Re:Definitely has uses but..
on
Oracle Linux?
·
· Score: 5, Insightful
I'd run Oracle on Oracle Linux instead of some other distro. (I wouldn't bother with Oracle for any database that didn't need its own server for disk bandwidth reasons; this applies to any server that runs a single service.)
The IT guy's main headache for a database server is going to be the interaction between the database and the OS. The issue is that the server is supposed to run best on a version of Red Hat with some weird extra things enabled. Red Hat doesn't entirely understand this stuff, because they don't use it for any other configurations. Oracle understands it (they wrote it), but they're not doing tech support for Red Hat. The OS is sufficiently different from a usual Linux box that the IT guy has no clue when things are breaking. When the company I was working for got one of these, it was further complicated because the hardware didn't come with anything set up, and came from a third vendor. So we got a machine from Dell, the OS from Red Hat, and the database program from Oracle, each shipped separately, and they couldn't be tested independantly.
I think it would make perfect sense for Oracle to distribute and support a Red Hat-derived Linux distribution exclusively for production servers. At least then there would be a vendor who would understand the thing.
People wouldn't really notice the difference if every time it started, the browser had a different element and animal. In fact, there's an extension to do just that (from when the browser was Phoenix^WFirebird^WFirefox). People are going to talk about the browser as Firefox in general regardless, and as long as the UI looks and acts as expected, people will recognize it as the same program.
As for (1), the binary packages from Mozilla work perfectly on all of the Linux systems I've tried any of them on, and having each user's home-directory install of Firefox update itself automatically works seemlessly (except that 1.5 doesn't keep your session across restarting, and you get a stupid release-notes page).
I hope if you actually do (2) that you make sure not to backport any security fixes from 2.0 to 1.5 or apply any 3rd-party bugfixes, because then you'd be violating the license on the Firefox logo and name.
Content providers can insist that Google be more proactive in making sure that their copyrighted work isn't available. But Google could respond by also not listing those providers as highly in searches. They have no legal obligation to not give preferential treatment in search results to people who don't sue them or complain too much (at least until they lose the inevitable anti-trust lawsuit, which will take ages).
Some news sources have asked that they not be quoted on Google News, too, I've heard. I think there were even some that I'd heard of and read at the time, but I can't remember any more what they were...
It seems like nobody likes any of the UI changes, which is entirely what I'd expect, because every change to a UI is a major hit to usability. This isn't to say that the new UI might not be more usable for new users. But experienced users will continue to try to use the UI the way that worked before, and it will cause problems for them. This is especially true if the improvement is in discoverability, because experienced users will only benefit in that, when the stupid computer refuses to work like it's supposed to, it's not quite as difficult to figure out what you have to do instead of the natural thing as it might be.
Of course, it's also good to offer improvements to the UI for users who decide to retrain themselves or for new users. But this should be done by adding configuration options (ideally with UI-driven configuration methods, like the Customize Toolbars dialog), and making the upgrade process configure these options based on what used to happen, not based on the current defaults. (Of course, if you're importing settings from a different program, set the options to match the default or configured behavior of that program, not the local defaults.) The ideal is that, when the user gets a new version of the program, everything looks the same as it did before, but new behavior is available when the user decides that it is desireable.
It doesn't matter what the US, South Korea, etc. actually care about. The North Korean government tells its people (and the world in general) that outsiders want to invade. The people accept the government in part because it prevents this invasion. (At least, the government seems to think this; whether the people actually buy it is a different matter.) So North Korea needs the deterant for domestic reasons. Furthermore, it needs the foreign attention for the same domestic reasons, so they have to elicit disapproving statements from the white house regularly while not going so far that anyone actually invades.
Working from home is a lot more plausible when you can do the actual work on the machine you have at home. I worked mostly from home for a while on a Java web app, and I got a lot more done at home than at the office, because the ergonomics were better, the physical environment less distracting, and the computer environment was pretty much the same. These days, I'm programming custom hardware, and it doesn't make sense to do it from home (and the office is a lot nicer than the other place was).
Re:This is the beginning of the end
on
The Next X Prize
·
· Score: 1
Your genes, Anonymous Coward, were sequenced by the Human Genome Project before anybody else's. But really, at $1m for 100 people at a prototype stage, this isn't hugely expensive (if anyone manages to make it cost-effective). Going into production from the prototype stage is commonly a factor of 100 drop in price, at which point, it's cheaper than going to the dentist.
