new user install of poc

new user install of poc

Post by gkap » Wed, 02 Jul 2003 01:46:17



following the instruction give in objc-3.1.32 I executed:
/home/poc/objc-3.1.32/]$ ./configure --prefix=$HOME
/home/poc/objc-3.1.32/]$ make 2> error.check.step5  # the content of the
file follows:-----------------------------------------------------------
byacc: 6 shift/reduce conflicts
byacc: 40 reduce/reduce conflicts
y.tab.c:1755: warning: definition of 'yylex' defaults to 'int'
y.tab.c:1796: warning: definition of 'yyerror' defaults to 'int'
y.tab.c:3187: warning: definition of 'yylex' defaults to 'int'
y.tab.c:3220: warning: definition of 'yyerror' defaults to 'int'        
cp: cannot stat `objcrtr.a': No such file or directory
make[2]: [bootstrap] Error 1 (ignored)
cp: cannot stat `objpakr.a': No such file or directyory        
make[2]: [bootstrap] Error 1 (ignored)
cp: cannot stat `cakitr.a': No such file or directory
make[2]: [bootstrap] Error 1 (ignored)
----------end-of-file-----------------
/home/poc/objc-3.1.32/]# make install 2> error.check.step6 # the content
of 2> follows------------------------------------------
cp: cannot stat `objcrt.a': no such file or directory
make[1]: [bootstrap] Error 1 (ignored)
cp: cannot stat `objpak.a': no such file or directory
make[1]: [bootstrap] Error 1 (ignored)
cp: cannot stat `cakit.a': no such file or directory
make[1]: [bootstrap] Error 1 (ignored)
----end-of-file-----------------------------------------
following the above I entered compiled and executed the hello.m program
illustrated in the file objc-3.1.32/manual.html under the title
" More Hello world ". It executed producing the expected results indicated.
I would like to know the cause of the error received during installation
and  whether or not the will come back to haunt me.
 
 
 

new user install of poc

Post by David Ste » Wed, 02 Jul 2003 02:45:16


Hello,

The yacc warnings are warnings, no errors, no need to worry about that.

The objpakr.a warning can be ignored, these are special versions of the
objpak 'Object Pak' toolkit.

This toolkit can be compiled with or without automatic reference counting
(objpakr.a is the version with 'r' reference counting).

However, if you have a warning that objpak.a itself is not found, then
you have a problem.

Obviously, you managed to compile the compiler (which platform?)

I would focus on the message about objpak.a, that's not normal.


> following the instruction give in objc-3.1.32 I executed:
> /home/poc/objc-3.1.32/]$ ./configure --prefix=$HOME
> /home/poc/objc-3.1.32/]$ make 2> error.check.step5  # the content of the
> file follows:-----------------------------------------------------------
> byacc: 6 shift/reduce conflicts
> byacc: 40 reduce/reduce conflicts
> y.tab.c:1755: warning: definition of 'yylex' defaults to 'int'
> y.tab.c:1796: warning: definition of 'yyerror' defaults to 'int'
> y.tab.c:3187: warning: definition of 'yylex' defaults to 'int'
> y.tab.c:3220: warning: definition of 'yyerror' defaults to 'int'        
> cp: cannot stat `objcrtr.a': No such file or directory
> make[2]: [bootstrap] Error 1 (ignored)
> cp: cannot stat `objpakr.a': No such file or directyory        
> make[2]: [bootstrap] Error 1 (ignored)
> cp: cannot stat `cakitr.a': No such file or directory
> make[2]: [bootstrap] Error 1 (ignored)
> ----------end-of-file-----------------
> /home/poc/objc-3.1.32/]# make install 2> error.check.step6 # the content
> of 2> follows------------------------------------------
> cp: cannot stat `objcrt.a': no such file or directory
> make[1]: [bootstrap] Error 1 (ignored)
> cp: cannot stat `objpak.a': no such file or directory
> make[1]: [bootstrap] Error 1 (ignored)
> cp: cannot stat `cakit.a': no such file or directory
> make[1]: [bootstrap] Error 1 (ignored)
> ----end-of-file-----------------------------------------
> following the above I entered compiled and executed the hello.m program
> illustrated in the file objc-3.1.32/manual.html under the title
> " More Hello world ". It executed producing the expected results indicated.
> I would like to know the cause of the error received during installation
> and  whether or not the will come back to haunt me.


