> Reply to every spam e-mail by going to their web site, > and filling out bogus info
Wow, how clueless. Your address is not, apparently, in the databases of any serious spamming rings.
If I quit my day job and work on this twelve hours a day, seven days a week for the next decade, I could probably just about get that little project done for all the spam I received last month.
> On the flip side, Microsoft software is easier to set up
*Some* Microsoft software is easier to set up; I personally would not place IIS in this category. The first time I ever installed Apache (on Windows 95, no less -- hey, it came with the computer) it took me about an hour, and most of that time was spent reading stuff I *probably* didn't need to know, just in case it should happen to contain anything important. I now have half a dozen more years of experience, and we've had an IIS server at work since August, and I've done quite a bit more than an hour's worth of reading about it, but if I had to set up IIS on my own, I'd still be pretty much lost. Setting up Apache on a new system takes me perhaps five minutes, unless I need to do something complicated.
It does, however, greatly depend on what you're accustomed to. When I went to set up Apache, I was already very used to the idea of editing a configuration file in a text editor, having done so for various other applications over the preceding several years; plus, I had enough background in programming to make the Apache config file format intuitive. The sample config file was very well laid out, thoroughly commented, and easy for me to understand, so I didn't have any trouble. If I hadn't been afraid of doing it wrong, it probably would have taken me fifteen minutes.
Basically, I'm one of those people who thinks these steps:
1. edit/etc/resolv.conf
are easier to perform than these:
1. Right-click on My Network Places and choose Properties.
2. Right-click on Local Area Connection and choose Properties.
3. Under the General tab, find TCP/IP in the list, and select it
4. Click the Properties button.
5. Edit the Preferred and Alternate DNS server entries.
That's not true for everyone. It depends on your experience. If you've never used a text editor before, for instance, you're going to have a hard time figuring out how to make the former procedure work for you. If you were trained in an IT shop that does everything the Microsoft way, you're going to prefer the second procedure.
So it depends on your background.
Frequently, the decision to use IIS is made by a third-party software vendor that chooses to build a product on ASPX (often in conjunction with either Oracle or MS SQL Server).
I'm not talking here about standard general-purpose software, such as office suites, but more in terms of specialized, field-specific software designed for a certain type of business or organization. Because of the economics of smaller scales (fewer customers per product), these things tend not to be held to anything like the same standards as general-purpose software that almost everyone uses (e.g., office suites). Typically, the usability is nightmarishly bad and the security worse. Oh, and for the joy of using such junk you get to pay... let's just say rather a lot more than it costs to license standard off-the-shelf software such as office suites and image editors.
In terms of power, California made its own problems. You want to use more power than any other state, but you don't want to have more power _plants_ than any other state, heck no. Heaven forfend you build any nuclear plants, and you _certainly_ can't build more plants that burn fossil fuels... but hey, let's just keep using as much power as is convenient, and if neighboring states won't sell you their power, well, then the federal government oughta *make* em, because there's no way California ought to have to take any responsibility for its own power consumption. That would be, like, totally wacked out, dude.
Eh. I said at the time (when the rolling blackouts were all over the news for weeks on end) that Californians needed to collectively get over themselves and solve their own problem. It doesn't take a rocket scientist to figure out that the possible solutions are to either use less power, or build more power plants.
As far as the technology sector, those industries don't *have* to be located in California. If you can't get the power you need to conduct your business in that state, there's real estate available elsewhere. (Not that I'm saying businesses should have picked up en masse and moved immediately at the first sign of trouble; moving is expensive, and moving on short notice is more expensive; I'm thinking more in terms of long-term planning and selecting where to put new offices and things, if there's an ongoing problem.)
> When I used IE7 last, I found it to be far from completion
Yeah, that was my take on the previous beta as well. It showed potential, but it clearly wasn't done. This one I haven't seen yet, but I'll definitely be looking at it.
> I do not understand how they can be pushing this as a viable solution for some.
Bear in mind, they're suggesting that people try it out who have been using IE6. They don't actually say that explicitely, but it's what I think they're implying.
The previous beta was mainly intended, I think, to pacify the people who had been clamoring for some evidence that work was being done on IE; they offered it up to skeptics as a demonstration that they'd been working on the problem.
*This* beta is presumably an actual beta, i.e., they're now ready to get feedback from testers and early adopters, so they want some people to try it out.
Wait, since when can an employer no longer terminate an employee for reading the newspaper and making personal phone calls on company time? What is this, the public sector? Sheesh.
If you think Vista disappoints, wait for Blackcomb
on
How Vista Disappoints
·
· Score: 1
Vista is a disappointment primarily because of what was promised, and because of the number of years that have passed. Vista, however, isn't where the feathers are going to fly. That'll be the next release. (Is it the next release after Longhorn still codenamed Blackcomb, or have they re-codenamed it now? Anyway, that one.) When XP was released, other systems were still catching up. Today, they've not so much caught up as zoomed past, and Vista is still next year's promise. Fine, it's normal for competing systems to go back and forth, with whichever one has the latest and greatest release out taking the lead. But the problem for Microsoft is, what's going to happen _after_ Vista. Vista right now is scrambling to get back in the lead. It's only got a couple of things going for it, that the other players haven't already got. Aero glass? Okay, nothing like that has yet made it onto the typical Linux user's desktop... yet. But the freedesktop.org guys have been playing for a while with the technology to do that sort of thing, and once the bugs get worked out, it's actually a better implementation. (You think translucency is nice in the titlebars of your windows? Try taking an entire window, such as a clock, and making it 30% opaque and always-on-top.) This will probably not hit the typical Linux user's desktop before Vista is released -- but it won't be long afterward in coming.
