[Thread Prev][Thread Next][Index]

RE: LAS Errors



Roland,

 

          Thank you for your responses.  The problem with the map applet does relate to a Sun Java plugin. I found that the IE internet options included Java (Sun) Use Java 2 v1.4.0. for <applet> with the box ticked. I removed the tick and the map applet now works properly.

 

 

Regards

 

Peter Turner

 

-----Original Message-----
From: Roland Schweitzer [mailto:Roland.Schweitzer@noaa.gov]
Sent: Saturday, 5 June 2004 2:51 AM
To: Turner, Peter (Marine, Hobart)
Cc: Jonathan.S.Callahan@noaa.gov; las_users@noaa.gov
Subject: Re: LAS Errors

 

Peter,

Peter.J.Turner@csiro.au wrote:

Jon and Roland,

 

      Thanks for your responses.  I located the problem yesterday which is why the server now works.  In the xml files that describe each dataset specified a latitude range as is in the data files:

 

<arange start=”10.039” step=”-0.036” size=”2085”>

 

it appears that the non-applet version of the map does not like this and exhibits the behavior I described.  There was an area in the Java script on the web source that had ylo = 10.039 and yhi = -64.9979. Once you selected the non applet map the browser you got the “ylo > yhi” message and the browser completely lockup for anything else you tried. The only way to stop the problem was to start a new browser.

 

I have fixed the problem by setting the range as:

 

<arrange start=”-64.9979” step=”0.036” size=”2085”>

 

I am not sure if this has any bad ramifications (not the order in the data files) and would be very grateful for suggestion about the right way to solve the problem.

This is the correct solution for the problem.  The configuration information contained in this axis definition is only used to initialize the map that appears in the user interface (either the applet or the server-side map) and this does not affect how the back-end analysis program "sees" the data and the fact that the storage order in the data file is different should not affect the operation of the product server.  I suppose we could consider having the servlet make some assumptions in the case where ylo>yhi and reverse them, but unless there is a great hue and cry from the users I am inclined to leave things as they are now.

 

The residual problems are that the “Select Region” options do not seem to do anything. In this case I can set that option to have no options or remove it all together. I suspect that it may no work because I am not working with a global dataset?

In general if you select a region the servlet checks to see if the region selected intersects the valid geographic range of the data.  Looking at you server, I suspect that none of the regions we have defined (except Australia) intersect the geographic region where your data are valid.  You seem to have hit upon a good solution, which is to eliminate the menu choices.  You could define your own regions contained within the area where your data is valid if need be.


I also have the minor problem that for most browsers the map applet shows the data region in colour and the rest of the world in grey. On my own computer (DELL D600, Windows XP and IE6) the whole world map is coloured.  I think we can live with this one, however, it does seems rather odd!

 

I have a similar set up (IE 6.0.28 on XP with the Java 1.4.1_01 Plug-in).  If indeed the problem occurs with the applet and not the server-side map I would be curious to know what JVM your browser is using.  Are you using the Sun Java plug-in with this browser?


Thank you for your comments and would particularly like to know it my adjustment of the “arrange” entry is the right thing to do!

 

Indeed it was.

Roland


Regards

 

Peter T

 

 

-----Original Message-----
From: Jonathan Callahan [mailto:Jonathan.S.Callahan@noaa.gov]
Sent: Friday, 28 May 2004 4:46 AM
To: Turner, Peter (Marine, Hobart)
Cc: las_users@noaa.gov
Subject: Re: LAS Errors

 

Peter,

I just had a look at your excellent server and was unable to reproduce the errors you describe.  I am using IE 6.0 (6.0.2800.1106) and Netscape 7.1 on a Windows XP box.

I wish we could offer a solution for you but it is nearly impossible to debug reproducable errors occurring on remote machines.  We would need to know exactly what the user was requesting to figure out why 'ylo must be less than yhi' was issued as an error message.

As for #3, there is already code in LAS that checks for browser incompatibilities and shunts the user to the non-applet interface when necessary.

Has anyone else in the LAS community seen these kinds of problems?


-- Jon


Peter.J.Turner@csiro.au wrote:

Hi,

 

   I am trying to setup a LAS server to serve Sea Surface Temperature data in the Australasian region. After some hand to hand combat with the system, particularly the ferret descriptor file I have a working system http://www.marine.csiro.au/las/servlets/dataset

I still need to customize the interface; however, there are a couple of issues relating to browsers:

 

1.     On some browsers the world map appears completely coloured in, and in others the area of the data is shown coloured and the rest grey. Mozzilla seems to work OK on a PC, but we have two “identical” versions of Internet Explorer 6 (6.0.2800.1106) which behave differently. One displays the applet correctly and the other does not. Why is this so?

2.     We have had an external client who was able to view and download some data using ie5, and then his system stopped working and he got the message:

 

 LAS server error

You have encountered an unexpected User Interface error! Please click here to return to the Datasets page. If this error persists please contact the LAS administrator for this site.


The following error occurred:

javax.servlet.ServletException: ylo must be less than yhi

etc

 

 

I understand that this might be a browser compatibility error but we cannot understand why it worked and then stopped. We have tried to replicate his problems with ie5 and cannot!

 

3.     I note that some sites using LAS check the browser and determine if it is OK to use.  Is there code available to do this?

(http://www.nerc-essc.ac.uk/las/)

I would appreciate some guidance as to the best way manage these problems.

Thanks in advance

 

Peter Turner

 

 

 

Dr Peter J. Turner

CSIRO Marine Research

Phone +61 3 6232 5094

Fax     +61 3 6232 5000

email Peter.J.Turner@csiro.au

 


[Thread Prev][Thread Next][Index]

Dept of Commerce / NOAA / OAR / PMEL / TMAP
Contact Us | Privacy Policy | Disclaimer | Accessibility Statement