You can, of course, deduce things about a person's linguistic background by noticing what kinds of mistakes they make. Some examples...
* Consistently confusing certain phonemes with one another even though they in fact sound nothing alike at all is a strong indicator that the native language doesn't contain either of them. Japanese people, for example, have a terrible time with English short vowels, particularly the a in "bad" and the u in "bus". This is because Japanese does not have these sounds. (Technically, the a in "bad" does occasionally occur in spoken Japanese, but it's a minority allophone in free variation for the a in "father". The u in "bus" does not occur at all.) Hence, "Pizza Hat". English speakers have a horrible time with glottal and pharyngeal stops, for the same reason.
* Consistently substituting certain words for certain other words in certain contexts where they don't make sense can mean that the native language has a word that is translated both ways. For example, using the word "teach" in place of "tell" ("Please, don't keep me in suspense. Teach me what happened!") is characteristic of Japanese. Saying "walk" instead of "live" (particularly when referring to living one's life a certain way) is characteristic of Greek (or of spending a lot of time reading old, sub-optimal English translations of things originally written in Greek; notably, the New Testament). English speakers often mix up ser and estar (in Spanish, and the corresponding verbs in other languages) and have trouble remembering which Japanese verbs to use to talk about wearing various kinds of clothing.
* Consistently confusing pairs of otherwise-related phonemes distinguished by the same type of contrast is a very strong indicator that the person's native language doesn't contrast on that feature. Koreans, for instance, have difficulty with the voiced/unvoiced distinction. English speakers have a similarly hard time hearing the distinction between aspirated and unaspirated.
* Having trouble with grammatical gender -- especially if it's inconsistent, not always getting the same words wrong in the same way, but different mistakes each time -- is an indication that the native language probably does not have grammatical gender. This doesn't apply much to English, since English has very little gender left in it.
* Attempting to pluralize adjectives when modifying plural nouns is a very strong indication that the person's native language does that. (This is characteristic of most European languages, with English being a notable exception.)
* If the target language is tonal, a native speaker can usually tell whether another speaker's native language is also tonal or not. This doesn't apply to English, obviously.
I'm sure there are others, but that's enough to illustrate the concept.
Preposition mistakes generally mean nothing, especially selecting the wrong preposition to introduce a prepositional phrase. Everyone makes them, and preposition usage varies so much even among dialects within a language that you can't usually draw firm conclusions. Frequent failure to correctly match up verbs with the complementary prepositions they prefer is a strong indication of a non-native speaker, but even this doesn't usually tell you very much about which other language they're coming from.
> The use of the awkward word "stuffs" has > been, in my experience, a strong (actually, > perfect) indicator that the speaker is Chinese.
Not Chinese specifically, just foreign in general. Foreign-language dictionaries and textbooks frequently don't bother to point out that certain words are mass nouns, and even if they do foreign-language learners don't always take note of it and mark it on their vocab cards, and even if they do they don't always remember.
Mass nouns that have a physical reason to be mass nouns (e.g., "water", "wood", "metal") are less of a problem, but the ones that refer collectively to obviously discrete things are a frequent source of trouble. "Staff" and "stuff" are among the ones that come up particularly often, but English has hundreds of mass nouns (some of which are pretty much always mass nouns, and others are only mass nouns when used with certain meanings; there are also some that can just be used either way). Native speakers don't usually think about these words as being significantly different from other nouns, but grammatically they really are. (Hang out with foreign language learners enough, and you will start to notice all sorts of interesting things about English. It's enlightening, actually.)
Admittedly, English mass nouns are particularly troublesome for native speakers of languages that ordinarily don't have grammatical number at all (e.g., Japanese), because they have had to get into a habit of forcing themselves, every single time they write a noun, to stop and think about whether it's just one or more than one (otherwise they end up writing things like "My two younger brother both goes to college."), something that comes natural to native speakers of languages that have a singular/plural distinction, and so keeping track of this can become an added distraction that makes it easier to forget about mass nouns. But native speakers of other languages (e.g., European languages, virtually all of which inflect nouns for case, number, and gender) do also make the mistake of pluralizing mass nouns inappropriately. It's not quite as frequent for them, but it's far from rare. It happens because people who have not yet achieved full fluency tend to think in their native language and then translate by components, adjusting the word order as necessary, translating the words into the other language, and so forth, whatever is needed to go from the one language to the other. If the English word that's a mass noun is a translation for a word in the person's native language that's not special in that way, this can easily result in an extraneous pluralization.
(I hang out on Lang-8, so I have had some exposure to English written by native speakers of various other languages.)
> Doing "not so well" doesn't mean that they can't do well enough.
I don't think you've grasped the magnitude of exactly how much atmosphere Mars doesn't have. The surface pressure is zero point something kPa. That's REALLY low pressure. Water at pressures that low is a vapor regardless of temperature. It can't solidify (on its own), let alone exist as a liquid. A standard college physics classroom vacuum pump can't create a vacuum with pressure that low.
For comparison, standard pressure on Earth at sea level is 101.325 kPa. On top of Mt. Everest, the pressure is about a third of standard, which is still somewhere around 50 or 100 times as much pressure as on Mars.
> I wouldn't rule out altogether an engineered species tailored for Mars.
To function in the near-vacuum of Mars, it would need to be engineered to be significantly different from anything found on Earth, except perhaps a small handful of anaerobic microbes, and I'm not even sure about those. Even most Earth life that can *survive* pressure that low can't perform metabolic functions to any significant extent without some source of metabolic gasses -- they just wait until they get some air and then start back up again. The ones that don't need oxygen generally need something else, and on Mars they won't have it -- unless you keep them in a pressurized container of some sort.
> It's worth noting that there is a huge gap > between the worst survivable environments > of Earth and the best of Mars.
Yes, exactly. It's not even really comparable.
> Carbon dioxide partial pressure is actually > higher for Mars than Earth
I can cut a thirty-inch super-extra-large pizza into six slices and give you just one slice -- only a sixth of the pizza. Or, if you prefer, I can make a pizza an eighth of an inch in diameter and use a microscope and precision laser to cut that eighth-inch-diameter pizza into twenty slices and give you nineteen of those slices -- almost the whole pizza. Which is more? This is admittedly a simplification. Partial pressure does also matter, up to a point -- but only up to a point. When you're comparing the atmosphere of Mars to Earth, any small advantage Mars might have in the partial pressure department is overwhelmingly dwarfed to nothingness by the spectacular disadvantage in absolute pressure.
I do not for one minute believe that Earth's lichens (let alone the plain old green algae that terraforming enthusiasts always think is magic) will grow on Mars (except in a pressurized container). There isn't any air. Plants don't grow unless they're immersed in air (or water or something).
> [Lichens] grow in the most inhospitable places on Earth.
