Laying aside the footprint of a 1st world, or 3rd world human (the two vary a lot) the actual land surface area is: 170,000,000,000,000 metres squared not 500,000,000,000,000 metres squared.
And this is on the assumption that every square inch of land is habitable. Or even if it is habitable (Phoenix), suitable for sustainable farming.
Heck, one entire continent is completely uninhabitable with our current technology, it'd be easier to live in the ocean as opposed to antarctica.
That may well be, I was just questioning the overly broad generalisation of the person on the other site. It may well be they put a quick black list in while working on a white list. Heck, the white list may be there now, have you checked?:) Actually, it could even be server side. Who knows.
The fellow questioning the failure to filter anything but http (and hopefully he meant https too) doesn't know his src tags. Not only can data URLs be in there, but absolutely any protocol is legal. ftp for example is in fact used, as are others. And of course the URLs can be relative. Sure it is possible to try and handle every combination, but I don't see anything wrong with the google fix.
That you were... *eases off the/. defensiveness trigger* You'd think my 10 years of linux use would allow me *some* familiarity with "stupid user tricks";)
No. I created/tmp/foo/bar/baz as root I then made baz world writeable I then created blah.txt as a restricted user. I then ran rm -rf foo as the restricted user, simulating an rm -rf of / Furthermore, I just reran this with a full directory tree/tmp/foo/bar(1-10)/baz(1-10) with random baz directories restricted user editable with files in them an rm -rf/tmp/foo as the restricted user returned about a dozen error messages. But *ALL* files and directories the user had access to were *REMOVED*
Odd. Before I did that I tested for myself. Created/tmp/foo/bar/baz/blah.txt with only baz and blah.txt being user writeable. Retesting, baz and blah.txt disappeared. I don't know what's up with your system or your permissions...
Running as a non-root user will still remove whatever it can in the tree, which will include the $HOME and any stuff the user created in/tmp. Some/var spools too.
"knew who he was" was exaggeration. I meant that he had probably searched for, then clicked on, his site a few dozen times, and googe figured out what he was looking for. But whatever. this subject is dead n buried.
And to emphasise, the UA sniffing was just a guess. The cookie thing is more certain though. And even if only cookie based, the point I was making re: parent still holds. Too many people assume they see the same google results as others. Like one commentator on NPR who noted that google thought he was the most important steven in the world. When I ran the same query as he did, he wasn't even on the first page. It never occurred to him, perhaps, that Google knew who he was, and that that skewed the results.
I strongly suspect you're wrong.;) These results were repeatable by me on the same computer, same path outwards to google. Just one was using Mozilla 1.7.12 with no cookies stored (my test browser) and the other was using Epiphany where I have made many previous searches in Google before.
Nope. As mentioned to the other person, platform was just a guess. (Cookie/IP tracking definitely is happening though). Nonetheless, why wouldn't google use the UA string as a clue? Obvious way to test would be to try the name of some linux project that is a general word. Do feel free to give it a shot.:)
I was just complaining more in general about the parent saying that Google results were tech heavy when it was likely that Google had decided based on past searches that he had a strong tech bias.
With cookies wiped using Mozilla 1.7.12 under Gentoo Linux I get wine.com as first hit as well. I qualified the cookie bit with OS/Browser since it seemed to me that I got more linux specific results when using Mozilla under Linux.
So in this case it seems the key factor is probably their cookie info. You might want to just keep an eye out and see if what is returned changes subtly for you later. Like the parent, you may just assume google is biased, when it is actually trying to sniff out your biases.:)
Also, it seems to affect more the ordering of results. I still get more or less the same search. Still. 134,000,000 results with cookies intact. 157,000,000 results without cookies. So obviously things are a little different.
I've noticed a lot of people make certain comments about Google's web search that others can't reproduce. I'd like to ask you to try running a Google search with cookies erased and blocked and compare. Platform matters as well. For example, on my machine, a search for "wine" returns WineHQ first and www.wineandco.com second. It knows I'm more likely to be interested in WINE vs the drink, and in french results versus english. Platform affects this as well. And probably browser.
On a Windows 2000 machine with Internet Explorer, wine.com is the first hit.
There is the also the fact that certain searches, local.google.com in particular, can rapidly identify a person, hashing of IP or no. While there are technical solutions, we can't even begin to step down this path. I hope the feds get smacked down for this.
I unfortunately do not have firefox to examine, but in the suite, the attachment list uses a listbox. The thunderbird screenshot looks similar. In the suite, this listbox has has rows="3" set in attributes. I see no reason the same could not be done in Thunderbird.
It is completely possible to set a max height on the list box, what is moderately annoying is that similar to the element in HTML, since the list is intended to be scrollable, this doesn't trigger scrollbars until the rowcount is hit.