The next release after Vista, and specifically the _timing_ of that release, is going to be the key. It needs to come out a year or two after Vista, circa 2009. It needs, desparately, to not be delayed into the 2012-2013 timeframe.
So here's the question: have the long delays for Vista been because of specific mistakes that were made for that particular version, and because of the needed rewrites (which hopefully will not have to be repeated yet for the very next release), or have they been because of inherent organizational issues within the Microsoft Windows team that are going to continue to dog subsequent releases? When Vista finally ships, can Microsoft turn around and build on it quickly, or have they lost their agility? This I consider to be the critical question for Microsoft right now, the one that will determine, in 2010 or so, whether they can continue to keep the OEMs on board, and whether Microsoft products will continue to ship on practically every new computer sold.
The OEMs are already testing the waters, playing a sort of game of footsie, as it were, with other operating systems in general and Linux in particular. That worries Microsoft, but it doesn't actually hurt them, yet. But it demonstrates a mindset in the OEMs: they're _looking_ at alternatives, one could say looking *for* alternatives. They're not entirely _satisfied_ with Microsoft. Microsoft needs to hold their interest. Vista will help, but it's not 2001 anymore, and Vista will not hold the OEMs' loyalty as long as XP has done. If 2011 rolls around and Microsoft is promising the next release after Vista for 2012, the OEMs are going to start defecting. (I mean really defecting, not playing with a model here and a model there with another OS, but switching over their entire product lines.)
One really long delay the market has tolerated, but the next product is going to have to ship on time, or the industry is going to turn elsewhere, and the bottom line is going to hurt.
Re:The problem of nerve impulse conduction
on
An Alternate Human
·
· Score: 1
This is why you want to add the ability to grow new eyes, like some creatures can grow new tails. So then if you lose an eye, it's inconvenient for a few months, but as long as you don't lose both at once, you're not blind, and even if you do, it's not permanent.
The last time I went to the third page, I was trying to figure out how much more a certain cite needed to climb to make it _into_ the results anyone might ever actually see. It was about halfway down the third page, so I figured it needed to climb another twenty spots or so to have any chance of being seen.
I don't know who was conducting this study, but from what I've seen, most users these days don't *scroll*. If it's too far down the results to be visible without scrolling, they don't see it. For that matter, a significant percentage of users never look past the first result, or perhaps the second if the first is clearly irrelevant.
Pegasus Mail.
When I moved from Windows to Linux (mainly because it's more configurable, and I like that, but also because the Windows versions available at the time had serious stability issues, and I was tired of having to close all my windows and reboot a couple of times a day), Pegasus Mail was the last app from which I had to tear myself away, and the only one I regretted having to leave. I searched high and low for another email application that even came CLOSE be being an adequate replacement for Pegasus Mail, and I never found one. Most people can't name an email app I haven't looked at. I ended up using Gnus, which has _most_ of the features that Pegasus has (plus many that it doesn't), but even that is in a number of ways inferior, and nothing else even comes close. Even Gnus, arguably the ultimate paragon of featurefulness, is missing _significant_ features that Pegasus Mail has, not least in the filtering system. Pegasus Mail also has a very easy learning curve; I routinely recommend it to end users who find Outlook Express too confusing.
I would rate Pegasus Mail as the best Windows-only application I've ever seen.
Another Windows app worth checking out is IrfanView, a graphics viewer and format converter, which also has a few other features (e.g., color adjustments). It doesn't do a lot, but it's a small download and terribly convenient to have around for quick format conversions, as it supports most of the bitmapped graphics formats I've run into, which is a fair number. It's also MUCH nicer for viewing than the quick-view thingy included in Windows.
There are, of course, also various cross-platform applications, some of which are a bit better on Windows than on the Mac, because they integrate better with the rest of the system, in terms of things like the widget set, the system clipboard, and so forth. A lot of the best software for Windows consists of Win32 ports of *nix software. Notably, many GTK applications work quite well on Win32 but have rather rougher edges on OS X.
A few years ago I would have said MS Word, but these days that seems less relevant, partly because OpenOffice.org has become so nice (and, notably, OO.o's frames are much easier to work with than MS Word's text boxes), and partly because in any case there is also the OS X version of Word, so if you've got a Mac, you don't really need BootCamp and Windows to get Word.
The number one Windows-only application you want to actively avoid is Visual Studio. The pain of trying to get anything done with its terrible interface can scar a man for life. Windows Media Player is almost as bad, and more difficult to avoid, since it is the default association for various common filetypes. You will want to associate IrfanView with the formats it supports, but that's mostly still images, so you'll also need something for audio and something for video, and I don't have good recommendations for those things for Win32. (Is there a Win32 port of xmms? I haven't seen it...)