 
 
 

new user install of poc

Post by gkap » Wed, 02 Jul 2003 04:03:16


compiled under two operating systems: rhl 7.2 with gcc 2.96-98 (intel
pIII), and rhl 8 gcc 3.2.(intel pII)

Is there some other script that I should have executed in preparation to
  the install i.e., such as bootSrc

following is a representation of my install:

HOME/]$ ls : objc-bootstarp-3.1.32.tar.gz , objc-3.1.32.tar.gz
HOME/]$ tar xvzf objc-bootstrap-3.1.32.tar.gz
HOME/]$ tar xvzf objc-3.1.32.tar.gz
HOME/]$ cd objc-bootstrap-3.1.32 ;./configure --prefix=$HOME
HOME/objc-bootstrap-3.1.32/]$ make
HOME/objc-bootstrap-3.1.32/]# make install
HOME/objc-bootstrap-3.1.32/]# exit ; cd ../objc-3.1.32
HOME/objc-3.1.32/]$ .configure --prefix=$HOME
HOME/objc-3.1.32/]$ make
HOME/objc-3.1.32/]$ su ...
HOME/objc-3.1.32/]# make install ; exit ; cd
HOME/]$ vi edit hello.m ; ./hello joe joe luke ...
....

Thank you.


> Hello,

> The yacc warnings are warnings, no errors, no need to worry about that.

> The objpakr.a warning can be ignored, these are special versions of the
> objpak 'Object Pak' toolkit.

> This toolkit can be compiled with or without automatic reference counting
> (objpakr.a is the version with 'r' reference counting).

> However, if you have a warning that objpak.a itself is not found, then
> you have a problem.

> Obviously, you managed to compile the compiler (which platform?)

> I would focus on the message about objpak.a, that's not normal.


>>following the instruction give in objc-3.1.32 I executed:
>>/home/poc/objc-3.1.32/]$ ./configure --prefix=$HOME
>>/home/poc/objc-3.1.32/]$ make 2> error.check.step5  # the content of the
>>file follows:-----------------------------------------------------------
>>byacc: 6 shift/reduce conflicts
>>byacc: 40 reduce/reduce conflicts
>>y.tab.c:1755: warning: definition of 'yylex' defaults to 'int'
>>y.tab.c:1796: warning: definition of 'yyerror' defaults to 'int'
>>y.tab.c:3187: warning: definition of 'yylex' defaults to 'int'
>>y.tab.c:3220: warning: definition of 'yyerror' defaults to 'int'        
>>cp: cannot stat `objcrtr.a': No such file or directory
>>make[2]: [bootstrap] Error 1 (ignored)
>>cp: cannot stat `objpakr.a': No such file or directyory        
>>make[2]: [bootstrap] Error 1 (ignored)
>>cp: cannot stat `cakitr.a': No such file or directory
>>make[2]: [bootstrap] Error 1 (ignored)
>>----------end-of-file-----------------
>>/home/poc/objc-3.1.32/]# make install 2> error.check.step6 # the content
>>of 2> follows------------------------------------------
>>cp: cannot stat `objcrt.a': no such file or directory
>>make[1]: [bootstrap] Error 1 (ignored)
>>cp: cannot stat `objpak.a': no such file or directory
>>make[1]: [bootstrap] Error 1 (ignored)
>>cp: cannot stat `cakit.a': no such file or directory
>>make[1]: [bootstrap] Error 1 (ignored)
>>----end-of-file-----------------------------------------
>>following the above I entered compiled and executed the hello.m program
>>illustrated in the file objc-3.1.32/manual.html under the title
>>" More Hello world ". It executed producing the expected results indicated.
>>I would like to know the cause of the error received during installation
>>and  whether or not the will come back to haunt me.