If it were an option as to whether you wanted the search retained, I suspect that just about anybody who chose to have their data kept wouldn't care if the government got it, and the government wouldn't have anything useful to do with it anyway. (Of course, there would be exceptions, but they'd end up on the "dumb criminal" news.)
Although what would be a lot better, both in terms of reducing unwanted disclosure and in terms of giving useful feedback, is if you could decide whether to retain the search when you see the results. There are plenty of instances of people messing up their profiles when they make a search that turns up the wrong things, and then the site keeps suggesting things they don't want. (The same is true of purchases, particularly with gifts for people who aren't like the customer.)
You're looking at the virtues wrong. Laziness, for a businessperson, is passing off tasks to other people or departments as soon as they're sufficiently specified to end up with the desired result; somebody who isn't trying to do everything else will have an easier time doing it (for an engineer, this is a computer, but that aspect isn't important to the virtue). Impatience, for a businessperson, is making sure there aren't any obstructions to progress (such as people waiting for things that could have been done sooner, or, worse, waiting for things to get done that nobody's working on). Hubris, for a businessperson, is thinking that you can solve any problem that anyone can solve.
Nobody should be so set on a particular method that they ignore other methods that could be more applicable. Nor should anybody come up with clever solutions without a need for them. For everybody, it is more effective for interfaces to be simple and modular, but communicate in sufficient detail to get things done (the executives should be able to pass some task to "engineering", and have it get allocated internally to appropriate people, rather than having to pick appropriate individuals who work well together and have time available in their schedules).
I'd tell Intel that I didn't buy their hardware because there wasn't open source support for it, except that, in point of fact, I bought their hardware because there's open source support for it. Sure, it's a pain that Intel doesn't let distributions redistribute unmodified firmware, because it means that I needed a wired network connection to install Gentoo, but getting the ipw3945 to work was a matter of doing "emerge ipw3945" (there was more of a hassle getting the newer-than-mainline ieee80211 module built than anything to do with the firmware, which was downloaded automatically. I didn't even need to go to the web page myself and agree to anything). This was all substantially easier than using other wireless, where appearantly identical devices have entirely different and incompatible hardware inside, which claims to be exactly the same, except that the manufacturer's flaky Windows drivers detect the difference and do the appropriate totally different thing.
That isn't to say that I don't think that Intel should release specs (it's not like they're hiding anything, when they're releasing open-source drivers anyway; I think they justdon't have enough tech writers or something), or that I don't think that Intel should interact more with the open-source distributors of their drivers (it's a pain that they don't push for inclusion of their Linux wireless drivers in the kernel, do the necessary coding-style work, and use the regular kernel development methodology), or that I'm not impatient with Intel not yet having a stable i915 driver, or even that I think there shouldn't be BSD-coding-style drivers created for Intel hardware in addition to the Linux-coding-style ones. But really, I don't have any problem with Intel doing only as much as is actually necessary to give me a good experience with their hardware.
I like this arrangement, too. I'd point out a couple of additional things, though:
It helps if unrelated people are sound-isolated from you, either by distance or by walls. Having too many people you can overhear is distracting, and being able to overhear people who are talking about stuff that isn't relevant is very distracting. It also helps if everybody you can hear can see you (cubes are terrible this way), because people get social cues about how many people they're distracting.
Expect people to be away from the office during the heavy development time. There's generally a period after all of the issues are settled, before integration and testing, when there's just a mass of code that has to get written, and some people will be a lot faster at this if they're not distracted by coworkers. People may also work funny hours for this reason, or listen to music on headphones. Insist that people overlap and pay attention at some point regularly, but not for most of the time.
I have no clue what people other than developers need. But the developers shouldn't be able to hear or see them, wherever they are.
Ed T. Rush, NBA manager of officials, announced plans to buy YouTube along with a group of referees and sports writers, including Jim Gray, Sam Smith, and Chad Ford.
Actually, this article was submitted by Yoda. It's actually about a Beer Age in Myst, and it should be "Doctor Who Makes Guinness Records Book of World". Obviously.
I think the defect is such that it depends on how battery packs are made from the cells. Sony probably kept their packs in safe configurations, while everybody else using their cells was less conservative. But note that laptop manufacturers don't seem to have recalled all batteries with Sony-produced cells from the relevant time period, so it seems like some pack designs happen to be safe.
