Compile Docs on RH 8.0

classic Classic list List threaded Threaded
4 messages Options
Reply | Threaded
Open this post in threaded view
|

Compile Docs on RH 8.0

Liguo Song
When trying to build the doc rpm package, I encountered a problem on the docs-ps
stage. After some investigation, I learned that it is due to the secure mode of
dvips that Redhat configured. In this mode, if no option is given, the output of
dvips will go to a printer and it is not allowed. Please read the following
discussion for details:

http://www.mail-archive.com/tetex@.../msg01782.html
https://sourceforge.net/tracker/index.php?func=detail&aid=717663&group_id=23164&atid=377580

One solution to avoid this problem is to invoke dvips with "-o file.ps". This
redirect the output to the stdin, and we can redirect it to the postscript file.
For example in the Docs/latex directory, /usr/bin/dvips -o refman.ps refman.

This is causing the problem of building docs on RH8.0, and I don't know about
RH9. I will test later.

Should we use this in our make files? If so, can anyone commit this change?

Thanks.


Liguo (Leo)



Reply | Threaded
Open this post in threaded view
|

Re: Compile Docs on RH 8.0

Luigi Ballabio-2
At 12:16 AM 10/3/03, Liguo Song wrote:
>if no option is given, the output of dvips will go to a printer.
>One solution to avoid this problem is to invoke dvips with "-o file.ps".
>Should we use this in our make files? If so, can anyone commit this change?

Yes, we should. I just committed the change.

Thanks,
         Luigi




Reply | Threaded
Open this post in threaded view
|

Re: Compile Docs on RH 8.0

Liguo Song
Another thing that I noticed when compiling the documentation of QuantLib-0.3.3
is that doxygen-1.3.4 (1.3.3 is required) will complain about obsolete tags,
such as:
   Warning: Tag `CGI_NAME' at line 220 of file .quantlib.doxy has become obsolete.
   To avoid this warning please update your configuration file using "doxygen -u"

Using "doxygen -u .quantlib.doxy" before "doxygen .quantlib.doxy" seems solve
this problem. Maybe we should also put this into the Makefile of Docs.

I am not familiar with doxygen. Any suggestion from people with more experienced
with doxygen?

Have a nice weekend.


Liguo (Leo)



Luigi Ballabio wrote:

> At 12:16 AM 10/3/03, Liguo Song wrote:
>
>> if no option is given, the output of dvips will go to a printer.
>> One solution to avoid this problem is to invoke dvips with "-o file.ps".
>> Should we use this in our make files? If so, can anyone commit this
>> change?
>
>
> Yes, we should. I just committed the change.
>
> Thanks,
>         Luigi
>
>
>
>
> -------------------------------------------------------
> This sf.net email is sponsored by:ThinkGeek
> Welcome to geek heaven.
> http://thinkgeek.com/sf
> _______________________________________________
> Quantlib-dev mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/quantlib-dev




Reply | Threaded
Open this post in threaded view
|

Re: Compile Docs on RH 8.0

Luigi Ballabio-2
At 1:20 PM -0500 10/3/03, Liguo Song wrote:

>Another thing that I noticed when compiling the documentation of
>QuantLib-0.3.3 is that doxygen-1.3.4 (1.3.3 is required) will
>complain about obsolete tags, such as:
>   Warning: Tag `CGI_NAME' at line 220 of file .quantlib.doxy has
>become obsolete.
>   To avoid this warning please update your configuration file using
>"doxygen -u"
>
>Using "doxygen -u .quantlib.doxy" before "doxygen .quantlib.doxy"
>seems solve this problem. Maybe we should also put this into the
>Makefile of Docs.
>
>I am not familiar with doxygen. Any suggestion from people with more
>experienced with doxygen?

Well, it's just that tags sometimes change between Doxygen releases.
Usually tags get added, and newer versions of Doxygen don't complain
upon processing older doxyfiles. This time, tags were removed, so
that the latest Doxygen issues obsolescence warnings. But: 1) this is
seldom the case; 2) they were only warnings, and the docs would have
been made succesfully anyway; 3) the doxyfile needs to be changed
only once per Doxygen release, and 4) the version in CVS is already
updated for 1.3.4, so I don't think a Makefile rule is needed...

Later,
        Luigi