All of which places have substantial quantities of atmosphere.
Mars does not.
So unless you set up a sealed, controlled environment (like a terrarium or a dome), lichens won't do so well on Mars. Neither will much of anything else, with the possible exception of some bacteria and maybe a couple of other esoteric microbes. Or robots.
More likely they've figured out how to substantially automate assembly so that there's very little labor involved per unit. (Labor is expensive in the first world. That's why so many labor-intensive products are made elsewhere and shipped in. On the other hand, power is relatively cheap in the first world, so if almost all of the work is being done by machines instead of people, locating manufacturing plants in the first world can make sense. You still need people to run the machines, of course, but the more automated your plant becomes the less labor you need per unit of product.)
Perhaps he has indeed been captured, but by someone *other* than legitimate law enforcement officers. For instance, maybe he is being held by the "enforcement" arm of Los Zetas. Perhaps they will hold him for ransom. Or perhaps they will accept a bribe from his enemies, to keep him or to eliminate him. Or perhaps they will hand him over to the USDA.
(Sorry. I think the conspiracy theory thread on that Korean unicorn story may have affected me in some way.)
Actually, the unicorn remains were _planted_ in Korea by the NKVD, back in the thirties, as part of a secret joint operation with the US Department of Agriculture. The operation was suspended during WWII and the Korean War, and then before they could get around to completing it the USDA had to pull out (because of the altered political situation in Korea), leaving the MVD to bungle it until the Sino-Soviet Split forced them to finally abandon the effort as well. Nobody trusted China to complete the project, so they were never informed.
The unicorns actually live in the Arctic, but the Russian, Scandinavian, Canadian, and American governments keep that covered up, because if it ever got out it would call the Reindeer Dogma into question.
I've given the notion quite a bit of thought, and I eventually concluded, somewhat reluctantly, that it's not practicable on a large scale, for fundamental reasons that have nothing to do with any specific design or protocol that might be attempted.
However, there are usually ways to get information in or out of a place that's disconnected from the internet for whatever reason. Ultimately, if *any* kind of communication in or out of the country is possible, somebody can use it indirectly to get information to or from the internet.
Communication mechanisms that would work for this include, but are not limited to, the following: traditional phone lines; cables of any kind that cross the border, even if they were intended to carry power; satellite phones, or any other kind of bi-directional satellite link; radio transceivers, short-wave or otherwise; any postal service that delivers international mail, whether it's private or public; any location where it's possible for a person to cross the border, legally or otherwise; carrier pigeon; heliography; smoke signals; slips of paper in sealed glass or plastic bottles floating across the Mediterranean Sea; slips of paper in metal containers hurled across the border with trebuchets; words spelled out in fields using rocks and sticks, viewable by anybody with satellites or planes; messages can be dropped into the country from planes, either in robust containers or with parachutes; I'm sure there are other ways not listed here.
None of this requires a networking protocol in any normal sense of that term. The message only has to get through to somebody who is willing to pass it along, and they can do so. Outbound messages, once picked up by anyone, can be spread via the regular internet, no problem. Is there one person anywhere outside Syria who's willing to post the stuff on Twitter? Similarly, inbound messages, once received within the country, can be communicated via whatever mechanisms people are using to communicate within the country, even if they can't reach the outside that way.
Traditionally, one really common way to get information _into_ a geopolitical area where the local government is uncooperative is via radio, because the hard part of radio is the transmitter, and you can use one that's well outside the jurisdiction in question and transmit across the border. People on the inside just need cheap battery-operated transistor radios, which are remarkably difficult for the government to hunt down and eliminate. Jamming is of limited practicality, especially if the sending station is fairly powerful: you can jam it in a small area, but trying to jam it nationwide, if it's a strong signal, is very difficult. VOA is one example of an organization that has been doing this for decades.
The harder part is getting information _out_ of the place. Traditional radio transmitters give away their position, so ones inside the country are relatively easy for the government to shut down (if they have effective control of the whole country, which admittedly may not currently be the case in Syria). However, it's possible to narrow-cast radio waves directionally, aiming for a particular known receiver, which I think may be approximately what satphones do. That would be harder to stop. In the case of Syria, though, I'm pretty sure both sides in this conflict have vehicular transport, so the easiest thing may just be to send couriers across the border (e.g., into Turkey or Jordan or maybe even Lebanon, which I think is not real far from Damascus) and send the message from there.
Just in case someone feels like being an idiot and taking my little allusion for more than the small bit of attempted humor that it is, I should probably clarify: I am not even remotely interested in taking sides in the current conflict in Syria. Please don't try to read anything political into my above post. It wasn't intended that way. It's just supposed to be silly and funny, that's all. Thanks.
Syria has always been dead. In no public or private utterance can it be admitted that situation has at any time been otherwise. Officially the change of status has never happened. Netcraft has confirmed that Syria is dead: therefore, Netcraft has always confirmed that Syria is dead, and thus Syria has always been dead. The failing third-world government of the moment always represents absolute evil, and thus it follows that any past or future situation in which that government might not be failing is impossible.
(8-character filenames regularly got mangled like that, for a variety of reasons. The precise details are a little fuzzy in my mind, since it has been a few years, but IIRC one possible cause was if the user gave the name in mixed case when saving the file. I think long extensions, such as.HTML, were another possible cause, although most Windows users didn't know the four-letter versions of.HTM and.JPG, so that was less common. Also, I believe some 8-bit and maybe even 16-bit software used old API calls or "interrupts" that sometimes resulted in LFNs being created for no really obvious reason. Oh, and I think copying or renaming a file in Windows Explorer sometimes caused it as well, though I'm not certain about that last one.)
> What was the last version of Windows you used? Windows 95?
Define "used".
On *my* computer, yeah, it's been a few versions.
However, I'm a network administrator, so I'm intimately familiar with the day-to-day behavior of versions up through Seven SP1. Admittedly, we haven't deployed Eight yet.
> You don't even have to reboot for driver installs.
It wouldn't matter very much if you had to reboot for a driver install, because that only happens when you first install Windows or when you add new hardware, neither of which is real often.
However, you frequently have to reboot Windows for application updates (antivirus software and browser plugins are particularly bad about this, and they're both things that have to be updated ANNOYINGLY often), and you virtually always have to reboot for Windows updates (which admittedly only happen once or occasionally twice a month).
> Aero offloads the UI to the GPU making the system faster, not slower.
That statement is very much at odds with my experience. I suppose it may depend on your graphics hardware. Almost all the systems I've worked with have onboard graphics, FWIW. But I'm not sure how you could get much faster than the Classic GUI; it is generally, as far as the user can notice without benchmarking tools, instantaneous. Aero is... not instantaneous. Also, it's at least twice as distracting as that horrible plastic-looking theme they tried to foist on us in XP.
> I have never seen explorer crash in either Vista or 7.
