On Mon, Nov 26, 2018 at 12:00 PM Marius Dumitru Florea <mariusdumitru.flo...@xwiki.com> wrote: > > On Thu, Nov 22, 2018 at 11:22 AM Thomas Mortagne <thomas.morta...@xwiki.com> > wrote: > > > On Thu, Nov 22, 2018 at 9:57 AM Vincent Massol <vinc...@massol.net> wrote: > > > > > > > > > > > > > On 21 Nov 2018, at 17:46, Adel Atallah <adel.atal...@xwiki.com> wrote: > > > > > > > > Hello, > > > > > > > > On Wed, Nov 21, 2018 at 5:36 PM Simon Urli <simon.u...@xwiki.com> > > wrote: > > > >> > > > >> Hi everyone, > > > >> > > > >> one of the most validation error we have with WCAG is about > > consecutive > > > >> line breaks: basically a <br /><br /> presents in a page. > > > >> > > > >> This happens mostly in our translation pages since the linebreaks in > > > >> plain syntax are translated in <br /> tags. > > > >> Caty provided a lot of details about this error on the related issue: > > > >> https://jira.xwiki.org/browse/XWIKI-15666. > > > >> > > > >> Currently we have around 140 validations failure because of this. > > > >> > > > >> Different proposal have been made in order to fix it, that I will try > > to > > > >> sum-up here: > > > >> > > > >> A. Remove completely this validation check > > > > > > > > -0, I think the validation can be useful at least to keep good > > practices. > > > > > > > >> B. Add an exception for the translation pages > > > > > > > > +1, simplest one. > > > > > > Note that the question is not so much about being simple (we can just > > remove WCAG for that and it’s the simplest ;)) but about being the right > > thing to do for people with disabilities. > > > > > > For me we have the following options: > > > > > > A) We don’t think that this check is useful, ie that it brings > > advantages for people with disabilities and then we can remove it. No need > > to add exceptions. > > > > > > B) We think the check is useful for people with disabilities and we > > should keep it, even for translations pages since I don’t see why people > > with disabilities shouldn’t be able to use translation pages. There are > > some ideas to fix this: I listed some in the jira issue and Thomas > > mentioned one too (it’s option D). > > > > > > > > C) Now I’m fine if we say the following: for technical reasons it’s > > already hard to ensure that we pass WCAG for user pages and thus FTM we > > focus only on those ones and we agree that we don’t pass WCAG for > > developer-oriented features, with the goal of improving on that aspect in > > the future. And thus we disable WCAG checks on technical pages (hidden > > pages) for now. > > > > > > > Excluding hidden page for now is indeed something that would make a > > lot of sense. > > > > I agree. The issue is that we can have hidden pages that: > > * are not displayed by default (I'm thinking panels)
Yes, we just need to have a filter a little bit more advanced than "hidden=false". We could imagine a white list of xclasses (which would have priority over "hidden=false" filter). Now for the specific panels use case we actually have a place where they are all displayed and validated all at once: the administration. > * are included / displayed dynamically with JavaScript so they won't appear > in the HTML downloaded by our WCAG tests (I'm thinking of async panels and > modals for instance) Validating DOM generated in javascript is indeed a problem we already have. Now for the few cases which are actually wiki pages included asynchronously and which don't fit in a generic xclass based filter we could add them trough the existing URL list property. About async panels, most of them are also cached which means they are really asyn only when loading the first page (when the framework find a cached content is return it instead of putting an async placeholder). > > So they won't be covered. > > > > > > > > > > Thanks > > > -Vincent > > > > > > > > > > >> C. Triggers the error only if more than 2 consecutive breaks is > > > >> encountered > > > > > > > > -1, it doesn't really makes sense to do that, it's like B. but badly > > done. > > > > > > > >> D. Create a rendering syntax dedicated to translation pages > > > > > > > > +1, could be a good idea but might be complicated. > > > > > > > >> > > > >> > > > >> A. Remove completely the validation check > > > >> > > > >> pros: > > > >> * the easiest one > > > >> * apparently the rule is not checked in other accessibility test, so > > > >> its real purpose for accessibility is unclear > > > >> > > > >> cons: > > > >> * IMO this rule is useful for checking the good practice of not > > using > > > >> <br /> > > > >> > > > >> B. Add an exception for the translation pages > > > >> > > > >> pros: > > > >> * same as for A > > > >> > > > >> cons: > > > >> * ? > > > >> > > > >> C. Triggers the error only if more than 2 consecutive breaks is > > encountered > > > >> > > > >> pros: > > > >> * ? > > > >> > > > >> cons: > > > >> * we would miss some consecutive <br /> that are used only for style > > > >> and we would catch some others in translations if we do 3 linebreaks > > > >> instead of 2. IMO it's only moving the problem > > > >> > > > >> D. Create a rendering syntax dedicated to translation pages > > > >> > > > >> pros: > > > >> * remove completely the problem of consecutive <br /> in > > translations > > > >> * can maybe be used to present them in another way? > > > >> > > > >> cons: > > > >> * need to develop/test/maintain a new rendering syntax > > > >> > > > >> I'd personnaly vote like this: > > > >> A: +0 > > > >> B: +1 > > > >> C: -1 > > > >> D: +0 > > > >> > > > >> WDYT? > > > >> > > > >> Simon > > > >> -- > > > >> Simon Urli > > > >> Software Engineer at XWiki SAS > > > >> simon.u...@xwiki.com > > > >> More about us at http://www.xwiki.com > > > > > > > > > -- > > Thomas Mortagne > > -- Thomas Mortagne
