Frameset scaling problem <span class="votesInfo" data-xf-init="tooltip" title="2 members like this thread!" >2

http://www.vivabit.com/atmedia2006/sessions/#mobile

ChrisHunt · Jan 26, 2006

how well framesets work in handhelds

As I recollect, your earth-shattering discovery was not that framesets "work" in handhelds, but that they degrade fairly gracefully in one particular brand of handheld.

Surely you can only say that framesets "work" in a particular browser if your instructions, that the screen be divided into seperate regions each displaying a different page, are actually observed. Given their severely truncated screen real-estate, mobile devices are clearly going to struggle with following such instructions - so they come up with what solutions they can. Blackberries (as I recollect) display a list of the pages that would appear in each of the frames. That's a fairly graceful way not to support framesets (though personally I think it should only do this if there's no <noframes> element to explicitly tell it what to do). What other brands of handheld do remains to be discovered.

It seems to me, given their patchy support for CSS and severely limited bandwidth, that the current generation of handhelds are always going to struggle to display pages designed for use on regular computers - whether they use frames or not. If handhelds are a significant part of your audience, you need to tailor lean, mean pages specifically for them.

-- Chris Hunt
Webmaster & Tragedian
Extra Connections Ltd

CliveC · Jan 26, 2006

Chris,

With all due respect I would suggest that you would be better equipped to talk about this once you have seen and used a Blackberry.

If handhelds are a significant part of your audience, you need to tailor lean, mean pages specifically for them.

You might want to attend the session on mobile computing while at@media2006:

The idea of mobile computing is to extract the INFORMATION from a website and leave behind images and fancy layout. Semantic mark-up already achieves much of this goal. Foamcow's site for example already "works" on a mobile device. The ONLY problem arises from having navigation in every page.

One solution is to wrap the menu in something like:
@media handheld {
.menu {display: none}
}
and then find a way of presenting just the menu on a separate page. Hello, that is what a frameset already does when the frames are viewed outside of the frameset.

I do not wish to argue back and forth on this. I just leave the information for open-minded readers to use or not use as they please.

I notice that you never commented on the Berea St thing in the other thread.

Kind regards,

Clive

http://tubularity.com/handheld

CliveC · Jan 26, 2006

For those who are interested try this:
view

in a regular browser.

In the Blackberry or in any lynx-type browser this is what you see:

SCREEN 1

Code:

[u] MENU [/u]

CLICK

SCREEN 2

Code:

MENU
[u]Page 1[/u]
[u]Page 2[/u]
[u]Page 3[/u]

CLICK Page 1

SCREEN 3

Code:

[i]Your page 1 content[/i]

Because of the nature of the Blackberry interface users know to click the back button when they are at a dead-end

BACK
Hey Presto! Back to the Menu

SCREEN 2

Code:

MENU
[u]Page 1[/u]
[u]Page 2[/u]
[u]Page 3[/u]

Clive

http://www.vivabit.com/atmedia2006/

ChrisHirst · Jan 27, 2006

I'll just throw this into discussion WRT Search Engines.

One client who had a framed site where the pages did end up as orphans in the SEs.
Yes they did have a "homepage" link,
Yes they did have related product links on pages,
and still little more than a third of pages had been indexed and the site had been around for 3 years.
When I took on the site some <noframes> content and site maps were implemented, this increased the indexed page count to around 3/4 of the total, took around 6 months to happen.
At the same time I added a self-referencing frameset script to all existing pages. The day after these pages were uploaded, sales increased by 10% and continued to rise for a few days. The first month showed a 20% increase overall and with a reduction in the PPC bill from ~£3000 to ~£900 per month (thats GBP NOT USD) once the noframes/sitemaps took effect the client was pretty happy.

The site has been redone since without frames (Not by me, I simply play a consulting role now) and SE referals and sales have increased, every product page is now indexed by the SEs and PPC is still relatively low and showing a very good ROI. That is all I need to know about frames.

They have their place, but a commercial site is not one of them anymore, if it ever actually was of course.

Chris.

Indifference will be the downfall of mankind, but who cares?
Woo Hoo! the cobblers kids get new shoes.
People Counting Systems

So long, and thanks for all the fish.

