On 20/09/2013 5:36 AM, Tim Abraldes wrote:
Separating and summarizing a little, it sounds like:
For alert/prompt/confirm, we're all (mostly?) agreed that the
"prevent this page from creating additional dialogs" checkbox should
have its meaning changed from "rate limit future
alert/prompt/confirm" to "prevent future alert/prompt/confirm." In
other words, bug 910501 is a valid bug.
That bug doesn't include the subtlety that only alert/prompt/confirm
should be blocked - it just mentions "dialogs". However, assuming that
additional constraint, then yeah, I do agree with it.
For window.print(), we may want to consider a way to prevent the
abuse of this function. A checkbox, similar to the one for
alert/prompt/confirm, is a potential option. I imagine that making
this a tab-modal dialog as opposed to an app-modal dialog would also
be beneficial.
For window.showModalDialog(), same comments as for window.print().
Note that these 2 functions already *do* hook into the same basic abuse
mechanism. As it is non-trivial to add checkboxes to these dialogs (eg,
IIUC, the "print" dialog is the one provided by the OS on Windows), it
uses a simple confirm-style dialog just to ask the abuse question -
however, it still hooks into the same basic flags and timer reset
capabilities used for prompt/alert/confirm.
So while I agree conceptually with bug 910501, I think we should be very
careful about a solution there that would cause print or modal dialogs
to stop working because a user selected "prevent future dialogs ..."
when an alert was up. IOW, I think we need to fix bug 910501 without
regressing the useful aspects of the current behaviour.
Mark
_______________________________________________
dev-platform mailing list
[email protected]
https://lists.mozilla.org/listinfo/dev-platform
