[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: R: [lwip-users] httpd question
|
From: |
Marek Matej |
|
Subject: |
Re: R: [lwip-users] httpd question |
|
Date: |
Wed, 20 Aug 2008 14:15:02 +0200 |
I removed all TAB character from page header and left rest of data
untouched, but FF behaves still the same. Then I change header to
"<!DOCTYPE html PUBLIC \"-//W3C//DTD HTML 4.01 Transitional//EN\">"
(generated in NVU html editor) and things start to work. (as well as
without header).
So I don't think it was only problem of bogus TAB.
Anyway, http server is not my main interrest.
It is the simplest application where I only need to
test my api (which I described at begining).
Marek
2008/8/20 Jonathan Larmour <address@hidden>:
> Marek Matej wrote:
>>
>> Thank You Jifl, Your suggestion was very usefull. Problem resides in
>> http header, but it wasnt TAB character (probably it was only partial
>> problem), but content of header, especial first line containing
>> "HTTP/1.0 200 OK Content-type: text/html"
>> When I remove this line everything was ok. Newly generated page by NVU
>> editor works as well.
>> So it was more http protocol probem rather then lwip problem.
>
> No, I think it really would be the tab characters. All you've done by
> removing the header entirely is make firefox believe it's talking to an
> ancient (HTTP 0.9 or earlier) web server. You are likely to run into
> problems without the ability to indicate properly about web pages not found,
> or be able to express content types.
>
> Jifl
>
>> 2008/8/19 Jonathan Larmour <address@hidden>:
>>
>>> Marek Matej wrote:
>>>
>>>> Ok, connection seems to be estabilished correctly in both cases.
>>>> Differences start after GET is received.
>>>> Attached are pcap files of FF and IE communication.
>>>> Please take a look, if You see some fundamental differences.
>>>
>>> The traffic and data sent appears to be the same. I think the difference
>>> is
>>> how FF and IE respond to spurious data: there are bogus hex 9 (tab)
>>> characters floating around in the data beside CRLFs. Look at this bit:
>>>
>>> 13:57:42.966190 IP 192.168.5.250.http > 192.168.5.74.3479: P 1:257(256)
>>> ack
>>> 257 win 1024
>>> 0x0000: 4500 0128 0007 0000 ff06 2e34 c0a8 05fa E..(.......4....
>>> 0x0010: c0a8 054a 0050 0d97 0000 19a1 f65d 074a ...J.P.......].J
>>> 0x0020: 5018 0400 6740 0000 4854 5450 2f31 2e30 address@hidden/1.0
>>> 0x0030: 2032 3030 204f 4b0d 0a09 436f 6e74 656e .200.OK...Conten
>>> 0x0040: 742d 7479 7065 3a20 7465 7874 2f68 746d t-type:.text/htm
>>> 0x0050: 6c0d 0a09 0d0a 093c 6874 6d6c 3e20 093c l......<html>..<
>>>
>>> See bytes 0x39, 0x53, 0x56, 0x5e.
>>>
>>> Most important I suspect is the one at 0x53 which should be an empty CRLF
>>> to
>>> indicate the end of HTTP headers but is in fact TAB, CR, LF. I suspect FF
>>> doesn't like that, and MSIE lazily accepts it.
>>>
>>> Jifl
>>> --
>>> eCosCentric Limited http://www.eCosCentric.com/ The eCos experts
>>> Barnwell House, Barnwell Drive, Cambridge, UK. Tel: +44 1223 245571
>>> Registered in England and Wales: Reg No 4422071.
>>> ------["The best things in life aren't things."]------
>>> Opinions==mine
>>>
>>>
>>> _______________________________________________
>>> lwip-users mailing list
>>> address@hidden
>>> http://lists.nongnu.org/mailman/listinfo/lwip-users
>>>
>>
>>
>>
>> _______________________________________________
>> lwip-users mailing list
>> address@hidden
>> http://lists.nongnu.org/mailman/listinfo/lwip-users
>>
>
>
> --
> eCosCentric Limited http://www.eCosCentric.com/ The eCos experts
> Barnwell House, Barnwell Drive, Cambridge, UK. Tel: +44 1223 245571
> Registered in England and Wales: Reg No 4422071.
> ------["The best things in life aren't things."]------ Opinions==mine
>
>
> _______________________________________________
> lwip-users mailing list
> address@hidden
> http://lists.nongnu.org/mailman/listinfo/lwip-users
>