Hi, I'm having some problems with SMAKE. I have a rather big makefile and SMAKE screws up compilation of one file in that SC doens't include the include files (I can see it in the list file). If I compile the file by hand from the command line everything works fine... strange thing.

This is happening while compiling Berkeley-DB 4.1.25 If someone would like to try it I can send over the makefile. Maybe I'm just blind...

--
Robert M. Münch
IT & Management Freelancer
Mobile: +49 (0)177 2452 802
Fax   : +49 (0)721 8408 9112
Web   : http://www.robertmuench.de

"Robert M. Münch" <robert.muench@robertmuench.de> wrote in message news:b4o485$289l$1@digitaldaemon.com...
> Hi, I'm having some problems with SMAKE. I have a rather big makefile and SMAKE screws up compilation of one file in that SC doens't include the include files (I can see it in the list file). If I compile the file by
hand
> from the command line everything works fine... strange thing.
>
> This is happening while compiling Berkeley-DB 4.1.25 If someone would like to try it I can send over the makefile. Maybe I'm just blind...

The only way to deal with those kinds of problems is to make a backup of the original makefile, and then start chopping off irrelevant parts of the problem makefile until it is reduced to just the syntax that causes the problem.

"Walter" <walter@digitalmars.com> schrieb im Newsbeitrag news:b4olfd$2mvq$4@digitaldaemon.com...

> The only way to deal with those kinds of problems is to make a backup of
the
> original makefile, and then start chopping off irrelevant parts of the problem makefile until it is reduced to just the syntax that causes the problem.

Hi, thanks I try this. I'm really wondering since years why it's mostly impossible to have a simple out-of-the-box makefile setup for bigger source-code libraries. On Linux it's mostly working but on Windows it's really a pain. I'm mostly fighting against such issues than programming problems... Robert

"Robert M. Münch" <robert.muench@robertmuench.de> wrote in message news:b4pcc4$4si$1@digitaldaemon.com...
> "Walter" <walter@digitalmars.com> schrieb im Newsbeitrag news:b4olfd$2mvq$4@digitaldaemon.com...
> > The only way to deal with those kinds of problems is to make a backup of
> the
> > original makefile, and then start chopping off irrelevant parts of the problem makefile until it is reduced to just the syntax that causes the problem.
> Hi, thanks I try this. I'm really wondering since years why it's mostly impossible to have a simple out-of-the-box makefile setup for bigger source-code libraries. On Linux it's mostly working but on Windows it's really a pain. I'm mostly fighting against such issues than programming problems... Robert

That's because makefile syntax & rules are a giant kludge. And nobody writes makefiles from scratch - they cut & paste & modify from others that someone else managed to get to work.

On Wed, 12 Mar 2003 17:02:46 -0800, Walter <walter@digitalmars.com> wrote:

> The only way to deal with those kinds of problems is to make a backup of the original makefile, and then start chopping off irrelevant parts of the
> problem makefile until it is reduced to just the syntax that causes the
> problem.

I further investigated the problem:

1. SMAKE always fails to compile one file.

2. If I cut & paste the compiler command to the command line it fails too (sometimes).

3. If I change the compile command the file compiles. Thats strange... Sometimes I need to change more, when the file compiles, and I use the original compile line it works too.

4. As posted SMAKE always compiles all files of a target. Is this normal? I though it checks which files changed and only compiles those.

-- 
Robert M. Münch

"Robert M. Münch" <robert.muench@robertmuench.de> wrote in message news:oprlzmliiyr6w2gz@news.digitalmars.com...
> On Wed, 12 Mar 2003 17:02:46 -0800, Walter <walter@digitalmars.com> wrote:
>
> > The only way to deal with those kinds of problems is to make a backup of
> > the original makefile, and then start chopping off irrelevant parts of
> > the
> > problem makefile until it is reduced to just the syntax that causes the
> > problem.
>
> I further investigated the problem:
>
> 1. SMAKE always fails to compile one file.
>
> 2. If I cut & paste the compiler command to the command line it fails too
> (sometimes).

? The compiler is pretty deterministic. Something it depends on is changing, like a .h or .sym file.

> 3. If I change the compile command the file compiles. Thats strange... Sometimes I need to change more, when the file compiles, and I use the original compile line it works too.

I have no idea what is happening. Again, try cutting the makefile down.

> 4. As posted SMAKE always compiles all files of a target. Is this normal?
I
> though it checks which files changed and only compiles those.

It depends on how you constructed the dependency list.