because replies to replies are only visible when score >=4 for unlogged in users
You can change thresholds without logging in. Since the claims were in context of having changed the display from default Flat/First into Nested/Highest, this will in fact happen automatically. (Try it). And that setting would be encoded in the URL to whoever she's emailing links to.
But whether said Score:5 is under said Score:1 or another Score:1, it's still posted after said Score:1, not before,
No. It'd look like this:
Raptors are cool because... (+5, Informative) 07:00
Try www.raptor.com (+1, Redundant) 09:00
For info, see http://raptor.com (+5,Informative) 08:00
The 08:00 reply was posted earlier, scored higher, and yet displayed (in Nested/HighestFirst) after the 09:00. But the 09:00 was clearly marked redundant in response to the 08:00 one, by a moderator in NewestFirst view.
BTW, in your last sentence: "s/the mod/some mod/" (or perhaps even "s/the mod/most mod/").
But, all moderators read -1/NewestFirst... a sig told me so!
In what life would you make yourself obsolete in your chosen profession for a weeks pay?!
It only takes one "traitor" to spoil it for all the rest. (Ok, you'd really need singers to represent 5-20 categories). It's unlikely that ALL singers are forward-thinking enough to understand that completing this one particular job might destroy their profession.
And even if vocal performers some how come to a shared, rational decision not to submit voices to machines, how long could they hold out? On one side, the TV industry will keep raising the payment offered for the week's work. On the other side, someday a singer is going to be facing bankrupcy and eviction, and turn to the only clear payout.
I write music and produce for TV series. I have never had to use a musician.
The submitter is a weatherman, BTW. A career that's ripe to be surrendered to machinery. All newreading "talking heads" might be replaced with CGI within a decade, but on-air "meteorologists" will be the first to go, since the content they read is the least varying.
As far as music, though, the first big use of this specific tech will be advertising jingles. Seriously, hardly any other kind of TV production desires to have any verbal lyrics. (Unless they're playing licensed pop music)
Mind you- this is a 5 minute walk from DOWNTOWN FINANCIAL CENTER in Boston. Not the boondocks.
That actually makes it MORE expensive. In the heart of an old, dense, poorly-planned city, the effort required to install any new cable is orders of magnitude bigger than just ripping through an empty field in a backhoe. And being in the financial district meant that all the other people inconvenienced by the work are high-wage broker-types whose time is expensive to waste. (Meaning the city will soak you on all the permits and police overtime)
Maybe, Boston is actually the worst US city for additional expense to infrastructure projects- big or small. The extreme example is how they just built $400,000,000 worth of tunnel for $10,800,000,000 (a 27x multiplier) because of all the uninterruptable activity above the dig site.
If you want cheap fiber to your building, try someplace more amenable, like Chicago.
Right now, I live out in the burbs near boston. 30 minutes away. I have ONE choice in internet access save dialup.
I've got relatives living in such areas (128 region), and they've got two cable companies selling ISP service. The effective speeds to in-state resources are an impressive 300Kb/s.
you've only got one week left to get a dashboard cradle and hands-free kit fitted - don't get caught out.
That's rather silly. It's a feel-good partial-solution... something which doesn't threaten cellphone business interests in any way (and only helps them sell more kits)
Studiesshow that the greatest danger of telephoning and driving is not from having one hand distracted with a machine (because many folks safely handle CDs and eat food while driving), but the mental load of being engaged in a conversation with somebody not in the vehicle.
But, I suppose that until we've got a few more years of fatality records as proof, the public won't believe this enough to influence law. (And even if they do believe it, prehaps talking and driving will be so important to them that safety will be voluntarily surrendered)
All consumer electronics devices should be manufactured - at base - with the lowest-common-denominator user in mind.
That's an ambiguous statement. If you mean "all designers should be aware of the possible existence of LCD users", then it is so trivially true that it hardly bears saying. But if you mean "all products should be designed so that an LCD can use them", then you are completely insane.
Frankly, this design strategy alone would revolutionize consumer product manufacture in many sectors (auto, electronics, etc), and solve many of the "user-unfriendly" problems that plague consumers today.
Yay! Now every single automobile can be a 2-person vehicle the size of a minivan, with a robotic lift to pull wheelchairs into the driver's seat. I can't wait until we're all driving those expensive, slow, bulky, and fuel-guzzling monsters just on the off-chance a few customers will find it convenient or important.
