> This is exactly what I've been waiting for them to do for...the last six months or something like that!
That would be pretty difficult seeing as the API hasn't been stable until very recently. Still, if you were that desperate, you could run the apache2filter module from CVS around the date of your chosen Apache 2 release with a release PHP.
> Ah crap, that means I can't put off upgrading any more.
Um, yes you can. The apache2filter SAPI module is still crufty and this isn't a release version. You've got a good few weeks yet while enough kinks are worked out to even warrant a.0 release.
By all means test it and throw it on your dev server, but don't be surprised when n random kiddie exploits a vulnerability or your httpd-error.log starts filling with tens of thousands of nice errors:)
Rather than filling the spider with a whole bunch of (potentially valid) addresses and loading your server with bogus clients you don't want, just make it difficult for them to extract the addresses.
I wrote a bit of PHP a few months ago that applied some spamproofing ala SlashDot (only a bit less agressive) that some might find useful.
ATM I have TV card, Sound Card, and NIC in my system; sure, this system's got on board NIC and sound, but so what? I have an SB Live that'll beat most on board stuff (well, maybe, I'm not a great fan of Live's anymore:), and needing another NIC isn't that unlikely; 3 is definately going to feel cramped.
> If it is someone you enjoy being with, i don't see why even a depressed person would have trouble talking to him
Never been clinically depressed then, I take it?
The problem with depression is it saps your self esteem; if you're around people you tend to feel you're poisoning their day, and if you're around people you *like*, it's even worse, since you'll care more for what they're feeling. At least if you avoid people they're not going to end up hating you for being around them.
> Now, if there is NOONE you feel comfortable with, then you have more serious problems than you even think yourself. SOL. sorry.
Now, if this is one of those areas where if you have no experience, you should just STUF. sorry.
> Come back up immediately after an unclean shutdown, but have a background task sucking up IO bandwidth for an hour or three.
FFS fsck's quite fast, actually. As for sucking up IO; background fsck can be ran at a higher nice value.
IO operations for niced tasks are reduced in favour of other tasks competing for IO; so, you *could* have fsck running for hours if your system's doing a lot of IO and fsck is running at a nice of +20, but you're unlikely to notice it on anything but an accurate IO benchmark.
Checking it on my FreeBSD box is easy too.. it's just rather unfortunate that it's an unsigned int, and hence overflows every 4.3GB; something I'm pretty sure happens more often than once a month.
Still, I can always set up a daemon to grab it, handle the overflows, plot it using rddtool, deduct icmp traffic, etc. Luckily ntl don't seem to care how much bandwidth users take provided they don't disrupt the service.
Except generic code that runs on any i386-i686 is going to be less optimizable and more verbose than code that runs only on i486-i686.
386 lacks certain instructions (e.g. cmpxchg); the compiler's not going to do a CPU check every time such an instruction will be useful, it'll just choose a solution that works everywhere, and you can be sure code that's good for a 16MHz 80386 that's just pulled the series out of 16bit hell is not going to be good for a 1600MHz Pentium 4 that doesn't even impliment the instruction set natively.
Besides which, anyone running i386 level hardware is hardly going to be interested in any of the new stuff in 5.0; they're probably still running 2.x, 3.x or maybe 4.x.
It will. It's just in it's own branch at the moment while everything's adapted to gcc3's new warnings stuff, IA-64 and the other new architectures gcc3 is needed for (IA-64, etc).
> What's the current status with mozilla? Is it as usable as Netscape 4.x (I hate 6.x)?
It's been more usable than 4.x for months. Recent releases are very stable. Startup time is about on par with Opera here (~3s when cached, next to ~2s for Opera).
4.x lacks usable CSS (and this is very important for modern sites.. the only reason most sites still work is because most sites still use techniques from 1995; I don't), and has laughable table layout code (it was made with basic HTML-for-tabular-data in mind, not triple nested layout tables); these alone make it pretty much useless to me.
If I didn't use Opera, I'd probably use Mozilla; at least I can trust it to Do The Right Thing (usually) when I'm developing sites; then I can go add my IE5/6/NS workarounds afterwards.
> Does it do any nasty or weird stuff on some sites?