 
 
 

new user install of poc

Post by David Ste » Wed, 02 Jul 2003 04:27:05



> following is a representation of my install:

> HOME/]$ ls : objc-bootstarp-3.1.32.tar.gz , objc-3.1.32.tar.gz
> HOME/]$ tar xvzf objc-bootstrap-3.1.32.tar.gz
> HOME/]$ tar xvzf objc-3.1.32.tar.gz
> HOME/]$ cd objc-bootstrap-3.1.32 ;./configure --prefix=$HOME
> HOME/objc-bootstrap-3.1.32/]$ make
> HOME/objc-bootstrap-3.1.32/]# make install
> HOME/objc-bootstrap-3.1.32/]# exit ; cd ../objc-3.1.32
> HOME/objc-3.1.32/]$ .configure --prefix=$HOME
> HOME/objc-3.1.32/]$ make
> HOME/objc-3.1.32/]$ su ...
> HOME/objc-3.1.32/]# make install ; exit ; cd
> HOME/]$ vi edit hello.m ; ./hello joe joe luke ...

Perfect.  No problems.  If the 'hello' test works, all is fine since that
test uses classes such as "Set" and "String", I believe.

There is also a test called "DependencyGraph" which is the example that is
in the book by Brad Cox and Andy Novobilski.

That test also uses ObjectPak, so I'm sure ObjectPak is present if you get that
to work.

The errors that you have posted seem to be related to the objpakr (refcnt)
variant not  yet being installed.

If you have them now, you'll find them in /usr/local/lib

 
 
 

new user install of poc

Post by gkap » Wed, 02 Jul 2003 05:15:42


I don't mean to beat a dead horse, but I don't as yet have a gestalt of
the objc system: the results from $HOME/]$ ls -l $HOME/lib/* are :
-rw-rw-r-- 1 me me 1489272 cakit.a
-rw-rw-r-- 1 me me 1836278 cakitr.a
-rw-r--r-- 1 me me     437 objcdlso.ld
-rw-r--r-- 1 me me    4699 objchelp.txt
-rw-rw-r-- 1 me me   55360 objcrt.a
-rw-rw-r-- 1 me me   62428 objcrtr.a
-rw-rw-r-- 1 me me  334992 objpak.a
-rw-rw-r-- 1 me me  406830 objpakr.a
-rw-r--r-- 1 me me     616 _predll.o
-rw-r--r-- 1 me me     617 _prelink.o
/usr/local/lib/  - is empty.  are these the expected values?


>>following is a representation of my install:

>>HOME/]$ ls : objc-bootstarp-3.1.32.tar.gz , objc-3.1.32.tar.gz
>>HOME/]$ tar xvzf objc-bootstrap-3.1.32.tar.gz
>>HOME/]$ tar xvzf objc-3.1.32.tar.gz
>>HOME/]$ cd objc-bootstrap-3.1.32 ;./configure --prefix=$HOME
>>HOME/objc-bootstrap-3.1.32/]$ make
>>HOME/objc-bootstrap-3.1.32/]# make install
>>HOME/objc-bootstrap-3.1.32/]# exit ; cd ../objc-3.1.32
>>HOME/objc-3.1.32/]$ .configure --prefix=$HOME
>>HOME/objc-3.1.32/]$ make
>>HOME/objc-3.1.32/]$ su ...
>>HOME/objc-3.1.32/]# make install ; exit ; cd
>>HOME/]$ vi edit hello.m ; ./hello joe joe luke ...

> Perfect.  No problems.  If the 'hello' test works, all is fine since that
> test uses classes such as "Set" and "String", I believe.

> There is also a test called "DependencyGraph" which is the example that is
> in the book by Brad Cox and Andy Novobilski.

> That test also uses ObjectPak, so I'm sure ObjectPak is present if you get that
> to work.

> The errors that you have posted seem to be related to the objpakr (refcnt)
> variant not  yet being installed.

> If you have them now, you'll find them in /usr/local/lib

 
 
 

new user install of poc

Post by gkap » Thu, 03 Jul 2003 05:26:08


I have run the code in DependencyGraph with the results shown below .
Did I in fact get the expected results?