How many Windows computers do you administer? The word "typically" in my sentence was important: the number of Explorer crashes does vary significantly from one computer to another. Usage patterns also matter, so it's possibly you're just not doing anything to trigger any of its bugs. Some of the computers I administer see between ten and thirty users a day, so they have to deal with a variety of user habits. Of course, all these users are logged in to a limited user account, so they can't generally crash the whole system, but they do manage to crash Windows Explorer sometimes. They crash the web browser even more often, but web browsers are another topic for another thread.
> I have also not seen Vista or 7 BSOD
I've only seen a Seven BSOD once. I've seen Vista do it several times, but then, Vista's been out longer, so I'm not sure whether that's really a difference. (It _is_ a difference from XP, which was undeniably crashier than Vista and Seven. Windows 9x, of course, was crashy to roughly the same extent that the Pacific Ocean is wet, and Windows 3 was even worse.)
> unless it was a bad driver or bad hardware, > which is not the fault of Windows.
Agreed (err, perhaps unless it's a bad driver that _came with_ Windows, arguably).
Bad hardware can crash any operating system (or, at least, any OS that doesn't run on a large cluster with fully redundant hardware in absolutely every category). Depending on the hardware, a really robust OS can potentially delay the crash for a while (I once had a system keep running, albeit with some errors, for several weeks after the hard drive had completely died -- everything it was actively using was already in memory, which of course would not work for a desktop system, but this was pretty much a glorified router; logfile writes were failing, but everything else was working... for a while), but that only flies for as long as the hardware in question isn't needed immediately and isn't screwing up anything else. When the wrong hardware does the wrong thing, the system is going to crash, period.
Yes, but they reacting emotionally rather than thinking the issue through. Do they really think if the treasury doesn't issue $1 bills people will carry coins around all the time? Stop and think about that for just a moment.
We already have $1 coins. We've had them since, I don't know, a long time ago. Something like 0.01% of the population carries and spends them. Everyone else considers them as either something to collect or just totally pointless, because $1 bills are so much more convenient. Most vendors don't even have a spot in their cash drawers for $1 coins, because they're not *used* enough to warrant it.
People dump coins in jars at the first opportunity and carry bills not because the coins aren't worth money but because they are annoying to carry around all the time. Americans have preferred bills over coins of the same denomination since sometime in the nineteenth century. The dollar is worth WAY less now than it was then, but the fact that people prefer to carry a $1 bill over a $1 coin given the choice has not changed at all. If we had fifty-cent bills and dollar coins, people would carry the fifty-cent bills and leave the dollar coins in a jar.
In other words, if we don't have $1 bills, the vending machines will be jamming with fives and tens instead of ones, but the usage of bills versus coins will barely change at all.
Anyway, the bills I'd want to get rid of are the $10 and the $50. There are only four fives in a twenty and only five twenties in a hundred. The $10 and the $50 aren't needed. The same is also true of the dime and the fifty-cent piece.
There are arguments to be made in favor of FAT16 or even FAT32, but I think I'd go with FAT12, just because it's simpler. You don't need LFNs for web hosting, do you?
No, it's not reasonable to set that aside. It's all part of the user experience. To the end user, it doesn't matter *why* the system has problems. It only matters that it *does* have problems. If the user needs to reboot because a patch was installed because of a vulnerability in IE (which the user hasn't even used in six months), it's still true that the user needs to reboot and therefore is interrupted and annoyed.
> XP and Win7 do not generally suffer from rebooting,
You mean besides every single time you install *anything*, EVER?
> slowness,
I'll give you that one (assuming you turn off Aero, which is reasonable).
> or "crashiness".
I suppose that depends what you compare it to. It doesn't crash nearly as often as XP, which in turn didn't crash nearly as much as 9x/Me. So they are making improvements in this area. It also depends what you count. The worst offender (excluding buggy third-party software, which wouldn't be fair to include) is generally Windows Explorer, which crashes typically every couple of months -- but if you know what you're doing it can be easily killed off and restarted. Serious system-level crashes (BSOD, auto-reboot without warning, hard lockup) are significantly less common on Vista and Seven (in the absence of hardware problems or buggy drivers) than on previous Microsoft OSes. I've seen them, but they're much more infrequent than on XP.
I can't meaningfully compare to iOS or Android, since I don't have either of those OSes. I can say that Vista and Seven are crashier than Debian, for example, but ordinary folks who don't work in IT are generally unlikely to run Debian on a phone or tablet, so that's really neither here nor there in the context of this thread.
I think the main issue for Windows Phone is that people associate Windows with hassle. They're willing to put up with hassle on PCs, for a variety of reasons, not least because most people associate PCs with hassle anyway, so it seems moot.
Phones are a different matter. They have hassles of their own (e.g., the tiny buttons or, worse, touchscreen-only interface makes typing a real pain), but they generally don't have the *same* hassles typically associated with a PC.
Just for example, people don't want their phone to interrupt them in the middle of a call or text message to tell them that Adobe Reader (or Java, or Norton, or Flash, or whatever) needs to be updated for the eleventeenth time this week (why can't all of that happen silently in the background while the user does other things?) or that they have unused icons on their desktop (why does Windows even care?) or that they need to restart (why?) or some other warning irrelevant to whatever they are doing at the moment. People don't want to have to hit Next eight times in a row on their phone just to do one simple thing. Basically, people don't want their phone to behave like a PC. Not all of this is entirely Microsoft's fault, but it is nonetheless their problem to deal with, if they want to sell a phone OS.
> It is always painful coding an arbitrary maximum > "I don't believe you" value though.
Don't I know it.
Once when I was setting up a very small mail server once (think: about a dozen users), I looked at the "max number of recipients per message" variable, thought about how many distinct recipients it might potentially be reasonable to have for a single (non-mailing-list) message, and set the value to five. It was overkill, of course...
Those of you who have ever run a mail server know what happened next. Within a day of deploying the thing I'd raised the limit at least three times. At that point the new limit lasted for a few months, but then...
What, seriously? You really have a legitimate work-related need to send exactly the same message word-for-word to more than fifty addresses at once, for real?
Yeah. More than fifty individually hand-typed addresses, even.
You can. There's an extension called NoScript. Page load times are very drastically reduced. In extreme cases, page load time can be reduced by a factor of a thousand or more. In the typical case, it's more like a factor of 20-50. (For some well-designed pages it barely makes any difference at all, but those aren't the pages that are giving you a problem.)
What is really surprising is how many sites continue to work just fine without all their precious awesome Javascript. It's somewhere in the 95% range. For the others, you do have the option to make an exception (either temporarily or permanently) and allow their scripts to run, if you want. In most cases it's not worth it: you can just find the same content on a different site.