IE6 does nasty and weird stuff on W3.org/Style (fixed positioning isn't supported, but it still processes the position: fixed; directive, meaning you can't do "position: absolute;position: fixed;" like you're supposed to. Argh.).
IE5 does nasty and weird stuff on every site that uses the CSS box model; it gets the sizes wrong on all boxes, meaning you need to exploit parser bugs to provide IE5 with tweaked sizes for it to work properly (and then provide Opera 5, which suffers the same parser bug, with real values).
NS4 does nasty and weird stuff when you specify an element should float: anywhere; it makes it completely unusable to use CSS layouts on it without spending months debugging an absolutely positioned workaround-nightmare.
Not seen Mozilla (or Opera) do anything this broken:)
> Yeah, but what does that say about Flight Sim 2002?
That flight sims are damn complex things to render when you have all the detail sliders set to Max. At least Q3 doesn't have a mindnumbing view depth, complex physics, dynamic weather systems across hundreds of square miles which can be viewed from any angle, and LOD stuff that needs to scale from 500km to 5'.
Re:Why there will never be a time machine
on
Time Travel
·
· Score: 2
The current frame of reference represents a place high up on the curve of probability; having people appear out of thin air is going to be on a much lower end of the curve, since every place and time isn't going to have time travelers, they'll be hugely spread out across space, time and dimensions.
It's quite possibe that someone from the future will appear out of nowhere, announce himself, and make the universe fork() us off into a different direction; but it's much more likely it WON'T happen; we tend to remain on a "default" higher probability fork, which also happens to lack flying sheep, nearby aliens and key cracking contests that always find the key after exhausting < 1% of the keyspace.
I don't necessarily subscribe to these ideas, btw:)
thttpd is good for static content, but..
on
Apache 2.0 Goes Gold!
·
· Score: 3, Informative
Plenty of people want dynamically driven sites; something httpd isn't aimed for.
Personally I use quite a few of the features of Apache; mod_php (thttpd has this, but each script blocks the server, Not Good[tm] unless your scripts are trivial), mod_proxy (Outside world -> FreeBSD/Apache -> WinXP/Apache, appears as part of my web tree, nice and clean), mod_rewrite (how anyone can put up with the crappy URL's dynamic sites like I don't know, a 1:1 mapping of URL's to the filesystem is bad enough) and mod_gzip (does thttpd support any content negotiation?) to name the main ones, and this is just a miniscule personal server:)
Re:Why there will never be a time machine
on
Time Travel
·
· Score: 2
> Alot of people have posted on this sort of theme. > I just don't get it - to continue the above > thread, what is to stop someone from Beta time > travelling back to Alpha? Or is the implication > that Beta (and Gamma, Delta, etc...) are all > different from Alpha?
Yes.
--A----------B
B is the point at which you originate on the timeline, A is your target. Except, by traveling to it, you're altering history (even just by displacing some air or bouncing into a hydrogen atom.. and, of course, actually being there in the first place:). This means you're no longer in *your* timeline, but you've split off a seperate one from that point:
_________B' / --A----------B
If you were to travel forward back to B, you'd actually end up at B'; where in the past you appeared and started altering things. Of course, if you didn't do anything, B' may be utterly identical to B as far as you can tell.
> Procmail's "language" is actually pretty logical, once you get past weird markers like "0:" and "*".
Yes, it's logical, but that doesn't make it good:)
Have a look at the code sometime; especially with whitespace highlighting on to see how it's not even been indented properly, never mind properly commented or refactored. It's riddled with goto's, magic numbers, and is optimized for parsing by removing as much whitespace as possible.
Just looking at it was enough to make me install maildrop (which has much cleaner code).. I've yet to actually start using it though:)
> Is there some open-source mail filter that doesn't reparse its rules on each run?
Exim's filter stuff probably fits into this category, along with the filter languages for the other MTA's.
Maildrop may be a reasonable replacement to the procmail part, since procmail's rather messy and has a filter language that would make Larry Wall blush.
> Drywalling? I thought it was sprayed-on fire-resistant coating on the steel trusses that was blown off by the explosion?