They have a slide that matches successive levels of application demand with: Text, Multimedia, Video & 3D, RMS.
Okay, so I understand that AI is more compute-intensive than video. And I understand that it could be easier (tera instead of peta) if social reasoning isn't included. But really, Intel, I just don't want RMS on my computer.
Also, the jump from nanoscale to terascale may be impressive, but I don't think it'd be useful to have a transistor with a 310-million-mile-wide gate. Your device isn't going to be useful if it doesn't fit inside earth's orbit.
Yeah, they always come out in the wrong font, and with the tab stops all screwed up. And there's a critical feature that never quite works correctly, if you know what I mean.
The next generation batteries will be safer, because it's relatively complicated to extract the energy from them. In order to get energy out of lithium polymer batteries, you need electrolyte and carefully arranged electrodes. This makes them a lot harder to make, but also means they're less dangerous, because, if something goes wrong, it tends to make the battery stop doing anything. Of course, the energy density means that if you manage to ignite it, it'll explode, but it doesn't spontaneously ignite.
The reason lithium ion batteries are particularly hazardous is that its internal resistance goes down when it gets hotter, which means that, if it gets shorted, it heats up rapidly and at an accelerating pace until it ignites and explodes. If you short a lithium polymer battery, it heats up until it fails, and then stops producing current and doesn't heat up further; you need an external source of energy to detonate them.
Essentially, the next generation of batteries is more explosive, but the current generation has a built-in detonator, which the next generation doesn't have. Of course, if you short a lithium polymer battery, it'll probably vaporize the thing that shorted it, heat up enough to burn you, and then melt down into toxic sludge. But it's not going to explode into a tower of flame.
They don't need to spend more time on QA (Q & A?). What they really need is some automated testing which keeps packages from going to the site. They do a pretty good job of fixing things; the fundamental issue is that problems aren't obvious until packages have been released and people complain. (E.g., people testing a package as to whether it should be marked as stable are using unstable packages, obviously, so they're unlikely to realize that it requires unstable packages, and they're likely to have other stuff that's under development available, too). The infrastructure should take care of consistancy-testing, and delay releases as appropriate, with the package maintainer able to leave the release queued to happen as soon as the dependancies are suitable.
"Star Trek" for it's time was cheap, and cheesy, and even silly - but they were serious about what they were trying to do: show a future of humanity where people of all races could work together. Sure, it had its flaws. Yes, the costumes and sets were, by today's and possibly even that time period, hokey.
So why can't a show feature both people of different races able work together and modern special effects? Is there some reason that a depiction of racial cooperation has to look dated and unrealistic?
It's not like the point of "Star Trek" is that the special effects are cheesy. Changing them is irrelevant to the actual content, and improves the appeal. It's not like they're suggesting turning Uhura into a white guy. Or, more in keeping with the original point, making Chekov an Arab who seduces the male aliens.
This seems to me to support the recent article protection system that the German edition is working on: allow anyone to make changes, but (for protected articles) require a logged-in user to make the changes visible to non-logged-in users. If the statistics are that most of the content is written by infrequent and unregistered authors, but that a group of 500 people touches everything before it gets very far, primarily doing copyediting, the new protection idea should be very effective. The expected result would be that the masses who generate the bulk of the content wouldn't notice a difference with respect to their own edits, since some registered user will integrate the changes before long anyway (i.e., before they're likely to be browsing in a new session), but they'd see better results for other people's content, because they don't see versions of contraversial articles before they're edited (under the assumption that the real issue with contraversial articles is that different people's versions are partially correct but biased and lack inconveniant details; between all of the versions, there exists the raw text for a good article).
You could always write a guide at some level, unless the game is so random as to be not worth playing. Even with a game that's generated on the fly, there must be a range of things the game knows to be legitimate clues that players will find satisfying, and that range can be written up. For example, there are extensive spoiler files for nethack, angband, and similar games, which explain useful strategies for dealing with the various random elements. In the murder-mystery game you describe, I suspect that the guide would list a variety of different possible forms of clues, and what each one tells you, and how trustworthy it is. Either there's some method which works for solving the mystery, and the guide can tell you that strategy, or there is no method, and the player will have some chance of being unable to solve the game.
