<hr>/IE problem...

Michael Fesser wrote:

.oO(GTalbot)

the page has huge problems to fix before even starting to investigate
the layout issue with the <hr

- 9,000 lines of code is insane
- hundreds of validation markup errors
- dozens of abs. pos. elements
- the page never starts to load and I'm on broadband
- no doctype declaration

- it loads all images at once (>5 MB)
- no chance without JS (except for turning off CSS - then you can see
all images on one page)
- thousands of empty lines in the markup
- relying on fixed font size (in my browsers the links to pictures 21
and 22 overlap with the photo)

Nice pictures, but completely broken presentation.

Micha

and the reason there are so many empty lines is that the code is
generated by the software code, not hard-coded (using java/JSP..)
that's how the code is generated...

..oO(maya)

Michael Fesser wrote:

- it loads all images at once (>5 MB)
- no chance without JS (except for turning off CSS - then you can see
all images on one page)
- thousands of empty lines in the markup
- relying on fixed font size (in my browsers the links to pictures 21
and 22 overlap with the photo)

Nice pictures, but completely broken presentation.

can u pls post a screenshot?

Not necessary. Just open Firefox and increase the font size one or two
steps. I have a minimum font size of 13px in all of my browsers, which
already causes a line break after the "20" in the top navigation bar.

(and yes, like 90% of websites out there, this one needs JavaScript
-- and CSS -- in order to function properly...)

There is _always_ a way to do it without JS. I consider sites that rely
on JS without alternative broken. In this case all it needs to fix it
would be a better server-side script to generate the gallery links and
to use a URL parameter to determine which image should be displayed:

/photoblog_new/india/page9/photos.jsp?img=1
/photoblog_new/india/page9/photos.jsp?img=2
/photoblog_new/india/page9/photos.jsp?img=3
....

Or something like that.

Micha

Michael Fesser wrote:

.oO(maya)

Michael Fesser wrote:
- it loads all images at once (>5 MB)
- no chance without JS (except for turning off CSS - then you can see
all images on one page)
- thousands of empty lines in the markup
- relying on fixed font size (in my browsers the links to pictures 21
and 22 overlap with the photo)

Nice pictures, but completely broken presentation.
can u pls post a screenshot?

Not necessary. Just open Firefox and increase the font size one or two
steps. I have a minimum font size of 13px in all of my browsers, which
already causes a line break after the "20" in the top navigation bar.

(and yes, like 90% of websites out there, this one needs JavaScript
-- and CSS -- in order to function properly...)

There is _always_ a way to do it without JS. I consider sites that rely
on JS without alternative broken. In this case all it needs to fix it
would be a better server-side script to generate the gallery links and
to use a URL parameter to determine which image should be displayed:

yes, technically of course, everything can be done server-side, but good
web-devedlopment practice says if you can do it client-side, in order to
not overburden the server, you should.. what is wrong with doing it
client-side? (a lot of this, in fact MOST of the stuff here, IS done on
the server, for example I count the no. of images in each 'images'
folder to generate the divs... THAT has to be done on the server, since
you need access to the 'images' folder.. I use java for that and for
generating all the divs-code, but the function to flip from photo to
photo is a JavaScript function.. sorry, but really would like to know
why this is so wrong..

(and what about sites that are purely HTML, that cannot run scripts on
the server??)

/photoblog_new/india/page9/photos.jsp?img=1
/photoblog_new/india/page9/photos.jsp?img=2
/photoblog_new/india/page9/photos.jsp?img=3

actually, I do have functionality in there that you can have links to
particular images...
http://www.francesdelrio.com/photobl...hotos.jsp?pn=7

I see what you're saying, but that would mean reloading entire page
every time to flip from photo to photo... ok, either way thank you for
taking the time..

...

Or something like that.

Micha

Jeff wrote:

maya wrote:
maya wrote:
dorayme wrote:
In article
95d82a6c-8ded-48e7-85ba-2d5c157927a6...ooglegroups.co
m>,
GTalbot <newsgroup (AT) gtalbot (DOT) org> wrote:

On 16 jan, 14:59, maya <maya778... (AT) yahoo (DOT) com> wrote:
amazing, the stupid things IE does sometimes...

inserted a simple <hr> tag in sidenav
here,http://www.francesdelrio.com/photobl...ge9/photos.jsp

in FF it looks fine, in IE (both 6 and 7) it's messed up.. <hr> gets
stretched out as far as the width of the window.. (in the
interests of
"fluid" design I have not given the sidenav div a fixed width.. now
this is causing me problems in IE... oh brother...)

thank you....
Maya,

the page has huge problems to fix before even starting to investigate
the layout issue with the <hr

- 9,000 lines of code is insane
- hundreds of validation markup errors
- dozens of abs. pos. elements
- the page never starts to load and I'm on broadband
- no doctype declaration

Just create a reduced testcase and upload it on a web-accessible
webpage. Make sure the page passes markup code validation and CSS code
validation. Then we will look and investigate the issue with the <hr>.