CliveC · Jan 27, 2006

Chris II,

With respect, your contribution is anecdotal and hence specious. We do not know the clients name... even if we did the framed site is long-gone.

You may note that my original post simply said:

I have yet to see any example of a frameset being less searchable. I wish someone could provide one.

This has yet to happen!

I would suggest that the only way that real evidence could be found would be by a study of site-logs.

I wish that you would understand that I do not give a rats-a** about whether you or anyone else uses framesets. I simply do not think that it is appropriate, in a technical forum, to have threads of the genre

Question: I have a problem with a frameset...
Answer: Don't use frames.

It is analogous to:

Question: I am having problem with a CSS layout
CliveC's Answer: Don't use CSS layouts.

In that case I might be justifiably flamed. However it might still be a more reasonable comment.

EXAMPLE:
View

in an IE browser at 800*600. The columns overlap. As you dig into the site it gets worse.

http://www.vivabit.com/atmedia2006/lastyear/.

Also the CSS does not validate. Great advertisement for modern web design no? Many CSS layouts can be broken. Simply by increasing font size.

Clive

CliveC · Jan 27, 2006

Guys,

This is getting ridiculous:

Code:

Date : Fri, 27 Jan 2006 [red]14:21:03 -0500[/red]
Subject : Tubularity - Visit 
Someone visited the site. 
Their IP address was: 86.140.238.48 
Their host was: host86-140-238-48.range86-140.btcentralplus.com 
Their referer is: [URL unfurl="true"]http://www.tek-tips.com/viewthread.cfm?qid=1178980&page=1[/URL] 
Their browser is: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-GB; rv:1.8) Gecko/20051111 Firefox/1.5

Date : Fri, 27 Jan 2006 [red]14:21:09 -0500[/red]
Subject : Tubularity - Visit 
Someone visited the site. 
Their IP address was: 128.30.52.13 
Their host was: lovejoy.w3.org 
Their referer is: 
Their browser is: W3C_Validator/1.432.2.5

This is not a site I even care much about since I don't want any more consulting work. I just use it mainly to experiment with ideas.

Pretty soon you will be able to go to Google and just enter "TU" and get your first hit as tubularity.com.

Clive

ca8msm · Jan 27, 2006

I have yet to see any example of a frameset being less searchable. I wish someone could provide one.

There are warnings from the search engines themselves that it could cause problems when they try to crawl your site. I'm not saying it does in all cases but it increases the risk. For example,

Google said:
It's also possible that we're not able to crawl your site due to technical reasons. A few of the most common ones are listed below:

Your pages use frames. Google supports frames to the extent that we can. Frames tend to cause problems with search engines, bookmarks, emailing links and so on, because frames don't fit the conceptual model of the web (every page corresponds to a single URL). If a user's query matches the page as a whole, Google returns the frame set. If a user's query matches an individual frame on the page, Google returns the URL for that frame. The page is not displayed in a frame because there may be no frame set corresponding to that URL.

Also, examples have been given of the three main problems that they list above (bookmarking, emailing links and orphan pages). Add to that their admission that they may have problems crawling framed sites, then developers should consider carefully if they are going to use frames for a website.

____________________________________________________________

Need help finding an answer?

Try the Search Facility or read FAQ222-2244 on how to get better results.

CliveC · Jan 27, 2006

ca8msm,

We have been here before. See post 6. and as far as the other "problems" we have been around and around in an endless loop in over 100 recent posts.

I think it is time to drop the subject and let "developers" decide for themselves or else expand the discussion.

If you want, here is something to think about over the weekend:

Imagine a three page website with the same 3 menu links in each page, linking to the other pages in the site. If you were programming a webbot, starting at the index page and continuing forward, would you revisit links that you had already been down?

Clive

ca8msm · Jan 28, 2006

Clive,

I know we've been here before but it's you who brought it back up! You posted your quote and then then said "This has yet to happen!". I simply replied to you, gave you an example piece of text from the biggest search engine that said they could have problems crawling a framed website.

If you think it's time the subject was dropped then don't post questions as someone will most probably answer them. There has been enough information on both threads that highlight the problems, so developers now should have enough information to make an informed decision.

