1 Nov 2004 01:01
1 Nov 2004 01:07
Re: Olivier Planson is out of the office.
On Mon, 1 Nov 2004 01:01:50 +0100, Olivier Planson <olivier.planson@...> wrote: > I will be out of the office starting 23/10/2004 and will not return until > 02/11/2004. Olivier, this it the 6th time you've had a holiday since June, you're making us feel bad.Jim.
1 Nov 2004 12:49
Re: Search controls & User Agent detection of
J. King wrote: > On Sun, 31 Oct 2004 11:23:12 +0000, Jim Ley <jim.ley@...> wrote: > >> I'm not sure of the point of requesting this here, the WHAT-WG appears >> to be dead [...] I don't really see the point spending too much >> effort on the list until some more info is coming through. > > > The request has been made. Whether he gets an answer now or later is > another matter, but at least this way Mr. O'Connor won't forget to > send it in later. Also, it's now possible to refer people to a URI > of this mailing if it's of relevance in a discussion, and people not > of the WHAT working group can reply and give their own thoughts on > the matter. > FWIW: "I'll be resuming work on replying to comments on the WHATWG <http://www.whatwg.org/> specs in a few days. As some of you may have noticed, the Web Apps <http://whatwg.org/specs/web-apps/current-work/> spec gained a few sections recently, mostly things I added while I was idling these last few days. As usual let the list <http://www.whatwg.org/mailing-list> know if you have any comments." http://ln.hixie.ch/?start=1099214002&count=1
7 Nov 2004 19:34
Submission progress
Submission progress
-------------------
Not sure if this would come under Web Controls or Web Forms, suspect
the former.
Modern browsers feature useful rich interfaces for monitoring the
progress of file downloads, but are not symmetrical, uploads are second
class citizens.
The browser UI should be extended so that similar information is
provided while transmitting large requests:
* Total size of data to be transmitted
* Percentage/amount of data sent so far, preferably with graphical
status bar
* Estimated time remaining
* Average rate of data transmission
Etc. The exact selection of information displayed would be
implementation defined.
This information should be displayed in a modal popup window, as the
browser window or tab is locked during submission.
The popup should only appear for form submissions where the total
quantity of data to be uploaded exceeds some relatively small
threshold, e.g. 1mb. This should be implementation defined.
(Continue reading)
7 Nov 2004 19:35
Enhanced data tables
Enhanced data tables
--------------------
Proposed addition to Web Controls 1.0.
Separate designation for data bearing tables that allow browsers to
provide extended data manipulation features such as:
* Sorting
Standard controls for sorting by each column/asc+desc
Native code for fast sorting of many data
* In-browser pagination
If the dataset is small, there is no reason why a larger number
of data can't be sent to the browser, to allow greater sorting
functionality there. Usable display of larger datasets would be
assisted by pagination.
Etc.
"Calculated cells" could be simply created with DOM and JS, but XPath
could be used for this function as well:
<calculated expression="../../colgroup[id=price]/td">
(Just a thought)
Ben
(Continue reading)
7 Nov 2004 19:59
Re: Submission progress
On Sun, 7 Nov 2004 18:34:45 +0000, Afternoon <afternoon@...> wrote: > This information should be displayed in a modal popup window, as the > browser window or tab is locked during submission. No, the browser window should not be locked, and no popup windows should be used, certainly not by specification. Form submissions have never been modal in HTML, and there's no reason why this should change now, users need the ability to cancel submissions, and by the simple mechanism of navigating somewhere else, not clicking cancel on a pointless dialog. Equally, the upload is only a component of the time taken to perform a post, with a countdown, the user will expect immediate response, even if the server is going to spend a number of minutes processing the data. > * Webmail applications would be able to give the user an estimated > time until mail attachments are uploaded to the server No they wouldn't, the modal window would be doing it, in any case rich applications like this don't use single submission queues, but have the submission going on in the background (by submitting a new frame, with the active frame available immediately for normal use) Popups annoy users. Jim
7 Nov 2004 20:18
Re: Submission progress