Yes; however, the core itself (and hence the stairwells) was protected by lots of ultra-lightweight fire resistant tiles; these were very weak and got blown away by the explosion; had they held in place, even in the case of the direct impact onto the core, at least one stairwell may have remained usable.
> > Those windows-inside-of-windows in Opera make me SICK.
Opera 6 has SDI support, although frankly MDI is one of the reasons I *prefer* Opera to other browsers. Routinely opening 20-30 browser windows just isn't funny otherwise. Sure, Mozilla has tabs, but they're buggy and bit of an afterthought to the interface.
> I like Opera because of the mouse gesture navigation, speed, superior page zooming capabilities, pre-filled hotlist... I could go on and on.
Author and User display modes make testing websites for graceful degredation easy (and double as a way to work around obnoxious sites with, e.g. blue on black colour schemes).
Printing support is second to none (although Mozilla is close).
Excellent CSS support (At least as good as, if not better than Mozilla in most places)
Full screen mode uses projection media which is nice for presentations.
Graceful crash handling (ala vim)
Concidering it cost me £13 to get rid of the adverts, I'm pleased with it.
Mozilla is a good browser, but it's custom GUI gadgets that act like nothing else on the system, large uncached load times, and feeble tab support mean it's not going to become my browser of choice for a while yet.
> The idea is simple and if implemented correctly > will be a huge deterrent to sending spam. > Specially if it takes you 2 seconds or so to > prepare the email!
That's great until someone legitimate wants to send lots of email; mailing lists etc. Sure, you can add whitelists, but all that does is add $MAX_INT maintainence costs.
> Face it SMTP is outdated and wholly inappropriate!
Under what criteria? How are you going to reliably and cheaply prevent some people from mass mailing and header forging while allowing others to mass mail?
Slashdot Mirror
> This is exactly what I've been waiting for them to do for...the last six months or something like that!
.0 release.
:)
That would be pretty difficult seeing as the API hasn't been stable until very recently. Still, if you were that desperate, you could run the apache2filter module from CVS around the date of your chosen Apache 2 release with a release PHP.
> Ah crap, that means I can't put off upgrading any more.
Um, yes you can. The apache2filter SAPI module is still crufty and this isn't a release version. You've got a good few weeks yet while enough kinks are worked out to even warrant a
By all means test it and throw it on your dev server, but don't be surprised when n random kiddie exploits a vulnerability or your httpd-error.log starts filling with tens of thousands of nice errors
Rather than filling the spider with a whole bunch of (potentially valid) addresses and loading your server with bogus clients you don't want, just make it difficult for them to extract the addresses.
I wrote a bit of PHP a few months ago that applied some spamproofing ala SlashDot (only a bit less agressive) that some might find useful.
Highlighted Source
Raw Source
It performs the following munging, depending on what you specify:
freaky@aagh.net
freaky (at) aagh (dot) net
freaky@aagh.N0SPAM.net.SPAMN0
freaky@aag&# 104;.net
random one of the above
random with entity encoding
all of the above
ATM I have TV card, Sound Card, and NIC in my system; sure, this system's got on board NIC and sound, but so what? I have an SB Live that'll beat most on board stuff (well, maybe, I'm not a great fan of Live's anymore :), and needing another NIC isn't that unlikely; 3 is definately going to feel cramped.
> It's also not really fair to compare it to Linux/Apache/MySQL, as SQL Server 2000 beats MySQL on MANY fronts, including speed and options.
That's not hard. MySQL's not exactly the most complete SQL implimentation around.
What about comparing it to PgSQL?
> If it is someone you enjoy being with, i don't see why even a depressed person would have trouble talking to him
Never been clinically depressed then, I take it?
The problem with depression is it saps your self esteem; if you're around people you tend to feel you're poisoning their day, and if you're around people you *like*, it's even worse, since you'll care more for what they're feeling. At least if you avoid people they're not going to end up hating you for being around them.
> Now, if there is NOONE you feel comfortable with, then you have more serious problems than you even think yourself. SOL. sorry.
Now, if this is one of those areas where if you have no experience, you should just STUF. sorry.
