ls *.tgz>myfilelist
for file in `cat myfilelist`; do
cp $file /home/username/;
done

I could also untar and process them too if I wanted – e.g.

ls *.tgz>myfilelist
for file in `cat myfilelist`; do
tar xzf /home/username/$file;
done

or something…
(corrected typo – in original I changed directory inside the script which would break finding the next file)

anyway…

if you want to copy data on a *nix machine from one place to the other and preserve permissions, you can do this…

From the source directory…

tar czf  – * | (cd /target ; tar xzfp -)

This creates a compressed tar stream from the current directory to the standard input, then changes to the target directory and uncompresses the stream, preserving ownership and permissions into the target directory.

there is an alternative form – not sure if it’s supported by all versions of tar but it runs like this…

tar czf -C/source/ – * |  tar xzfp -C/target/ -

where -C says change to directory blah. I need to check the syntax – e.g. absolute versus relative, and if there is a space b/w the C and the name… if in doubt, check ‘man tar’…

My 6 rules for doing a PhD…

1) don’t do it.

2) don’t do it.

3) don’t do it.

4) don’t do it.

5) don’t do it.

6) don’t do it.

OK. if you really want to know, don’t do it…

1) unless you really really really really really really want to.

Academia is a strange beast, feircely competitive, not much actual research, no money and usually you don’t get to choose what you research. So saving the world is usually not on the cards. Getting noticed is a vague possibility, but you’d do better on a game show.

2) You have a super-fantastic supervisor.

Make that 2 really super fantastic supervisors. In 2 subject areas. That way when one leaves for a better position elsewhere (because why would they stick around anywhere near here if they were any good) you will still have one good supervisor. Also try and get them to have different areas of expertise. This way a) they can’t disagree with eachother much, and 2) you have to learn each of their areas good enough to explain it to the other. You will therefore learn both areas and do well.  Also if you are looking at something that covers two disciplines you’ll probably be doing something interesting…

3) DO NOT give up your life to do it.

Make sure you have things other than the PhD to do. Give up the PhD before you give up the hobbies.

4) DO NOT do the PhD in the same country as your undergrad.

Even the same state is bad. The same uni is asking for trouble. The whole point of the PhD is to grow. Familiarity breeds contempt. Besides, it should be hard.

5) The subject is so clearly an award winning idea it make you sit up at night wondering how no one else saw it before.

Odds on they have… this means that at the last 3 weeks before subittion you will find 5 years of papers in an obscure Finnish publication a) refuting your work and/or b) giving exactly the same experiments in greater detail, etc. The alternative is something so boring people’s eyes will glaze over upon recieving an email with your CV in it… without even turning on their computer.  If it truely is a new idea work hard and fast because it won’t stay new for long…  Most cutting edge research seems to be discovered simultaneously at two or three different places around the world.

6) join a research group.

Make sure you have collegues working on similar things. share results with collegues, go to as many
conferences as you can. Network. Email friends OS. Be active in your community, whatever form that takes. Then it does not matter if you ever complete.

OK. if at the end of all that you still want to do it, good luck. now read phillip greenspun’s Career guide for PhD’s

Thank you Ubuntu!

*sheesh*  what a mess.

I started with Ubuntu 6.04 and have used the upgrade path each time – either apt-get dist-upgrade or the upgrade tool. I’ve had to rebuild my machine atleast 4 times as a consequence.  You’d think I’d've learned by now…

I’m rather frustrated. Each time I do a fresh install I have to re install and configure everything.

This time I was using the upgrade tool – which does not show you what it is doing – could not get the console to show anything – so when it hung downloading (on er… about day 2 of the download), I could not see why…

so I shut down the GUI and ran apt-get dist-upgrade instead

fine.

installed, rebooted… er. no dual screen. OK, check the nvidia drivers. Ah, need to activate the  nvidia 180 driver… OK, reboot… no gdm. bummer. OK, reconfigure xorg – xorg not installed. install xorg… can’t – can’t remove nvidia driver. remove nvidia driver – does not exist. install then remove nvidia driver- no install candidate. insert several more hours of frustraction then *scream*

so I’ve burned the ISO to CD and installed on a fresh partition…

said it before… what a mess.

addendum – As an example – things that don’t work now – no pdf viewer from firefox, no youtube vids.  Isn’t it about time they got over their culture war on closed source?

A red capsicum, chopped

a couple of tomatoes – 3 I think – roughly chopped

a small amount of spagetti

3 mushrooms

spash of red wine, splash of olive oil

3 v thin slices of salami

1 kangaroo fillet – sliced into 1cm slices

heat pan, boil jug, defrost roo.

fry capsicum, add tomato

add wine and oil

splash of chilli if that’s what you like.

add the mushrooms and shredded salami

turn off heat and cover with lid

boil spagetti, grill roo – do not over cook – should be bloody

assemble.

spagetti in a flat pile.

sprincke some cheese

roo next.

then sauce…

rest of the bottle of wine…

Drool….

what beautiful prose – how could I delete that?

*sigh*