On 7 Nov 2004, at 18:59, Jim Ley wrote: > No, the browser window should not be locked, and no popup windows > should be used, certainly not by specification. Form submissions have > never been modal in HTML, and there's no reason why this should change > now, users need the ability to cancel submissions, and by the simple > mechanism of navigating somewhere else, not clicking cancel on a > pointless dialog. > > Equally, the upload is only a component of the time taken to perform a > post, with a countdown, the user will expect immediate response, even > if the server is going to spend a number of minutes processing the > data. The browser could sidestep this problem by displaying an upload complete message. Applications which spend a great deal of time processing in this way should notify the user in advance, again so they do not give up before a response is sent. > No they wouldn't, the modal window would be doing it, in any case rich > applications like this don't use single submission queues, but have > the submission going on in the background (by submitting a new frame, > with the active frame available immediately for normal use) Using a frames to submit in the background is a solution which tries to work around browser deficiencies. My understanding is that this working group is about developing "ideal" solutions from the ground up. Though I may have misunderstood WHAT WG's intentions.(Continue reading)
7 Nov 2004 20:35
Re: Submission progress
On Sun, 7 Nov 2004 19:18:39 +0000, Afternoon <afternoon@...> wrote: > > Equally, the upload is only a component of the time taken to perform a > > post, with a countdown, the user will expect immediate response, even > > if the server is going to spend a number of minutes processing the > > data. > > The browser could sidestep this problem by displaying an upload > complete message. as a web application developer, I don't the browser doing this sort of stuff, I want the browser doing as little as possible, giving me the ability to do more with extra events, sure, that seems reasonably - a submissionProgress event would be welcome (not sure it's essential, but it would be nice) but forcing the UA to throw up dialogs and messages are annoying. > Using a frames to submit in the background is a solution which tries to > work around browser deficiencies. No it's not! it's so that users can carry on working during the submission, whilst it's not the most elegant method, it does work well, the user can continue interacting with the application, dead time in web-applications is the biggest problem, users don't normally expect to not be able to do anything after pressing save. > Though I may have misunderstood WHAT WG's intentions. Who knows, there's only been 2 posts from the WG in the last 2 months, I'm pretty sure this is just a useful dumping ground for ideas now, nothing seems to be going anywhere.(Continue reading)
8 Nov 2004 03:25
Re: Submission progress
On 8 Nov, 2004, at 7:34 AM, Afternoon wrote: > ... > Modern browsers feature useful rich interfaces for monitoring the > progress of file downloads, but are not symmetrical, uploads are > second class citizens. > ... That's the fault of the browsers; it's nothing to do with any What-WG spec. Mozilla doesn't show upload progress, for example <https://bugzilla.mozilla.org/show_bug.cgi?id=257975>, but it used to <https://bugzilla.mozilla.org/show_bug.cgi?id=24197>. > ... > This information should be displayed in a modal popup window, as the > browser window or tab is locked during submission. > ... As Jim Ley said, there is no need for it to be shown in a modal window. Doing that would actually be counterproductive, for three reasons: (1) trying to cancel a mistaken submission the usual way -- by clicking the Stop button in the toolbar -- wouldn't work, so mistaken submissions would be *more* likely to happen; (2) some window managers don't allow a window to be moved or minimized if a modal child window is open (so you wouldn't be able to minimize the browser window while waiting for the upload to finish); (3) it would be inconsistent with other progress windows, which are almost always non-modal. But again, that's nothing to do with What-WG.(Continue reading)
8 Nov 2004 04:19
Re: Enhanced data tables
On 8 Nov, 2004, at 7:35 AM, Afternoon wrote: > ... > Proposed addition to Web Controls 1.0. > > Separate designation for data bearing tables All HTML tables are already data-bearing, unless they're non-conformant. (<td> stands for Table Data.) You might make a case that <table> is more often used incorrectly than correctly, but you would then also need to make a case that creating a "separate designation for data-bearing tables" would solve the problem. The latter would have two main pitfalls. Firstly, authors who had used <table> properly in good faith would be annoyed that *they* were the ones having to change their markup, rather than the authors who had used <table> wrongly to begin with. Secondly, the mistaken authors might *also* start using the new syntax just because it's the cool thing to do, even when it's inappropriate (just like they jumped from using <b> to using <strong> even when it was inappropriate, or from using <i> to using <em> even when it was inappropriate, or from producing "HTML" to producing "XHTML" even when it was not well-formed). > that allow browsers to provide extended data manipulation features > such as: > > * Sorting > Standard controls for sorting by each column/asc+desc There's no reason browsers couldn't have implemented this already,(Continue reading)
RSS Feed