On Wed, Apr 6, 2011 at 6:05 PM, Nathan Sweet <[email protected]> wrote:
> Diego,
> Do you mind if I convert your response into a comment on my blog? Thanks for
> your feedback.
>

Sure use it at your discretion, spread the voice ;-)

I thought some of these traps could help others see why these old
browsers should be dismissed asap for the benefit of the community.
You may have a look at the Feature Testing section of
http://github.com/dperini/nwmatcher/src/nwmatcher.js for references
over these bugs, it will be a better source than my limited embedded
memory :=)

--
Diego


> On Wed, Apr 6, 2011 at 7:21 AM, Diego Perini <[email protected]> wrote:
>>
>> Nathan,
>> good collection of informations about native DOM API. It would be
>> great to have some site work as a repository for all these known bugs.
>>
>> There are more bugs in getElementById and getElementsByTagName (on IE
>> normally). There are so many bugs that I can hardly remember them all.
>>
>>
>> Known problems with id/name and "getElementById / getElementsByName":
>>
>> - id/name conflicts, as you explained, but with more repercussions
>>  the worst problem being with forms elements but not only there
>>
>> - an element with id="length" will shadow all DOM length properties
>>  <div id="length"></div> document.getElementsByTagName('div').length; //
>> boom
>>
>> - any leaked global variable is a potential clash with element having
>> id with the same name
>>  <div id="max"></div>  max = document.getElementById('max'); // boom
>>
>>
>> Known problems with "getElementsByTagName":
>>
>> - universal selector "*" not supported on IE < 9, returns text nodes
>>
>> - wrong results with universal selector when context is an <object>
>> element
>>  <object ... ><param name="loop" value="y"><param name="movie"
>> src="a.swf"></object>
>>  object.getElementsByTagName('*') // 0 elements found
>>  using object.childNodes can help with this in some circumstance
>>
>> - html5 elements are not recognized properly, needs html5 shim
>>  document.getElementsByTagName("abbr") will return two items
>>  for each element found, one being <abbr> and the other being </abbr>
>>  closing tag is being interpreted as a different element
>>
>> In conclusion, if you can, avoid using IE < 9  ;-)
>>
>> --
>> Diego
>>
>>
>> On Wed, Apr 6, 2011 at 2:03 AM, nathanJsweet <[email protected]>
>> wrote:
>> > I wouldn't normally try to use this mailing list to shamelessly
>> > promote something, but I'm doing this to ask any new-comers to
>> > JavaScript and those advanced in their years, to consider my newly
>> > newly launched blog, which can be found at "nathansweet.me". Any
>> > feedback or interest you might show would be greatly appreciated.
>> >
>> > --
>> > To view archived discussions from the original JSMentors Mailman list:
>> > http://www.mail-archive.com/[email protected]/
>> >
>> > To search via a non-Google archive, visit here:
>> > http://www.mail-archive.com/[email protected]/
>> >
>> > To unsubscribe from this group, send email to
>> > [email protected]
>> >
>>
>> --
>> To view archived discussions from the original JSMentors Mailman list:
>> http://www.mail-archive.com/[email protected]/
>>
>> To search via a non-Google archive, visit here:
>> http://www.mail-archive.com/[email protected]/
>>
>> To unsubscribe from this group, send email to
>> [email protected]
>
>
>
> --
> Nathan Sweet
> Web Developer
> 206.588.6137
> nathansweet.me
>
> --
> To view archived discussions from the original JSMentors Mailman list:
> http://www.mail-archive.com/[email protected]/
>
> To search via a non-Google archive, visit here:
> http://www.mail-archive.com/[email protected]/
>
> To unsubscribe from this group, send email to
> [email protected]
>

-- 
To view archived discussions from the original JSMentors Mailman list: 
http://www.mail-archive.com/[email protected]/

To search via a non-Google archive, visit here: 
http://www.mail-archive.com/[email protected]/

To unsubscribe from this group, send email to
[email protected]

Reply via email to