You do NOT have to start speaking corporatese yourself, at least, not mostly. You may occasionally have to demonstrate the ability to do so, but on the whole you should express yourself clearly in plain English.
However, not being able to decipher the unclear speech of the higher-ups would be a significant problem. However lame their lingo may be, you'll nonetheless want to learn enough of it to be able to understand them at least as well as they understand one another.
And, I should note, this sort of language is not very hard to learn. It's much easier to deciper than e.g. legal verbiage, or even some of the weirder corners of academia.
> Having Wikipedia on your hdd is all well and good, but > if you can't easily search it, what's the point?
Searching is no sweat. Anything on your hard drive is much easier (in terms of computer resources) to search than something online. Certainly, searching a local HD-based copy of Wikipedia should be faster, on any modern computer, with a reasonable search facility (involving an index, I imagine), than waiting for the results to a Wikipedia search query to come back over a dialup connection.
Keeping up-to-date is a bigger problem, but if you don't have convenient access to the net all the time, a six-month-old copy of Wikipedia would be considerably better than no Wikipedia. Similarly, a slightly out of date copy of the CPAN would be considerably better than nothing.
Some things just won't work, though. Perlmonks would be a good example of a site for which an out-of-date local copy just isn't going to be useful. Slashdot is another example. What would be the point?
If they're really thinking of this as "the whole web on a disk", they're barking up the wrong tree. Large parts of the web would not be useful to include. However, I'm sure there's 80GB worth of stuff out there that *would* be useful to have. If they're thinking in terms of carefully selecting the stuff that would be really useful to have in a local copy, that might actually prove useful. Potentially.
I would expect ntp.commercial-isp.net to usually be stratum-3 (assuming it exists).
> somebody running an international high-energy physics experiment can be excused for going
> to a level-2 or level-1 server. everybody else is wrong to do that.
My understanding is that a stratum-3 NTP server, all else being equal, should get its time from stratum-2 NTP servers. That's what stratum-2 servers are *for*. There is such a thing as a stratum-4 server (i.e., if I put in an NTP server at work (which I would like to do), on our small network that has some 25 systems on it, it could be stratum-4 and pull time from stratum-3 sources, and would still have greater accuracy than we need), but ordinary users would ordinarily not have such a thing, and would use a stratum-3 source. A large business or an ISP would want its own stratum-3 time server, presumably.
It's a stratum-1 NTP server. Stratum-1 NTP servers are *ONLY* supposed to be used by other stratum-1 NTP servers and by stratum-2 NTP servers, *not* by any random device on the internet.
A LAN router should *NEVER* be using a stratum-1 NTP server; it should be using a stratum-3 NTP server if possible, or *maybe* a stratum-2 server, with special permission, under unusual circumstances, if there is no stratum-3 server available.
If D-Link won't do anything, this guy's going to have to notify everyone who runs a stratum-1 or stratum-2 server in Denmark, give them time to reconfigure, and then shut down the service.
> on date X, send bogus packets in response... not just wrong time, > but seriously wrong time, like a packet with time of 9s in all > fields, which would be most seriously wrong.
It would be better, on date X, to just stop the service (at the old, hardcoded-in-the-routers address, leaving the new service at the new address). This is both kinder to end users (who did not know about this when they bought the hardware and probably still don't) and also a better use of network resources.
Anyway, shouldn't stratum-1 NTP servers reject (or drop) all requests except from known stratum-1 and stratum-2 NTP servers (and maybe stratum-3 NTP servers on certain approved networks)? I thought stratum 2 was where publically open NTP servers were supposed to live, with private ones for local networks on stratum 3 using a stratum-2 server.
> a lack of real training in how to use the system, and more importantly, why it is better
The problem is, in the case of most collaboration software, it's _not_ better than email. The key and most important aspect of collaboration is that it involves working with other people. Other people have email. Everybody has email -- well, everybody who uses computers and has an internet connection. Collaboration software that only works for collaborating with the 2% of people who use the same collaboration software is worthless for collaborating with the other 98% of people. So whatever other nifty features it may have, it's nevertheless fundamentally worse than email. A lot worse.
There is a solution to this, however. The web is just as widespread as email. Everybody's got it. Collaboration software _can_ be better than email if it can be accessible via the web, run on standard web browsers, and require little or no additional training beyond what is required to use the web in general. That, and the interface has to be at least as convenient as a typical mailreader, which probably means a carefully-thought-through application of AJAX technology. With XMLHttpRequest now supported by all the major browsers, the key thing standing in the way of widespread adoption is an interface that's simultaneously as easy to learn to use as email, and as extensible/customizeable for advanced users as email is (which fundamentally means the ability to use a different client by a different vendor and still collaborate with everyone else). In other words, it needs to be email 2.0, not some off-the-wall proprietary system with a half-baked and completely different paradigm.
Key improvements that are needed over the extisting email system include the following:
* A transport system designed to be fundamentally and reliably two-way, so that when someone contacts you you are able to know reliably things like how the message came to you, how to contact the sender, and who his service provider is. The Received: header system in email is not adequate, but experience with email, and the _ways_ in which it is inadequate, need to be taken into account when architecting a replacement system.