> - lowered price points mean poorer readers (like me, for years) can afford to assemble a decent library without paying $30/hardback
:(
Is it just me who only see's about and average of £1 off a used book compared with a new one? Some of them are even more expensive than new
> Come back up immediately after an unclean shutdown, but have a background task sucking up IO bandwidth for an hour or three.
FFS fsck's quite fast, actually. As for sucking up IO; background fsck can be ran at a higher nice value.
IO operations for niced tasks are reduced in favour of other tasks competing for IO; so, you *could* have fsck running for hours if your system's doing a lot of IO and fsck is running at a nice of +20, but you're unlikely to notice it on anything but an accurate IO benchmark.
See Running "fsck" in the Background, section 7.
Checking it on my FreeBSD box is easy too.. it's just rather unfortunate that it's an unsigned int, and hence overflows every 4.3GB; something I'm pretty sure happens more often than once a month.
Still, I can always set up a daemon to grab it, handle the overflows, plot it using rddtool, deduct icmp traffic, etc. Luckily ntl don't seem to care how much bandwidth users take provided they don't disrupt the service.
> I sure wish they'd figure out how to issue binary diffs instead of complete rpm packages.
:)
rsync is your friend for binaries that haven't changed much. You just need to find an RPM server that runs it
Except generic code that runs on any i386-i686 is going to be less optimizable and more verbose than code that runs only on i486-i686.
386 lacks certain instructions (e.g. cmpxchg); the compiler's not going to do a CPU check every time such an instruction will be useful, it'll just choose a solution that works everywhere, and you can be sure code that's good for a 16MHz 80386 that's just pulled the series out of 16bit hell is not going to be good for a 1600MHz Pentium 4 that doesn't even impliment the instruction set natively.
Besides which, anyone running i386 level hardware is hardly going to be interested in any of the new stuff in 5.0; they're probably still running 2.x, 3.x or maybe 4.x.
> so FreeBSD 5.0 won't be using gcc3?
It will. It's just in it's own branch at the moment while everything's adapted to gcc3's new warnings stuff, IA-64 and the other new architectures gcc3 is needed for (IA-64, etc).
> What's the current status with mozilla? Is it as usable as Netscape 4.x (I hate 6.x)?
:)
It's been more usable than 4.x for months. Recent releases are very stable. Startup time is about on par with Opera here (~3s when cached, next to ~2s for Opera).
4.x lacks usable CSS (and this is very important for modern sites.. the only reason most sites still work is because most sites still use techniques from 1995; I don't), and has laughable table layout code (it was made with basic HTML-for-tabular-data in mind, not triple nested layout tables); these alone make it pretty much useless to me.
If I didn't use Opera, I'd probably use Mozilla; at least I can trust it to Do The Right Thing (usually) when I'm developing sites; then I can go add my IE5/6/NS workarounds afterwards.
> Does it do any nasty or weird stuff on some sites?
IE6 does nasty and weird stuff on W3.org/Style (fixed positioning isn't supported, but it still processes the position: fixed; directive, meaning you can't do "position: absolute;position: fixed;" like you're supposed to. Argh.).
IE5 does nasty and weird stuff on every site that uses the CSS box model; it gets the sizes wrong on all boxes, meaning you need to exploit parser bugs to provide IE5 with tweaked sizes for it to work properly (and then provide Opera 5, which suffers the same parser bug, with real values).
NS4 does nasty and weird stuff when you specify an element should float: anywhere; it makes it completely unusable to use CSS layouts on it without spending months debugging an absolutely positioned workaround-nightmare.
Not seen Mozilla (or Opera) do anything this broken
> Yeah, but what does that say about Flight Sim 2002?
That flight sims are damn complex things to render when you have all the detail sliders set to Max. At least Q3 doesn't have a mindnumbing view depth, complex physics, dynamic weather systems across hundreds of square miles which can be viewed from any angle, and LOD stuff that needs to scale from 500km to 5'.
The current frame of reference represents a place high up on the curve of probability; having people appear out of thin air is going to be on a much lower end of the curve, since every place and time isn't going to have time travelers, they'll be hugely spread out across space, time and dimensions.
:)
It's quite possibe that someone from the future will appear out of nowhere, announce himself, and make the universe fork() us off into a different direction; but it's much more likely it WON'T happen; we tend to remain on a "default" higher probability fork, which also happens to lack flying sheep, nearby aliens and key cracking contests that always find the key after exhausting < 1% of the keyspace.
