrmann <derhoermi@gmx.net>
Date: Mon, 06 Sep 2004 02:35:43 +0200
To: olivier Thereaux <ot@w3.org>
Cc: QA Dev <public-qa-dev@w3.org>
Message-ID: <41639877.176287467@smtp.bjoern.hoehrmann.de>

* olivier Thereaux wrote:
>Let's not forget, however, that the use of the validator outside of the 
>wvo/qa-dev platform has apparently increased recently, and is bringing 
>us a big user base, and although I cannot quantify it, I would be 
>almost certain that it also brings us *quality* participation. All of 
>this because we've made efforts to make the validator easier to install 
>on various platforms, and I do not wish to cancel these efforts without 
>a very good reason to do so.

I am just saying that we should not blue-sky portability requirements,
especially since `check`'s portability depends heavily on that of its
dependencies. I don't want to run into arguments about whether we can
use version W of X that not work with version Y of Z without a very
good reason to do so either. Win32 support is a good example, OpenSP
makes it something between very difficult and impossible to use -R on
Win32, if `check` MUST work without modifications on Win32, what would
we do?

  a) [_] ignore the requirement and release a new version

  b) [_] work around the issue in our code checking $^O

  c) [_] wait until someone fixes OpenSP and expect Win32
         users to compile OpenSP from CVS

  d) [_] wait until proper Win32 binaries are released

  e) [_] drop -R for all platforms

We currently operate in terms of a) just that there is no requirement
for Win32 support yet. With such a requirement, we could no longer do
a) and I seriously doubt that we will agree to do c), d), or e) which
leaves us with doing b) -- but as that would most likely just turn -R
off for Win32, and considering that Win32 user have to change at least
the shebang line anyway (until we rename `check` to `check.pl` or let
them do that), a) would probably be a better option as that would
provide a better chance to make Win32 users aware that there is a
potential security risk when using the Validator on Win32. But then
we do not really "support" Win32, rendering the introduction of such a
Win32 portability requirement quite pointless.
Received on Monday, 6 September 2004 00:36:30 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 16:54:46 UTC