Thanks,

I'm trying

#ifndef socklen_t
#define socklen_t size_t
#endif

which works ok for me.

I have to leave town suddenly for about a week and won't be able to work on
this further until I get back. There is a problem with the "spaces" change
you propose that I can't figure out -- it looks like calling a procedure a
second time is removing the result of calling it the first. I wonder if I
made a stupid pointer/reference error of some kind.

I put a copy of the current version of my ant-based sql tool /db regression
test tool on the anonymous ftp at firebird as sqlxml.zip.prealpha, in case
anyone is interested in trying it out. ( with very brief instructions) I
hope this isn't rude, if anyone doesn't want it there please take it away
again.


Thanks
David Jencks

On 2001.02.07 09:51:13 -0500 Ken Richard wrote:
> 1. MSVC can perform any cast. The only problem was the lack of
> definition
> for the socklet_t. You could alternativly do something like:
>
> #ifdef _MSC_VER
> #define socket_t size_t
> #endif
>
>
> You may want to use the _MSC_VER instead of the WIN32 symbol in case
> people
> are using different compilers on windows.
>
>
>
> -----Original Message-----
> From: David Jencks [mailto:davidjencks@...]
> Sent: Wednesday, February 07, 2001 1:27 AM
> To: IB-Java@yahoogroups.com
> Subject: RE: [IB-Java] Interclient 20 changes
>
>
> Thanks for checking this out.
>
> 2. works fine for me in rh7.
>
> 1. does not- this is why I changed it in the first place- socklen_t is
> defined as a uint, and can't be cast to an int.
>
> Does
> // socklen_t clilen;
> uint clilen;
> at line 148 work for you, leaving line 171 as I had it? My prototype for
> the last argument to accept is socklen_t = uint. Can you cast this in
> msvc?
>
> I wonder if my change will compile on earlier linices-- perhaps the
> prototype for accept changed from int to uint in recent glibc.
>
> Thanks
> David Jencks
>
>
>
> On 2001.02.06 13:51:43 -0500 Ken Richard wrote:
> > There were a few compile problems with msvc:
> >
> > 1. socklet_t is not defined in the msvc headers and the last argument
> to
> > accept is prototypes as a int*.
> >
> > Index: NetTCP.cpp
> > ===================================================================
> > RCS file: /cvsroot/firebird/interclient/20/dev/interserver/NetTCP.cpp,v
> > retrieving revision 1.2
> > diff -r1.2 NetTCP.cpp
> > 148c148
> > < socklen_t clilen;
> > ---
> > > size_t clilen;
> > 171c171
> > < tmpSockfd = accept (sockfd, (struct sockaddr *) &tcpCliAddr,
> > &clilen);
> > ---
> > > tmpSockfd = accept (sockfd, (struct sockaddr *) &tcpCliAddr,
> > (int*)&clilen);
> >
> >
> >
> >
> > 2. Msvc does not allow static initializers within the class definition.
> >
> > Index: IB_Catalog.h
> > ===================================================================
> > RCS file: /cvsroot/firebird/interclient/20/dev/interserver/IB_Catalog.h,v
> > retrieving revision 1.2
> > diff -r1.2 IB_Catalog.h
> > 279c279,280
> > < static const char* const spaces = "
> ";
> > ---
> > > //ken richard 2-6-2001
> > > static const char* const spaces;
> >
> > Index: IB_Catalog.cpp
> > ===================================================================
> > RCS file:
> /cvsroot/firebird/interclient/20/dev/interserver/IB_Catalog.cpp,v
> > retrieving revision 1.2
> > diff -r1.2 IB_Catalog.cpp
> > 82a83,85
> > > const char* const
> > > IB_Catalog::spaces = " ";
> > >
> >
> >
> >
> >
> > I didn't get to run anything against the server on Win32 - but I wanted
> > to
> > try a compile with your changes. I should get a chance sometime this
> > week
> > to run some stuff through it.
> >
> >
> > -Ken
> >
> >
> >
> >
> > -----Original Message-----
> > From: David Jencks [mailto:davidjencks@...]
> > Sent: Tuesday, February 06, 2001 12:25 PM
> > To: firebird-devel@...
> > Cc: ib-java@yahoogroups.com
> > Subject: [IB-Java] Interclient 20 changes
> >
> >
> > (sorry about the previous blank message)
> >
> > Hi!
> >
> > I have (I think) after waiting a week for CVS access and many struggles
> > with CVS contributed an ant build script and many changes to
> > DatabaseMetaData.
> >
> > The ant build script can be used to build interclient, interserver
> (gcc)
> > or
> > both. ( on linux, rh7). You may need to adjust some directory specs to
> > get
> > output where you want. If you haven't tried ant, do so now.
> > (http://jakarta.apache.org/ant/index.html)
> >
> >
> > The changes to DatabaseMetaData make getTables work reliably and make
> > methods taking wildcards in parameters accept escaped wildcards
> properly.
> >
> > I can only build/test this on linux, I don't know if these changes
> break
> > windows/other os versions. If you can test.. please do so.
> >
> > ***********
> > CVS problem.
> >
> > I am having a big prblem with CVS, any advice/help would be greatly
> > appreciated. I connect with ssh, and after every command it asks me
> for
> > my
> > password. After I enter it, cvs says
> >
> > Could not chdir to home directory /home/users/d_jencks No such file or
> > directory
> >
> > I think as a result of this I can only commit one file at a time. Any
> > ideas? I created a /home/users/d_jencks directory on my machine, it
> > didn't
> > help ( my local username is david, not d_jencks)
> >
> > **********
> >
> > After the cvs problem is sorted out I will add my ant based sql tool
> that
> > can be used for regression testing.
> >
> > Thanks
> > David Jencks
> >
> >
> >
> > To unsubscribe from this group, send an email to:
> > IB-Java-unsubscribe@egroups.com
> >
> >
> >
> >
> >
> > To unsubscribe from this group, send an email to:
> > IB-Java-unsubscribe@egroups.com
> >
> >
> >
> >
> >
>
>
>
> To unsubscribe from this group, send an email to:
> IB-Java-unsubscribe@egroups.com
>
>
>
>
>
> To unsubscribe from this group, send an email to:
> IB-Java-unsubscribe@egroups.com
>
>
>
>
>