I don't necessarily subscribe to these ideas, btw
Plenty of people want dynamically driven sites; something httpd isn't aimed for.
:)
Personally I use quite a few of the features of Apache; mod_php (thttpd has this, but each script blocks the server, Not Good[tm] unless your scripts are trivial), mod_proxy (Outside world -> FreeBSD/Apache -> WinXP/Apache, appears as part of my web tree, nice and clean), mod_rewrite (how anyone can put up with the crappy URL's dynamic sites like I don't know, a 1:1 mapping of URL's to the filesystem is bad enough) and mod_gzip (does thttpd support any content negotiation?) to name the main ones, and this is just a miniscule personal server
> I just don't get it - to continue the above
> thread, what is to stop someone from Beta time
> travelling back to Alpha? Or is the implication
> that Beta (and Gamma, Delta, etc...) are all
> different from Alpha?
Yes.B is the point at which you originate on the timeline, A is your target. Except, by traveling to it, you're altering history (even just by displacing some air or bouncing into a hydrogen atom.. and, of course, actually being there in the first place
> Procmail's "language" is actually pretty logical, once you get past weird markers like "0:" and "*".
:)
:)
Yes, it's logical, but that doesn't make it good
Have a look at the code sometime; especially with whitespace highlighting on to see how it's not even been indented properly, never mind properly commented or refactored. It's riddled with goto's, magic numbers, and is optimized for parsing by removing as much whitespace as possible.
Just looking at it was enough to make me install maildrop (which has much cleaner code).. I've yet to actually start using it though
> Is there some open-source mail filter that doesn't reparse its rules on each run?
Exim's filter stuff probably fits into this category, along with the filter languages for the other MTA's.
Maildrop may be a reasonable replacement to the procmail part, since procmail's rather messy and has a filter language that would make Larry Wall blush.
Actually it's the same without -O. I suck, whatever. Maybe you should have posted the diffs so we could actually see what you were trying to say :P
> % gcc -S add.c
.file "add1.c" .file "add2.c"
> % gcc -S add2.c
Let's try gcc -O -S:
-% diff add1.s add2.s
1c1
<
---
>
*cough*
> since I never saw FreeBSD backtraces I can't say if it is of a similar detail level.
:)
It's effectively a big-ass core file you can run gdb on. Probably a tad more detailed than anything that will fit in syslog
> Drywalling? I thought it was sprayed-on fire-resistant coating on the steel trusses that was blown off by the explosion?
:)
Yes; however, the core itself (and hence the stairwells) was protected by lots of ultra-lightweight fire resistant tiles; these were very weak and got blown away by the explosion; had they held in place, even in the case of the direct impact onto the core, at least one stairwell may have remained usable.
According to Horizon, anyway
Opera 6 has SDI support, although frankly MDI is one of the reasons I *prefer* Opera to other browsers. Routinely opening 20-30 browser windows just isn't funny otherwise. Sure, Mozilla has tabs, but they're buggy and bit of an afterthought to the interface.
> I like Opera because of the mouse gesture navigation, speed, superior page zooming capabilities, pre-filled hotlist... I could go on and on.
Concidering it cost me £13 to get rid of the adverts, I'm pleased with it.
Mozilla is a good browser, but it's custom GUI gadgets that act like nothing else on the system, large uncached load times, and feeble tab support mean it's not going to become my browser of choice for a while yet.
> Good, and then you'll have to wait 4 hours for your HDD to write them ;-).
This transmits 320GB/s; food for thought is that most system MEMORY architectures would struggle to reach 1/400th of that. Yummy.
> The idea is simple and if implemented correctly
> will be a huge deterrent to sending spam.
> Specially if it takes you 2 seconds or so to
> prepare the email!
That's great until someone legitimate wants to send lots of email; mailing lists etc. Sure, you can add whitelists, but all that does is add $MAX_INT maintainence costs.
> Face it SMTP is outdated and wholly inappropriate!
Under what criteria? How are you going to reliably and cheaply prevent some people from mass mailing and header forging while allowing others to mass mail?