110104 - trefsgar.com - Directly accessing elements with "id" attribute fails in Javascript

Status: RESOLVED FIXED

(Tech Evangelism Graveyard :: English US, defect, P1)

Description

[Alex Weyers]

From Bugzilla Helper:
User-Agent: Mozilla/4.0 (compatible; MSIE 5.5; Windows NT 5.0)
BuildID:    20011019

The scrolling text on the above site fails to display and scroll utilizing 
Netscape 6.1-6.2.  This is because it contains javascript that references a DIV 
element by id tag directly.

For a more base case example, see 
http://www.geocities.com/debugweyers/news.html.

Instead of utilizing document.getElementById("news") (which works on both IE 
and NS), I simply try to access the div element directly 
("news.style.letterSpacing").  Attempts to reference an element by id without a 
call to getElementById fail.

Reproducible: Always
Steps to Reproduce:
1.go to http://www.trefsgar.com in Navigator

Actual Results:  Page loads, but there is a blank spot in the table.

Expected Results:  There should be javascript-driven scrolling text in the 
blank spot of the table.

It appears that IE exposes ids of elements to the namespace of javascript.  
Adam Locke and I poked around and could not find any clear spec as to whether 
or not this violates standards (hopefully, some DOM expert will materialize 
with an instant answer).  Obviously, if the only way to retrieve an element is 
using the getElement methods, then this bug needs to be evangelized to the site.

Comment 1

[Johnny Stenback (:jst)]

Over to evangelism, polluting the global namespace with element id's is a IE:ism
that won't be supported in mozilla.

Comment 2

[Bob Clary [:bc] (inactive)]

yes, all element references need to be obtained from the DOM via getElementById
or other means. 

Comment 3

[Boris Zbarsky [:bzbarsky]]

*** Bug 110281 has been marked as a duplicate of this bug. ***

Comment 4

[Bob Clary [:bc] (inactive)]

Alex, why are you putting edt and topembed keywords on this bug? This is an
evangelism bug and as such, those keywords have no meaning. If you wish to drive
changes in the browser, then file a bug against the DOM and lobby to have it
included in the edt/topembed process.

Removing edt0.9.4, topembed keywords

Comment 5

[Marek Z. Jeziorek]

The issue is the from our embedding customer point of view it has high priority
(it is tracked as such). How do we mark high priority evangelism bugs?

Comment 6

[Alex Weyers]

Sorry about that.

I was copying over the keywords from Bug 110281 (which was closed and marked a 
duplicate of this one).  As duplicate bugs, I figured they should share the 
same priority (and they do in fact bear a high priority from the standpoint of 
an embedding customer).

Comment 7

[Bob Clary [:bc] (inactive)]

If it is indeed solely an evangelism issue then priority p1 and severity will
get it on my radar. Per your comments I will bump the priority / severity for
this bug. I will contact the site as soon as I can. When I do I will mark the
bug assigned to me and set the target milestone for a follow testing date to see
if they have fixed their site.  See http://mozilla-evangelism.bclary.com/ for
more details on bugzilla/mozilla evangelism and you can of course contact me
directly.

If you really intend to drive product changes then you need to file product bugs
since that will get the engineer's attention. FWIW, there will be great (and
justifiable) opposition to reproducing IE's behavior of exposing all Elements
with ID attributes as global objects.

Comment 8

[Bob Clary [:bc] (inactive)]

Also, if you need to query on these bugs we can agree on something you can add
to the Status whiteboard. Then you can query on that to locate your high
priority evangelism issues. However when you use product related keywords in
evangelism bugs, it messes up the planning/queries that the product team uses to
track the product issues they need to fix. Therefore product related keywords
are not appropriate in evangelism bugs. I hope this clarifies things.

Comment 9

[Bob Clary [:bc] (inactive)]

Can we just use bugscape 10967 rather than the entire url in the status
whiteboard? It is not not clickable and really horks the display in a buglist
and is not even completely visible when looking that the entire bug. Or even
just bugscape with the link pasted into the comments. How do you want to use
this info? in queries or to get the actual link?