There are really only two improvements I want to Javascript:
First, and most important, clicking the stop button should IMMEDIATELY cause all scripts in the tab, as well as all other loading activity (including automatic refresh), to desist, be canceled, and NOT start up again (unless the user takes action specifically to reload the page). In other words, stop should mean stop and I should NOT have to click it over and over again on the same page. The stop button should Actually Work. Right away. At once. Without delay. The first time. Every time. I've wanted this since Navigator 3, IIRC. (Oh, and on a related note, I shouldn't have to put a third toolbar button between stop and reload just to prevent the stop button from turning into the reload button when clicked. Those are completely opposite buttons. They should never change into one another, EVER. Talk about user-hostile. Whoever implemented that conflation should have his commit privileges revoked forever, no matter how many other contributions he's made.)
Second, I would really like the ability to allow scripts to run for some fixed (but user-configurable) amount of time, and after the designated amount of time they are suspended indefinitely and do not run any more (unless I make an exception for that site). If they complete during the designated amount of time, fine, but they do NOT automatically run again when a time runs out (again, unless I make an exception for that site). This could be an improvement to the NoScript extension. I think I'd set it to about a tenth of a second of CPU time per script up to a maximum of maybe two wallclock seconds per page. That's plenty of time for virtually all *legitimate* scripts and a great many superfluous ones besides, and it would barely have an impact on perf at all (as compared to not running the scripts, which is what I'm doing now).
A mean rise of 3.2mm per year times two decades, at 10 years per decade (assuming arithmetic mean -- if it was geometric mean the summary should say so), comes out to more than 25 inches. That's more than two feet. That's a significant claim, and one we should be able to easily verify.
If sea level had risen by more than two feet, a lot of coastlines should have moved. Significantly. We should be able to see this change on *maps*. Can you name one place in the world where this has happened, where it cannot be easily and obviously explained via more traditional mechanisms (river delta deposits and so on)?
I am not aware of any such example. I'm going to try to keep an open mind here, global warming enthusiasts: feel free to chime in with *specific*, *verifiable* examples of particular geographical locations where a coastline has shifted because of a sea level rise in excess of two feet. If the claim were true, I'd expect it to be quite easy to come up with examples from several different continents. Feel free to list as many examples as you can.
> its properties starting to resemble the deployed IPv4 network
Oh, so you mean there are a number of popular services, which are so well known as to be basically household names, that are only available via this protocol and no other, right? And the number of people who use the network both at work and at home rivals that of any other network, including the phone network? Right?
No?
Because that is what it would mean for the deployed IPv6 network to resemble the deployed IPv4 network.
That's the main number one key thing (related to software) that organizations are willing to pay money for and keep paying year after year. It's pretty hard to sell a support contract that doesn't include phone support.
Now, when they call, you can use other tools (ssh, VNC, whatever) in the process of providing support, and customers don't mind that at all, generally speaking. (This is assuming what you're providing is a whole package, so you configured the server before shipping it to them.) But they want to call somebody on the phone the moment they have a problem. That's what they pay you for. If they don't have somebody to call, they aren't getting good support, as far as they're concerned.
And they DO NOT want to be put on hold.
> Live chat is available for paying customers.
Yeah, a few customers might go for that, possibly. Add phone support, and you can multiply your revenue by at least five.
> What are you doing during these phone calls > that could not be done better with email?
You know that and I know that, but not everyone sees it that way. When the customer is frustrated with the computer because it WON'T WORK RIGHT and they just want somebody to FIX IT before they throw it OUT THE WINDOW, the absolute last thing they want to do is try to use the computer to get support. In technical terms, the computer may be working just fine, and the email or live chat would work just fine if they calmed down and used it, but the frustrated customer is usually not thinking rationally at that point. (If they were thinking rationally, nine times out of ten, they wouldn't NEED support.)
The major exception here is if you're supporting IT professionals. In that case, email support is fine, as long as you respond to it promptly[1], and you can ask them to do things like attach a screenshot and put the version number they're using in the subject line. IT professionals can handle that sort of thing. They may even be able to provide you with a list of steps to reproduce the problem.
But if your users are primarily in some other industry (doctors, nurses, librarians, school teachers, lawyers, retailers, etc. -- pretty much anything outside IT) you have to make some allowances, and that generally means phone support.
Footnote: [1] The precise definition of "promptly" is of course dependent on exactly how much money is changing hands in connection with the support contract. For a couple hundred bucks a year, you don't expect same-day responses every time. At ten grand a year, on the other hand, you don't want to be kept waiting very many minutes.
Most Americans perceive vanilla as plain, because as a rule vanilla products here have exactly the same amount of vanilla in them as other products. A white cake may have a teaspoon of vanilla in it. A chocolate cake will have a teaspoon of vanilla in it plus a quarter cup or more of cocoa powder. A carrot cake will also have a teaspoon of vanilla, plus carrot shavings and an assortment of spices. Similarly, vanilla icing has the same amount of vanilla in it as chocolate icing or strawberry icing or mint icing or lemon icing. Vanilla ice cream has the same amount of vanilla in it as chocolate or fudge ripple or strawberry swirl or rocky road or moose tracks.
There is a flavor of ice cream called "french vanilla", which has several times as much vanilla in it as other flavors. However, it also has other differences (notably, more heavy cream in lieu of milk), so on the whole it's not a very good reflection of actual vanilla flavor. Also, cheap brands of "french vanilla" may not actually have that much vanilla in them.
If you take a recipe for vanilla something-or-another and substitute additional vanilla extract in lieu of some other liquid (ideal would be to replace water, because in that case the change won't have much impact on texture; but most recipes don't call for water, so you may have to adapt), you can find out what vanilla actually tastes like. But most Americans have never done this.
I can see how odors might be described as having texture, in the loose sense of the word "texture". The timbre of sounds could also be described this way. A "smooth" sound might have relatively little variation in pitch from microsecond to microsecond and relatively few or relatively harmonically compatible frequencies at any given instant. A "rough" sound in contrast might have rapidly shifting and/or dissonant overtones and undertones. Something similar (albeit probably not time-based) could be said of smells: the different components might go well together or not, might be balanced well against one another or not, and so on.
However, if you think odors have color, I'm pretty sure you're experiencing something that isn't inherent in the odor itself. You might have a mild case of synesthesia in addition to your strong sense of smell (or else an over-active imagination).
The other poster was assuming that the individual olfactory components of an individual person's smell (as distinct from other humans), on a per-compound basis, change from day to day.
Admittedly, this is not common knowledge, so it would have been nice to see a citation -- but this is Slashdot, so sometimes we have to make some allowances.
Slashdot Mirror
You can, of course, deduce things about a person's linguistic background by noticing what kinds of mistakes they make. Some examples...
