phunni

You shouldn't need to do much string manipulation - I think there may well be a method for pulling the about of the URL anyway... There is in java - although I can't remember what it is off the top of my head...

Basically if www.mypage.com/ maps to a particular page then you can simply strip off the entire URL (minus query string) after than point.

But - wouldn't it be easier to have it as a query parameter:

www.mypage.com?style=about

You could then simply pull the "style" parameter off the incoming request parameters... To my mind it would be a much simpler approach...

Out of interest - does the -i switch work on your Sun box for sed?

On the original sed command:

sed -i -e "s/localhost:1433/164.11.132.104:1433/g" myuwe/apps/uPortal/properties/PersonDirs.xml

I've made two very minor changes:

perl-i -pe "s/localhost:1433/164.11.132.104:1433/g" myuwe/apps/uPortal/properties/PersonDirs.xml

and it seems to be working a treat. I was trying to avoid perl because I was actually porting from a perl script and it did a whole bunch of complicated stuff... The fact, however, that I can use perl as a simple command - like sed (almost exactly like sed as it turns out...) means that it's still a bash script and I can understand and maintain it easily

That's proving hard to get working in a script for some reason...

Is there a better option for find/replace than sed? The arse of it is that sed with a -i was perfect, but the soalris version doesn't have it... :sad:

It would appear that -i is not an argument all versions of sed will accept. It works quite happily in cygwin for example, but not on solaris where only -n -e and -f are supported

Can anyone come up with a simple alternative?

Thanks neddie - you're my bash guru!

I'm trying to run the following line in a bash script:

sed -n s/localhost:1433/164.11.132.104:1433/g myuwe/apps/uPortal/properties/PersonDirs.xml

My understanding is that ant instances of the string "localhost:1433" should be replaced by "164.11.132.104:1433", but this doesn't seem to be happening...

What am I getting wrong...?

Amazing what small, but crucial, things you can so easily miss...

Thanks - that's solved it...

OK, I'm trying to write a bash script - so I googled a couple of tutorials and got started. The first thing I wanted to do was to write a simpel function - all the tutorials seemed to suggest I could.

So, to test the concept I wrote the following:

#!/bin/bash
echo "Thanks for coming..."
finish

function finish() {
 echo "done..."
 EOF
}

When I attempt to run this I get:

Thanks for coming...
./myuwefit.sh: line 8: finish: command not found

So, what am I missing here...?

I'm not 30 yet... :P

Dansguardian is the daddy for content filtering...

I'm an arch user also - and you do find a lot of helpful friendly people on the arch forums, but you also encounter a lot of elitism where people only want a few "experts" to discover their precious little distro.

For me it's great because it's fast, lightweight and has the best (imho) package manager around - in fact it's the package manager that really makes it...

I had arch running on a laptop without any significant problems...

Boredom - pure and simple. I was bored one day and decided to try and see if I could get Linux working.

Haven't really looked back since...

Mosst of your questions I can't answer off the top of my head - but I will say that you should change your root password straight away - whether you think it's been compromised or not.

Also - see if any services are running that you know nothing about and if any other ports have been opened...

It looks like SQLLight might be a good option - there are drivers for mono, so it should be fine on both windows and Linux...

I'm writing in C#. So, if I used an access DB file, would that work happily on Linux if I had the right drivers?

I've had trouble with a Belkin KVM and Linux (despite it claiming to be Linux compatible on the box) - it made the mouse go crazy until I unplugged it and plugged it back in again...

I'm currently messing around with developing a simpl game. At present all the startup data is based on an XML file, but it struck me that it would faster and easier to code if it was all in a database file.

Now, the only databases I've ever worked with are enterprise server types - and some testing with access databases. What I need is a database that can be distributed as a file with the app - a la Access, but is platform independant since I want the game to be for Windows and Unix...

Any suggestions? Is there such a thing?

The GPL afaik doesn't stop you selling anything - but you have to make the source code available...

The entry can be purely a logo - or it can include a background and therefore be a "banner" Either is fine...

Fair enough then! :D

I hadn't noticed the windows logo!

I was commenting on the fact that your background is from yetisports - I wondered if there was an issue there...

Are there not copyright issues with those images...?

Linux is fine for gaming - I've done it loads as have others here.

Many games have native linux ports and for those that haven't you hould check out cedega