[cvsnt] bug in cvs(nt)?

Wed Jun 9 21:17:18 BST 2004

Greetings,

	Not sure what the policy is re: attachments, etc.

	I have already been interfacing with the WinCVS folks, and they tell me that this looks to be a cvs.exe bug.

	I am using NT4.

	What is necessary to do in order to confirm that this performance is a bug?

	This "possible bug" occurs when "cvs info" is called from WinCVS.

	Here is a brief recap:

> I received an error.  Understand, the module I am attempting to checkout, in
> this case, already exists on my box. 
> 
>  (I captured the screen image, and I am not sure if it is ok to attach the
>  .jpg images to this reply or not.  They are  approximately 98kb and 48kb.) 

First Error:

  (Generated after "cvs info" was sent by WinCVS to CVSNT)

	"Something bad happened to CVSNT, and it crashed.  Unfortunately DBGHLP.DLL is missing so we can't 
produce a crash dump."

	Typing "cvs info" at the command line for WinCVS produced an identical result.

Example:

cvs info 
Available protocols:

local               (internal)
ext                 ext 2.0.41a
fork                fork 2.0.41a
ntserver            ntserver 2.0.41a
pserver             pserver 2.0.41a
server              server 2.0.41a
sserver             sserver 2.0.41a
ssh                 ssh 2.0.41a

End Example.

	> 
>  After I acknowledged the first error, cvs.exe threw this exception (from
>  cvs.exe application error): 
> 
>   "The exception unknown software exception (0xc06d007e) ocurred in the
>   application at location  0x77f1d642." 

I have a JIT debugger (DrMingw) if that would be helpful.  The one time I did catch the information generated from 
the JIT debugger, "get_protocol"... was referenced.

>
{
	Next is what occurred on the WinCVS (client) side.  Not necessarily applicable to cvsnt.

	[>  After acknowledging that particular error, I was given a similar CVSROOT
>  dialog box to what you, Oliver, were  referencing in the link above
>  (http://people.freenet.de/ogiesen/cvsroot.png). 

	(See link for example:  CVSROOT)

> 
>  The first thing I noticed then was that there were no values assigned to
>  any of the CVSROOT descriptions. 
	]

	Last bit indicates that CVSNT did not return the appropriate values to WinCVS.

}
> 

	I hope this is helpful.

		Paul G.