* The way bounce messages have traditionally been handled by email is unusably bad in the face of spam that forges From: headers. The above requirement may help, but on the whole a better system for handling notification of delivery success or failure is needed. Frankly, the user should know right away, when sending a note, whether the receiving server is going to accept it for delivery to the inteded recipient, or not. There can be a few seconds of delay while it decides, but the connection shouldn't close until the matter is settled, and the sender's interface should immediately make it evident (e.g., by color-coding the message in the list, or by some other means) if there is any problem, or if the message was accepted.
* Reliable and consistent handling of types of information other than plain text. MIME is only good enough if you know what software the user at the other end has installed, or if you are sending something highly standardized (e.g.,.jpg images). As a transport mechanism, it works, but nothing about the interface or any particular content type is specified, and that's a problem.
* The burden of storing large attachments must be shifted from the receiving server to the sending server -- among other reasons, because this allows the user to know _at sending time_ whether there's a size-limit problem.
* Threading needs to be handled much better -- among other things, it should NOT be necessary to have every message from the conversation in order to construct the general flow. Even something like the References: header used on usenet would be better than the In-Reply-To: header of standard email, but preferably even more reliable. Among other things, there should be a mechanism for a user who is brought into Cc: partway through a conversation to (by just clicking a button) request the rest of the messages from t
> I don't use email filters because I don't trust them to not block important content.
I don't use third-party filters, because I don't trust them, and my own filters, which I write myself, don't so much block as sort. However...
> When one email address starts to attract spam, I just delete it and create a new one.
> I put an auto-responder on the old account that says, "To my friends: this account has
> attracted too much spam - please contact me offline for my new email address". Within
> a month, everyone important has my new email. I do this ritual about once every six months.
I had a friend once who did that. After about the third time I quit bothering. At this point, I no longer have any idea what his email address is, and I'm not sure why I should care.
What, you're still relying on webmasters to provide a comfortable set of colors for you?
Wow, how... sad.
I haven't surfed the web with page-specified colors enabled since 1998. If webmasters' tastes in colors are anything like as bad now as they were then, I don't want any part of it. All pages have #FFE6BC text on a #294D4A background as far as I'm concerned.
> A company may put in its EULA that it owns the data in your account, characters included, > and may do whatever it pleases with it, including deleting it and display it on a cross > on Golgotha. But that ownership may extend only to the bytes and bytes in that data, not > to the player who "created" that data.
I think altering the gameplay experience for the player falls under the category of owning the game, not owning the player. I don't think the other poster intended to indicate that the player would be strapped into a chair with his hands superglued to a controller and _forced_ to play the altered game. The way I read his suggestion, was that when the offending player logged into the game, the game experience would be altered to resemble a less-fun game, probably for some specified amount of play time. This would be a way of allowing the player to earn his way back into the main game, rather than banning him forever.
Although, I think it might be more effective to combine that with the crucifixion idea, i.e., when the player logs into the game he finds his character stuck up on a cross, unable to do much, for a specified amount of time -- not a specified amount of real-world time while the player can just not log in, but a specified amount of _playing_ time, so that the player must sit through the punishment -- if he wants to get back into the game, that is. How to ensure that the time only counts if the player is sitting at the keyboard is left as an exercise.
It wasn't good enough to _think_ he was dead. They had to _know_. There are several reasons for this, some more obvious than others, but the most obvious reason for those who were there at the time is that if he wasn't dead, they needed to break his legs so he'd die quickly, because they needed to get the dying men off the crosses before the holiday.
So they stuck a spear in to see what came out. The blood had already separated, so blood and water came out separate, strong evidence that he wasn't just unconcious (or faking), but actually dead.
> especially if he's going to spray blood and water all over your nice clean tunic?
It was a soldier who did the stabbing; presumably he was in uniform, and laundry was presumably not a major consideration -- probably he was accustomed to carrying out his duties despite such trivial personal concerns. In any case as someone else pointed out nothing was going to "spray" out with any great force if the body was in fact a corpse.
If there had been a spray of just blood, rather than a flow of blood and water, they would have broken his legs, so that he would suffocate.
Yeah, but they're going to take a significant PR hit, unless I've missed my guess. Vista will sell, sure. What's going to happen in ten years, though, or twenty? After Vista ships, if they deliver what they need to on the service packs, and put together a passable Blackcomb/Vienna... maybe they will keep their position. But what if Blackcomb/Vienna are just as late and underfeatured? What if their internal organization only gets *worse*? Sure, Vista can suck, and they can sell millions of units, but at some point they've got to turn that around and start delivering again.
Monopolies don't ordinarily last very many decades. There are exceptions, mainly when there's an artificial barrier preventing them from failing -- e.g., in the case of local phone companies, nobody else has right of way to run cable -- but in the absense of such artificial considerations making the field a natural monopoly, a 90% market share is not long-term sustainable. Sooner or later, something will give.
So don't upgrade until the service pack comes out. Fight the urge to have the latest and greatest. Just think how much smoother your experience would have been if you'd stayed on Windows 2000 Pro until XP SP2 came out. Okay, so keep XP SP2 now until Vista hits its second or third service pack.
Or, of course, you could migrate to BSD. It's been around long enough now that they should have it just about right pretty soon:-)
Slashdot Mirror
> Reply to every spam e-mail by going to their web site,
> and filling out bogus info
Wow, how clueless. Your address is not, apparently, in the databases of any serious spamming rings.