____________________________________________________________

Need help finding an answer?

Try the Search Facility or read FAQ222-2244 on how to get better results.

CliveC · Jan 28, 2006

QAS - Question Avoidance Syndrome

cas8msm,

The question is directed to you.
It is a simple question.
It has a yes/no answer.
It does not require much thought.

Answer Yes or No and that can be the last word. Promise!

Clive

http://tubularity.com/bookmark/

ca8msm · Jan 28, 2006

OK Clive, let's play it your way then. I won't comment on anything, I'll simply answer your question.

My Answer: No

Now, I have some questions on the three page website with the same 3 menu links in each page, linking to the other pages in the site for you.

Same rules, Yes or No - simple 3 page website, simple questions, simple answers. If you were to create the website using frames, what would your answers be to the following questions:

1) Would your user's have a problem with bookmarking individual pages?
2) Would your user's have a problem with emailing links for individual pages?
3) Do search engines support your site fully?
4) Would a user, using a search engine, receive results for individual pages that were not meant to be viewed on their own?
5) Can a user without javascript go to any URL on your site and view the page in the context it was meant to be viewed?

Now, you build the site again, this time without frames. What are your answers to the same questions:

6) Would your user's have a problem with bookmarking individual pages?
7) Would your user's have a problem with emailing links for individual pages?
8) Do search engines support your site fully?
9) Would a user, using a search engine, receive results for individual pages that were not meant to be viewed on their own?
10) Can a user without javascript go to any URL on your site and view the page in the context it was meant to be viewed?

____________________________________________________________

Need help finding an answer?

Try the Search Facility or read FAQ222-2244 on how to get better results.

CliveC · Jan 28, 2006

ca8msm, I see that you were not content to have the last word.

Perhaps you see the conundrum. If you had answered incorrectly then your bot would be in an endless loop. Answering correctly is an ipso facto acknowledgement that a bot cannot crawl a frameset any differently from the way it crawls a non-framed site.

Since your questions have been asked and answered dozens of times, and since you will not accept the answers, look at this instead and maybe you can arrive at your own conclusions. No need to post them if you don't want to.

This is a site that was googled the day I put it up (as most of my sites are whether framed or unframed).

Here are the relevant logs from a re-googling on the 24th:

Code:

#Software: Microsoft Internet Information Services 6.0
#Version: 1.0
#Date: 2006-01-24 00:32:31
#Fields: date time  cs-method cs-uri-stem cs-uri-query s-port cs-username c-ip cs-version cs(User-Agent) cs(Cookie) cs(Referer) cs-host sc-status sc-substatus sc-win32-status sc-bytes cs-bytes time-taken 
2006-01-24 00:32:31 GET /robots.txt - 80 - 66.249.65.174 HTTP/1.1 Mozilla/5.0+(compatible;+Googlebot/2.1;++[URL unfurl="true"]http://www.google.com/bot.html)[/URL] - - tubularity.com 404 0 2 1795 247 31
2006-01-24 00:32:31 GET /bookmark/index.html - 80 - 66.249.65.174 HTTP/1.1 Mozilla/5.0+(compatible;+Googlebot/2.1;++[URL unfurl="true"]http://www.google.com/bot.html)[/URL] - - tubularity.com 200 0 0 423 239 78
2006-01-24 00:34:00  GET /bookmark/about.html - 80 - 66.249.65.174 HTTP/1.1 Mozilla/5.0+(compatible;+Googlebot/2.1;++[URL unfurl="true"]http://www.google.com/bot.html)[/URL] - - tubularity.com 200 0 0 423 239 31
2006-01-24 00:34:22 GET /bookmark/contact.html - 80 - 66.249.65.174 HTTP/1.1 Mozilla/5.0+(compatible;+Googlebot/2.1;++[URL unfurl="true"]http://www.google.com/bot.html)[/URL] - - tubularity.com 200 0 0 426 241 46
#Software: Microsoft Internet Information Services 6.0
#Version: 1.0

QED - quod erat demonstrandum

Clive

ca8msm · Jan 28, 2006

QAS - Question Avoidance Syndrome

Suffering from the same problem are you Clive?