And just wait for the hunting rifle for the blind...
leading me back to the parent - that no - products are not being crippled by the including the needs of the 'crippled' - in fact it probably makes life better for the rest of us
Fine, say they're not "crippled" then. Instead, we could say they are "six times as expensive to account for features fewer than 1% of customers will ever need".
The reason these systems don't work for blind people is because of friends like you.
His blind friends shouldn't have to beg for someone to come over and help them when signing up for a new website.
By providing assistance, he wouldn't fix the problem- just perpetuate it. Sometimes providing a workaround just helps the underlying problem survive and grow ever-larger.
so it's unlikely that there'd be a Score:5, Redundant
Of course you'll never see +5 Redundant. He never said anything like that. Read the post again. The paragraph you just quoted clearly says that readers of the page will see one comment at +1 Redundant, go down a little, and see another (from a different author, but providing the same info) as +5 Insightful.
That can happen because sorting by something other than time posted can put an early post below a later one, swapping the apparent redundancy between them.
He's argueing against the entire concept of non-chronological sorting, because it causes readers to view comments in a different order than the moderators and actual commentors did.
I might be incredibly naive, but it seems something more like page 1.5.
This may be an effect of the rapid rate of new comments. By the time you finish reading 100 responses on page 1, dozens more messages can have been added to page 1. So many things that you already read on 1 have been pushed to 2 by the time you get there.
A better CGI would supply a "Read more" link at the bottom of the page, which tells the server the ID# of the final comment you read, so it can start printing more from exactly after it.
(The core problem is use of relative indexing rather than specific DB keying, and is a well-known stumbling-block for programmers)
REPORTING AND AUDIT. If Customer wishes to increase the number of Installed System,
That document doesn't appear to be internally consistent, as scrolling down to Appendix 1, Section 1 gets to this:
the license terms for the components permit Customer to copy, modify, and redistribute the component, in both source code and binary code forms.
It goes on to mention a few specific files not redistributable (basically providing a recipe for "Pink Tie" groups to do their work)
The only way to reconile the appendix statement with the apparent contradiction in REPORTING AND AUDIT is to take it that the "Installed System" as mentioned there applies only to systems under the service agreement. This viewpoint is not at direct odds with other parts of the document, particularly the initial definitions.
Again I ask: how does adding an alternative inhibit functionality that is already there?
You are suggesting that gestures are somehow confined to only replacing history navigation functions. That is patently untrue, as you well know. Why do you ask if you already answered it?
I call bullshit. Name any investigation that has determined that. I don't believe you can.
What are you talking about? It's bloody obvious: the statement "we feed and house them better than if they weren't prisoners" is trivially true.
In the US, prisoners are given 100% free food, housing, and medical care. Non-prisoner US citizens don't get that. Non-citizens certainly don't get it.
The government wouldn't be feeding them at all if they weren't prisoners.
No, "agents" would be simply doing their jobs and following orders. Those civilians are actually making the decisions! The Israeli government obeys the votes of its Jewish residents.
Democracies find themselves the targets of terrorism for the same reason that dictators tend to be assasinated: someone who violently disagrees with a government's actions will target the people in charge. In modern "first world" nations, that means every random voter.
learning PER-SITE navigation is simply ridiculous.
It'll be especially cool when someone who's already using mouse-gestures at the OS level visits a website which implements its own gesture interpretation! Every movement can have 2 or 3 separate effects!
Ideally, someone using Links (as opposed to Lynx) gets the same information as anyone else. That's the beauty of having a text only, no frames, no scripts browser around.
You seem to be implying that links doesn't support frames, and that lynx does. But the reverse is true. Links doesn't count as a no-frames browser.
My initial question stands: how are mouse guestures any less accessable than other browser features that require you to locate and click a button or click and drag on a scrollbar?
Because they're different and weird. Oddity by itself creates inaccessibility.
Traditional, well-known techniques are something that the disabled person can plan ahead for. Her client-side software can recognize common navigational features of a web page, and translate them into a form the user can understand. This is impossible if some complicated Jscript is going to generate all links in realtime by running weird code.
(Maybe, with time, a client which understands the specific trickery can be created.)
Mouse guestures just add another technique of navigation that is still bound internally to the same page-hyperlink model of HTML.
