Re: [twsocket] THttpCli bug -- reported earlier on another server, repeating behavior again

2009-05-14 Thread Fastream Technologies
Strange, why do I get the exception then?! Did you PostMessage from OnRequestDone as Francois said? I think there is a need for a check for double onrequestdones after a get/head/post. Regards, SZ On 5/14/09, Arno Garrels arno.garr...@gmx.de wrote: Fastream Technologies wrote: I tried to

Re: [twsocket] THttpCli bug -- reported earlier on another server, repeating behavior again

2009-05-14 Thread Fastream Technologies
Hi, The event handler function is already in a mesage posted by triggerrequestdone. I think the problem is not too less messages but instead too much of it. (SetReady also posts message!) Don't you think we need a mechanism for assuring that Setready and onrequestdone is called just once per

Re: [twsocket] THttpCli bug -- reported earlier on another server, repeating behavior again

2009-05-14 Thread Fastream Technologies
BTW, I test with httpStst not httptst as that is (SSL) is what we need in our program. Maybe you can send me your test code so that I can be sure? I am ok with delphi. BR, SZ On 5/14/09, Arno Garrels arno.garr...@gmx.de wrote: Fastream Technologies wrote: I tried to provide. In the ICS

Re: [twsocket] THttpCli bug -- reported earlier on another server, repeating behavior again

2009-05-14 Thread Arno Garrels
Fastream Technologies wrote: BTW, I test with httpStst not httptst as that is (SSL) is what we need in our program. I tested the HttpTst demo successfully as well. Maybe you can send me your test code so that I can be sure? I am ok with delphi. You can be sure! I can send you my Delphi

Re: [twsocket] THttpCli bug -- reported earlier on another server, repeating behavior again

2009-05-14 Thread Fastream Technologies
Please send me your modified HTTPsTST source code so that I can have a look. I really wonder why we are not in sync... Yes I am using Icsv7 but with my own defines which is legal. I wrote that to you before: NOFORMS;NO_DEBUG_LOG;USE_SSL;NO_ADVANCED_HTTP_CLIENT_FEATURES;SECURITY_WIN32 Please try

Re: [twsocket] THttpCli bug -- reported earlier on another server, repeating behavior again

2009-05-14 Thread Arno Garrels
Fastream Technologies wrote: Please send me your modified HTTPsTST source code so that I can have a look. I really wonder why we are not in sync... Yes I am using Icsv7 but with my own defines which is legal. I wrote that to you before:

Re: [twsocket] THttpCli bug -- reported earlier on another server, repeating behavior again

2009-05-14 Thread Fastream Technologies
Try this in debug mode: http://www.fastream.com/alpha/HttpsTst.zip On 5/14/09, Arno Garrels arno.garr...@gmx.de wrote: Fastream Technologies wrote: Please send me your modified HTTPsTST source code so that I can have a look. I really wonder why we are not in sync... Yes I am using

Re: [twsocket] THttpCli bug -- reported earlier on another server, repeating behavior again

2009-05-14 Thread Arno Garrels
Fastream Technologies wrote: Try this in debug mode: http://www.fastream.com/alpha/HttpsTst.zip This demo calls HEAD the sync method!! And does not compile due to missing files. -- To unsubscribe or change your settings for TWSocket mailing list please goto

Re: [twsocket] THttpCli bug -- reported earlier on another server, repeating behavior again

2009-05-13 Thread Fastream Technologies
Did you do what I wrote below? With ICS v7, it raises an exception (Component not ready!). On 5/12/09, Fastream Technologies ga...@fastream.com wrote: Hello, Please open the C++ or Delphi Httpstst demo and add these lines to the end of the onrequestdone event: if(RqType == httpHEAD)

Re: [twsocket] THttpCli bug -- reported earlier on another server, repeating behavior again

2009-05-13 Thread Fastream Technologies
I notice in email exchanges sometimes people tend to answer each one separately without reading the entire messages sent by me in whole. Strange... Anyway, the problem with THttpCli is it issues OnRequestDone in a non-sync way! Uses multiple message posts (first for setready and then for

Re: [twsocket] THttpCli bug -- reported earlier on another server, repeating behavior again