If I quit my day job and work on this twelve hours a day, seven days a week for the next decade, I could probably just about get that little project done for all the spam I received last month.
> On the flip side, Microsoft software is easier to set up *Some* Microsoft software is easier to set up; I personally would not place IIS in this category. The first time I ever installed Apache (on Windows 95, no less -- hey, it came with the computer) it took me about an hour, and most of that time was spent reading stuff I *probably* didn't need to know, just in case it should happen to contain anything important. I now have half a dozen more years of experience, and we've had an IIS server at work since August, and I've done quite a bit more than an hour's worth of reading about it, but if I had to set up IIS on my own, I'd still be pretty much lost. Setting up Apache on a new system takes me perhaps five minutes, unless I need to do something complicated. It does, however, greatly depend on what you're accustomed to. When I went to set up Apache, I was already very used to the idea of editing a configuration file in a text editor, having done so for various other applications over the preceding several years; plus, I had enough background in programming to make the Apache config file format intuitive. The sample config file was very well laid out, thoroughly commented, and easy for me to understand, so I didn't have any trouble. If I hadn't been afraid of doing it wrong, it probably would have taken me fifteen minutes. Basically, I'm one of those people who thinks these steps: 1. edit /etc/resolv.conf
are easier to perform than these:
1. Right-click on My Network Places and choose Properties.
2. Right-click on Local Area Connection and choose Properties.
3. Under the General tab, find TCP/IP in the list, and select it
4. Click the Properties button.
5. Edit the Preferred and Alternate DNS server entries.
That's not true for everyone. It depends on your experience. If you've never used a text editor before, for instance, you're going to have a hard time figuring out how to make the former procedure work for you. If you were trained in an IT shop that does everything the Microsoft way, you're going to prefer the second procedure.
So it depends on your background.
Frequently, the decision to use IIS is made by a third-party software vendor that chooses to build a product on ASPX (often in conjunction with either Oracle or MS SQL Server). I'm not talking here about standard general-purpose software, such as office suites, but more in terms of specialized, field-specific software designed for a certain type of business or organization. Because of the economics of smaller scales (fewer customers per product), these things tend not to be held to anything like the same standards as general-purpose software that almost everyone uses (e.g., office suites). Typically, the usability is nightmarishly bad and the security worse. Oh, and for the joy of using such junk you get to pay... let's just say rather a lot more than it costs to license standard off-the-shelf software such as office suites and image editors.
In terms of power, California made its own problems. You want to use more power than any other state, but you don't want to have more power _plants_ than any other state, heck no. Heaven forfend you build any nuclear plants, and you _certainly_ can't build more plants that burn fossil fuels... but hey, let's just keep using as much power as is convenient, and if neighboring states won't sell you their power, well, then the federal government oughta *make* em, because there's no way California ought to have to take any responsibility for its own power consumption. That would be, like, totally wacked out, dude. Eh. I said at the time (when the rolling blackouts were all over the news for weeks on end) that Californians needed to collectively get over themselves and solve their own problem. It doesn't take a rocket scientist to figure out that the possible solutions are to either use less power, or build more power plants. As far as the technology sector, those industries don't *have* to be located in California. If you can't get the power you need to conduct your business in that state, there's real estate available elsewhere. (Not that I'm saying businesses should have picked up en masse and moved immediately at the first sign of trouble; moving is expensive, and moving on short notice is more expensive; I'm thinking more in terms of long-term planning and selecting where to put new offices and things, if there's an ongoing problem.)
Yeah, but do you use Ball mason jars or Kerr mason jars? It's a question almost as important as big-endian versus little-endian...
> When I used IE7 last, I found it to be far from completion Yeah, that was my take on the previous beta as well. It showed potential, but it clearly wasn't done. This one I haven't seen yet, but I'll definitely be looking at it. > I do not understand how they can be pushing this as a viable solution for some. Bear in mind, they're suggesting that people try it out who have been using IE6. They don't actually say that explicitely, but it's what I think they're implying. The previous beta was mainly intended, I think, to pacify the people who had been clamoring for some evidence that work was being done on IE; they offered it up to skeptics as a demonstration that they'd been working on the problem. *This* beta is presumably an actual beta, i.e., they're now ready to get feedback from testers and early adopters, so they want some people to try it out.
Wait, since when can an employer no longer terminate an employee for reading the newspaper and making personal phone calls on company time? What is this, the public sector? Sheesh.
Vista is a disappointment primarily because of what was promised, and because of the number of years that have passed. Vista, however, isn't where the feathers are going to fly. That'll be the next release. (Is it the next release after Longhorn still codenamed Blackcomb, or have they re-codenamed it now? Anyway, that one.) When XP was released, other systems were still catching up. Today, they've not so much caught up as zoomed past, and Vista is still next year's promise. Fine, it's normal for competing systems to go back and forth, with whichever one has the latest and greatest release out taking the lead. But the problem for Microsoft is, what's going to happen _after_ Vista. Vista right now is scrambling to get back in the lead. It's only got a couple of things going for it, that the other players haven't already got. Aero glass? Okay, nothing like that has yet made it onto the typical Linux user's desktop... yet. But the freedesktop.org guys have been playing for a while with the technology to do that sort of thing, and once the bugs get worked out, it's actually a better implementation. (You think translucency is nice in the titlebars of your windows? Try taking an entire window, such as a clock, and making it 30% opaque and always-on-top.) This will probably not hit the typical Linux user's desktop before Vista is released -- but it won't be long afterward in coming.