Comment 10

[Bob Clary [:bc] (inactive)]

See http://bugscape.netscape.com/show_bug.cgi?id=10967

contact info: [e-mail]

Error: news is not defined
Source File: http://www.trefsgar.com/
Line: 72

Dreamweaver 3.0 code

function MM_findObj(n, d) { //v3.0

<!-- Original:  news scroller v2.0 rev 2 30mar00 -->
<!-- Original:  Dion ([e-mail]) -->
<!-- Web Site:  http://www.iinet.net.au/~biab -->

Contacted.

Comment 11

[Michael Dunn]

I was just following someone elses pattern; the shorter version is fine.

Comment 12

[Jennifer Kobayashi]

Adding keywords topembed and edt0.9.4, based on bugscape bug 10967

Comment 13

[Bob Clary [:bc] (inactive)]

how many times to I have to explain that product related keywords have no
meaning for evangelism bugs? removing keywords: edt0.9.4, topembed

Comment 14

[Jennifer Kobayashi]

Sorry Bob, I had blanket instructions... so just following orders. My apolgoies 
on the confusion. 

Comment 15

[Bob Clary [:bc] (inactive)]

contacted again

Comment 16

[magus1011ca]

This bug is a violation of the W3C specification for the id attribute as found 
in section 7.5.2 of specification 4.01 specifically: (note stars)

The id attribute has several roles in HTML:

As a style sheet selector. 
As a target anchor for hypertext links. 
As a means to reference a particular element from a script. *****
As the name of a declared OBJECT element. 
For general purpose processing by user agents (e.g. for identifying fields when 
extracting data from HTML pages into a database, translating HTML documents 
into other formats, etc.).

Please stop attempting to avoid correcting it via evangelism and fix it. 

Thank You.

Comment 17

[Bob Clary [:bc] (inactive)]

The HTML 4.01 recommendation does not specify how a script should access an
element by id value. The DOM Core Level 2 recommendation does however. The use
of the id attribute as a global variable to access an element with that id
attribute is not a standard. document.getElementById('idvalue') is the standard
way to access an element by id value. Please get a clue before you quote standards.

Thank you.

Comment 18

[Bob Clary [:bc] (inactive)]

->default

Comment 19

[philip mccaffrey]

as i say i am extremely naive in this sphere but possess an intuitive
intelligence (I was informed by a competitive mentor at university,the only
one I've had;and a place i had no ambition to go to - my family 'forced"
me to take a "mature student entry exam",an experiment of Margaret Thatcher to
project an egalitarian image.He,the mentor urged me to take philosophy &
literature jointly,as it would be easiest option as regards requisite study
for one so disinclined.he would read German philosophers untranslated,& at
!st tutorial I'd experienced (on an extremely complex & to me bizarre and 
largely incomprehensible poem by Louis MacNeice, 'Brother Fire')i was so amazed at his interpretation & erudition that i very eagerly enquired,"Are you English
Man?"Looking haughtily insulted,he said "Me:I'm fae the Gorbals"Caught @
the act i believe!For a mentor he was more than somewhat "parasitic".debating philosophy(which I hadn't read ,just got from him;& he'd often get very angry!He also remarked earnestly,but with a smirk,that i was the most " honest" guy he'd ever met.In the end I graduated  with  the same level of degree as he,which was unfair.He should have got a 1st but no one did that year.I developed a,not for
1moment anticipated mental illness.Sincerity was my downfall.Don't have a clue
how I got started on this or what is it's relevance!My apologies to cyberspace
from cogitative dimension.How the hell is Magnus known:it's my middle name & I
can''t recall using it.Scusi ,if this is pointless drivel.Mentor got very
upset when i argued vehemently for the inclusion of philosophical validity.for"coincidence"as a significantly meaningful phenomenon.

Comment 20

[RNicoletto]

All test case URL provided in this bug are no more valid.

Mark as INVALID.

Comment 21

[Chris Lawson (gone)]

That would be FIXED by death of site.