2009-05-13 Thread Arno Garrels
Fastream Technologies wrote: Did you do what I wrote below? With ICS v7, it raises an exception (Component not ready!). if(RqType == httpHEAD) SslHttpCli1-GetASync(); Not for me, I used the Delphi HttpTst demo and added the to lines as the last in OnRequestDone. That means, the

Re: [twsocket] THttpCli bug -- reported earlier on another server, repeating behavior again

2009-05-13 Thread Fastream Technologies
Hello, HEAD must be first then GET, is that what you did? Regards, SZ On 5/13/09, Arno Garrels arno.garr...@gmx.de wrote: Fastream Technologies wrote: Did you do what I wrote below? With ICS v7, it raises an exception (Component not ready!). if(RqType == httpHEAD)

Re: [twsocket] THttpCli bug -- reported earlier on another server, repeating behavior again

2009-05-13 Thread Arno Garrels
Fastream Technologies wrote: Hello, HEAD must be first then GET, is that what you did? I just tried it again: Not using proxy cmd HEAD /otherroots/www.suckersluck.com/scripts HTTP/1.1 cmd Accept: image/gif, image/x-xbitmap, image/jpeg, image/pjpeg, */* cmd Connection: Keep-Alive cmd

Re: [twsocket] THttpCli bug -- reported earlier on another server, repeating behavior again

2009-05-13 Thread Fastream Technologies
Hello, I found something VERY STRANGE in your log: On 5/13/09, Arno Garrels arno.garr...@gmx.de wrote: Fastream Technologies wrote: Hello, HEAD must be first then GET, is that what you did? I just tried it again: Not using proxy cmd HEAD /otherroots/www.suckersluck.com/scripts

Re: [twsocket] THttpCli bug -- reported earlier on another server, repeating behavior again

2009-05-13 Thread Arno Garrels
Fastream Technologies wrote: HEAD Failed ! WHY İS THIS HERE? This is the exception message from DoRequestSync() and displayed in the first ButtonClick-HEAD request which is a sync request. So what you see is a good example, why it should avoided to mix async and sync methods. Either use

Re: [twsocket] THttpCli bug -- reported earlier on another server, repeating behavior again

2009-05-13 Thread Fastream Technologies
I never told I use sync methods at all! I use both async. I call getasync in the onrequestdone of the head! I think this bug exists in that situation as well... Regards, SZ On 5/13/09, Arno Garrels arno.garr...@gmx.de wrote: Fastream Technologies wrote: HEAD Failed ! WHY İS THIS HERE?

Re: [twsocket] THttpCli bug -- reported earlier on another server, repeating behavior again

2009-05-13 Thread Arno Garrels
Fastream Technologies wrote: I never told I use sync methods at all! I use both async. I call getasync in the onrequestdone of the head! I think this bug exists in that situation as well... What bug? Please provide a simple test application that shows the bug in action, it has to be as simple

Re: [twsocket] THttpCli bug -- reported earlier on another server, repeating behavior again

2009-05-13 Thread Fastream Technologies
I tried to provide. In the ICS Httpstst demo, which is async, I added the lines SslHttpCli1-GetAsync(); to onrequestdone of head and it shows component not ready error on the url I provided (68/scripts, NOT www.suckersluck...). If the component is not ready in onrequestdone, when will it be

Re: [twsocket] THttpCli bug -- reported earlier on another server, repeating behavior again

2009-05-13 Thread Francois Piette
I tried to provide. In the ICS Httpstst demo, which is async, I added the lines SslHttpCli1-GetAsync(); to onrequestdone of head and it shows component not ready error on the url I provided (68/scripts, NOT www.suckersluck...). If the component is not ready in onrequestdone, when will it

Re: [twsocket] THttpCli bug -- reported earlier on another server, repeating behavior again

2009-05-12 Thread Fastream Technologies
I am now using v7! Upgraded yesterday, forgot to let you know... On 5/11/09, Arno Garrels arno.garr...@gmx.de wrote: Fastream Technologies wrote: In my case, I get data from first head being pending in second head/GET. I know it is weird/rare case yet this customer was annoyed and he led

