+1 to Anne's statement. If you put them in the tests, you can't ever easily remove them - you've created a legacy of content that relies on the v1 standard (not just someone's temporary implementation). Better to rip off the bandaid now, unless you're really signed up to support init*Event in perpetuity. On Thu, Sep 29, 2011 at 6:27 AM, Anne van Kesteren <annevk@opera.com> wrote: > On Thu, 29 Sep 2011 10:02:15 +0200, Olli Pettay <Olli.Pettay@helsinki.fi> > wrote: > >> On 09/22/2011 12:25 AM, Anne van Kesteren wrote: >> >>> 1) In your earlier email you indicated it was not implemented >>> consistently. >>> 2) It seems rather hard to remove if you have a test suite that requires >>> it. >>> >> >> I don't buy this. We're removing/changing all sorts of stuff now from DOM >> which are in DOM 2 test suite. The same way v1 testsuite could have tests >> for init*Event, but it could be removed from v2 testsuite. >> > > So you are saying that even though we know we are going to remove it, we > should add it anyway? The DOM2 features are vastly different. When those > were implemented it was not with the expectation they would be removed soon. > And removing ill-designed DOM2 features is certainly not a cost-free effort. > It is rather cumbersome. > > > > But I don't care too much if we have init*Event or not. >> It doesn't really cause any harm to have it, if we can have it >> implemented consistently. >> > > Further propagating known bad APIs does cause harm in that authors will be > confused. > > > > (For certain testing having some way to initialize event would be good.) >> > > So implement event constructors. > > > -- > Anne van Kesteren > http://annevankesteren.nl/ > >
Received on Thursday, 29 September 2011 16:40:50 UTC