]$ cd objc-3.1.32/DependencyGraph
]$ make
...
ojbc -g -c mGraph.m -o mGraph.o
mGraph.m:29: warning: difinition of 'yyparse' default to 'int'
Syntax.m:13: warning: difinition of 'yylex' default to 'int'
Syntax.m:16: warning: difinition of 'yylex' default to 'int'
]$ ./dg main INPUT
Reachable from main
terminate phaseTwo closeInputFile openInputFile closeOutputFile
initialize main
phaseOne compute

Not reachable from main:
phaseThree

Undefined nodes:
closeInputFile openIputFile closeOutputFile
--------end-of-output--------
I wonder, is there a validation suite for the poc system?
Tkank you.



>>following is a representation of my install:

>>HOME/]$ ls : objc-bootstarp-3.1.32.tar.gz , objc-3.1.32.tar.gz
>>HOME/]$ tar xvzf objc-bootstrap-3.1.32.tar.gz
>>HOME/]$ tar xvzf objc-3.1.32.tar.gz
>>HOME/]$ cd objc-bootstrap-3.1.32 ;./configure --prefix=$HOME
>>HOME/objc-bootstrap-3.1.32/]$ make
>>HOME/objc-bootstrap-3.1.32/]# make install
>>HOME/objc-bootstrap-3.1.32/]# exit ; cd ../objc-3.1.32
>>HOME/objc-3.1.32/]$ .configure --prefix=$HOME
>>HOME/objc-3.1.32/]$ make
>>HOME/objc-3.1.32/]$ su ...
>>HOME/objc-3.1.32/]# make install ; exit ; cd
>>HOME/]$ vi edit hello.m ; ./hello joe joe luke ...

> Perfect.  No problems.  If the 'hello' test works, all is fine since that
> test uses classes such as "Set" and "String", I believe.

> There is also a test called "DependencyGraph" which is the example that is
> in the book by Brad Cox and Andy Novobilski.

> That test also uses ObjectPak, so I'm sure ObjectPak is present if you get that
> to work.

> The errors that you have posted seem to be related to the objpakr (refcnt)
> variant not  yet being installed.

> If you have them now, you'll find them in /usr/local/lib

 
 
 

new user install of poc

Post by David Ste » Fri, 04 Jul 2003 03:14:12



> I have run the code in DependencyGraph with the results shown below .
> Did I in fact get the expected results?
> ]$ ./dg main INPUT

...

Yes.  You got the expected results.  To verify that it is indeed so, see
for example the book that contains the DependencyGraph example.

It is coded there in two versions : there is a C version (producing the
same output) and an Objective-C version.

The book discusses the differences (and similarities) in coding for both
versions, and what the use of Objective-C is in this.

During development of POC, I had a suite of test cases for POC, but POC
is stable already for a long time, I think the best test case is the software
that is  being compiled with POC ! See for example http://nmhds.sourceforge.net

 
 
 

new user install of poc

Post by gkap » Fri, 04 Jul 2003 11:12:18


I want to thank you for the time and help that you have given to me. As
you can probably tell from my cautious approach I have dropped in to
objc from an alternate universe. It isn't the stability of the code that
  I question rather I would seek confirmation that I have downloaded a
"certified" source copy. For example if I could verify the md5sum.


>>I have run the code in DependencyGraph with the results shown below .
>>Did I in fact get the expected results?

>>]$ ./dg main INPUT

> ...

> Yes.  You got the expected results.  To verify that it is indeed so, see
> for example the book that contains the DependencyGraph example.

> It is coded there in two versions : there is a C version (producing the
> same output) and an Objective-C version.

> The book discusses the differences (and similarities) in coding for both
> versions, and what the use of Objective-C is in this.

> During development of POC, I had a suite of test cases for POC, but POC
> is stable already for a long time, I think the best test case is the software
> that is  being compiled with POC ! See for example http://nmhds.sourceforge.net

 
 
 

new user install of poc

Post by David Ste » Fri, 04 Jul 2003 15:14:54


The md5sum of the objc package on the one hand and the result of dg main INPUT,
if you save it into a file /tmp/foo on the other hand:

2188bfead367ef1e0e56f0735641f819  /tmp/foo
ee713974b44d6bf8894b4ce8e8db914e  /usr/local/src/objc-3.1.32.tar.gz


> I want to thank you for the time and help that you have given to me. As
> you can probably tell from my cautious approach I have dropped in to
> objc from an alternate universe. It isn't the stability of the code that
>  I question rather I would seek confirmation that I have downloaded a
> "certified" source copy. For example if I could verify the md5sum.



