I thought I might add to my analysis of the accidental reboot earlier,
with a few opinions about the nature of mistakes.
Jeronim's mistake, if there was one, was relying on "being careful",
usually pictured as a furrowing of the eyebrows and some sort of
commitment to double-checking. The problem is, you can only keep your
eyebrows furrowed for so long. When a task becomes routine, mistakes are
inevitable. For example, when you say a tongue-twister, you can often
get it right the first time. But say it 10 times quickly, and mistakes
are almost impossible to avoid. In general, it's impossible to maintain
concentration when performing a routine task. Add fatigue to the mix,
and mistakes become inevitable.
The objective, then, is not to avoid mistakes. It is to analyse the task
at hand to determine the mistakes that might be made. If you can see a
potential for a dangerous, damaging mistake, change your work practices
to avoid the possibility altogether.
Recognising the potential for mistakes requires experience --
unfortunately we often have to make a mistake before we realise the
dangerous situation we were in. But after we make a mistake, we're in a
position to think of ways to avoid it in the future, and that does NOT
mean resolving to furrow your eyebrows more often.
For example, I once attempted to run this on geoffrin:
% chmod -R wikidev .
but I made a typo and did this instead:
% chmod -R wikidev /
I was being careful at the time, it's unreasonable to expect yourself to
pause for a few seconds every time you want to press enter. I resolved
to never again perform a recursive write operation on "." -- I now make
a habit of using * instead. There are no dangerous keys near *.
Another example of a mistake mitigating device is the automatic syntax
check in sync-file and scap. Coloured bash prompts are a third type -- I
started using them myself so that I would know the difference between my
home computer and a remote computer. Colours and movement aid in fast
recognition. They require less concentration than reading, at a time
when concentration is at a premium.
So in summary, there are three types of mistake mitigation which I use
on a regular basis:
* Good habits -- adopt work practices where common mistakes are not harmful
* Automated double-checking
* Cognitive aids, such as colours and movement
Finally, remember to consider your mental state. Avoid logging in when
fatigued. Take regular rest breaks. *Never* log in when drunk.
Also at http://en.wikipedia.org/wiki/User:Tim_Starling/Mistakes
-- Tim Starling
Hey everyone,
Some time ago I've wrote a proposed CheckUser Policy
(http://meta.wikimedia.org/wiki/Proposed_CheckUser_Policy), and since
I can't find anyway to improve it, I'm posting it to Meta. I hope to
get some feedback from everyone on this. If someone could also fix the
grammar on that page, I'd be very grateful.
Of course, [[CheckUser]] will get changed in time, after the policy
will be established.
I think that's all, hope to see some comments from the community.
--
Pozdrawiam,
Dariusz "Datrio" Siedlecki
On de.wikipedia, repeatedly:
Fatal error: Call to a member function on a non-object in
/usr/local/apache/common-local/php-1.5/extensions/LuceneSearch.php on
line 400
Notice to Wikimedia server admins: I've rearranged the Apache
configuration files in the pmtpa cluster.
First, the master files have moved from /h/w/conf to /h/w/conf/httpd to
keep things more distinct from the other programs' files we have in
there. Additionally, instead of symlinking the NFS-shared directory into
/usr/local/apache/conf, each Apache now has a local copy of the config
files, which should be more robust.
If you need to change the apache config files, run the sync-apache
script to rsync the files out before restarting the daemons.
Also, I've switched over from RewriteRule+ampescape patch to an Alias
for the /wiki/Foo aliasing. This should make it easier for us to use
standard Apache packages in the future.
-- brion vibber (brion @ pobox.com)
Hello, i ve one big problem with my MediaWiki.
I ve installed it @ my localhost (an IIS from Microsoft, with PHP and MySQL)
and it works fine.
But even when i search for an title that doesnt exist i get the Fatal Error:
"Fatal error: Only variables can be passed by reference in
c:\Inetpub\wwwroot\includes\SearchEngine.php on line 36"
What must i do to get an page like "Nothing where found with this text."
e.g.
I havnŽt changed this dokument.
LG
Sascha Brockmeyer
Adler crashed this evening; I had the colo reboot it, but it crashed
again within a couple hours.
We're running on samuel plus the smaller servers for now, but some of
the smaller boxes also seem a bit flaky as well (suda's been responding
oddly, ariel is out).
-- brion vibber (brion @ pobox.com)
Hi,
Now that on Meta the
Approved requests for new languages
page includes the Ladino language
I'd like to request the help of one
of our developers in setting up the
subdomains:
http://lad.wikipedia.org/
and
http://lad.wiktionary.org/
I'd also like to request admin access
as things get underway in those wikis.
Could someone help out with these
tasks.
With much appreciation and many thanks,
Jay B.
[[meta:User:ILVI]]
ilooy.gaon(a)gmail.com
I often use subpages in my wiki. Therefor the subpages shall be edited by
clicking [edit]. This works for pages with enough headings but not for pages
without them.. I also know solutions, using a template like "edit
subpage". But the next problem the [edit]-buttons are hidden and the
[edit-subpage] buttons are still visible when a user is not logged in.
So i'm looking for a php-solution when using the {{:subpage}} syntax.
Any ideas?
Thanks Heinz
I ran across this code after looking for a way to force certain pages to
not cache:
global $wgTitle;
if ($wgTitle)
{
$ts = mktime();
$now = gmdate("YmdHis", $ts + 60);
$ns = $wgTitle->getNamespace();
$ti = wfStrencode($wgTitle->getDBkey());
$sql = "UPDATE cur SET cur_touched='$now' WHERE cur_namespace=$ns
AND
cur_title='$ti'";
wfQuery($sql, DB_WRITE);
}
Where do I put this code? In the extension I am needing it for?
(dynamicpages2)
Hello ye of the Wiki
I am trying to solve a problem and hope that someone might have already
considered or actually done this already. I was wondering if any one had
created any code that would collect My Watchlist data and published it as an
RSS feed. If no one has can someone please point me in the right direction
in the code to accomplish this myself?
Thank you,
___________________________________________________
John Anthony Hartman
website: http://www.websage.org
personal media outlet: Multi-Media Me- http://pmo.websage.org
"Any sufficiently advanced technology is indistinguishable from magic. "
--Arthur C. Clarke