willisoften

Bumping this to the top in the hope that someone can give me some current information. I've been a bit out of touch. I appear to be having the same problem in both Ubuntu 5.10 and Slackware 10.2 it appears firefox browser related but occasionally elsewhere. AMD processor nforce 2 chipset SATA drives Both systems are pretty much up to date and on different drives. Installing Firefox 1.5 made no difference. Nor did the other fixes suggested in this post. I'm busy doing coursework at the moment so I was hoping someone else would do the hard work for me :) I'll probably move back to Debian proper in the near future so if theres anything to do to fix this in a scratch install that would be good to know too.

Using transformToURI() to create a text file $xslt_parse->transformToURI($xml_obj,$uploadDir.'output.txt'); Some users will be using Linux and some Windows. Currently using Apache on a Windows XP machine as specified. File is created but is full of the little square characters (linefeed / carriage return) Has anyone got a code snippett that will convert a file from unix to dos? I've racked my brains and searched like a mad searching thing and never yet twigged it. Help appreciated - online tutorial is good too. I'm a bit brain-fnookered at the moment so something pretty specific is required. Thanks Will

Edited - badly put! 1. Yes there are forums and mailing lists. You'll have to go to the slackware site and look about to get instructions for subscribing to a slackware mailing list - I haven't subscribed - I've no idea if it's working. http://www.linuxquestions.org/questions/fo....php?forumid=14 is the approved slackware forum but it's not part of slackware it just has Patrick Volkerdings approval. 2. Not as such. Take a look at swaret http://swaret.sourceforge.net/index.php 3. Config files! and a few scripts and utilities like netconfig 4. Two of the cds are source. So you've installed from the other two. Good :) If you installed KDE or Gnome or both you should have the various configuration options that come with the desktop(s). Book: http://www.slackbook.org/ I like slackware - but it can be a bit intimidating. I've been moving a lot of stuff over to slackware on the 2.6 kernel ( a pain to get working on my box) as Debian has really disapointed of late. Sarge is a year behind where it should be and seems to be a bit shy of modern hardware support, while Etch is a bit wobbly. If you need a machine for working-on rather than recreational experimentation - I'd encourage you try Ubuntu or Mepis. Both very competent distros. I'd lean towards Ubuntu as it's more true to Debian and has much more community support. Hardware support is very good. The Mepis community annoys the hell out of me. Can't help it I'm old and cranky.

I switched repositories but all that happened was that I got into a loop that cited "Some packages failed to down load would you like to try another repository?" I started the install again and it ran through without problems - just a huge amount of stuff to configure now.....

I tried Arch before and just wasn't all that impressed but the install was relatively problem free. I went back to Debian - then tried Ubuntu. Then decided to try Arch again. Pacman continually freezes at 99% - the only reference I can find to this on the Arch sites is a bug report - marked fixed. Theres a reference to a workaround but no link. So can anyone tell me how to get round this? I was looking to find out how to restart pacman but don't see anything.

True. Mepis is pretty good though. I often recommend it came across it last year when living in Dublin. Always mix it up with morphix.........

From the Ubuntu Wiki This is now working for us but the save password for this session option was not working. The gnome-keyring is now installed and working and the session option appears to be working for some users. SOLVED - sort of but we haven't got a "fix" as such.

I've been impressed with Ubuntu's latest release. So much so that I've been promoting it at work. We've run into a huge snag - we can't access our network shares. I've tried Gnome's Connect to server smbclient - mount -t smbfs ...... Gnome's Connect to server worked perfectly for me in Debian Sarge. (Gnome 2.8.3?) This has appeared in the wiki but none of the wiki solutions are working for us. Any ideas?

Thanks for the replies, for some reason I didn't get email notifications or I'd have checked back sooner. Maybe they got spam filtered by my ISP? I'll investigate the promise and adaptec cards. Most of the cards I've seen so far mention every version Windows and nothing else. I was a promise controller on my last mobo that allowed me to do this before so I should have thought of them - but...........

We do - "Manlycodriva Linux" just rolls off the tounge.

Currently looking to put all the devices on my new system on seperate ID channels. I'm not looking for raid. Can anyone recommend a card?

I did a few experiements last night and it does create a temporary iso ( I set it to use my home). Seems like burning on the fly would usually do ~ but I think I'll have to dig out an old hard drive and set up a temporary partition on that. I'm going to set up a new system soon anyway and I'll be able to be a bit more promiscuous with my hd space. Thanks ramfree - SOLVED

I watch movies on the pc and I've ripped a few, to watch on the laptop - no dvd rom at one time. At the moment I'm talking about burning an 8GB backup. It was a pretty interesting article anyway. Thanks! Actually I usually just go apt-get install libdvdcss and thats the end of the problem. Any information pertaining to temporary files and DVD burning would be most welcome.

