ARgghh--smarter browser??
in [Mac Applications]
|
Prev: iPhone Question?
Next: Using iPhoto 09
From: Barry Margolin on 13 Feb 2010 14:15 In article <130220101012394218%nospam(a)nospam.invalid>, nospam <nospam(a)nospam.invalid> wrote: > In article <roger-B30442.11134413022010(a)freenews.netfront.net>, Roger > <roger(a)roger.net> wrote: > > > Does anyone else here get tired of loading a web page, starting to read > > what you want, then as the page continues to load other elements like > > ads or graphics it jumps all around as it accommodates the newly loading > > elements, and you have to again (maybe more than twice) scroll around to > > relocate the text you were trying to read? > > > > It seems to me it would be easy to program some stability into the > > process, in a browser. I'd love if Safari were the first to do so. > > it doesn't always know the size of all the parts until they're actually > downloaded, notably images. I think a common reason for this is AJAX scripts. After things are done downloading, the script rewrites things on the fly, changing the layout. -- Barry Margolin, barmar(a)alum.mit.edu Arlington, MA *** PLEASE post questions in newsgroups, not directly to me *** *** PLEASE don't copy me on replies, I'll read them in the group ***
From: dorayme on 13 Feb 2010 17:05 In article <roger-B30442.11134413022010(a)freenews.netfront.net>, Roger <roger(a)roger.net> wrote: > Does anyone else here get tired of loading a web page, starting to read > what you want, then as the page continues to load other elements like > ads or graphics it jumps all around as it accommodates the newly loading > elements, and you have to again (maybe more than twice) scroll around to > relocate the text you were trying to read? > > It seems to me it would be easy to program some stability into the > process, in a browser. I'd love if Safari were the first to do so. > One notable reason for this is HTML authors do not employ best authoring practice. To take just one example: if the author has a lot of images and merely legally goes: <img src="..." alt="..."> instead of <img src="..." width="..." height="..." alt="..."> then the page does not know how big the images are until they are fetched. Images have a native size and this is picked up eventually. If the author puts in the widths and heights, then provision is made for the spaces in advance. As the page loads, the images get loaded into prepared correct spaces, no jerkiness. Without the preknowledge of the spaces required, the server to browser exchanges do the best they can, the text is loaded along with the beginnings of the first images mentioned in the HTML and everything is constantly adjusted on the fly, every image is a total surprise - like children coming into class having to bring in desks to suit themselves and arranging them as the frazzled teacher says where and where... -- dorayme
From: dorayme on 13 Feb 2010 17:19 In article <7to278Fuf1U1(a)mid.individual.net>, TaliesinSoft <taliesinsoft(a)me.com> wrote: > On 2010-02-13 10:13:45 -0600, Roger said: > > > Does anyone else here get tired of loading a web page, starting to read > > what you want, then as the page continues to load other elements like > > ads or graphics it jumps all around as it accommodates the newly loading > > elements, and you have to again (maybe more than twice) scroll around to > > relocate the text you were trying to read? > > > > It seems to me it would be easy to program some stability into the > > process, in a browser. I'd love if Safari were the first to do so. > > I've long wanted browsers (I'm using Safari) to have the option of not > displaying any part of a page until the downloading is complete. I hate > what I've dubbed the splip-splap-splop display as the page appears > piece by piece. Quite some time ago I posted my wanting in this regard > and the responses were mainly negative, people feeling that delaying > the presentation would make the browser seem slow. What exactly would seem slow is one question. But what would be slow is simply you would have to wait. You might not be as happy with this as you think, T. I am often getting the information I want and I am been and gone before all is loaded. So I would never use that option. You might not want to wait for a long article to download if you could start reading immediately. The problem with your option is that it is good maybe for some website pages but bad for many others and it is hard to know in advance. But I take your point, maybe browsers could have a button on them so that you can press it when you notice a slow and irritatingly loading page! But hang on! There is such a button/option and I used it extensively for years when on dial up. It is called Command Tab or click the plus sign or double click in the browser space to bring up a new tab. The irritating site loads in the tab you are not watching. Is something irritating when it is not actually being observed? Trees continue to exist when unobserved and even make sounds when falling in deep forests with no one around to hear them. <g> -- dorayme
From: Warren Oates on 13 Feb 2010 20:23 In article <doraymeRidThis-8FB856.09050014022010(a)news.albasani.net>, dorayme <doraymeRidThis(a)optusnet.com.au> wrote: > One notable reason for this is HTML authors do not employ best > authoring practice. To take just one example: if the author has a > lot of images and merely legally goes: True. Also, Ajax generally writes to inner HTML, and the auteur should have the width/height of the container set up in advance. Also, one can dynamically assemble one's page with PHP in the output buffer, and flush the whole thing all at once. It all comes down to good auteurship. -- Very old woody beets will never cook tender. -- Fannie Farmer
From: Jeffrey Goldberg on 13 Feb 2010 22:46
On 2010-02-13 10:13 AM, Roger wrote: > Does anyone else here get tired of loading a web page, starting to read > what you want, then as the page continues to load other elements like > ads or graphics it jumps all around as it accommodates the newly loading > elements, and you have to again (maybe more than twice) scroll around to > relocate the text you were trying to read? > > It seems to me it would be easy to program some stability into the > process, in a browser. I'd love if Safari were the first to do so. This was do-able with HTML 3.2, which could be rendered before completing the first pass (as long as the page author included size information about inline images and the like). But the DOM model in HTML 4.01 often requires the whole document to be read before even the beginning of it could be rendered. And then, of course, there is JavaScript, which adds a whole other dimension to what can happen to a page. (After all, things can move around on the page even after the whole thing is loaded is the included JavaScript tells it to). There are great improvements from HTML 3.2 to HTML 4. Things that are very good ideas. But stable "render as you load" ability got lost. Few browsers made use of the ability to provide stable rendering with HTML 3.2 (lynx certainly did, and possibly Opera), so I guess the guys at w3c decided that it could be sacrificed in a move to better CSS control. -j -- Jeffrey Goldberg http://goldmark.org/jeff/ I rarely read HTML or poorly quoting posts Reply-To address is valid |