Threadsafety/Exceptions for Java Mapscript?

[Daniel Caldeweyher]

Sean, Umberto,

thank you for your prompt replies and help.

For the past couple of days I've been testing using synchronized blocks and
methods:

mapObj mapobj = null;
synchronized(mapobj) {
  mapobj = new mapObj(path);
}
......
synchronized(map) {
  mapURL = mapbean.renderImage(map.draw(), "_mainmap");
}
......
public synchronized String renderImage(imageObj img, String suffix) {
  long rnd = Math.abs(new java.util.Random().nextLong());
  String imgfile = rnd + suffix + '.' + img.getFormat().getExtension();
  img.save(img.getImagepath() + imgfile, map);
  return img.getImageurl() + imgfile;
}

Using synchronized has certainly improved stability. However I haven't yet
compiled Java Mascript using the latest CVS because I am developing on
Windows XP and and there exists no java mapscript make file for Win32.

Once we put the first version of my project up for public testing it will
be on a Linux machine, I will then be able to test the latest CVS.

Until then I will hope that a pre-compiled Java mapscript will be uploaded
to http://hobu.stat.iastate.edu/mapserver/build_output;

Thanks,

Daniel


On Mon, 11 Apr 2005 10:01:38 -0600, Sean Gillies <sgillies at FRII.COM> wrote:

>On Apr 11, 2005, at 3:49 AM, Umberto Nicoletti wrote:
>
>> On Mon, 2005-04-11 at 03:33 -0500, Daniel Caldeweyher wrote:
>>> Hi everyone,
>>>
>>> first of all I would like to thank all the developers for this great
>>> piece
>>> of software.
>>>
>>> I am working on a GIS project for my University using JavaMascript on
>>> Tomcat 5.5.7. My supervisor would to like to go online with my
>>> solution in
>>> a month or two for public testing. In the beginning 'public' means
>>> open to
>>> all University members, other Stakeholders and for presentations.
>>> However
>>> in the (near) future it is very likely that my project will be made
>>> available to the general public, initially targetted at South East-
>>> Queensland only, but the whole of the state and an Australian-wide
>>> installation are likely to follow. That is why I gave scalablity and
>>> maintainability a high priority and chose Java Servlet/JSP/struts as
>>> the
>>> development basis.
>>>
>>> When I started Umberto Nicoletti's website indicated that "A patch
>>> that
>>> enables exceptions in mapscript, so that mapscript errors don't crash
>>> your
>>> JVM (and Tomcat) anymore." will come soon. Well its two-three months
>>> later
>>> now and I was wondering when it will be released? It is very
>>> important to
>>> me as I would not be very happy if i had to port my entire site to
>>> php or
>>> python.
>>>
>>
>> The patch is in mainstream mapserver, I'll update my web site asap.
>> There has been some discussion on thread safety lately so make sure you
>> check out the archives: search the subject for 'Java MapScript issues
>> Servlet'.
>>
>> You could find this link interesting (or search bugzilla for threads)
>>
>> http://mapserver.gis.umn.edu/bugs/buglist.cgi?
>> bug_status=UNCONFIRMED&bug_status=NEW&bug_status=ASSIGNED&bug_status=RE
>> OPENED&field0-0-0=product&type0-0-0=substring&value0-0
>> -0=thread&field0-0-1=component&type0-0-1=substring&value0-0
>> -1=thread&field0-0-2=short_desc&type0-0-2=substring&value0-0
>> -2=thread&field0-0-3=status_whiteboard&type0-0-3=substring&value0-0
>> -3=thread
>>
>> Regards,
>> Umberto
>>
>
>Daniel, Umberto,
>
>I just committed a change to CVS related to bug 1203
>
>   http://mapserver.gis.umn.edu/bugs/show_bug.cgi?id=1203
>
>I'd appreciate it very much if you would read the discussion there, and
>if possible, update your CVS mapserver and help me test.
>
>cheers,
>Sean
>
>--
>Sean Gillies
>sgillies at frii dot com
>http://users.frii.com/sgillies