Hello everybody
The following emails for some reaseon never made to the list
so I am resending them.
---------------- Email 1 ----------------------------------------
James C wrote:
>> src/table.hh might be a reasonable place to start thinking
>> about the requirements for integrating a new set of tags.
Don't know. I looked at it. Are you sure you meant table.hh?
I'm not sure that any new tags would actually be added anyway.
If anything I would look at
Html_tag_open_script(...) {
Html_tag_close_script(...) {
as likely candidates for a starting point.
I did get binaries for iceweasel that worked. 285 files @ 21M and
all I do is click:
"accept"
Thats it. Close the browser(and the chroot, since i can't get it
working otherwise) and start dillo. I don't need javascript again
until the lease expires about 20hrs later.
Bruce
--------------- Email 2 ----------------------------------------
James C wrote
> Do you have dtrace, dtruss or strace, to look for open() to see
> where JavaScript is writing its state? That's my guess about
> why you'd need a chroot or similar.
I have strace, but the reason I need the chroot is that the
library versions for iceweasel don't match the versions on my
system and the files are all over the place. Iceweasel will
eventually go away. Right now getting the glue between dillo and
duktake is beginning to look pretty good but I have run into a
puzzle of a different kind.
I put a print statement in Html_tag_open_script and
Html_tag_close_script and indeed you see "in Html_tag_open_script"
and "Html_tag_close_script" every time the script tag is
encountered. Now if I load that signon page I get a blank screen
and the messages are -> not <- displayed, and the stop Icon stays
permanently on. Looking at the code I find the tag
<noscript>
<meta HTTP-EQUIV="REFRESH" content="0;
url=
https://guestportal.burnaby.ca/login_noscript.html">
</noscript>
If I delete these lines dillo loads the page normally and every
thing is fine. Looking in the table in html.cc I find:
const TagInfo Tags[] =
... ...
/* noscript 1011 */
... ...
blindly replaceing this line with:
{"script", B8(111001),'R',2, Html_tag_open_script,NULL,Html_tag_close_script},
and changing "script" to "noscript" on the assumtion that when
dillo hits the tag </noscript> it will then just continue until it
hits </noscript>, throw away the data and then continue on,
which it apparently does. The question is what to do about it. If
it is something that is going to be implemented but nobody has
got around to it then I would at least put in a:
Html_tag_open_noscript() and a
Html_tag_close_noscript()
so that web pages don't stall when dillo hits that tag.
Bruce