Regards, Gérard

Amazing also - in that case - that it loaded fine for me in
Australia when I mentioned the hundreds of errors hours ago...

This is the url:

http://www.francesdelrio.com/photobl...ge9/photos.jsp



thank you all for yr responses... believe it or not, what happened
is that literally minutes after I posted my question I started having
problems w/my site, browser would hang, etc.. have no idea what
happened, I emailed my webhosting, the guy said there had been a
little problem and he had had to restart my server.. as to the code:

this is GENERATED code, it's generated with JSP's.. yes, there are
many divs... each photo has its own div, and I use JavaScript to
move from div to div (i.e., from photo to photo..), so yes, there a
lot of divs... I don't write the divs, they're generated with java
code (even some of the CSS is genearted with java..)

as for the CSS, I just checked above url in W3C CSS validator, it
validates.. (can't post url, it's too long..)

as for HTML validation, find no errors when check in HomeSite... in
the W3C validator, however, it shows lots of errors that I don't get
(has problems with all special chars (&nbsp;...etc) for one thing...
so where does this leave me??

thanks for comments about the photos.. my photo equipment, to answer
one of the respondents, is a digital Nikon D80, with with a 18-135mm
lens...

thank you..

also when I do give it a fixed width, either in px or in em, it gets a
different width on IE from FF... man, what a pain....

Well, you are running this in quirksmode when you don't have a strict
doctype with a path.

Browsers then have different interpretation of whether to include any
padding and borders in the width.

Jeff


thank you...

point well-taken about doctype.. I HAD neglegted that... just
inserted the "transitional" one.. (never sure whether to use that one or
the "strict" one..)

thank you..

On 2008-01-17, maya wrote:
....

and the reason there are so many empty lines is that the code is
generated by the software code, not hard-coded (using java/JSP..)
that's how the code is generated...

Then the code generator needs to be fixed. (Or the output piped
through a filter to remove consecutive blank lines.)


--
Chris F.A. Johnson <http://cfaj.freeshell.org>
================================================== =================
Author:
Shell Scripting Recipes: A Problem-Solution Approach (2005, Apress)

..oO(maya)

Michael Fesser wrote:

There is _always_ a way to do it without JS. I consider sites that rely
on JS without alternative broken. In this case all it needs to fix it
would be a better server-side script to generate the gallery links and
to use a URL parameter to determine which image should be displayed:

yes, technically of course, everything can be done server-side, but good
web-devedlopment practice says if you can do it client-side, in order to
not overburden the server, you should..

That's what a server is for - to serve content to a client. On a normal
website the server CPU is usually idling around anyway most of the time.

The point is: if the server does it, it works always, it's reliable. If
the client has to do it, it doesn't work always because it relies on the
client's capabilities, which you can't control.

what is wrong with doing it
client-side?

It's unreliable. I also use client-side scripting from time to time, but
always only as an extension, for example to provide some nice-to-have or
convenience features. But on my web sites JS is always added _after_ the
main things work as expected with purely server-side scripting.

(a lot of this, in fact MOST of the stuff here, IS done on
the server, for example I count the no. of images in each 'images'
folder to generate the divs... THAT has to be done on the server, since
you need access to the 'images' folder.. I use java for that and for
generating all the divs-code, but the function to flip from photo to
photo is a JavaScript function.. sorry, but really would like to know
why this is so wrong..

Said above. And besides the unreliability another problem mentioned is
the load on the client. You're worried about overburden the server, but
what about the clients? With your current approach _all_ the images are
downloaded, that's more than 5 MB in the given gallery. Even on a fast
broadband connection this takes some time, not to mention slower ones or
even dial-up, which is still used in many, many regions.

You're moving the load from the server to the network and the clients,
which wastes much more resources than some simple server scripts. And,
ironically, it might also increase the load on your server again because
of the increased traffic. Even if a user just wants to have a quick look
through your galleries, he might unintentionally download all the 5 MB,
even if he isn't really interested in pictures from India. 5 MB wasted
on your server, on the network and the client's connection for nothing.

(and what about sites that are purely HTML, that cannot run scripts on
the server??)

Get another host? LAMP hosting is really cheap. Here in Germany for
example I can get a good one for 2.99 EUR/month (including cronjobs,
SSH, .htaccess, my own php.ini, 15 G traffic, log file analyzer etc.)

