From HTTP-WG (I expect most people here will be on the list). More fodder for tests... ... Date: Wed, 7 Feb 2001 21:59:01 +0100 From: "Joris Dobbelsteen"...
Mark Nottingham
mnot@...
Feb 8, 2001 1:00 am
24
Apologies to anyone who's already seen this, but, seeing as there's an interoperability issue here, I thought it'd be relevant to this list too ... From: Miles...
Miles Sabin
msabin@...
Mar 14, 2001 9:55 pm
25
... HTTP is not TCP-specific. This fact probably explains why half-closed connections, being a very subtle TCP-specific issue, are not mentioned in HTTP specs....
Alex Rousskov
rousskov@...
Mar 15, 2001 12:09 am
26
Alex Rousskov wrote, ... That's true in principle, but TCP implementations predominate, and it makes sense to be sure that HTTP gets on well with TCP. We have...
Miles Sabin
msabin@...
Mar 15, 2001 8:28 am
27
... IMHO, the straightforward and simple behavior (always proceed with the connection to the server, fetch the object, try to respond to the client) will be...
Alex Rousskov
rousskov@...
Mar 15, 2001 2:54 pm
28
I'd like to know what current tools there are for doing HTTP compliance testing. At this point, I don't care whether they're free or not. I don't care...
Karr, David
david.karr@...
Jul 27, 2001 7:44 pm
29
Hi David, You might want to look at http://www.measurement-factory.com/products/co-advisor/ While our tool does not cover everything you want, it may be a...
Alex Rousskov
rousskov@...
Jul 27, 2001 9:04 pm
30
Note that I'm very new to exploration of the details of the HTTP spec, much less its implementation. I would assume that, like many protocol specifications...
Karr, David
david.karr@...
Aug 1, 2001 4:27 pm
31
... Yes. AFAIK, several Squid-based projects use Transfer-Encoding to compress entities in transfer between cooperating intermediaries. I would imagine that...
Alex Rousskov
rousskov@...
Aug 1, 2001 5:59 pm
32
Hi, I would like to know if there is a tabularised checklist kind of a thing available anywhere, so that we can verify the http compliance of a component...
gaya377@...
Aug 10, 2001 9:14 am
33
... Yes. RFC 2616 is, essentially, a list of MUST and SHOULD (and MAY) cases. A table would have to consist of pointers to RFC paragraphs/sections. Some...
Alex Rousskov
rousskov@...
Aug 10, 2001 1:33 pm
34
One of the 2616 authors used to have a checklist of requirements sitting around... I think I have a reference, will have to dig around. Anyway, it was really...
Mark Nottingham
mnot@...
Aug 10, 2001 10:53 pm
35
Here is a feature checklist: http://www.agranat.com:1998/ Also, see http://www.w3.org/Protocols/HTTP/Forum/Reports/ for reports on existing implementations. ...
Carl Kugler
kugler@...
Aug 13, 2001 4:46 pm
36
... This is a cool compilation! While the items are just "features" and many are not really testable protocol items, it is very handy indeed. The second URL...
Alex Rousskov
rousskov@...
Aug 13, 2001 5:09 pm
37
Note that these reports are from around 1998, so the implementations might be more compliant now. -Carl Alex Rousskov <rousskov@measurement-fa To:...
Carl Kugler
kugler@...
Aug 13, 2001 6:21 pm
38
Actually, I was refering to the attached draft (just found it; got it from Ross about a year ago). Ross wanted to make sure it's understood it isn't official,...
Mark Nottingham
mnot@...
Aug 16, 2001 1:11 am
39
Just FYI: At the IETF Apps area open meeting in London, there was brief discussion of revising the HTTP spec to clarify, drop unused portions, etc. I don't...
Mark Nottingham
mnot@...
Aug 16, 2001 1:22 am
40
http://www.w3.org/QA/Activity/ <-- overview http://www.w3.org/QA/WG/ <-- working group http://www.w3.org/QA/IG/ <-- interest group The working group does...
Mark Nottingham
mnot@...
Aug 17, 2001 12:30 am
41
See: http://cvs.apache.org/viewcvs.cgi/httpd-test/perl-framework/ HTTP/1.1 testing is pretty basic, but there's code there... -- Mark Nottingham ...
Mark Nottingham
mnot@...
Sep 1, 2001 4:59 am
42
... I use a package in tcl which implements the client-side of the http protocol. The functionality works except for one url which keeps reporting a 404 - Not...
feit@...
Sep 20, 2001 1:59 pm
43
... You will always have erroneous or otherwise unexpected responses if you use real URLs. There are a lot of odd servers out there. There is no way around it....
Alex Rousskov
rousskov@...
Sep 20, 2001 3:03 pm
44
How do browsers figure out a response is "Complete" if the origin server doesn't send a "Content-Length" header? What is the normal strategy for that? The...
Karr, David
david.karr@...
Oct 4, 2001 6:40 pm
45
... RFC 2616 defines browsers behavior in this case. There are several scenarios like Chunked Encoding and end-of-file/connection detection. See a lengthy...
Alex Rousskov
rousskov@...
Oct 4, 2001 7:01 pm
46
In HTTP/1.1, the response could be sent with "Transfer-Encoding: chunked" in which case a zero-length chunk (possibly followed by a trailer) terminates the...
Carl Kugler
kugler@...
Oct 4, 2001 7:27 pm
47
I would also add that the server must terminate the connection nicely (e.g.., a FIN/ACK in TCP). Any other kind of termination should be regarded as an...
Hu, James
jxh@...
Oct 4, 2001 7:46 pm
48
I started to read that earlier, but I missed the short "closing the connection" point. So, how exactly do you detect the closing of the connection? If you're ...
Karr, David
david.karr@...
Oct 4, 2001 7:48 pm
49
Typically, your select() will return that the socket is readable, and when you read, you will get back 0 bytes. This signals to you that the connection has...
Hu, James
jxh@...
Oct 4, 2001 7:52 pm
50
... You wait for the read(2) call on a socket to return zero bytes, indicating an end-of-file condition. Unfortunately, the situation is complicated by...
Alex Rousskov
rousskov@...
Oct 4, 2001 8:06 pm
51
... ... but only if the error number tells you so. Negative read size may simply mean that select(2) lied and no data is really available yet. This has very...
Alex Rousskov
rousskov@...
Oct 4, 2001 8:10 pm
52
... So after you've finished reading the last real block of data and get back to the "select" loop, the server will have closed the connection. At the ...
Having problems with message search?
Fill out this form to ensure your group is one of the first to be migrated to the new message search system.
