> As usual you forgot about the SSL side, right?
> The .inc' Arno produced for
> me seems not compatible with the new buffering system!
It is exact that I didn't checked the SSL stuff. I was thinking there was no
impact on it.
Anyway, ICS-V6-SSL is alpha code and is not even available.
> Also the
Try using an Agent property value the same as the one used by IE.
Contribute to the SSL Effort. Visit http://www.overbyte.be/eng/ssl.html
--
[EMAIL PROTECTED]
Author of ICS (Internet Component Suite, freeware)
Author of MidWare (Multi-tier framework, freeware)
http://www.overbyte.be
- Origina
As usual you forgot about the SSL side, right? The .inc' Arno produced for
me seems not compatible with the new buffering system! Also there was no
BCB2006 package included!!
Regards,
SubZero
- Original Message -
From: "Francois PIETTE" <[EMAIL PROTECTED]>
To: "ICS support mailing"
S
Please offer suggestion for this circumstance -->
If I use httpcli to get this page:
http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=6286283929
everything is perfect.
If I try to get non existent auction at same site with INTERNET EXPLORER for
example:
http://cgi.ebay.com/ws/eBayISAPI.dl
I updated ICS-V6 with the ZLib stuff Angus sent to me. See his message
below.
The DLL are available separately in http://www.overbyte.be/arch/ZLibDll.zip
Thanks to Angus.
--
Contribute to the SSL Effort. Visit http://www.overbyte.be/eng/ssl.html
--
[EMAIL PROTECTED]
http://www.overbyte.be
> To test the new buffering, is it sufficient to copy on the wsockbuf.pas?
> My
> copy here uses ICSv6SSL and I do not want to change the entire package due
> to time constraints...
Defenitely not. There are changes in WSocket.pas also. All to buffer
handling has been moved to WSockBuf?.pas. Pre
To test the new buffering, is it sufficient to copy on the wsockbuf.pas? My
copy here uses ICSv6SSL and I do not want to change the entire package due
to time constraints...
Regards,
SZ
- Original Message -
From: "Francois PIETTE" <[EMAIL PROTECTED]>
To: "ICS support mailing"
Sent: S
Forgot to say that multithread application should also benefit of the buffer
scheme change because of better concurrency. There is now one critical
section per component instead of one global critical section.
In the future I plan to use a global list of free buffers to reduce memory
allocation
I've just uploaded a new ICS-V6 beta to my website.
This version include:
- Latest Arno Garrels fixes and changes (Mostly to support ICS-SSL)
- Latest fastream changes (Mostly to support CBuilder 2006)
- My own changes (A completely revamped TWSocket buffer scheme).
The new buffer scheme (Located
Francois PIETTE wrote:
>> Can you please help me? Something like below,
>> or do I still haven't got it fully?
>
> Looks partially good.
> The problem, if I understand your code correctly, is that for the new
> window, the message numbers have not the same numbers. You must map
> old message numbe
> Can you please help me? Something like below,
> or do I still haven't got it fully?
Looks partially good.
The problem, if I understand your code correctly, is that for the new
window, the message numbers have not the same numbers. You must map old
message numbers to the new message numbers bef
Francois PIETTE wrote:
>> In V6, how can I extract messages of the to be detached socket only?
>> Is it Peekmessage(Msg, OldHWnd, MsgLow, MsgLow + MsgCnt, PM_REMOVE)?
>
> No. This would retrieve all messages for all component sharing the
> same TIcsWndHandler.
> It is necessary to iterate thru FMs
> In V6, how can I extract messages of the to be detached socket only?
> Is it Peekmessage(Msg, OldHWnd, MsgLow, MsgLow + MsgCnt, PM_REMOVE)?
No. This would retrieve all messages for all component sharing the same
TIcsWndHandler.
It is necessary to iterate thru FMsgMap to find all occurence of se
> Ah. I see. Thank you. Before I posted my question I read as much
> information on all this as I could including my old postings from last
> year
> and the wsocket.hlp file. As I understand it then, the Wsocket.hlp file
> is
> wrong or even worse, vague giving the impression that something
Francois PIETTE wrote:
>> AFAIK winsock API function WSAAsyncSelect() is a common, blocking
>> function. In this case it's called to disable winsock notifications.
>> Because the window is detached/destroyed in subsequent lines.
>> BTW: Same is done in V5.
>> So for a short while the detached socke
15 matches
Mail list logo