To fix this in Thunderbird would probably require editing the XUL file. Just like in web development, not *everything* can be done in CSS. In this case, requires ability to tweak a little "HTML" too.:)
Trivial for anyone with marginal knowledge of CSS, and ability to follow the blinking squares in the DOM inspector. Thus, assuming there are a few more people besides this guy who actually are bothered by this bit of the UI, I'm sure someone can give him a CSS file that has what he wants.
Re:hope they fixed some of the more glaring bugs
on
Thunderbird 1.5 Arrives
·
· Score: 5, Informative
Trivial to fix with a custom userChrome.css stylesheet. I don't have Thunderbird, but I had problem with expanded headers so I simply did the following:/* if full headers are enabled, trigger a scrollbar after ten lines */ #msgHeaderView { max-height: 10em; overflow: auto; }/* keep the enigmail box from creating a scrollbar - annoying */ #expandedEnigmailBox { max-width: 80em; }
I don't know if the DOM Inspector is available for Thunderbird, but every time I want to tweak the suite a little, I actually edit it. No harder than editing a web page.
... I mean, crashed the web server but not the pop mail server - still think it is more likely they are ducking and covering for now. Also. +OK lake.stark.k12.oh.us Cyrus POP3 v2.2.12-OS X 10.4.0 server ready
OS X 10.4 - that's kind of neat Fortunately their version of Cyrus is fairly up to date - some old ones had some fairly serious security holes - I think 2.2.12 is probably ok.
Starting nmap 3.83.DC13 ( http://www.insecure.org/nmap/ ) at 2006-01-07 00:06 EST Interesting ports on 66.144.97.98: (The 1665 ports scanned but not shown below are in state: filtered) PORT STATE SERVICE 80/tcp closed http 110/tcp open pop3
Interesting. Look like they just shut down HTTP. Guess they are waiting for this to blow over. I mean - it could just be on different machines, or have crashed the server, but POP was still quite responsive.
Right. That was my point. * it has been established this is due to a file format decision goes back to Windows 3.0 * it allows arbitrary code execution
For Microsoft to call it non-critical on older platforms just because no one has tried exploiting it yet is just irresponsible.
Slashdot Mirror
Laying aside the footprint of a 1st world, or 3rd world human (the two vary a lot) the actual land surface area is:
170,000,000,000,000 metres squared
not
500,000,000,000,000 metres squared.
And this is on the assumption that every square inch of land is habitable. Or even if it is habitable (Phoenix), suitable for sustainable farming.
Heck, one entire continent is completely uninhabitable with our current technology, it'd be easier to live in the ocean
as opposed to antarctica.
http://kb.mozillazine.org/Reducing_memory_usage_(F irefox)#Settings_that_reduce_memory_usage
That may well be, I was just questioning the overly broad generalisation of the person on the other site. :)
It may well be they put a quick black list in while working on a white list.
Heck, the white list may be there now, have you checked?
Actually, it could even be server side. Who knows.
yeah, src attribute.
should've previewed.
The fellow questioning the failure to filter anything but http (and hopefully he meant https too) doesn't know his src tags.
Not only can data URLs be in there, but absolutely any protocol is legal.
ftp for example is in fact used, as are others.
And of course the URLs can be relative.
Sure it is possible to try and handle every combination, but I don't see anything wrong with the google fix.
I wonder if Google could employ some sort of bayesian filter to make your life harder.
That you were... /. defensiveness trigger* ;)
*eases off the
You'd think my 10 years of linux use would allow me *some* familiarity with "stupid user tricks"
No. /tmp/foo/bar/baz as root /tmp/foo/bar(1-10)/baz(1-10) with random baz directories restricted user editable with files in them /tmp/foo as the restricted user returned about a dozen error messages.
I created
I then made baz world writeable
I then created blah.txt as a restricted user.
I then ran rm -rf foo as the restricted user, simulating an rm -rf of /
Furthermore, I just reran this with a full directory tree
an rm -rf
But *ALL* files and directories the user had access to were *REMOVED*
Odd. Before I did that I tested for myself. /tmp/foo/bar/baz/blah.txt
Created
with only baz and blah.txt being user writeable.
Retesting,
baz and blah.txt disappeared.
I don't know what's up with your system or your permissions...
Running as a non-root user will still remove whatever it can in the tree, which will include the $HOME and any stuff the user /tmp. /var spools too.
created in
Some
"knew who he was" was exaggeration.
I meant that he had probably searched for, then clicked on, his site a few dozen times, and googe figured out what he was looking for.
But whatever. this subject is dead n buried.
And to emphasise, the UA sniffing was just a guess. The cookie thing is more certain though. And even if only cookie based, the point I was making re: parent still holds. Too many people assume they see the same google results as others.