* Consistently confusing certain phonemes with one another even though they in fact sound nothing alike at all is a strong indicator that the native language doesn't contain either of them. Japanese people, for example, have a terrible time with English short vowels, particularly the a in "bad" and the u in "bus". This is because Japanese does not have these sounds. (Technically, the a in "bad" does occasionally occur in spoken Japanese, but it's a minority allophone in free variation for the a in "father". The u in "bus" does not occur at all.) Hence, "Pizza Hat". English speakers have a horrible time with glottal and pharyngeal stops, for the same reason.
* Consistently substituting certain words for certain other words in certain contexts where they don't make sense can mean that the native language has a word that is translated both ways. For example, using the word "teach" in place of "tell" ("Please, don't keep me in suspense. Teach me what happened!") is characteristic of Japanese. Saying "walk" instead of "live" (particularly when referring to living one's life a certain way) is characteristic of Greek (or of spending a lot of time reading old, sub-optimal English translations of things originally written in Greek; notably, the New Testament). English speakers often mix up ser and estar (in Spanish, and the corresponding verbs in other languages) and have trouble remembering which Japanese verbs to use to talk about wearing various kinds of clothing.
* Consistently confusing pairs of otherwise-related phonemes distinguished by the same type of contrast is a very strong indicator that the person's native language doesn't contrast on that feature. Koreans, for instance, have difficulty with the voiced/unvoiced distinction. English speakers have a similarly hard time hearing the distinction between aspirated and unaspirated.
* Having trouble with grammatical gender -- especially if it's inconsistent, not always getting the same words wrong in the same way, but different mistakes each time -- is an indication that the native language probably does not have grammatical gender. This doesn't apply much to English, since English has very little gender left in it.
* Attempting to pluralize adjectives when modifying plural nouns is a very strong indication that the person's native language does that. (This is characteristic of most European languages, with English being a notable exception.)
* If the target language is tonal, a native speaker can usually tell whether another speaker's native language is also tonal or not. This doesn't apply to English, obviously.
I'm sure there are others, but that's enough to illustrate the concept.
Preposition mistakes generally mean nothing, especially selecting the wrong preposition to introduce a prepositional phrase. Everyone makes them, and preposition usage varies so much even among dialects within a language that you can't usually draw firm conclusions. Frequent failure to correctly match up verbs with the complementary prepositions they prefer is a strong indication of a non-native speaker, but even this doesn't usually tell you very much about which other language they're coming from.
> The use of the awkward word "stuffs" has
> been, in my experience, a strong (actually,
> perfect) indicator that the speaker is Chinese.
Not Chinese specifically, just foreign in general. Foreign-language dictionaries and textbooks frequently don't bother to point out that certain words are mass nouns, and even if they do foreign-language learners don't always take note of it and mark it on their vocab cards, and even if they do they don't always remember.
Mass nouns that have a physical reason to be mass nouns (e.g., "water", "wood", "metal") are less of a problem, but the ones that refer collectively to obviously discrete things are a frequent source of trouble. "Staff" and "stuff" are among the ones that come up particularly often, but English has hundreds of mass nouns (some of which are pretty much always mass nouns, and others are only mass nouns when used with certain meanings; there are also some that can just be used either way). Native speakers don't usually think about these words as being significantly different from other nouns, but grammatically they really are. (Hang out with foreign language learners enough, and you will start to notice all sorts of interesting things about English. It's enlightening, actually.)
Admittedly, English mass nouns are particularly troublesome for native speakers of languages that ordinarily don't have grammatical number at all (e.g., Japanese), because they have had to get into a habit of forcing themselves, every single time they write a noun, to stop and think about whether it's just one or more than one (otherwise they end up writing things like "My two younger brother both goes to college."), something that comes natural to native speakers of languages that have a singular/plural distinction, and so keeping track of this can become an added distraction that makes it easier to forget about mass nouns. But native speakers of other languages (e.g., European languages, virtually all of which inflect nouns for case, number, and gender) do also make the mistake of pluralizing mass nouns inappropriately. It's not quite as frequent for them, but it's far from rare. It happens because people who have not yet achieved full fluency tend to think in their native language and then translate by components, adjusting the word order as necessary, translating the words into the other language, and so forth, whatever is needed to go from the one language to the other. If the English word that's a mass noun is a translation for a word in the person's native language that's not special in that way, this can easily result in an extraneous pluralization.
(I hang out on Lang-8, so I have had some exposure to English written by native speakers of various other languages.)
> Doing "not so well" doesn't mean that they can't do well enough.
I don't think you've grasped the magnitude of exactly how much atmosphere Mars doesn't have. The surface pressure is zero point something kPa. That's REALLY low pressure. Water at pressures that low is a vapor regardless of temperature. It can't solidify (on its own), let alone exist as a liquid. A standard college physics classroom vacuum pump can't create a vacuum with pressure that low.
For comparison, standard pressure on Earth at sea level is 101.325 kPa. On top of Mt. Everest, the pressure is about a third of standard, which is still somewhere around 50 or 100 times as much pressure as on Mars.
> I wouldn't rule out altogether an engineered species tailored for Mars.
To function in the near-vacuum of Mars, it would need to be engineered to be significantly different from anything found on Earth, except perhaps a small handful of anaerobic microbes, and I'm not even sure about those. Even most Earth life that can *survive* pressure that low can't perform metabolic functions to any significant extent without some source of metabolic gasses -- they just wait until they get some air and then start back up again. The ones that don't need oxygen generally need something else, and on Mars they won't have it -- unless you keep them in a pressurized container of some sort.
> It's worth noting that there is a huge gap
> between the worst survivable environments
> of Earth and the best of Mars.
Yes, exactly. It's not even really comparable.
> Carbon dioxide partial pressure is actually
> higher for Mars than Earth
I can cut a thirty-inch super-extra-large pizza into six slices and give you just one slice -- only a sixth of the pizza. Or, if you prefer, I can make a pizza an eighth of an inch in diameter and use a microscope and precision laser to cut that eighth-inch-diameter pizza into twenty slices and give you nineteen of those slices -- almost the whole pizza. Which is more? This is admittedly a simplification. Partial pressure does also matter, up to a point -- but only up to a point. When you're comparing the atmosphere of Mars to Earth, any small advantage Mars might have in the partial pressure department is overwhelmingly dwarfed to nothingness by the spectacular disadvantage in absolute pressure.
I do not for one minute believe that Earth's lichens (let alone the plain old green algae that terraforming enthusiasts always think is magic) will grow on Mars (except in a pressurized container). There isn't any air. Plants don't grow unless they're immersed in air (or water or something).
> [Lichens] grow in the most inhospitable places on Earth.
All of which places have substantial quantities of atmosphere.
Mars does not.
So unless you set up a sealed, controlled environment (like a terrarium or a dome), lichens won't do so well on Mars. Neither will much of anything else, with the possible exception of some bacteria and maybe a couple of other esoteric microbes. Or robots.
