RE: TEC 3.7.1 Information Button

Yes, I realized that after Thomas sent me a suggestion.  Everything is working 
fine now.  The problem was, we wanted to display the web page based on 
hostname.  On the web page itself, we only wanted to display the message and 
time stamp of the Tec Event.  Then we needed a link to how the event would be 
handled and in the case where someone was to be paged, our on-call list 
(Exceed) was supposed to come up.  

I got it all figured out and working if anyone is interested in a similar 
situation and doesn't know much about HTML.

Regards,

"Mina Merali, IBM Certified Deployment Professional
- Tivoli Enterprise Console v3.9, Tivoli Data warehousing v2.1."
(905)513-5023



-----Original Message-----
From: owner-tme10@xxxxxxxxxxxxxxxx [mailto:owner-tme10@xxxxxxxxxxxxxxxx] On 
Behalf Of michael.staats@xxxxxxx
Sent: Thursday, November 25, 2004 3:17 AM
To: tme10@xxxxxxxxxxxxxxxx
Subject: RE: [tme10] TEC 3.7.1 Information Button



>> cgi-lib.pl does a GET and POST which basically
>> get the data and posts it into key=value format.

>you mixed up the HTTP functions GET and POST, I'm afraid.
>
>GET means that the URL contains all of the parameters
>(like http://bla.foo.com/cgi-bin/hallo?a=x1&b=x2&c=%22%54%53%22)
>and the CGI script has to read the internal env variable QUERY_STRING 
>to parse the webserver data.

>[...]

>if ($ENV{REQUEST_METHOD} eq "GET") {
>  $data=$ENV{QUERY_STRING};
>}
>elsif ($ENV{REQUEST_METHOD} eq "POST") {  
>read(STDIN,$data,$ENV{CONTENT_LENGTH});
>}
>%var=&decode($data);

If you are not restricted to perl4 I strongly recommend using the CGI module 
from perl5.

You just say "use CGI; my $query = new CGI;" and can access the data via 
convienient functions and hashes without thinking about POST or GET or decoding 
the data (which is more complicated than you'd think if you implement the 
standard, not just what gets sent from most browsers at a first glance). CGI 
just works.

A completely different problem I ran into when writing my own tec_help.pl was 
that you must not print 8 bit characters back für your HTML output, althoug it 
is completely html-legal to use them if you declare the code set (e. g. ISO 
8859-1). If you use them, the console says that it can't get the data back from 
the server, be it the builtin Spider (not recommended for production) or a 
standalone one. So you must encode all data coming from your tec slots when you 
want to print them, since you cannot know in advance what is in them. 
Especially when localized OS versions are in use for languages that use 
accented cahracters (i. e. basically all non-english versions... or is there a 
latin OS somewhere out there, I don't remember accented characters in latin :-).

Bye,
        Michael
-- 
Dr. Michael Staats
RWE Systems Computing GmbH
Data Center
SIC-PS Systems Management
Altenessener Str. 37/39
45141 Essen

T intern 70-26919
T extern +49(0)201/12-26919
F extern +49(0)201/12-24751
mailto:michael.staats@xxxxxxx