The performance problem in using typelib comes in initially loading the
typelib by the oleauto dll. Richard Grimes has empirically found it to
be 10 times higher than the proxy/stub marshalling.

Once the typelib has been loaded, the method call time difference b/w
the two is not much.

Syed

-----Original Message-----
From: Brad Wilson [mailto:dotnetguyPOBOX.COM]
Sent: Wednesday, January 16, 2002 10:43 AM
To: DCOMDISCUSS.MICROSOFT.COM
Subject: Re: Puzzling... Exe Server not needing Proxy/Stud???

Brian Muth wrote:

>> Don Box did some tests several years back... search the archives.
>> If memory serves it was orders of magnitude different.

> orders of magnitude????

> Can you back this up with a reference, Phil? I couldn't find it.

The only numbers I remember seeing from Don were in support of using
automation marshalling, saying that it was approximately as fast as the
"MIDL /Oicf" marshalling code.

Brad

--
Read my web log at http://www.quality.nu/dotnetguy/
----------------------------------------------------------------
Users Guide http://discuss.microsoft.com/archives/mailfaq.asp
contains important info. Save time, search the archives at
http://discuss.microsoft.com/archives/index.html .
To unsubscribe, mailto:DCOM-signoff-requestDISCUSS.MICROSOFT.COM
----------------------------------------------------------------
Users Guide http://discuss.microsoft.com/archives/mailfaq.asp
contains important info. Save time, search the archives at
http://discuss.microsoft.com/archives/index.html .
To unsubscribe, mailto:DCOM-signoff-requestDISCUSS.MICROSOFT.COM

• Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]