Re: [twsocket] THttpCli bug -- reported earlier on another server, repeating behavior again

2009-05-12 Thread Arno Garrels
Fastream Technologies wrote: I am now using v7! Upgraded yesterday, forgot to let you know... And still the same problem? If true than please check value of FReceiveLen before the request is it zero? -- Arno Garrels On 5/11/09, Arno Garrels arno.garr...@gmx.de wrote: Fastream Technologies

Re: [twsocket] THttpCli bug -- reported earlier on another server, repeating behavior again

2009-05-12 Thread Fastream Technologies
Hello, Please open the C++ or Delphi Httpstst demo and add these lines to the end of the onrequestdone event: if(RqType == httpHEAD) SslHttpCli1-GetASync(); and click head. You will see that even the onrequestdone is not the final event before becoming ready as it gives component

Re: [twsocket] THttpCli bug -- reported earlier on another server, repeating behavior again

2009-05-12 Thread Fastream Technologies
For normal traffic (RFC-compliant GET/HEAD), I posted a message from OnRequestDone for real work (as a workaround) yet this is not a real solution for this case as data keeps coming... On 5/12/09, Fastream Technologies ga...@fastream.com wrote: Hello, Please open the C++ or Delphi Httpstst

Re: [twsocket] THttpCli bug -- reported earlier on another server, repeating behavior again

2009-05-11 Thread Fastream Technologies
Hello, Ok the server is indeed IIS 5.1 and the direct url is: http://69.28.242.152:7979/otherroots/www.suckersluck.com/scripts which returns body with the HEAD command. ICS client handles this in a somewhat bad way--it seem to work with GUI but when you programmatically send GET after HEAD, it

Re: [twsocket] THttpCli bug -- reported earlier on another server, repeating behavior again

2009-05-11 Thread Arno Garrels
Fastream Technologies wrote: Here is the problem routine: Sorry I do not see a problem, since it is the server (or proxy) not responding. That's what my packet log showed yesterday. -- Arno Garrels -- To unsubscribe or change your settings for TWSocket mailing list please goto

Re: [twsocket] THttpCli bug -- reported earlier on another server, repeating behavior again

2009-05-11 Thread Fastream Technologies
Arno, Yesterday I sent you a wrong url! This new url is the direct IIS 5.1 url which returns body in HEAD response and it confuses thttpcli. Regards, SZ On 5/11/09, Arno Garrels arno.garr...@gmx.de wrote: Fastream Technologies wrote: Here is the problem routine: Sorry I do not see a

Re: [twsocket] THttpCli bug -- reported earlier on another server, repeating behavior again

2009-05-11 Thread Arno Garrels
Fastream Technologies wrote: Arno, Yesterday I sent you a wrong url! This new url is the direct IIS 5.1 url which returns body in HEAD response and it confuses thttpcli. It does not, just tested with V7 Httptst demo. Multiple GET/HEAD requests w/o any problem! -- Arno Garrels -- To

Re: [twsocket] THttpCli bug -- reported earlier on another server, repeating behavior again

2009-05-11 Thread Fastream Technologies
In my case, I get data from first head being pending in second head/GET. I know it is weird/rare case yet this customer was annoyed and he led us to panic even more... On 5/11/09, Arno Garrels arno.garr...@gmx.de wrote: Fastream Technologies wrote: Arno, Yesterday I sent you a wrong url!

Re: [twsocket] THttpCli bug -- reported earlier on another server, repeating behavior again

2009-05-10 Thread Arno Garrels
It's a server (or proxy) problem! Use a sniffer such as Wireshark if you don't believe it. After a long delay (timeout?) the server answers with a 404, and closes the connection. -- Arno Garrels Fastream Technologies wrote: Hello, Could you guys try this url and the below actions

Re: [twsocket] THttpCli bug -- reported earlier on another server, repeating behavior again

2009-05-10 Thread Fastream Technologies
Ok. There was no proxy in the lastest configuration I sent. So the issue is the Mystery server dislikes the HEAD command which ordinary browsers do not send and takes it as a hack attempt. Maybe there is a configuration option for this which the customer can tweak. Thank you, SZ On 5/10/09,