how about having all changes hidden (viewable from wikipedia.com/latest or whatever), allowing the mod's to merge the changes with the visible version of the article (or have them merge after a week if theyre not deleted by a mod, or something)
It could mean an unlimited, free account of google adwords, in which case, thats not polluting searches, thats just google 'paying' for aol's adwords account
ie makes it really hard to make a single stylesheet for all browsers, but all i do (and many others) is develop in a real browser, then create an ie specific sheet using conditional comments. As css is cascading, just declare the sheet second and ie will use all the values in your ie hacks sheet, like this (its saturday, forgive me for any bad code:
just remember to put only overides in default_ie.css (dont duplicate everything). thats probably not the exact syntax - google for 'conditional comments'.
Hope this helps you stop using tables for non-tabular data.
javascripts.style represents css, the only difference is you remove the - from css, and uppercase the next letter (i dont think javascript likes -'s in object/variable/etc names):
What about incremental backups. unless your going to be deleting and recreating every file every day/week/whatever, then backing this thing up shouldn't be much trouble.
there's also ssh as a filesystem using fuse (in 2.6.12 or later). sshfs iirc. That can obviously be used over the interweb much more securely than samba or nfs.
from your list, you missed two very important english foods:
Mushy pea butties (plenty of vinegar) meat 'n' 'tato pie butties with brown sauce (bread, whole pie, brown sauce, bread) tea
do americans even know what brown sauce is? do you have a different word for it? it cant be that only us northerners (north england) eat brown sauce. (bloody southerners dont)
statically generate your HTML and seperate your code into purely generative (stuff that just reads the same old stuff from the db, and generates the same old html for all visitors) and your unique per-visitor logic and you could do miracles. Alot of people don't do this though.
that works, until you decide that you want to change something on all your pages, then it becomes messy. if your not doing that much, the overhead of a database isn't going to hurt that much anyway.
yes, i was scarred by windows:(, but ive been using linux for about 3 years and freebsd for about 2 months now (my gf loves debian !)
the reason i say reboot at the end is so you can put a new kernel on your internal repository. it also means you dont have to write a shell script to restart any services that were updated. It also forces applications to reload any dynamic libraries that may have been updated.
updating using this method should be perfectly safe, presuming you properly test the packages on your internal repository before you leave at the end of the day (or dont put them in the repository).
i wouldn't do this if i was updating from the official debian repository, even for stable, as things are more likely to go wrong.
ssh wouldn't be useful in the situation i was suggesting a fix for, as it would mean doing each of 10000 machines individually.
Just to clarify, when i say internal repository, i mean one source in sources.list, pointing to an internal server, containing only packages which were placed there manually and properly tested. If all the machines are the same, there should be absolutely no problems (as i say, properly test for any differences in the machines).
i'd just like to say, you can pipe dd through gzip (or if you have a stupid amount of time, bzip), so ghost's compression isn't an extra feature over dd
Slashdot Mirror
how about having all changes hidden (viewable from wikipedia.com/latest or whatever), allowing the mod's to merge the changes with the visible version of the article (or have them merge after a week if theyre not deleted by a mod, or something)
It could mean an unlimited, free account of google adwords, in which case, thats not polluting searches, thats just google 'paying' for aol's adwords account
if somebody rewrites this in java, for my phone - i'l donate money to the project / buy the product
yes:
http://www.cssplay.co.uk/layouts/fixed.html
it will look proper in ie, you just need to use the hacks file like i said.
ie makes it really hard to make a single stylesheet for all browsers, but all i do (and many others) is develop in a real browser, then create an ie specific sheet using conditional comments. As css is cascading, just declare the sheet second and ie will use all the values in your ie hacks sheet, like this (its saturday, forgive me for any bad code:
/> />
<style type='text/css' src='/css/default.css'
<!-- [if lte 6.5]
<style type='text/css' src='/css/default_ie.css'
-->
just remember to put only overides in default_ie.css (dont duplicate everything). thats probably not the exact syntax - google for 'conditional comments'.
Hope this helps you stop using tables for non-tabular data.
javascripts .style represents css, the only difference is you remove the - from css, and uppercase the next letter (i dont think javascript likes -'s in object/variable/etc names):
background-color
backgroundColor
the yahoo toolbar is installed by default with acrobat reader?
let me hide it in the system tray, and display a number over the icon.
until which time, i'l be forced to use kmail.
how come system tray isn't a standard feature for modern gui email clients?
All the development team at the company I work for recently migrated to ODF.
(btw, I am the dev team. Before, I used to use whatever Koffice's default format was)
What about incremental backups. unless your going to be deleting and recreating every file every day/week/whatever, then backing this thing up shouldn't be much trouble.
yes, but i dont see any go faster stripe [*]
[*] I didn't actually look at the picture, so there may be a very sharp looking racing stripe
$501,000 for a 1.6tb nas?
look carefully at this:
n g
http://download.freshmeat.net/screenshots/27914.p
thats fluxbox. KDE's default colour scheme is just that - a colour sheme (and some dont like the widget style, which is also configurable).
You can make KDE look like a CDE ripoff (not that you'd want to)
there's also ssh as a filesystem using fuse (in 2.6.12 or later). sshfs iirc. That can obviously be used over the interweb much more securely than samba or nfs.
Its not that fast though
i'd also be interested in what is the sql query for 'click this button with the mouse'.
thanks.
heh, i added tea at the last minute. dont bother picking me up on that i said 'two'
from your list, you missed two very important english foods:
Mushy pea butties (plenty of vinegar)
meat 'n' 'tato pie butties with brown sauce (bread, whole pie, brown sauce, bread)
tea
do americans even know what brown sauce is? do you have a different word for it? it cant be that only us northerners (north england) eat brown sauce. (bloody southerners dont)
brown sauce:
http://en.wikipedia.org/wiki/Brown_sauce
hmm, that post became quite OT, ah well.
that works, until you decide that you want to change something on all your pages, then it becomes messy. if your not doing that much, the overhead of a database isn't going to hurt that much anyway.
Afaik InnoDB isn't the default, MyISAM is
or if your classroom doubles as a faraday cage
i wonder if its possible to write a 'reload shared libs' app (i wouldn't beable to, but if its possible, somebody could).
It would be a very cool feature of the hurd, if they adopted it (as hurd can have more than one concurrent kernel)
yes, i was scarred by windows :(, but ive been using linux for about 3 years and freebsd for about 2 months now (my gf loves debian !)
the reason i say reboot at the end is so you can put a new kernel on your internal repository. it also means you dont have to write a shell script to restart any services that were updated. It also forces applications to reload any dynamic libraries that may have been updated.
updating using this method should be perfectly safe, presuming you properly test the packages on your internal repository before you leave at the end of the day (or dont put them in the repository).
i wouldn't do this if i was updating from the official debian repository, even for stable, as things are more likely to go wrong.
ssh wouldn't be useful in the situation i was suggesting a fix for, as it would mean doing each of 10000 machines individually.
Just to clarify, when i say internal repository, i mean one source in sources.list, pointing to an internal server, containing only packages which were placed there manually and properly tested. If all the machines are the same, there should be absolutely no problems (as i say, properly test for any differences in the machines).
ive changed my mind:
i think it would be more fair for a large multinational to run the internet. Yes, Microsoft would be perfect:
whois google.com
No match for "google.com"
hmmmm
i'd just like to say, you can pipe dd through gzip (or if you have a stupid amount of time, bzip), so ghost's compression isn't an extra feature over dd