Ok I've just bought a cheap dual layer dvd burner. Lite-on. Does anyone know under Windows it apparently requires 9GB of free space to create a DVD image file. (only for a full disk I hope!) I do have at least 9GB free in /home. BUT how does burning work in Linux ( I use Gnome Baker & Nautilus ) previously I've used /tmp - 1.9GB as the temporary directory to burn disks I'm positive were in excess of 2GB. Does anybody really know how this works. I could set up another partition and call it /burn or something but I don't really want too, it would eat space somewhere else. But if everyone was sure it was the only option..........

Well absolutely every single day? Not just on Tuesday or once or twice a week? It's a short list for everyday. Gnome My usual desktop. Nautilus Gnome Terminal Evolution check my personal e-mail everyday Firefox - The web is a lot of my job. XMMS - Entertainment while I work. man - if that can be called a program. Any text editor. It's Vim at the moment as the fact I use it seems to annoy colleagues. To be truthful I'm perverse enough to enjoy arguing with them about it's obvious superiority. In truth I'm a bad person. I'm not actually fussy when it comes to text editors. Very frequently more than twice a week say: GImageView Gimp I'll have a number of images to manage and / or edit every week. LyX There's almost always documentation of some sort to produce. LyX makes a really nice job of it, with minimum effort on my part. Nearly everything else is done in Windows for Company wide compatibility. I can probably get my Lotus Notes to run under wine if I put in the effort as well.

Thanks, for that I would never have though to use " rather than ' force of habit I suppose.

Is there a way to find an ' in a regular expression apart from [[:punct:]] I want to check a number of files for correct use of s' and 's \' doesn't work but [[:punct:]] does but I can't make the ? operator work when using it. Any thoughts out there? egrep -o '[A-Za-z]*s[[:punct:]]|[A-Za-z]*[[:punct:]]s' filename is currently working for me but it seems incredibly clumsy.

Well good for the schools that can make it work. I've There are other problems related to IT in British schools though not least of which is the lack of any sort of IT department in quite a few smaller schools. My nephew goes to a school with 200 pupils and IT is taught by what seems to be a slightly confused Maths teacher. Theres a belief in UK Education that anyone can teach IT and maintain a network. IT technicians with expertise who are employed by schools are often paid less than teachers with little IT experience or qualification who are teaching it. While MS software may be the cause of many problems and Open source the solution to some of those - there are other problems in IT in UK schools. And General adoption of open-source would perhaps free up the resources to pay for someone qualified to manage and maintain a school network. Though I can't really see that being the case - you need a substanial amount of time and money to make the change. Even long-term savings may not benough to pay someone £20.000 + a year. Not to mention finding a teacher bright enough / motivated enough to learn some new tricks. Theres also the problem of being tied into contracts with MS service providers and justifying the ditching of software you may have already licensed. MS Office is still a business standard. You don't have to like it, it's just true. Thats all the leverage a company needs to sell a school an MS solution. I'm not sure this is really a valid argument but then these companies aren't selling to particularly tech savvy people. AND why sell a system that needs minimum maintenance? Unfortunately many schools would need to be in a stronger position to adopt open source. Coming to the end of service contracts. Proper IT teacher. In a strong position financially. Availability of a supplier who will provide a local solution - and crucially the supplier has to be the cheapest bidder. So as for "Orwell High School in Suffolk" they've done a hell of a job on many levels.

Actually I forgot a couple of things should have been: egrep '^[A-Za-z]{8,9}$' examples which works - just shows you your memory goes as you get older..........

I expected this to find words in a word-list which had at least 8 characters and no more than 9. egrep '[A-Za-z]{8,9}' wordlist My problem is that the 9 doesn't seem to make any difference. So I guess I don't really understand how this works. A number of books and tutorials seem to suggest that n is finite Match only n times {n} but this clearly doesn't seem to be the case. Seems really to be match at least n times. Could some regex guru explain this in small words? I did this last year and don't remember being surprised by the results!

Mandriva is good first time round, recently though I know a number of new users who have had good results with morphix. Why not take a look at a number of live cd's and play around a little? That's what I've been advising people new to our company lug to do. At work Fedora is a popular choice with our first timers. Admittedly theres a fair amount of support available but I'm not sure they need much even though many aren't really tech people. I haven't used it much but it too seems like a nice clean and easy distro. As for morphix the install isn't very flexible though not actually difficult. As it's debian based it's also easy to manitain / upgrade. Cofiguration tools are not as comprehensive as Mandriva though. Best of luck!