More likely they've figured out how to substantially automate assembly so that there's very little labor involved per unit. (Labor is expensive in the first world. That's why so many labor-intensive products are made elsewhere and shipped in. On the other hand, power is relatively cheap in the first world, so if almost all of the work is being done by machines instead of people, locating manufacturing plants in the first world can make sense. You still need people to run the machines, of course, but the more automated your plant becomes the less labor you need per unit of product.)
Perhaps he has indeed been captured, but by someone *other* than legitimate law enforcement officers. For instance, maybe he is being held by the "enforcement" arm of Los Zetas. Perhaps they will hold him for ransom. Or perhaps they will accept a bribe from his enemies, to keep him or to eliminate him. Or perhaps they will hand him over to the USDA.
(Sorry. I think the conspiracy theory thread on that Korean unicorn story may have affected me in some way.)
Actually, the unicorn remains were _planted_ in Korea by the NKVD, back in the thirties, as part of a secret joint operation with the US Department of Agriculture. The operation was suspended during WWII and the Korean War, and then before they could get around to completing it the USDA had to pull out (because of the altered political situation in Korea), leaving the MVD to bungle it until the Sino-Soviet Split forced them to finally abandon the effort as well. Nobody trusted China to complete the project, so they were never informed.
The unicorns actually live in the Arctic, but the Russian, Scandinavian, Canadian, and American governments keep that covered up, because if it ever got out it would call the Reindeer Dogma into question.
I've given the notion quite a bit of thought, and I eventually concluded, somewhat reluctantly, that it's not practicable on a large scale, for fundamental reasons that have nothing to do with any specific design or protocol that might be attempted.
However, there are usually ways to get information in or out of a place that's disconnected from the internet for whatever reason. Ultimately, if *any* kind of communication in or out of the country is possible, somebody can use it indirectly to get information to or from the internet.
Communication mechanisms that would work for this include, but are not limited to, the following: traditional phone lines; cables of any kind that cross the border, even if they were intended to carry power; satellite phones, or any other kind of bi-directional satellite link; radio transceivers, short-wave or otherwise; any postal service that delivers international mail, whether it's private or public; any location where it's possible for a person to cross the border, legally or otherwise; carrier pigeon; heliography; smoke signals; slips of paper in sealed glass or plastic bottles floating across the Mediterranean Sea; slips of paper in metal containers hurled across the border with trebuchets; words spelled out in fields using rocks and sticks, viewable by anybody with satellites or planes; messages can be dropped into the country from planes, either in robust containers or with parachutes; I'm sure there are other ways not listed here.
None of this requires a networking protocol in any normal sense of that term. The message only has to get through to somebody who is willing to pass it along, and they can do so. Outbound messages, once picked up by anyone, can be spread via the regular internet, no problem. Is there one person anywhere outside Syria who's willing to post the stuff on Twitter? Similarly, inbound messages, once received within the country, can be communicated via whatever mechanisms people are using to communicate within the country, even if they can't reach the outside that way.
Traditionally, one really common way to get information _into_ a geopolitical area where the local government is uncooperative is via radio, because the hard part of radio is the transmitter, and you can use one that's well outside the jurisdiction in question and transmit across the border. People on the inside just need cheap battery-operated transistor radios, which are remarkably difficult for the government to hunt down and eliminate. Jamming is of limited practicality, especially if the sending station is fairly powerful: you can jam it in a small area, but trying to jam it nationwide, if it's a strong signal, is very difficult. VOA is one example of an organization that has been doing this for decades.
The harder part is getting information _out_ of the place. Traditional radio transmitters give away their position, so ones inside the country are relatively easy for the government to shut down (if they have effective control of the whole country, which admittedly may not currently be the case in Syria). However, it's possible to narrow-cast radio waves directionally, aiming for a particular known receiver, which I think may be approximately what satphones do. That would be harder to stop. In the case of Syria, though, I'm pretty sure both sides in this conflict have vehicular transport, so the easiest thing may just be to send couriers across the border (e.g., into Turkey or Jordan or maybe even Lebanon, which I think is not real far from Damascus) and send the message from there.
Just in case someone feels like being an idiot and taking my little allusion for more than the small bit of attempted humor that it is, I should probably clarify: I am not even remotely interested in taking sides in the current conflict in Syria. Please don't try to read anything political into my above post. It wasn't intended that way. It's just supposed to be silly and funny, that's all. Thanks.
Syria has always been dead. In no public or private utterance can it be admitted that situation has at any time been otherwise. Officially the change of status has never happened. Netcraft has confirmed that Syria is dead: therefore, Netcraft has always confirmed that Syria is dead, and thus Syria has always been dead. The failing third-world government of the moment always represents absolute evil, and thus it follows that any past or future situation in which that government might not be failing is impossible.
> Because "OR" uses more characters than "~1".
.HTML, were another possible cause, although most Windows users didn't know the four-letter versions of .HTM and .JPG, so that was less common. Also, I believe some 8-bit and maybe even 16-bit software used old API calls or "interrupts" that sometimes resulted in LFNs being created for no really obvious reason. Oh, and I think copying or renaming a file in Windows Explorer sometimes caused it as well, though I'm not certain about that last one.)
Never used Windows 95 much, didja?
(8-character filenames regularly got mangled like that, for a variety of reasons. The precise details are a little fuzzy in my mind, since it has been a few years, but IIRC one possible cause was if the user gave the name in mixed case when saving the file. I think long extensions, such as
> FAT12 does support long file names, you know...
Oh, huh.
No, for some reason I was not aware of that, or perhaps I knew at one time but had forgotten.
> What was the last version of Windows you used? Windows 95?
Define "used".
On *my* computer, yeah, it's been a few versions.
However, I'm a network administrator, so I'm intimately familiar with the day-to-day behavior of versions up through Seven SP1. Admittedly, we haven't deployed Eight yet.
> You don't even have to reboot for driver installs.
It wouldn't matter very much if you had to reboot for a driver install, because that only happens when you first install Windows or when you add new hardware, neither of which is real often.
However, you frequently have to reboot Windows for application updates (antivirus software and browser plugins are particularly bad about this, and they're both things that have to be updated ANNOYINGLY often), and you virtually always have to reboot for Windows updates (which admittedly only happen once or occasionally twice a month).
> Aero offloads the UI to the GPU making the system faster, not slower.
That statement is very much at odds with my experience. I suppose it may depend on your graphics hardware. Almost all the systems I've worked with have onboard graphics, FWIW. But I'm not sure how you could get much faster than the Classic GUI; it is generally, as far as the user can notice without benchmarking tools, instantaneous. Aero is... not instantaneous. Also, it's at least twice as distracting as that horrible plastic-looking theme they tried to foist on us in XP.
