On Tue, Sep 29, 2015 at 4:54 PM, Guillaume Lerouge <[email protected]> wrote:
> Hi Denis, > > thanks for your message. Please see my thoughts below. > > On Tue, Sep 29, 2015 at 11:12 AM, Denis Gervalle <[email protected]> wrote: > > > Hi devs, > > > > I am opening this discussion to move the discussion started in an issue > > (XWIKI-12628) to its proper place, the ML. (Discussing into JIRA is very > > bad since it lower the audience, we tend to have done it too much > > recently). > > > > So the issue is that, in the Nested Document concept, we can create > > children to inexistent (WebHome) page. This introduce some abnormalities: > > - Administering inexistent page (including page rights, see > > http://jira.xwiki.org/browse/XWIKI-12629) > > - Children of inexistent page (see > > http://jira.xwiki.org/browse/XWIKI-12628 > > ) > > > > Since we agree that having nodes in the tree of pages is fine without > > having the intermediate document created, > > > I know that's what has been decided so far, but I was wondering what was > the rationale behind it? Is there a specific benefit that comes from > allowing this behavior? > > > > we definitely need to better show > > this concept to the final user. It would be a pain to explain to a user > > that access rights for page children of an inexistent page is > administrable > > from that inexistent page. It is also very poor when you navigate from > the > > breadcrumb to a inexistent page as a normal user to reach "The requested > > page could not be found." with no way to navigate further. > > > > Agreed. > > These inexistent page are not necessarily invisible, there are links to > > them, and showing that as "error" page looks inappropriate IMO. > > > Agreed. > > > > So there are probably a couple of things we could do to improve this: > > > > A) display the list of children in addition to the actual message > > > > As you point out below, this doesn't fully solve the navigation problem > since you can only go down, not further up. > That's not fully true, since you have the breadcrumb. What I do not like, is that it looks first as an error page, then propose a way out. IMO those "empty" node are not errors, just nodes without content, just there for supporting their children. > > > > B) display a "default" and nice (not an error) dashboard on them that > allow > > navigation (with a button for those that can edit, to create the page) > > C) create them all the time, with the dashboard above by default > > > > To me, B) and C) are almost the same in the end, though I'm not sure a > dashboard is best. I'd rather create an empty page, with the list of > children available from the menu and/or in a tab in the footer, as on every > other standard page. > > > > D) do not propose links to those page in the breadcrumb to users without > > edit right on them ? > > > > How would you then display the breadcrumb? With holes in it? It would also > make the breadcrumb inconsistent with the URL. > All I was proposing here is to remove the link, not the text, so you cannot click on them to reach them. I agree, this is far from a perfect solution, but since we have the Sibling viewer, it is not really breaking the navigability of the site. I was just suggesting here a way that limit the situation where a normal user reach those page, that could also be seen as useless to reach (else the user should had create it). Any other place with a link to those page should be treated the same. So only user with the ability to really act on those pages (administering or creating them) will have link to access them. > > > E) ... (please add more idea) > > > > I am in favor of B) currently, since I do not think we are in an "error" > > case like A) seems to expose. > > I am wondering if D) could be useful as well to limit navigation to those > > page or not for user that will not find them so useful... but it may > exists > > opposite UC. > > > > WDYT ? > > > > I think my favorite one is C). I don't really like having pages that exist > from a rights management and hierarchy standpoint but not from a content > standpoint - but maybe there's a good reason for this that I'm missing? > > Thanks, > > Guillaume > > -- > > Denis Gervalle > > SOFTEC sa - CEO > > > _______________________________________________ > devs mailing list > [email protected] > http://lists.xwiki.org/mailman/listinfo/devs > -- Denis Gervalle SOFTEC sa - CEO _______________________________________________ devs mailing list [email protected] http://lists.xwiki.org/mailman/listinfo/devs