The danger of any fancy new site-specific navigation technique is that if it becomes popular enough, the site designer might forget to include the old-faithful A HREF clickable html links. Someone who uses gestures primarily will add a new features, attach it to the gesture list, and call it a day, leaving accessibility-challenged persons (and also robots, webcrawlers, etc) unable to reach the content.
This has already happened in those sites which use a Shockwave Flash bar for primary navigation.
The best way to keep your PK secure is to burn it to a CD, put the CD in a locked drawer, and only EVER put it in your machine when you want to sign anything.
That approach is still unsafe. If your machine has been 0wned, then all the PK software on it is untrustworthy, and could copy down the key the first time you insert that CD to sign something.
Safer is to keep the key and PK software on an independent machine, which is non-networked and completely barred from running any new executables. Any files needing signing must be hand-carried to this box.
(Of course, such an expensive and slow process may be too cumbersome for someone like Debian to use, as they release scores of new packages each day)
I will say that the private key should be kept on the personal machine of the person doing the signing, so they can dl the package, and sign it locally then upload it.
Well of course. When I said "developer machine", that was a synonym for "personal machine" like you're talking about. The developer's machine is naturally where the key will reside, so that is what hackers should aim at.
Now before you go all 'keylogger' on me
Although a keylogger can work, that's not the smartest approach for a hacker. It's too convoluted. The wise hacker will actually replace the gpg executable with a trojaned version, so that whenever certain filenames are signed they are first replaced with an evil variant.
Tainting gpg in that way will finish the hacker's job... from then on, the developer will do all the work of uploading and distributing the infected package under his own trust reputation.
It is you who is mistaken about a great many things.
I concur that dh003i is whacko. However on the following, he isn't completely wrong.
But to ascribe MS's achievement of monopoly status solely to the government "granting it these priviledges" is ridiculous.
A brief study of history and economics will reveal a simple truth - a capitalist free market, without an inpartial entity with the authority, power and will to maintain a level playing field, will trend towards consolidation and bigness.
It's true that free markets always tend towards consolidation. But, that consolidation will only succeed if supported by a government. Once a single entity starts to claim an excessive share of wealth (say 30+% of the money supply), other participants in the marketplace will always take steps to reduce their influence. These steps may start small, but will quickly escalate to reclaiming the money by any possible means, including violent ones.
Those people who are harmed by a true monopoly will initiate force against the monopolists, unless their is some powerful coercer (the government) blocking this path. And if the government sides with the monopolist, then it too faces the risk of destruction by an all-out revolution.
The MS monopoly could've only suceeded if the government deployed troops to fend off angry customers in the guise of "protecting intellectual property rights". Similarly, Standard Oil could only have persisted as a complete monopoly if the government had acted to protect their physical property rights. In both cases, the government was smart enough (or merely democratic enough) not to favor a monopolist over the bulk of their population.
Currency is established voluntarily by individuals acting of their own free will.
And you know what else was created voluntarily by individuals acting of their own free will? Government
That's the central fallacy of capital-L Libertarianism: the idea that given freedom, people will naturally assemble into a nearly-optimal system, and that this system will differ substantially from what we have now.
Those two statements are contradicted by a simple glance at the current world. People had freedom, and they created the system we have now. There is no reason to expect the same won't happen again, if by some miracle all laws were magically replaced with the injunction "Never initiate force".
With this "Public Key Cryptography" you could conceivably sign software in such a way that it could not be altered without breaking the signature,
No... the way to alter software is easy to conceive.
You simply have to hack into the computer holding the private keys used for the signing (very likely the same computer holding the source code as well, and the system which normally uploads new packages to the distribution point). Once there, you can make changes and sign them just as if they were official.
Since attackers of this type have already demonstrated an ablity to hack into computers, PK signing doesn't add any true security. It adds some defensive obscurity, since it's more difficult for the attackers to locate a developer's machine than a distribution one. But dev systems will be more vulnerable to hacking, since they're not likely to apply patches as quickly as a public server. (And I won't recite the old line about "security through obscurity")
The only true benefit from PK signing is that end-users are protected from poor security at mirror sites. Suppose your ISP offers a Debian package mirror as a high-speed convenience, but doesn't secure it well. If it's compromised, trojan packages could be sent to you on the next "apt-get". Comparing those packages against signatures from an official debian.org site will protect you. But that assumes the official servers are sufficiently well-run to avoid being hacked. And as we've seen today, that's not the case.