Like one commentator on NPR who noted that google thought he was the most important steven in the world. When I ran the same query as he did, he wasn't even on the first page.
It never occurred to him, perhaps, that Google knew who he was, and that that skewed the results.
I strongly suspect you're wrong. ;)
These results were repeatable by me on the same computer, same path outwards to google.
Just one was using Mozilla 1.7.12 with no cookies stored (my test browser) and the other was using Epiphany where I have made many previous searches in Google before.
Nope. As mentioned to the other person, platform was just a guess. (Cookie/IP tracking definitely is happening though). :)
Nonetheless, why wouldn't google use the UA string as a clue?
Obvious way to test would be to try the name of some linux project that is a general word.
Do feel free to give it a shot.
I was just complaining more in general about the parent saying that Google results were tech heavy when it was likely that Google had decided based on past searches that he had a strong tech bias.
With cookies wiped using Mozilla 1.7.12 under Gentoo Linux I get wine.com as first hit as well.
:)
I qualified the cookie bit with OS/Browser since it seemed to me that I got more linux specific results when using Mozilla under Linux.
So in this case it seems the key factor is probably their cookie info.
You might want to just keep an eye out and see if what is returned changes subtly for you later.
Like the parent, you may just assume google is biased, when it is actually trying to sniff out your biases.
Also, it seems to affect more the ordering of results. I still get more or less the same search.
Still.
134,000,000 results with cookies intact.
157,000,000 results without cookies.
So obviously things are a little different.
I've noticed a lot of people make certain comments about Google's web search that others can't reproduce.
I'd like to ask you to try running a Google search with cookies erased and blocked and compare.
Platform matters as well.
For example, on my machine, a search for "wine" returns WineHQ first and www.wineandco.com second.
It knows I'm more likely to be interested in WINE vs the drink, and in french results versus english.
Platform affects this as well. And probably browser.
On a Windows 2000 machine with Internet Explorer, wine.com is the first hit.
There is the also the fact that certain searches, local.google.com in particular, can rapidly identify a person, hashing of IP or no.
While there are technical solutions, we can't even begin to step down this path. I hope the feds get smacked down for this.
Try this link instead
... er. Don't have Thunderbird to examine. :)
But, don't have have Firefox handy either.
I unfortunately do not have firefox to examine, but in the suite, the attachment list uses a listbox. The thunderbird screenshot looks similar.
:)
In the suite, this listbox has has rows="3" set in attributes.
I see no reason the same could not be done in Thunderbird.
It is completely possible to set a max height on the list box, what is moderately annoying is that similar to the
element in HTML, since the list is intended to be scrollable, this doesn't trigger scrollbars until the rowcount is hit.
To fix this in Thunderbird would probably require editing the XUL file. Just like in web development, not *everything* can be done in CSS. In this case, requires ability to tweak a little "HTML" too.
Trivial for anyone with marginal knowledge of CSS, and ability to follow the blinking squares in the DOM inspector.
Thus, assuming there are a few more people besides this guy who actually are bothered by this bit of the UI,
I'm sure someone can give him a CSS file that has what he wants.
Trivial to fix with a custom userChrome.css stylesheet. /* if full headers are enabled, trigger a scrollbar after ten lines */ /* keep the enigmail box from creating a scrollbar - annoying */
I don't have Thunderbird, but I had problem with expanded headers so I simply did the following:
#msgHeaderView
{
max-height: 10em;
overflow: auto;
}
#expandedEnigmailBox
{
max-width: 80em;
}
I don't know if the DOM Inspector is available for Thunderbird, but every time I want to
tweak the suite a little, I actually edit it. No harder than editing a web page.
... I mean, crashed the web server but not the pop mail server - still think it is more likely they are ducking and covering for now.
Also.
+OK lake.stark.k12.oh.us Cyrus POP3 v2.2.12-OS X 10.4.0 server ready
OS X 10.4 - that's kind of neat
Fortunately their version of Cyrus is fairly up to date - some old ones had some fairly serious security holes - I think 2.2.12 is probably ok.
Starting nmap 3.83.DC13 ( http://www.insecure.org/nmap/ ) at 2006-01-07 00:06 EST
Interesting ports on 66.144.97.98:
(The 1665 ports scanned but not shown below are in state: filtered)
PORT STATE SERVICE
80/tcp closed http
110/tcp open pop3
Interesting. Look like they just shut down HTTP.
Guess they are waiting for this to blow over.
I mean - it could just be on different machines, or have crashed the server, but POP was still quite responsive.
Right. That was my point.
* it has been established this is due to a file format decision goes back to Windows 3.0
* it allows arbitrary code execution
For Microsoft to call it non-critical on older platforms just because no one has
tried exploiting it yet is just irresponsible.