The next release after Vista, and specifically the _timing_ of that release, is going to be the key. It needs to come out a year or two after Vista, circa 2009. It needs, desparately, to not be delayed into the 2012-2013 timeframe.
So here's the question: have the long delays for Vista been because of specific mistakes that were made for that particular version, and because of the needed rewrites (which hopefully will not have to be repeated yet for the very next release), or have they been because of inherent organizational issues within the Microsoft Windows team that are going to continue to dog subsequent releases? When Vista finally ships, can Microsoft turn around and build on it quickly, or have they lost their agility? This I consider to be the critical question for Microsoft right now, the one that will determine, in 2010 or so, whether they can continue to keep the OEMs on board, and whether Microsoft products will continue to ship on practically every new computer sold.
The OEMs are already testing the waters, playing a sort of game of footsie, as it were, with other operating systems in general and Linux in particular. That worries Microsoft, but it doesn't actually hurt them, yet. But it demonstrates a mindset in the OEMs: they're _looking_ at alternatives, one could say looking *for* alternatives. They're not entirely _satisfied_ with Microsoft. Microsoft needs to hold their interest. Vista will help, but it's not 2001 anymore, and Vista will not hold the OEMs' loyalty as long as XP has done. If 2011 rolls around and Microsoft is promising the next release after Vista for 2012, the OEMs are going to start defecting. (I mean really defecting, not playing with a model here and a model there with another OS, but switching over their entire product lines.)
One really long delay the market has tolerated, but the next product is going to have to ship on time, or the industry is going to turn elsewhere, and the bottom line is going to hurt.
This is why you want to add the ability to grow new eyes, like some creatures can grow new tails. So then if you lose an eye, it's inconvenient for a few months, but as long as you don't lose both at once, you're not blind, and even if you do, it's not permanent.
So, then, about that Plutonium 239 I wanted to get for, umm, completely peaceful, civilian, umm, personal use, ...
The last time I went to the third page, I was trying to figure out how much more a certain cite needed to climb to make it _into_ the results anyone might ever actually see. It was about halfway down the third page, so I figured it needed to climb another twenty spots or so to have any chance of being seen.
I don't know who was conducting this study, but from what I've seen, most users these days don't *scroll*. If it's too far down the results to be visible without scrolling, they don't see it. For that matter, a significant percentage of users never look past the first result, or perhaps the second if the first is clearly irrelevant.
Pegasus Mail. When I moved from Windows to Linux (mainly because it's more configurable, and I like that, but also because the Windows versions available at the time had serious stability issues, and I was tired of having to close all my windows and reboot a couple of times a day), Pegasus Mail was the last app from which I had to tear myself away, and the only one I regretted having to leave. I searched high and low for another email application that even came CLOSE be being an adequate replacement for Pegasus Mail, and I never found one. Most people can't name an email app I haven't looked at. I ended up using Gnus, which has _most_ of the features that Pegasus has (plus many that it doesn't), but even that is in a number of ways inferior, and nothing else even comes close. Even Gnus, arguably the ultimate paragon of featurefulness, is missing _significant_ features that Pegasus Mail has, not least in the filtering system. Pegasus Mail also has a very easy learning curve; I routinely recommend it to end users who find Outlook Express too confusing. I would rate Pegasus Mail as the best Windows-only application I've ever seen. Another Windows app worth checking out is IrfanView, a graphics viewer and format converter, which also has a few other features (e.g., color adjustments). It doesn't do a lot, but it's a small download and terribly convenient to have around for quick format conversions, as it supports most of the bitmapped graphics formats I've run into, which is a fair number. It's also MUCH nicer for viewing than the quick-view thingy included in Windows. There are, of course, also various cross-platform applications, some of which are a bit better on Windows than on the Mac, because they integrate better with the rest of the system, in terms of things like the widget set, the system clipboard, and so forth. A lot of the best software for Windows consists of Win32 ports of *nix software. Notably, many GTK applications work quite well on Win32 but have rather rougher edges on OS X. A few years ago I would have said MS Word, but these days that seems less relevant, partly because OpenOffice.org has become so nice (and, notably, OO.o's frames are much easier to work with than MS Word's text boxes), and partly because in any case there is also the OS X version of Word, so if you've got a Mac, you don't really need BootCamp and Windows to get Word. The number one Windows-only application you want to actively avoid is Visual Studio. The pain of trying to get anything done with its terrible interface can scar a man for life. Windows Media Player is almost as bad, and more difficult to avoid, since it is the default association for various common filetypes. You will want to associate IrfanView with the formats it supports, but that's mostly still images, so you'll also need something for audio and something for video, and I don't have good recommendations for those things for Win32. (Is there a Win32 port of xmms? I haven't seen it...)