____________________________________________________________

Need help finding an answer?

Try the Search Facility or read FAQ222-2244 on how to get better results.

CliveC · Jan 31, 2006

CA8MSM,

Sorry, I thought my example answered your questions.
NNYNY & NNYNY

Clive

http://www.co.wise.tx.us/Default.htm

ca8msm · Jan 31, 2006

Clive,

I must admit, I'm a little baffled at some of your answers. If I go through the questions maybe you could shed some light on your choice of answers:

1) Would your user's have a problem with bookmarking individual pages?

You answered No. I would have thought that the correct answer would be Yes. For example, the first website that I found by searching google for "this site uses frames" came up with

which I'm going to use as an example. If I click on the "Directory" menu item from the navigation frame, the content changes to show the relevant data. However, if I bookmark this page, when I click on my bookmark, I'm actually sent to the default home page and not the page I thought I was bookmarking. Isn't this a problem bookmarking individual pages?

2) Would your user's have a problem with emailing links for individual pages?
Again, taking the same example as above where I am on the "directory" page, if I go to the File menu on my browser and choose to send the link via email, the URL that is put in the email is for the default homepage and not the directory. Isn't this a problem with emailing individual page links?

3) Do search engines support your site fully?
This is something that seems to be in the "gray" area. You've posted logs that show google crawling your site and you don't seem to have any problems with search engines picking up your content. You may very well be correct and there may be no problems with SE's crawling framed sites, it's just that when SE's such as google say things like "Google supports frames to the extent that we can. Frames tend to cause problems with search engines" it tends to make me a little nervous and think that they may be telling the truth.

4) Would a user, using a search engine, receive results for individual pages that were not meant to be viewed on their own?
Another one that you answered No with where my opinion would be that it should be a Yes. For example, if I search google for "Wise County Information Technology Directory" the 5th entry is for the page I was viewing in my earlier example, however, when I click on the link, I see the directory but not in the context it was meant to be viewed as there is no frameset and therefore no navigation menu for me to use.

5) Can a user without javascript go to any URL on your site and view the page in the context it was meant to be viewed?
I thought my example in thread253-1180038 (dated 23 Jan 06 7:19) highlighted a problem that non-javascript users would get and therefore, in my opinion, I'd say the answer should be No.

I do, however, agree with your answers on questions 6-10.

I hope you didn't take my post the wrong way, but I thought that this would be a good way of highlighting what I think are problems.

____________________________________________________________

Need help finding an answer?

Try the Search Facility or read FAQ222-2244 on how to get better results.

CliveC · Jan 31, 2006

Did you look at the example?
I am at the Clinton library today.
Do you have any questions for Bill & Hillary?

http://tubularity.com/wjclib.jpg

Clive

ca8msm · Jan 31, 2006

Sorry, yes I did look at the example. I guess it's actually quite a good workaround in that it does allow the users to bookmark/email links.

It does create additional work for the developer though as they will have to create a frameset page for every page on their site though and depending on the size of your site that may be a lot of additional work. It is also a hack involving a lot more effort than if you hadn't used frames (i.e. you wouldn't have had to anything).

____________________________________________________________

Need help finding an answer?

Try the Search Facility or read FAQ222-2244 on how to get better results.

CliveC · Jan 31, 2006

ca8msm,

With respect, it is difficult to have a rational argument with someone who keeps wanting to change the rules as they go along.

In web development almost everything beyond simple mark-up needs some extra work to be cross-browser compliant.

When it comes to the subject of framesets it seems that you only want to discuss perceived disadvantages and ignore advantages and then pretend that framesets are the only part of web development that have problems that cannot possibly be solved.

If I use a frameset when needed, I don't even need to test it on all the different browsers. The reason, because if one of the most popular browsers did not support framesets surely someone would have reported it on tek-tips or elsewhere by now.

As I understand it from your comments, any solution to any problem must NOT rely on Javascript, neither must it cause any extra work from the developer and it can't be a workaround.

OK! solve this problem. The user's requirement is that one portion of the screen remains stationary while the rest can scroll. No JavaScript, no extra work, no workarounds.

Kind regards,

Clive