Slashdot Mirror
You can change thresholds without logging in. Since the claims were in context of having changed the display from default Flat/First into Nested/Highest, this will in fact happen automatically. (Try it). And that setting would be encoded in the URL to whoever she's emailing links to.
But whether said Score:5 is under said Score:1 or another Score:1, it's still posted after said Score:1, not before,
No. It'd look like this:
The 08:00 reply was posted earlier, scored higher, and yet displayed (in Nested/HighestFirst) after the 09:00. But the 09:00 was clearly marked redundant in response to the 08:00 one, by a moderator in NewestFirst view.
BTW, in your last sentence: "s/the mod/some mod/" (or perhaps even "s/the mod/most mod/").
But, all moderators read -1/NewestFirst... a sig told me so!
In what life would you make yourself obsolete in your chosen profession for a weeks pay?!
It only takes one "traitor" to spoil it for all the rest. (Ok, you'd really need singers to represent 5-20 categories). It's unlikely that ALL singers are forward-thinking enough to understand that completing this one particular job might destroy their profession.
And even if vocal performers some how come to a shared, rational decision not to submit voices to machines, how long could they hold out? On one side, the TV industry will keep raising the payment offered for the week's work. On the other side, someday a singer is going to be facing bankrupcy and eviction, and turn to the only clear payout.
I write music and produce for TV series. I have never had to use a musician.
The submitter is a weatherman, BTW. A career that's ripe to be surrendered to machinery. All newreading "talking heads" might be replaced with CGI within a decade, but on-air "meteorologists" will be the first to go, since the content they read is the least varying.
As far as music, though, the first big use of this specific tech will be advertising jingles. Seriously, hardly any other kind of TV production desires to have any verbal lyrics. (Unless they're playing licensed pop music)
"And in the middle of the stage, a beige computer tower"
(That picture is a recreation from Macross Plus. Although the idea of computerized singers is not uncommon in scifi)
Mind you- this is a 5 minute walk from DOWNTOWN FINANCIAL CENTER in Boston. Not the boondocks.
That actually makes it MORE expensive. In the heart of an old, dense, poorly-planned city, the effort required to install any new cable is orders of magnitude bigger than just ripping through an empty field in a backhoe. And being in the financial district meant that all the other people inconvenienced by the work are high-wage broker-types whose time is expensive to waste. (Meaning the city will soak you on all the permits and police overtime)
Maybe, Boston is actually the worst US city for additional expense to infrastructure projects- big or small. The extreme example is how they just built $400,000,000 worth of tunnel for $10,800,000,000 (a 27x multiplier) because of all the uninterruptable activity above the dig site.
If you want cheap fiber to your building, try someplace more amenable, like Chicago.
Right now, I live out in the burbs near boston. 30 minutes away. I have ONE choice in internet access save dialup.
I've got relatives living in such areas (128 region), and they've got two cable companies selling ISP service. The effective speeds to in-state resources are an impressive 300Kb/s.
you've only got one week left to get a dashboard cradle and hands-free kit fitted - don't get caught out.
That's rather silly. It's a feel-good partial-solution... something which doesn't threaten cellphone business interests in any way (and only helps them sell more kits)
Studies show that the greatest danger of telephoning and driving is not from having one hand distracted with a machine (because many folks safely handle CDs and eat food while driving), but the mental load of being engaged in a conversation with somebody not in the vehicle.
But, I suppose that until we've got a few more years of fatality records as proof, the public won't believe this enough to influence law. (And even if they do believe it, prehaps talking and driving will be so important to them that safety will be voluntarily surrendered)
All consumer electronics devices should be manufactured - at base - with the lowest-common-denominator user in mind.
That's an ambiguous statement. If you mean "all designers should be aware of the possible existence of LCD users", then it is so trivially true that it hardly bears saying. But if you mean "all products should be designed so that an LCD can use them", then you are completely insane.
Frankly, this design strategy alone would revolutionize consumer product manufacture in many sectors (auto, electronics, etc), and solve many of the "user-unfriendly" problems that plague consumers today.
Yay! Now every single automobile can be a 2-person vehicle the size of a minivan, with a robotic lift to pull wheelchairs into the driver's seat. I can't wait until we're all driving those expensive, slow, bulky, and fuel-guzzling monsters just on the off-chance a few customers will find it convenient or important.
And just wait for the hunting rifle for the blind...
leading me back to the parent - that no - products are not being crippled by the including the needs of the 'crippled' - in fact it probably makes life better for the rest of us
Fine, say they're not "crippled" then. Instead, we could say they are "six times as expensive to account for features fewer than 1% of customers will ever need".