You do NOT have to start speaking corporatese yourself, at least, not mostly. You may occasionally have to demonstrate the ability to do so, but on the whole you should express yourself clearly in plain English. However, not being able to decipher the unclear speech of the higher-ups would be a significant problem. However lame their lingo may be, you'll nonetheless want to learn enough of it to be able to understand them at least as well as they understand one another. And, I should note, this sort of language is not very hard to learn. It's much easier to deciper than e.g. legal verbiage, or even some of the weirder corners of academia.
> Having Wikipedia on your hdd is all well and good, but
> if you can't easily search it, what's the point?
Searching is no sweat. Anything on your hard drive is much easier (in terms of computer resources) to search than something online. Certainly, searching a local HD-based copy of Wikipedia should be faster, on any modern computer, with a reasonable search facility (involving an index, I imagine), than waiting for the results to a Wikipedia search query to come back over a dialup connection.
Keeping up-to-date is a bigger problem, but if you don't have convenient access to the net all the time, a six-month-old copy of Wikipedia would be considerably better than no Wikipedia. Similarly, a slightly out of date copy of the CPAN would be considerably better than nothing.
Some things just won't work, though. Perlmonks would be a good example of a site for which an out-of-date local copy just isn't going to be useful. Slashdot is another example. What would be the point?
If they're really thinking of this as "the whole web on a disk", they're barking up the wrong tree. Large parts of the web would not be useful to include. However, I'm sure there's 80GB worth of stuff out there that *would* be useful to have. If they're thinking in terms of carefully selecting the stuff that would be really useful to have in a local copy, that might actually prove useful. Potentially.
I would expect ntp.commercial-isp.net to usually be stratum-3 (assuming it exists). > somebody running an international high-energy physics experiment can be excused for going > to a level-2 or level-1 server. everybody else is wrong to do that. My understanding is that a stratum-3 NTP server, all else being equal, should get its time from stratum-2 NTP servers. That's what stratum-2 servers are *for*. There is such a thing as a stratum-4 server (i.e., if I put in an NTP server at work (which I would like to do), on our small network that has some 25 systems on it, it could be stratum-4 and pull time from stratum-3 sources, and would still have greater accuracy than we need), but ordinary users would ordinarily not have such a thing, and would use a stratum-3 source. A large business or an ISP would want its own stratum-3 time server, presumably.
It's a stratum-1 NTP server. Stratum-1 NTP servers are *ONLY* supposed to be used by other stratum-1 NTP servers and by stratum-2 NTP servers, *not* by any random device on the internet. A LAN router should *NEVER* be using a stratum-1 NTP server; it should be using a stratum-3 NTP server if possible, or *maybe* a stratum-2 server, with special permission, under unusual circumstances, if there is no stratum-3 server available. If D-Link won't do anything, this guy's going to have to notify everyone who runs a stratum-1 or stratum-2 server in Denmark, give them time to reconfigure, and then shut down the service.
> on date X, send bogus packets in response... not just wrong time,
> but seriously wrong time, like a packet with time of 9s in all
> fields, which would be most seriously wrong.
It would be better, on date X, to just stop the service (at the old, hardcoded-in-the-routers address, leaving the new service at the new address). This is both kinder to end users (who did not know about this when they bought the hardware and probably still don't) and also a better use of network resources.
Anyway, shouldn't stratum-1 NTP servers reject (or drop) all requests except from known stratum-1 and stratum-2 NTP servers (and maybe stratum-3 NTP servers on certain approved networks)? I thought stratum 2 was where publically open NTP servers were supposed to live, with private ones for local networks on stratum 3 using a stratum-2 server.
> a lack of real training in how to use the system, and more importantly, why it is better The problem is, in the case of most collaboration software, it's _not_ better than email. The key and most important aspect of collaboration is that it involves working with other people. Other people have email. Everybody has email -- well, everybody who uses computers and has an internet connection. Collaboration software that only works for collaborating with the 2% of people who use the same collaboration software is worthless for collaborating with the other 98% of people. So whatever other nifty features it may have, it's nevertheless fundamentally worse than email. A lot worse. There is a solution to this, however. The web is just as widespread as email. Everybody's got it. Collaboration software _can_ be better than email if it can be accessible via the web, run on standard web browsers, and require little or no additional training beyond what is required to use the web in general. That, and the interface has to be at least as convenient as a typical mailreader, which probably means a carefully-thought-through application of AJAX technology. With XMLHttpRequest now supported by all the major browsers, the key thing standing in the way of widespread adoption is an interface that's simultaneously as easy to learn to use as email, and as extensible/customizeable for advanced users as email is (which fundamentally means the ability to use a different client by a different vendor and still collaborate with everyone else). In other words, it needs to be email 2.0, not some off-the-wall proprietary system with a half-baked and completely different paradigm. Key improvements that are needed over the extisting email system include the following: * A transport system designed to be fundamentally and reliably two-way, so that when someone contacts you you are able to know reliably things like how the message came to you, how to contact the sender, and who his service provider is. The Received: header system in email is not adequate, but experience with email, and the _ways_ in which it is inadequate, need to be taken into account when architecting a replacement system. * The way bounce messages have traditionally been handled by email is unusably bad in the face of spam that forges From: headers. The above requirement may help, but on the whole a better system for handling notification of delivery success or failure is needed. Frankly, the user should know right away, when sending a note, whether the receiving server is going to accept it for delivery to the inteded recipient, or not. There can be a few seconds of delay while it decides, but the connection shouldn't close until the matter is settled, and the sender's interface should immediately make it evident (e.g., by color-coding the message in the list, or by some other means) if there is any problem, or if the message was accepted. * Reliable and consistent handling of types of information other than plain text. MIME is only good enough if you know what software the user at the other end has installed, or if you are sending something highly standardized (e.g., .jpg images). As a transport mechanism, it works, but nothing about the interface or any particular content type is specified, and that's a problem.