>>>I have run the code in DependencyGraph with the results shown below .
>>>Did I in fact get the expected results?

>>>]$ ./dg main INPUT

>> ...

>> Yes.  You got the expected results.  To verify that it is indeed so, see
>> for example the book that contains the DependencyGraph example.

>> It is coded there in two versions : there is a C version (producing the
>> same output) and an Objective-C version.

>> The book discusses the differences (and similarities) in coding for both
>> versions, and what the use of Objective-C is in this.

>> During development of POC, I had a suite of test cases for POC, but POC
>> is stable already for a long time, I think the best test case is the software
>> that is  being compiled with POC ! See for example http://nmhds.sourceforge.net

 
 
 

new user install of poc

Post by gkap » Sat, 05 Jul 2003 06:17:15


62fe18ed5caf288c4373b115e81e6367 objc-bootstrap-3.1.32.tar.gz

I am using the sources in the examples subdirectory as a means of
learning the objc conventions, I realize that this is not an objc
question; but if you don't mind : what is it that I need to install so
as to be able to compile and test esql.em ? Upon running the Makefile
from examples it return a message "Problem running SQLCC( ).


> The md5sum of the objc package on the one hand and the result of dg main INPUT,
> if you save it into a file /tmp/foo on the other hand:

> 2188bfead367ef1e0e56f0735641f819  /tmp/foo
> ee713974b44d6bf8894b4ce8e8db914e  /usr/local/src/objc-3.1.32.tar.gz

 
 
 

new user install of poc

Post by David Ste » Sat, 05 Jul 2003 06:39:35


Quote:> 62fe18ed5caf288c4373b115e81e6367 objc-bootstrap-3.1.32.tar.gz

> what is it that I need to install so
> as to be able to compile and test esql.em ? Upon running the Makefile
> from examples it return a message "Problem running SQLCC( ).

This requires Embedded SQL (esql), from IBM Informix.

It's an Objective-C with embedded SQL statements in there, just as you can
have embedded SQL in C source.

 
 
 

new user install of poc

Post by gkap » Tue, 08 Jul 2003 23:57:01


It is obvious that I am a naive user, and so I hope the you understand
that I pose this question with good will: In the thread "Field
Initialization" would the use of gcov or gprof provide information
relevant to the discussion?

>>62fe18ed5caf288c4373b115e81e6367 objc-bootstrap-3.1.32.tar.gz

>>what is it that I need to install so
>>as to be able to compile and test esql.em ? Upon running the Makefile
>>from examples it return a message "Problem running SQLCC( ).

> This requires Embedded SQL (esql), from IBM Informix.

> It's an Objective-C with embedded SQL statements in there, just as you can
> have embedded SQL in C source.

 
 
 

new user install of poc

Post by gkap » Wed, 09 Jul 2003 23:57:55


If you don't mind, what is the cause of the difference between the
source in the src subdirectory as opposed to the include subdirecotry?
 
 
 

1. File access after user install/ finding path where user installed your program?

I dont install everything to /program files, and I dont think I should
expect other users too either.  I want to do this without making a reg key
if possible.  I have found 2 Win file API function that might help.  If
someone else has any experience with these I would like to know.

SearchPath
GetFullPath

If I understand correctly, I could use these to find the path to a file (ie
search for: my.exe) and reciev a string (d:/program/) that I could combine
to make a string (d:/program/ + filename.000) for use with CreatFile to open
it.

Am I doing it the wrong way?  Maybe the uses intsalled to c:\temp\ .  I dont
want the user to get a lot of "file not found" errors because they installed
to a different drive/directory than the one I compiled in.

2. laserwriter pro driver

3. MSDN/VB Studio Install-New User

4. Please help -unable to view .chm files

5. Does POC have an internal identifier for POC?

6. M98 does not comply with OFX specs

7. Installing for All Users (and not only currently logged on user)

8. OODB for VLSI Design

9. new POC package for watcom

10. new upload poc 1.9.13

11. how to assign user Dial In access right after add new user using NetUserAdd?

12. w2k: runas command doesn't work with /user:user but work with /user:domain\user

13. runas command doesn't work with /user:user but work with /user:domain\user