Great!
The reason these systems don't work for blind people is because of friends like you.
His blind friends shouldn't have to beg for someone to come over and help them when signing up for a new website.
By providing assistance, he wouldn't fix the problem- just perpetuate it. Sometimes providing a workaround just helps the underlying problem survive and grow ever-larger.
so it's unlikely that there'd be a Score:5, Redundant
Of course you'll never see +5 Redundant. He never said anything like that. Read the post again. The paragraph you just quoted clearly says that readers of the page will see one comment at +1 Redundant, go down a little, and see another (from a different author, but providing the same info) as +5 Insightful.
That can happen because sorting by something other than time posted can put an early post below a later one, swapping the apparent redundancy between them.
He's argueing against the entire concept of non-chronological sorting, because it causes readers to view comments in a different order than the moderators and actual commentors did.
I might be incredibly naive, but it seems something more like page 1.5.
This may be an effect of the rapid rate of new comments. By the time you finish reading 100 responses on page 1, dozens more messages can have been added to page 1. So many things that you already read on 1 have been pushed to 2 by the time you get there.
A better CGI would supply a "Read more" link at the bottom of the page, which tells the server the ID# of the final comment you read, so it can start printing more from exactly after it.
(The core problem is use of relative indexing rather than specific DB keying, and is a well-known stumbling-block for programmers)
REPORTING AND AUDIT. If Customer wishes to increase the number of Installed System,
That document doesn't appear to be internally consistent, as scrolling down to Appendix 1, Section 1 gets to this:
the license terms for the components permit Customer to copy, modify, and redistribute the component, in both source code and binary code forms.
It goes on to mention a few specific files not redistributable (basically providing a recipe for "Pink Tie" groups to do their work)
The only way to reconile the appendix statement with the apparent contradiction in REPORTING AND AUDIT is to take it that the "Installed System" as mentioned there applies only to systems under the service agreement. This viewpoint is not at direct odds with other parts of the document, particularly the initial definitions.
Again I ask: how does adding an alternative inhibit functionality that is already there?
You are suggesting that gestures are somehow confined to only replacing history navigation functions. That is patently untrue, as you well know. Why do you ask if you already answered it?
I call bullshit. Name any investigation that has determined that. I don't believe you can.
What are you talking about? It's bloody obvious: the statement "we feed and house them better than if they weren't prisoners" is trivially true.
In the US, prisoners are given 100% free food, housing, and medical care. Non-prisoner US citizens don't get that. Non-citizens certainly don't get it.
The government wouldn't be feeding them at all if they weren't prisoners.
but civilians are not agents of their government.
No, "agents" would be simply doing their jobs and following orders. Those civilians are actually making the decisions! The Israeli government obeys the votes of its Jewish residents.
Democracies find themselves the targets of terrorism for the same reason that dictators tend to be assasinated: someone who violently disagrees with a government's actions will target the people in charge. In modern "first world" nations, that means every random voter.
Gates' believes that everyone should be self made, and build their own wealth by themselves
Hmmm... Gates' net worth has never been less than $1 million. His wealthy parents had prepared him a large trust fund before he was born.
When he was 20, that fund became the startup financing for Microsoft.
learning PER-SITE navigation is simply ridiculous.
It'll be especially cool when someone who's already using mouse-gestures at the OS level visits a website which implements its own gesture interpretation! Every movement can have 2 or 3 separate effects!
Ideally, someone using Links (as opposed to Lynx) gets the same information as anyone else. That's the beauty of having a text only, no frames, no scripts browser around.
You seem to be implying that links doesn't support frames, and that lynx does. But the reverse is true. Links doesn't count as a no-frames browser.
My initial question stands: how are mouse guestures any less accessable than other browser features that require you to locate and click a button or click and drag on a scrollbar?
Because they're different and weird. Oddity by itself creates inaccessibility.
Traditional, well-known techniques are something that the disabled person can plan ahead for. Her client-side software can recognize common navigational features of a web page, and translate them into a form the user can understand. This is impossible if some complicated Jscript is going to generate all links in realtime by running weird code.
(Maybe, with time, a client which understands the specific trickery can be created.)
Mouse guestures just add another technique of navigation that is still bound internally to the same page-hyperlink model of HTML.