But for such simple gallery sites you don't even need full-powered
server-side scripting capabilities. You could also generate all the
pages on your local machine and then just upload the static HTML pages.
With some shell scripts and command line tools this can be automated
quite well, at least on a *nix box (with using make or Ant).

/photoblog_new/india/page9/photos.jsp?img=1
/photoblog_new/india/page9/photos.jsp?img=2
/photoblog_new/india/page9/photos.jsp?img=3

actually, I do have functionality in there that you can have links to
particular images...
http://www.francesdelrio.com/photobl...hotos.jsp?pn=7

OK, but I can't bookmark it from just seeing it in the gallery when
clicking through. The address bar never changes to reflect the current
position. So where does this link come from? Would I have to manually
add the "pn" parameter?

I see what you're saying, but that would mean reloading entire page
every time to flip from photo to photo...

Correct, but this is not a problem! The size of the HTML would be maybe
4-5 KB, if at all. In fact this is how it works on billions of websites!
It's common, it's reliable, it's usable and accessible. It also makes
your images _easily_ bookmarkable. Of course then one could add client-
side scripting to improve the user experience, for example by preloading
the next image in the background like it's done in some JS-based gallery
scripts (like Lightbox 2 and its derivates).

BTW: Some browsers like Firefox already do something similar. If you use
a 'link' element in the head of your documents to indicate the next
document, FF will automatically download it in the background. If you
then click a link to go to the next page, it's already available in the
browser cache. Quite nice.

Micha

..oO(maya)

point well-taken about doctype.. I HAD neglegted that... just
inserted the "transitional" one.. (never sure whether to use that one or
the "strict" one..)

For new sites always "Strict", since you're not transitioning from
anything.

Micha

maya wrote:

Jeff wrote:
maya wrote:
maya wrote:
dorayme wrote:
In article
95d82a6c-8ded-48e7-85ba-2d5c157927a6...ooglegroups.co
m>,
GTalbot <newsgroup (AT) gtalbot (DOT) org> wrote:

On 16 jan, 14:59, maya <maya778... (AT) yahoo (DOT) com> wrote:
amazing, the stupid things IE does sometimes...

inserted a simple <hr> tag in sidenav
here,http://www.francesdelrio.com/photobl...ge9/photos.jsp

in FF it looks fine, in IE (both 6 and 7) it's messed up.. <hr
gets
stretched out as far as the width of the window.. (in the
interests of
"fluid" design I have not given the sidenav div a fixed width.. now
this is causing me problems in IE... oh brother...)

thank you....
Maya,

the page has huge problems to fix before even starting to investigate
the layout issue with the <hr

- 9,000 lines of code is insane
- hundreds of validation markup errors
- dozens of abs. pos. elements
- the page never starts to load and I'm on broadband
- no doctype declaration

Just create a reduced testcase and upload it on a web-accessible
webpage. Make sure the page passes markup code validation and CSS
code
validation. Then we will look and investigate the issue with the
hr>.

Regards, Gérard

Amazing also - in that case - that it loaded fine for me in
Australia when I mentioned the hundreds of errors hours ago...

This is the url:

http://www.francesdelrio.com/photobl...ge9/photos.jsp



thank you all for yr responses... believe it or not, what happened
is that literally minutes after I posted my question I started
having problems w/my site, browser would hang, etc.. have no idea
what happened, I emailed my webhosting, the guy said there had been
a little problem and he had had to restart my server.. as to the code:

this is GENERATED code, it's generated with JSP's.. yes, there are
many divs... each photo has its own div, and I use JavaScript to
move from div to div (i.e., from photo to photo..), so yes, there a
lot of divs... I don't write the divs, they're generated with java
code (even some of the CSS is genearted with java..)

as for the CSS, I just checked above url in W3C CSS validator, it
validates.. (can't post url, it's too long..)

as for HTML validation, find no errors when check in HomeSite... in
the W3C validator, however, it shows lots of errors that I don't get
(has problems with all special chars (&nbsp;...etc) for one
thing... so where does this leave me??

thanks for comments about the photos.. my photo equipment, to
answer one of the respondents, is a digital Nikon D80, with with a
18-135mm lens...

thank you..

also when I do give it a fixed width, either in px or in em, it gets
a different width on IE from FF... man, what a pain....

Well, you are running this in quirksmode when you don't have a strict
doctype with a path.

Browsers then have different interpretation of whether to include
any padding and borders in the width.

Jeff


thank you...


point well-taken about doctype.. I HAD neglegted that... just
inserted the "transitional" one.. (never sure whether to use that one or
the "strict" one..)