Following an LPi course at work (IBM). I didn't choose between it and the RHCE it's just free :) The material isn't that hard - but there's quite a lot of it. So I'd guess the exams could be difficult enough. (Two exams as far as I know: 101 and 102 I think). It tends to be based on rpm based distros. But a lot of the material would apply to any distro. There are books available for each course why not see if you can beg borrow or steal some and take a look and see if there is much difference and which appeals to you more? Good luck whatever you decide to do.

Background: I had a large number of files to check for the presence of a particular legal clause. It's been renamed Bongo. The files are all in different languages but luckily the legal clause refers to a particular named product which always appears in English so in a way I was using grep to search for a keyword. Round here a couple of days was scheduled to open and scan through every file to look for the title! Then a list of files containing the clause and those not containing the clause was to be compiled. You gotta love proprietry software :) Anyway I didn't need a script but there are a lot of people round here are using Linux but are a bit command-line shy. The sript was written to work for them and to demonstrate for the lug just how powerful the various commands can be. It's my first attempt at a script for anything usefull. It's not finished, A lot of it is still hard coded -I'm trying to make it a bit more portable and a bit more useful and perhaps even a bit more elegant. But for the moment it seems to do what it was intended to do. So for what it's worth here it is: #!/bin/bash DEFRESULTS=search_results echo -e "\nThis script will find the word \"Bongo\" in" echo -e "any text file in which it appears.\n" echo -e "Enter the path to the directory where licence files are stored." echo -e "If directory is in $HOME the directory name will be enough:\n" read DRCTRY #check directory exists if [ -n "$DRCTRY" ] && [ -d "$DRCTRY" ] then echo Directory exists else echo Directory does not exist echo Continue Y / n ? read CONT if [ "$CONT" = "y" ] || [ "$CONT" = "Y" ] then until [ -n "$DRCTRY" ] && [ -d "$DRCTRY" ] do echo Enter the directory where licence files are stored: read DRCTRY done echo Directory Exists! else exit 0 fi fi cd $DRCTRY pwd echo -e "Enter the name of output file or results will be written to ~/$DEFRESULTS" read RESULTS if [ -z $RESULTS ] then RESULTS=$DEFRESULTS fi #echo used first with > to create or overwrite output file. echo -e "The Bongo Clause appears in the following files:\n" > ~/$RESULTS #Search for Bongo -l filename instead line including text. grep -l '\<Bongo\>' * | tee -a ~/$RESULTS #count $RESULTS and assign to NUMY NUMY=`grep -l '\<Bongo\>' * | wc -l` #write to standard output and a file #use echo as tee is not getting input from a file or command echo -e "\n$NUMY Files contain the Bongo Clause\n\n\n" | tee -a ~/$RESULTS echo -e "list files which do not contain \"Bongo\" y/n" read CONT1 #If y continue to list files not containing "Bongo" #Anything else and exit if [ "$CONT1" = "y" ] || [ "$CONT1" = "Y" ] then echo -e "The Bongo Clause absent from the following files:\n" >> ~/$RESULTS grep -L '\<Bongo\>' * | tee -a ~/$RESULTS NUMN=`grep -L '\<Bongo\>' * | wc -l` echo Results writen to ~/$RESULTS echo -e "\n $NUMN Files do not contain the Bongo Clause\n" | tee -a ~/$RESULTS else echo Results writen to ~/$RESULTS fi exit 0

Hi Steve, $DIRECTORY was a typo I didn't paste from the original script. I have now got this to work based on the information you gave me this morning. (Friday) The if clause wasn't really the BIG problem. I had a following clause which evaluated to true - even if the string was null This was an until loop which asked the user to read DIRECTORY until the directory existed. Basically hit the Enter key and you had satisfied the condition - heh presto your in your HOME directory. Had to power down all my machines at work to day at 3.45pm or I'd have posted earlier. Solution is in work (Dublin) While I'm currently at home in Belfast. However I can certainly post it on Monday if the example is of any use to anyone? Seems like good manners to make this offer even if it's not a particularly sophisticated script. Once again Steve many thanks. Much appreciated.

Hi Steve, thanks for reply. Alas Alack & Thrice woe but ain't a workin either. I feel the logic itself is wrong somewhere in my script so perhaps these type of tests are just not going to work.