> I have never seen explorer crash in either Vista or 7.
How many Windows computers do you administer? The word "typically" in my sentence was important: the number of Explorer crashes does vary significantly from one computer to another. Usage patterns also matter, so it's possibly you're just not doing anything to trigger any of its bugs. Some of the computers I administer see between ten and thirty users a day, so they have to deal with a variety of user habits. Of course, all these users are logged in to a limited user account, so they can't generally crash the whole system, but they do manage to crash Windows Explorer sometimes. They crash the web browser even more often, but web browsers are another topic for another thread.
> I have also not seen Vista or 7 BSOD
I've only seen a Seven BSOD once. I've seen Vista do it several times, but then, Vista's been out longer, so I'm not sure whether that's really a difference. (It _is_ a difference from XP, which was undeniably crashier than Vista and Seven. Windows 9x, of course, was crashy to roughly the same extent that the Pacific Ocean is wet, and Windows 3 was even worse.)
> unless it was a bad driver or bad hardware,
> which is not the fault of Windows.
Agreed (err, perhaps unless it's a bad driver that _came with_ Windows, arguably).
Bad hardware can crash any operating system (or, at least, any OS that doesn't run on a large cluster with fully redundant hardware in absolutely every category). Depending on the hardware, a really robust OS can potentially delay the crash for a while (I once had a system keep running, albeit with some errors, for several weeks after the hard drive had completely died -- everything it was actively using was already in memory, which of course would not work for a desktop system, but this was pretty much a glorified router; logfile writes were failing, but everything else was working... for a while), but that only flies for as long as the hardware in question isn't needed immediately and isn't screwing up anything else. When the wrong hardware does the wrong thing, the system is going to crash, period.
Ah, okay, yeah, my bad. 2.5 inches is *significantly* more plausible than 25 inches.
Yes, but they reacting emotionally rather than thinking the issue through. Do they really think if the treasury doesn't issue $1 bills people will carry coins around all the time? Stop and think about that for just a moment.
We already have $1 coins. We've had them since, I don't know, a long time ago. Something like 0.01% of the population carries and spends them. Everyone else considers them as either something to collect or just totally pointless, because $1 bills are so much more convenient. Most vendors don't even have a spot in their cash drawers for $1 coins, because they're not *used* enough to warrant it.
People dump coins in jars at the first opportunity and carry bills not because the coins aren't worth money but because they are annoying to carry around all the time. Americans have preferred bills over coins of the same denomination since sometime in the nineteenth century. The dollar is worth WAY less now than it was then, but the fact that people prefer to carry a $1 bill over a $1 coin given the choice has not changed at all. If we had fifty-cent bills and dollar coins, people would carry the fifty-cent bills and leave the dollar coins in a jar.
In other words, if we don't have $1 bills, the vending machines will be jamming with fives and tens instead of ones, but the usage of bills versus coins will barely change at all.
Anyway, the bills I'd want to get rid of are the $10 and the $50. There are only four fives in a twenty and only five twenties in a hundred. The $10 and the $50 aren't needed. The same is also true of the dime and the fifty-cent piece.
There are arguments to be made in favor of FAT16 or even FAT32, but I think I'd go with FAT12, just because it's simpler. You don't need LFNs for web hosting, do you?
> Vulnerabilities aside,
No, it's not reasonable to set that aside. It's all part of the user experience. To the end user, it doesn't matter *why* the system has problems. It only matters that it *does* have problems. If the user needs to reboot because a patch was installed because of a vulnerability in IE (which the user hasn't even used in six months), it's still true that the user needs to reboot and therefore is interrupted and annoyed.
> XP and Win7 do not generally suffer from rebooting,
You mean besides every single time you install *anything*, EVER?
> slowness,
I'll give you that one (assuming you turn off Aero, which is reasonable).
> or "crashiness".
I suppose that depends what you compare it to. It doesn't crash nearly as often as XP, which in turn didn't crash nearly as much as 9x/Me. So they are making improvements in this area. It also depends what you count. The worst offender (excluding buggy third-party software, which wouldn't be fair to include) is generally Windows Explorer, which crashes typically every couple of months -- but if you know what you're doing it can be easily killed off and restarted. Serious system-level crashes (BSOD, auto-reboot without warning, hard lockup) are significantly less common on Vista and Seven (in the absence of hardware problems or buggy drivers) than on previous Microsoft OSes. I've seen them, but they're much more infrequent than on XP.
I can't meaningfully compare to iOS or Android, since I don't have either of those OSes. I can say that Vista and Seven are crashier than Debian, for example, but ordinary folks who don't work in IT are generally unlikely to run Debian on a phone or tablet, so that's really neither here nor there in the context of this thread.
I think the main issue for Windows Phone is that people associate Windows with hassle. They're willing to put up with hassle on PCs, for a variety of reasons, not least because most people associate PCs with hassle anyway, so it seems moot.
Phones are a different matter. They have hassles of their own (e.g., the tiny buttons or, worse, touchscreen-only interface makes typing a real pain), but they generally don't have the *same* hassles typically associated with a PC.
Just for example, people don't want their phone to interrupt them in the middle of a call or text message to tell them that Adobe Reader (or Java, or Norton, or Flash, or whatever) needs to be updated for the eleventeenth time this week (why can't all of that happen silently in the background while the user does other things?) or that they have unused icons on their desktop (why does Windows even care?) or that they need to restart (why?) or some other warning irrelevant to whatever they are doing at the moment. People don't want to have to hit Next eight times in a row on their phone just to do one simple thing. Basically, people don't want their phone to behave like a PC. Not all of this is entirely Microsoft's fault, but it is nonetheless their problem to deal with, if they want to sell a phone OS.
> It is always painful coding an arbitrary maximum
> "I don't believe you" value though.
Don't I know it.
Once when I was setting up a very small mail server once (think: about a dozen users), I looked at the "max number of recipients per message" variable, thought about how many distinct recipients it might potentially be reasonable to have for a single (non-mailing-list) message, and set the value to five. It was overkill, of course...
Those of you who have ever run a mail server know what happened next. Within a day of deploying the thing I'd raised the limit at least three times. At that point the new limit lasted for a few months, but then...
What, seriously? You really have a legitimate work-related need to send exactly the same message word-for-word to more than fifty addresses at once, for real?
Yeah. More than fifty individually hand-typed addresses, even.
You can. There's an extension called NoScript. Page load times are very drastically reduced. In extreme cases, page load time can be reduced by a factor of a thousand or more. In the typical case, it's more like a factor of 20-50. (For some well-designed pages it barely makes any difference at all, but those aren't the pages that are giving you a problem.)
What is really surprising is how many sites continue to work just fine without all their precious awesome Javascript. It's somewhere in the 95% range. For the others, you do have the option to make an exception (either temporarily or permanently) and allow their scripts to run, if you want. In most cases it's not worth it: you can just find the same content on a different site.