What does a doctype do when it comes to rendering html? For years they
did nothing at all, all it did was make the validator happy. That's it.

Then came a generation of browsers that switched from the old box model
they were using to a standards compliant box model if and only if the
doctype was strict and it had a url.

Now, some may care if the validator complains about errors and use a
transitional doctype instead just to get rid of errors. I say, why care
about what the validator thinks. The browser doesn't. Use the strict
with url if you want a box model more or less standards compliant and
the same across browsers. This doesn't mean you shouldn't endeavor to
write clean html, just don't get anal about it.

As pointed out elsewhere, use strict for all new html. You might as
well not have a doctype if use transitional (or strict without a url),
you won't notice any rendering differences.

If there's a doctype other than strict that actually does something
with a browser, someone tell me what it is and what it does. I don't
keep up with everything. Perhaps the xhtml, but there's no real reason
to use those for plain html.

Somebody that knows something current about getting *out* of quirks
mode should follow up.

Jeff

thank you..

In article <13p3bi0a61r7pfd (AT) corp (DOT) supernews.com>,
Jeff <jeff (AT) spam_me_not (DOT) com> wrote:

maya wrote:
Jeff wrote:
maya wrote:
maya wrote:
dorayme wrote:
GTalbot wrote:
maya wrote:
everyman's dog wrote:

point well-taken about doctype.. I HAD neglegted that... just
inserted the "transitional" one.. (never sure whether to use that one or
the "strict" one..)

What does a doctype do when it comes to rendering html? For years they
did nothing at all, all it did was make the validator happy. That's it.

Then came a generation of browsers that switched from the old box model
they were using to a standards compliant box model if and only if the
doctype was strict and it had a url.

....
If there's a doctype other than strict that actually does something
with a browser, someone tell me what it is and what it does. I don't
keep up with everything. Perhaps the xhtml, but there's no real reason
to use those for plain html.

Somebody that knows something current about getting *out* of quirks
mode should follow up.

Well, Jeff, as I understand this stuff, it is not a simple matter
of the strictness of doctype but in some ways the very presence
of a doctype (there are complicated tables of necessary and
sufficient conditions - see the bottom of page <url:
http://hsivonen.iki.fi/doctype/>) does the triggering. I avoid
having to understand all the arcane details by a simple strategy,
I just use one necessary condition I know, namely 4.01 Strict.

See best perhaps:

<http://www.quirksmode.org/css/quirksmode.html>

and there have been discussions, of course, on this ng about the
matter.

--
dorayme

dorayme wrote:

In article <13p3bi0a61r7pfd (AT) corp (DOT) supernews.com>,
Jeff <jeff (AT) spam_me_not (DOT) com> wrote:

maya wrote:
Jeff wrote:
maya wrote:
maya wrote:
dorayme wrote:
GTalbot wrote:
maya wrote:
everyman's dog wrote:
point well-taken about doctype.. I HAD neglegted that... just
inserted the "transitional" one.. (never sure whether to use that one or
the "strict" one..)
What does a doctype do when it comes to rendering html? For years they
did nothing at all, all it did was make the validator happy. That's it.

Then came a generation of browsers that switched from the old box model
they were using to a standards compliant box model if and only if the
doctype was strict and it had a url.

...
If there's a doctype other than strict that actually does something
with a browser, someone tell me what it is and what it does. I don't
keep up with everything. Perhaps the xhtml, but there's no real reason
to use those for plain html.

Somebody that knows something current about getting *out* of quirks
mode should follow up.


Well, Jeff, as I understand this stuff, it is not a simple matter
of the strictness of doctype but in some ways the very presence
of a doctype (there are complicated tables of necessary and
sufficient conditions - see the bottom of page <url:
http://hsivonen.iki.fi/doctype/>) does the triggering. I avoid
having to understand all the arcane details by a simple strategy,
I just use one necessary condition I know, namely 4.01 Strict.

Thanks,

I see that my doctype knowledge is outdated, at least as far as the
sniffing goes. Trouble with html is that you have to throw away part of
what you know every few years.

I see that they are working on html 5 again:
<URL: http://www.whatwg.org/specs/web-apps/current-work/>

At least strict with a URL still works and covers the widest range of
browsers.

I seem to be initially too far ahead of the curve and then I fall
behind. I was working with streaming video in '96, with flash
predecessors back in '97, AJAX predecessors in '98.

All of those are big time now, and I know very little about the
current technology, same with the doctypes until you updated me.

I've spent the last couple days working through equal height
columns... and using some of your container height tricks.

Jeff

See best perhaps:

http://www.quirksmode.org/css/quirksmode.html

and there have been discussions, of course, on this ng about the
matter.