The fundamental part of making a game interesting is to make determining the correct strategy or following it enjoyable. I think, actually, that the rise of strategy guides is due to the increase in arbitrary elements of games. A strategy guide is a lot less tempting if it is reasonable for players to figure out the tricks themselves, rather than having to guess something unlikely or try a lot of possibilities.
Slashdot Mirror
I'd run Oracle on Oracle Linux instead of some other distro. (I wouldn't bother with Oracle for any database that didn't need its own server for disk bandwidth reasons; this applies to any server that runs a single service.)
The IT guy's main headache for a database server is going to be the interaction between the database and the OS. The issue is that the server is supposed to run best on a version of Red Hat with some weird extra things enabled. Red Hat doesn't entirely understand this stuff, because they don't use it for any other configurations. Oracle understands it (they wrote it), but they're not doing tech support for Red Hat. The OS is sufficiently different from a usual Linux box that the IT guy has no clue when things are breaking. When the company I was working for got one of these, it was further complicated because the hardware didn't come with anything set up, and came from a third vendor. So we got a machine from Dell, the OS from Red Hat, and the database program from Oracle, each shipped separately, and they couldn't be tested independantly.
I think it would make perfect sense for Oracle to distribute and support a Red Hat-derived Linux distribution exclusively for production servers. At least then there would be a vendor who would understand the thing.
People wouldn't really notice the difference if every time it started, the browser had a different element and animal. In fact, there's an extension to do just that (from when the browser was Phoenix^WFirebird^WFirefox). People are going to talk about the browser as Firefox in general regardless, and as long as the UI looks and acts as expected, people will recognize it as the same program.
As for (1), the binary packages from Mozilla work perfectly on all of the Linux systems I've tried any of them on, and having each user's home-directory install of Firefox update itself automatically works seemlessly (except that 1.5 doesn't keep your session across restarting, and you get a stupid release-notes page).
I hope if you actually do (2) that you make sure not to backport any security fixes from 2.0 to 1.5 or apply any 3rd-party bugfixes, because then you'd be violating the license on the Firefox logo and name.
Content providers can insist that Google be more proactive in making sure that their copyrighted work isn't available. But Google could respond by also not listing those providers as highly in searches. They have no legal obligation to not give preferential treatment in search results to people who don't sue them or complain too much (at least until they lose the inevitable anti-trust lawsuit, which will take ages).
Some news sources have asked that they not be quoted on Google News, too, I've heard. I think there were even some that I'd heard of and read at the time, but I can't remember any more what they were...
It seems like nobody likes any of the UI changes, which is entirely what I'd expect, because every change to a UI is a major hit to usability. This isn't to say that the new UI might not be more usable for new users. But experienced users will continue to try to use the UI the way that worked before, and it will cause problems for them. This is especially true if the improvement is in discoverability, because experienced users will only benefit in that, when the stupid computer refuses to work like it's supposed to, it's not quite as difficult to figure out what you have to do instead of the natural thing as it might be.
Of course, it's also good to offer improvements to the UI for users who decide to retrain themselves or for new users. But this should be done by adding configuration options (ideally with UI-driven configuration methods, like the Customize Toolbars dialog), and making the upgrade process configure these options based on what used to happen, not based on the current defaults. (Of course, if you're importing settings from a different program, set the options to match the default or configured behavior of that program, not the local defaults.) The ideal is that, when the user gets a new version of the program, everything looks the same as it did before, but new behavior is available when the user decides that it is desireable.
It doesn't matter what the US, South Korea, etc. actually care about. The North Korean government tells its people (and the world in general) that outsiders want to invade. The people accept the government in part because it prevents this invasion. (At least, the government seems to think this; whether the people actually buy it is a different matter.) So North Korea needs the deterant for domestic reasons. Furthermore, it needs the foreign attention for the same domestic reasons, so they have to elicit disapproving statements from the white house regularly while not going so far that anyone actually invades.
Working from home is a lot more plausible when you can do the actual work on the machine you have at home. I worked mostly from home for a while on a Java web app, and I got a lot more done at home than at the office, because the ergonomics were better, the physical environment less distracting, and the computer environment was pretty much the same. These days, I'm programming custom hardware, and it doesn't make sense to do it from home (and the office is a lot nicer than the other place was).