The danger of any fancy new site-specific navigation technique is that if it becomes popular enough, the site designer might forget to include the old-faithful A HREF clickable html links. Someone who uses gestures primarily will add a new features, attach it to the gesture list, and call it a day, leaving accessibility-challenged persons (and also robots, webcrawlers, etc) unable to reach the content.
This has already happened in those sites which use a Shockwave Flash bar for primary navigation.
The best way to keep your PK secure is to burn it to a CD, put the CD in a locked drawer, and only EVER put it in your machine when you want to sign anything.
That approach is still unsafe. If your machine has been 0wned, then all the PK software on it is untrustworthy, and could copy down the key the first time you insert that CD to sign something.
Safer is to keep the key and PK software on an independent machine, which is non-networked and completely barred from running any new executables. Any files needing signing must be hand-carried to this box.
(Of course, such an expensive and slow process may be too cumbersome for someone like Debian to use, as they release scores of new packages each day)
I will say that the private key should be kept on the personal machine of the person doing the signing, so they can dl the package, and sign it locally then upload it.
Well of course. When I said "developer machine", that was a synonym for "personal machine" like you're talking about. The developer's machine is naturally where the key will reside, so that is what hackers should aim at.
Now before you go all 'keylogger' on me
Although a keylogger can work, that's not the smartest approach for a hacker. It's too convoluted. The wise hacker will actually replace the gpg executable with a trojaned version, so that whenever certain filenames are signed they are first replaced with an evil variant.
Tainting gpg in that way will finish the hacker's job... from then on, the developer will do all the work of uploading and distributing the infected package under his own trust reputation.
It is you who is mistaken about a great many things.
I concur that dh003i is whacko. However on the following, he isn't completely wrong.
But to ascribe MS's achievement of monopoly status solely to the government "granting it these priviledges" is ridiculous.
A brief study of history and economics will reveal a simple truth - a capitalist free market, without an inpartial entity with the authority, power and will to maintain a level playing field, will trend towards consolidation and bigness.
It's true that free markets always tend towards consolidation. But, that consolidation will only succeed if supported by a government. Once a single entity starts to claim an excessive share of wealth (say 30+% of the money supply), other participants in the marketplace will always take steps to reduce their influence. These steps may start small, but will quickly escalate to reclaiming the money by any possible means, including violent ones.
Those people who are harmed by a true monopoly will initiate force against the monopolists, unless their is some powerful coercer (the government) blocking this path. And if the government sides with the monopolist, then it too faces the risk of destruction by an all-out revolution.
The MS monopoly could've only suceeded if the government deployed troops to fend off angry customers in the guise of "protecting intellectual property rights". Similarly, Standard Oil could only have persisted as a complete monopoly if the government had acted to protect their physical property rights. In both cases, the government was smart enough (or merely democratic enough) not to favor a monopolist over the bulk of their population.
Currency is established voluntarily by individuals acting of their own free will.
And you know what else was created voluntarily by individuals acting of their own free will? Government
That's the central fallacy of capital-L Libertarianism: the idea that given freedom, people will naturally assemble into a nearly-optimal system, and that this system will differ substantially from what we have now.
Those two statements are contradicted by a simple glance at the current world. People had freedom, and they created the system we have now. There is no reason to expect the same won't happen again, if by some miracle all laws were magically replaced with the injunction "Never initiate force".
With this "Public Key Cryptography" you could conceivably sign software in such a way that it could not be altered without breaking the signature,
No... the way to alter software is easy to conceive.
You simply have to hack into the computer holding the private keys used for the signing (very likely the same computer holding the source code as well, and the system which normally uploads new packages to the distribution point). Once there, you can make changes and sign them just as if they were official.
Since attackers of this type have already demonstrated an ablity to hack into computers, PK signing doesn't add any true security. It adds some defensive obscurity, since it's more difficult for the attackers to locate a developer's machine than a distribution one. But dev systems will be more vulnerable to hacking, since they're not likely to apply patches as quickly as a public server. (And I won't recite the old line about "security through obscurity")
The only true benefit from PK signing is that end-users are protected from poor security at mirror sites. Suppose your ISP offers a Debian package mirror as a high-speed convenience, but doesn't secure it well. If it's compromised, trojan packages could be sent to you on the next "apt-get". Comparing those packages against signatures from an official debian.org site will protect you. But that assumes the official servers are sufficiently well-run to avoid being hacked. And as we've seen today, that's not the case.