313217 - document.open('replace') now opens text/plain document where it used to open text/html

Status: RESOLVED FIXED

(Core :: DOM: Core & HTML, defect)

Description

[Martijn Wargers (dead)]

This was mentioned here:
http://forums.mozillazine.org/viewtopic.php?t=326309&highlight=

I'll attach a testcase.

This used to work in 2004-05-20 build, but fails in 2004-05-22 build, so very
likely a regression from bug 73409.

Comment 1

[Martijn Wargers (dead)]

Attachment: testcase

IE6 passes for this testcase.

Comment 2

[Peter van der Woude [:Peter6]]

Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.8b5) Gecko/20051020
Firefox/1.5 ID:2005102014

Confirmed

also works in Opera 8.5

Comment 3

[Boris Zbarsky [:bzbarsky]]

This was purposeful, actually.  Any type that's not text/html is treated as
text/plain.  Do Opera or IE actually replace in this case?  Or just treat
"replace" as a MIME type and fall back to HTML?  I can do either one, assuming
this is common enough for us to need to fix...

Comment 4

[Boris Zbarsky [:bzbarsky]]

Attachment: Patch to make a type of "replace" mean HTML and replace

Comment 5

[Boris Zbarsky [:bzbarsky]]

Attachment: Patch to make a type of "replace" mean HTML but not replace

Comment 6

[Boris Zbarsky [:bzbarsky]]

bc, do you think you can give me an idea of how common this is in time for 1.8?

Comment 7

[Brendan Eich [:brendan]]

http://lxr.mozilla.org/classic/source/lib/libmocha/lm_doc.c#1210

"replace" is an optional second argument.  The first argument to document.open,
if given, is always a MIME type.  If not recognized, the default was
"text/html", if I recall correctly.  I did not try to wade through the classic
netlib's uses of content_type to prove this.

/be

Comment 8

[Boris Zbarsky [:bzbarsky]]

Right.  What I implemented is that we default to text/plain for unrecognized
types, since rendering random data as HTML sounds like a recipe for disaster to
me...  Of course the only type we recognize is text/html.  ;)

Comment 9

[Bob Clary [:bc] (inactive)]

(In reply to comment #6)
> bc, do you think you can give me an idea of how common this is in time for 1.8?

Yes, I can try but I'm not sure how representative a scan would be of the full
universe of pages. Would a scan of the top sites 1 level deep be enough? That
will take about a day. A 2 deep scan would take several days. 

Comment 10

[Boris Zbarsky [:bzbarsky]]

Yeah, that's not going to be representative enough, probably.  OK.  So can
someone tell me which of those two patches we want?  That is, whether 'replace'
in the first slot actually replaces in IE/Opera?

Comment 11

[Martijn Wargers (dead)]

(In reply to comment #10)
> That is, whether 'replace'
> in the first slot actually replaces in IE/Opera?
No, 'replace' in the first slot doesn't replace in IE6. Another history item in
the list is added.
But I see now that any random string other than 'replace' is doing the same as
'replace' in IE6. Only 'text/plain' is opening the document in text/plain.

Comment 12

[Boris Zbarsky [:bzbarsky]]

Right.  I don't think I want that behavior.

Comment 13

[Boris Zbarsky [:bzbarsky]]

Comment on attachment 200309 [details] [diff] [review]
Patch to make a type of "replace" mean HTML but not replace

Just treat "replace" in the first slot as if nothing were set.

Comment 14

[Johnny Stenback (:jst)]

Comment on attachment 200309 [details] [diff] [review]
Patch to make a type of "replace" mean HTML but not replace

r+sr=jst

Comment 15

[Boris Zbarsky [:bzbarsky]]

Fixed on trunk.  Probably not worth messing with this on branch, but if someone thinks we should, please tell me why!

Comment 16

[JetLi]

(In reply to comment #15)
> Fixed on trunk.  Probably not worth messing with this on branch, but if someone
> thinks we should, please tell me why!
> 

Hi
wil this still fixed in Firefox 1.5 RC1 or Final too?

Comment 17

[Boris Zbarsky [:bzbarsky]]

Firefox 1.5 is on the branch.  So no, I have no plans to fix this in 1.5 unless this is a seriously major issue on the web.

Comment 18

[JetLi]

(In reply to comment #17)
> Firefox 1.5 is on the branch.  So no, I have no plans to fix this in 1.5 unless
> this is a seriously major issue on the web.
> 

thnx for reply. i think that would be great if fix this in branch too, because in branch i had tested too, it fixed my another issue in YaBB 2 Software.This wil help my developer on www.YaBBforum.com

Comment 19

[Boris Zbarsky [:bzbarsky]]

Again, if there is significant breakage from this, please describe it in detail.  Right now, as things stand, I doubt drivers would take this on branch.