Finding path on CCGI

[#1]

Quote from: nac

Quote from: Bob have you tried naming it "web_access.cgi"? Aha! That worked. Thanks! Also, my original problem (broken PERL WIKI) was fixed simply by adding .cgi to the filename.

The script filename I suggested was bashexample.cgi which would have worked. Extensions of .php and .pl also work for PHP and Perl scripts respectively.
Pity that crontab can't be manipulated though.

Quote from: nac

Would it be helpful if I was to summarise all this as an edit to http://usertools.plus.net/tutorials/id/18 ; if so, is that file just raw html (it looks like it) or is it rendered from some other source?

An updated tutorial would be helpful. That Usertools one is written in bbcode, but with some vISP specific keywords so that (logged in) Plusnet, Force9 and Free-online customers see text appropriate to themselves. The text is editable only by UserGroup members, though if you'd like to post or PM an update to me I'll add it.
Alternatively you could write an article for the Community Support Library - see "how-to-write-an-article" guide. I'll add the Usertools guide to the library if using that as a starter would be easier - anyone can edit any article (subject to moderation by library moderators).
David

[#2]

Hi David,
It seems that the most valuable thing for me to do would be to update http://usertools.plus.net/tutorials/id/18. I googled bbcode and found that I should be able to render it locally using code from http://www.bbcode-to-html.com/ (ie, I can write markup and then view the resultant html to make sure my syntax etc. is good).
If you agree, please can you tell me how I can grab the (bbcode/pre-rendered) source of the article.. or post it here (then I'll post the edited version back here for you to upload).
If you disagree, and think that an article for the community support library would be more useful, then please add the existing usertools guide to the library and I'll use it as a starting point.
In either case, this should give me at least 24hrs breathing space before I'm called upon to do anything..
thanks,
Neal.

[#3]

Hi Neal,
Thanks for your offer. I've PMed the raw text of the tutorial article to you, wouldn't want that to be attached to a post.
Obviously please let me know if you need any information or help.
Regards
David

[#4]

Hello chaps,
You may want to look at http://usertools.plus.net/tutorials/id/5 too.
It talks about 'Preventing Directory Listing' using IndexIgnore *. I've read that that isn't supported any more.
Best regards,
Barry.

[#5]

Thanks. As a temporary measure I've added a note at the top of that tutorial drawing attention to potential problems caused by some of the commands and cross-referring to the update FAQ for more information.
David

[#6]

Quote from: spraxyt

file paths follow, with 'home1', 'home2', … depending on which storage device your account is assigned to.

The /files/homeX/username paths are a bit of a legacy fix. Back in olden times, those were our absolute paths, but they are now (and were on the previous platform) a symlink of their former selves. Our absolute paths are now in the form /share/storage/0X/us/username and can be found using

#!/bin/bash
echo "Content-type: text/plain"
echo
pwd

whatever.cgi, chmod 700 and call in browser.
Or to illustrate:

<?php
echo realpath($_SERVER["SCRIPT_FILENAME"]);
?>

whatever.php, chmod 700 and call in browser.
The symlinked paths work for most cases, but not all, so it's worth knowing. The absolute path is marginally more efficient where useable. Just to confuse things, suexec thinks we're all under /var/www.
Gabe

[#7]

Quote from: Bob

Access to crontab will not be possible via the conventional methods, although I'm fairly confident you're already aware of this jelv? I'm not sure if it's possible to use something akin to this which I know Gabe has paid mention to before when I've been chatting with him over PM. I can't really see why we wouldn't allow crontab modifications via ticket, although we'd need some guidelines to make sure people submitted all of the relevant information. I'll put this suggestion to the Net-Ops Team later on and see what they say... I'll also look to get something in the FAQ about finding out the path name and how to request password changes etc. Once traffic has died down here, we'll look to port what we've put in the Community Library across to the portal support pages. Remember, *anyone* can create Community Library articles. Edit: Have spoken to Networks and am just awaiting the go-ahead that we can make crontab modifications following a request made via the ticketing system. Assuming the process gets signed off then I'll add it to the FAQ as well...

Hi Bob Can the Path to Crontab be supplied I have written a script to set crontab in a fool-proof way, basically it has a front-end that reads all scripts files on a users ccgi web space, and will allow only these to be used as the link in the crontab file, it uses dropdown selectors to set the trigger times, I have added locks to make sure it only allows times not intervals, there are other options included to make it  easy to use while stopping silly errors. I have tested it and it works fine apart from scripts being called sometimes more than once. I need conformation of the correct paths I should be using as I may be calling crontab wrong. can you please advise me, if the small problem of repeat calls can be sorted I will make this script available to plus net, it could save users and support staff a lot of time and streamline the setting of crontab. If you wish to see this working please let me know I will send you the access url so you can see it working for yourself.

[#8]

Quote from: Gabe

The /files/homeX/username paths are a bit of a legacy fix. Back in olden times, those were our absolute paths, but they are now (and were on the previous platform) a symlink of their former selves. Our absolute paths are now in the form /share/storage/0X/us/username and can be found using #!/bin/bash echo "Content-type: text/plain" echo pwd whatever.cgi, chmod 700 and call in browser. Or to illustrate: <?php echo realpath($_SERVER["SCRIPT_FILENAME"]); ?> whatever.php, chmod 700 and call in browser. The symlinked paths work for most cases, but not all, so it's worth knowing. The absolute path is marginally more efficient where useable. Just to confuse things, suexec thinks we're all under /var/www. Gabe

FWIW I found that paths on the web server were something similar, wish it had been explained properly when I was writing my "rewrite" rules as the extra folders broke many of the examples I found on the web.