Your genes, Anonymous Coward, were sequenced by the Human Genome Project before anybody else's. But really, at $1m for 100 people at a prototype stage, this isn't hugely expensive (if anyone manages to make it cost-effective). Going into production from the prototype stage is commonly a factor of 100 drop in price, at which point, it's cheaper than going to the dentist.
If it were an option as to whether you wanted the search retained, I suspect that just about anybody who chose to have their data kept wouldn't care if the government got it, and the government wouldn't have anything useful to do with it anyway. (Of course, there would be exceptions, but they'd end up on the "dumb criminal" news.)
Although what would be a lot better, both in terms of reducing unwanted disclosure and in terms of giving useful feedback, is if you could decide whether to retain the search when you see the results. There are plenty of instances of people messing up their profiles when they make a search that turns up the wrong things, and then the site keeps suggesting things they don't want. (The same is true of purchases, particularly with gifts for people who aren't like the customer.)
You're looking at the virtues wrong. Laziness, for a businessperson, is passing off tasks to other people or departments as soon as they're sufficiently specified to end up with the desired result; somebody who isn't trying to do everything else will have an easier time doing it (for an engineer, this is a computer, but that aspect isn't important to the virtue). Impatience, for a businessperson, is making sure there aren't any obstructions to progress (such as people waiting for things that could have been done sooner, or, worse, waiting for things to get done that nobody's working on). Hubris, for a businessperson, is thinking that you can solve any problem that anyone can solve.
Nobody should be so set on a particular method that they ignore other methods that could be more applicable. Nor should anybody come up with clever solutions without a need for them. For everybody, it is more effective for interfaces to be simple and modular, but communicate in sufficient detail to get things done (the executives should be able to pass some task to "engineering", and have it get allocated internally to appropriate people, rather than having to pick appropriate individuals who work well together and have time available in their schedules).
I'd tell Intel that I didn't buy their hardware because there wasn't open source support for it, except that, in point of fact, I bought their hardware because there's open source support for it. Sure, it's a pain that Intel doesn't let distributions redistribute unmodified firmware, because it means that I needed a wired network connection to install Gentoo, but getting the ipw3945 to work was a matter of doing "emerge ipw3945" (there was more of a hassle getting the newer-than-mainline ieee80211 module built than anything to do with the firmware, which was downloaded automatically. I didn't even need to go to the web page myself and agree to anything). This was all substantially easier than using other wireless, where appearantly identical devices have entirely different and incompatible hardware inside, which claims to be exactly the same, except that the manufacturer's flaky Windows drivers detect the difference and do the appropriate totally different thing.
That isn't to say that I don't think that Intel should release specs (it's not like they're hiding anything, when they're releasing open-source drivers anyway; I think they justdon't have enough tech writers or something), or that I don't think that Intel should interact more with the open-source distributors of their drivers (it's a pain that they don't push for inclusion of their Linux wireless drivers in the kernel, do the necessary coding-style work, and use the regular kernel development methodology), or that I'm not impatient with Intel not yet having a stable i915 driver, or even that I think there shouldn't be BSD-coding-style drivers created for Intel hardware in addition to the Linux-coding-style ones. But really, I don't have any problem with Intel doing only as much as is actually necessary to give me a good experience with their hardware.
I like this arrangement, too. I'd point out a couple of additional things, though:
It helps if unrelated people are sound-isolated from you, either by distance or by walls. Having too many people you can overhear is distracting, and being able to overhear people who are talking about stuff that isn't relevant is very distracting. It also helps if everybody you can hear can see you (cubes are terrible this way), because people get social cues about how many people they're distracting.
Expect people to be away from the office during the heavy development time. There's generally a period after all of the issues are settled, before integration and testing, when there's just a mass of code that has to get written, and some people will be a lot faster at this if they're not distracted by coworkers. People may also work funny hours for this reason, or listen to music on headphones. Insist that people overlap and pay attention at some point regularly, but not for most of the time.
I have no clue what people other than developers need. But the developers shouldn't be able to hear or see them, wherever they are.
Ed T. Rush, NBA manager of officials, announced plans to buy YouTube along with a group of referees and sports writers, including Jim Gray, Sam Smith, and Chad Ford.
Actually, this article was submitted by Yoda. It's actually about a Beer Age in Myst, and it should be "Doctor Who Makes Guinness Records Book of World". Obviously.