* The burden of storing large attachments must be shifted from the receiving server to the sending server -- among other reasons, because this allows the user to know _at sending time_ whether there's a size-limit problem.
* Threading needs to be handled much better -- among other things, it should NOT be necessary to have every message from the conversation in order to construct the general flow. Even something like the References: header used on usenet would be better than the In-Reply-To: header of standard email, but preferably even more reliable. Among other things, there should be a mechanism for a user who is brought into Cc: partway through a conversation to (by just clicking a button) request the rest of the messages from t
> I don't use email filters because I don't trust them to not block important content. I don't use third-party filters, because I don't trust them, and my own filters, which I write myself, don't so much block as sort. However... > When one email address starts to attract spam, I just delete it and create a new one. > I put an auto-responder on the old account that says, "To my friends: this account has > attracted too much spam - please contact me offline for my new email address". Within > a month, everyone important has my new email. I do this ritual about once every six months. I had a friend once who did that. After about the third time I quit bothering. At this point, I no longer have any idea what his email address is, and I'm not sure why I should care.
What, you're still relying on webmasters to provide a comfortable set of colors for you? Wow, how... sad. I haven't surfed the web with page-specified colors enabled since 1998. If webmasters' tastes in colors are anything like as bad now as they were then, I don't want any part of it. All pages have #FFE6BC text on a #294D4A background as far as I'm concerned.
Indeed.
However, I've discovered that this feature, in making replies too easy, can cause us to inadvertently run into slashdot's anti-spam features.
> A company may put in its EULA that it owns the data in your account, characters included,
> and may do whatever it pleases with it, including deleting it and display it on a cross
> on Golgotha. But that ownership may extend only to the bytes and bytes in that data, not
> to the player who "created" that data.
I think altering the gameplay experience for the player falls under the category of owning the game, not owning the player. I don't think the other poster intended to indicate that the player would be strapped into a chair with his hands superglued to a controller and _forced_ to play the altered game. The way I read his suggestion, was that when the offending player logged into the game, the game experience would be altered to resemble a less-fun game, probably for some specified amount of play time. This would be a way of allowing the player to earn his way back into the main game, rather than banning him forever.
Although, I think it might be more effective to combine that with the crucifixion idea, i.e., when the player logs into the game he finds his character stuck up on a cross, unable to do much, for a specified amount of time -- not a specified amount of real-world time while the player can just not log in, but a specified amount of _playing_ time, so that the player must sit through the punishment -- if he wants to get back into the game, that is. How to ensure that the time only counts if the player is sitting at the keyboard is left as an exercise.
> Why stab someone who's already dead,
It wasn't good enough to _think_ he was dead. They had to _know_. There are several reasons for this, some more obvious than others, but the most obvious reason for those who were there at the time is that if he wasn't dead, they needed to break his legs so he'd die quickly, because they needed to get the dying men off the crosses before the holiday.
So they stuck a spear in to see what came out. The blood had already separated, so blood and water came out separate, strong evidence that he wasn't just unconcious (or faking), but actually dead.
> especially if he's going to spray blood and water all over your nice clean tunic?
It was a soldier who did the stabbing; presumably he was in uniform, and laundry was presumably not a major consideration -- probably he was accustomed to carrying out his duties despite such trivial personal concerns. In any case as someone else pointed out nothing was going to "spray" out with any great force if the body was in fact a corpse.
If there had been a spray of just blood, rather than a flow of blood and water, they would have broken his legs, so that he would suffocate.
> they're going to sell millions of units anyway
Yeah, but they're going to take a significant PR hit, unless I've missed my guess. Vista will sell, sure. What's going to happen in ten years, though, or twenty? After Vista ships, if they deliver what they need to on the service packs, and put together a passable Blackcomb/Vienna... maybe they will keep their position. But what if Blackcomb/Vienna are just as late and underfeatured? What if their internal organization only gets *worse*? Sure, Vista can suck, and they can sell millions of units, but at some point they've got to turn that around and start delivering again.
Monopolies don't ordinarily last very many decades. There are exceptions, mainly when there's an artificial barrier preventing them from failing -- e.g., in the case of local phone companies, nobody else has right of way to run cable -- but in the absense of such artificial considerations making the field a natural monopoly, a 90% market share is not long-term sustainable. Sooner or later, something will give.
> I'd rather they wait and get it right
:-)
So don't upgrade until the service pack comes out. Fight the urge to have the latest and greatest. Just think how much smoother your experience would have been if you'd stayed on Windows 2000 Pro until XP SP2 came out. Okay, so keep XP SP2 now until Vista hits its second or third service pack.
Or, of course, you could migrate to BSD. It's been around long enough now that they should have it just about right pretty soon