There are really only two improvements I want to Javascript:
First, and most important, clicking the stop button should IMMEDIATELY cause all scripts in the tab, as well as all other loading activity (including automatic refresh), to desist, be canceled, and NOT start up again (unless the user takes action specifically to reload the page). In other words, stop should mean stop and I should NOT have to click it over and over again on the same page. The stop button should Actually Work. Right away. At once. Without delay. The first time. Every time. I've wanted this since Navigator 3, IIRC. (Oh, and on a related note, I shouldn't have to put a third toolbar button between stop and reload just to prevent the stop button from turning into the reload button when clicked. Those are completely opposite buttons. They should never change into one another, EVER. Talk about user-hostile. Whoever implemented that conflation should have his commit privileges revoked forever, no matter how many other contributions he's made.)
Second, I would really like the ability to allow scripts to run for some fixed (but user-configurable) amount of time, and after the designated amount of time they are suspended indefinitely and do not run any more (unless I make an exception for that site). If they complete during the designated amount of time, fine, but they do NOT automatically run again when a time runs out (again, unless I make an exception for that site). This could be an improvement to the NoScript extension. I think I'd set it to about a tenth of a second of CPU time per script up to a maximum of maybe two wallclock seconds per page. That's plenty of time for virtually all *legitimate* scripts and a great many superfluous ones besides, and it would barely have an impact on perf at all (as compared to not running the scripts, which is what I'm doing now).
A mean rise of 3.2mm per year times two decades, at 10 years per decade (assuming arithmetic mean -- if it was geometric mean the summary should say so), comes out to more than 25 inches. That's more than two feet. That's a significant claim, and one we should be able to easily verify.
If sea level had risen by more than two feet, a lot of coastlines should have moved. Significantly. We should be able to see this change on *maps*. Can you name one place in the world where this has happened, where it cannot be easily and obviously explained via more traditional mechanisms (river delta deposits and so on)?
I am not aware of any such example. I'm going to try to keep an open mind here, global warming enthusiasts: feel free to chime in with *specific*, *verifiable* examples of particular geographical locations where a coastline has shifted because of a sea level rise in excess of two feet. If the claim were true, I'd expect it to be quite easy to come up with examples from several different continents. Feel free to list as many examples as you can.
> its properties starting to resemble the deployed IPv4 network
Oh, so you mean there are a number of popular services, which are so well known as to be basically household names, that are only available via this protocol and no other, right? And the number of people who use the network both at work and at home rivals that of any other network, including the phone network? Right?
No?
Because that is what it would mean for the deployed IPv6 network to resemble the deployed IPv4 network.
Holding my breath I am not.
> Why do you offer phone support at all?
That's the main number one key thing (related to software) that organizations are willing to pay money for and keep paying year after year. It's pretty hard to sell a support contract that doesn't include phone support.
Now, when they call, you can use other tools (ssh, VNC, whatever) in the process of providing support, and customers don't mind that at all, generally speaking. (This is assuming what you're providing is a whole package, so you configured the server before shipping it to them.) But they want to call somebody on the phone the moment they have a problem. That's what they pay you for. If they don't have somebody to call, they aren't getting good support, as far as they're concerned.
And they DO NOT want to be put on hold.
> Live chat is available for paying customers.
Yeah, a few customers might go for that, possibly. Add phone support, and you can multiply your revenue by at least five.
> What are you doing during these phone calls
> that could not be done better with email?
You know that and I know that, but not everyone sees it that way. When the customer is frustrated with the computer because it WON'T WORK RIGHT and they just want somebody to FIX IT before they throw it OUT THE WINDOW, the absolute last thing they want to do is try to use the computer to get support. In technical terms, the computer may be working just fine, and the email or live chat would work just fine if they calmed down and used it, but the frustrated customer is usually not thinking rationally at that point. (If they were thinking rationally, nine times out of ten, they wouldn't NEED support.)
The major exception here is if you're supporting IT professionals. In that case, email support is fine, as long as you respond to it promptly[1], and you can ask them to do things like attach a screenshot and put the version number they're using in the subject line. IT professionals can handle that sort of thing. They may even be able to provide you with a list of steps to reproduce the problem.
But if your users are primarily in some other industry (doctors, nurses, librarians, school teachers, lawyers, retailers, etc. -- pretty much anything outside IT) you have to make some allowances, and that generally means phone support.
Footnote:
[1] The precise definition of "promptly" is of course dependent on exactly how much money is changing hands in connection with the support contract. For a couple hundred bucks a year, you don't expect same-day responses every time. At ten grand a year, on the other hand, you don't want to be kept waiting very many minutes.
Most Americans perceive vanilla as plain, because as a rule vanilla products here have exactly the same amount of vanilla in them as other products. A white cake may have a teaspoon of vanilla in it. A chocolate cake will have a teaspoon of vanilla in it plus a quarter cup or more of cocoa powder. A carrot cake will also have a teaspoon of vanilla, plus carrot shavings and an assortment of spices. Similarly, vanilla icing has the same amount of vanilla in it as chocolate icing or strawberry icing or mint icing or lemon icing. Vanilla ice cream has the same amount of vanilla in it as chocolate or fudge ripple or strawberry swirl or rocky road or moose tracks.
There is a flavor of ice cream called "french vanilla", which has several times as much vanilla in it as other flavors. However, it also has other differences (notably, more heavy cream in lieu of milk), so on the whole it's not a very good reflection of actual vanilla flavor. Also, cheap brands of "french vanilla" may not actually have that much vanilla in them.
If you take a recipe for vanilla something-or-another and substitute additional vanilla extract in lieu of some other liquid (ideal would be to replace water, because in that case the change won't have much impact on texture; but most recipes don't call for water, so you may have to adapt), you can find out what vanilla actually tastes like. But most Americans have never done this.
I can see how odors might be described as having texture, in the loose sense of the word "texture". The timbre of sounds could also be described this way. A "smooth" sound might have relatively little variation in pitch from microsecond to microsecond and relatively few or relatively harmonically compatible frequencies at any given instant. A "rough" sound in contrast might have rapidly shifting and/or dissonant overtones and undertones. Something similar (albeit probably not time-based) could be said of smells: the different components might go well together or not, might be balanced well against one another or not, and so on.
However, if you think odors have color, I'm pretty sure you're experiencing something that isn't inherent in the odor itself. You might have a mild case of synesthesia in addition to your strong sense of smell (or else an over-active imagination).
The other poster was assuming that the individual olfactory components of an individual person's smell (as distinct from other humans), on a per-compound basis, change from day to day.
Admittedly, this is not common knowledge, so it would have been nice to see a citation -- but this is Slashdot, so sometimes we have to make some allowances.