I think the defect is such that it depends on how battery packs are made from the cells. Sony probably kept their packs in safe configurations, while everybody else using their cells was less conservative. But note that laptop manufacturers don't seem to have recalled all batteries with Sony-produced cells from the relevant time period, so it seems like some pack designs happen to be safe.
They have a slide that matches successive levels of application demand with: Text, Multimedia, Video & 3D, RMS.
Okay, so I understand that AI is more compute-intensive than video. And I understand that it could be easier (tera instead of peta) if social reasoning isn't included. But really, Intel, I just don't want RMS on my computer.
Also, the jump from nanoscale to terascale may be impressive, but I don't think it'd be useful to have a transistor with a 310-million-mile-wide gate. Your device isn't going to be useful if it doesn't fit inside earth's orbit.
Yeah, they always come out in the wrong font, and with the tab stops all screwed up. And there's a critical feature that never quite works correctly, if you know what I mean.
That's why certain well-known analysts work for Cheerful Bull instead.
The next generation batteries will be safer, because it's relatively complicated to extract the energy from them. In order to get energy out of lithium polymer batteries, you need electrolyte and carefully arranged electrodes. This makes them a lot harder to make, but also means they're less dangerous, because, if something goes wrong, it tends to make the battery stop doing anything. Of course, the energy density means that if you manage to ignite it, it'll explode, but it doesn't spontaneously ignite.
The reason lithium ion batteries are particularly hazardous is that its internal resistance goes down when it gets hotter, which means that, if it gets shorted, it heats up rapidly and at an accelerating pace until it ignites and explodes. If you short a lithium polymer battery, it heats up until it fails, and then stops producing current and doesn't heat up further; you need an external source of energy to detonate them.
Essentially, the next generation of batteries is more explosive, but the current generation has a built-in detonator, which the next generation doesn't have. Of course, if you short a lithium polymer battery, it'll probably vaporize the thing that shorted it, heat up enough to burn you, and then melt down into toxic sludge. But it's not going to explode into a tower of flame.
They don't need to spend more time on QA (Q & A?). What they really need is some automated testing which keeps packages from going to the site. They do a pretty good job of fixing things; the fundamental issue is that problems aren't obvious until packages have been released and people complain. (E.g., people testing a package as to whether it should be marked as stable are using unstable packages, obviously, so they're unlikely to realize that it requires unstable packages, and they're likely to have other stuff that's under development available, too). The infrastructure should take care of consistancy-testing, and delay releases as appropriate, with the package maintainer able to leave the release queued to happen as soon as the dependancies are suitable.
"Star Trek" for it's time was cheap, and cheesy, and even silly - but they were serious about what they were trying to do: show a future of humanity where people of all races could work together. Sure, it had its flaws. Yes, the costumes and sets were, by today's and possibly even that time period, hokey.
So why can't a show feature both people of different races able work together and modern special effects? Is there some reason that a depiction of racial cooperation has to look dated and unrealistic?
It's not like the point of "Star Trek" is that the special effects are cheesy. Changing them is irrelevant to the actual content, and improves the appeal. It's not like they're suggesting turning Uhura into a white guy. Or, more in keeping with the original point, making Chekov an Arab who seduces the male aliens.
Only if it's actually yours...
This seems to me to support the recent article protection system that the German edition is working on: allow anyone to make changes, but (for protected articles) require a logged-in user to make the changes visible to non-logged-in users. If the statistics are that most of the content is written by infrequent and unregistered authors, but that a group of 500 people touches everything before it gets very far, primarily doing copyediting, the new protection idea should be very effective. The expected result would be that the masses who generate the bulk of the content wouldn't notice a difference with respect to their own edits, since some registered user will integrate the changes before long anyway (i.e., before they're likely to be browsing in a new session), but they'd see better results for other people's content, because they don't see versions of contraversial articles before they're edited (under the assumption that the real issue with contraversial articles is that different people's versions are partially correct but biased and lack inconveniant details; between all of the versions, there exists the raw text for a good article).
Remember where those $100 bills that Hezbollah is handing out come from. Hint: they do not originate in Iran.
Of course not. Everybody knows that $100 bills come from North Korea.
You could always write a guide at some level, unless the game is so random as to be not worth playing. Even with a game that's generated on the fly, there must be a range of things the game knows to be legitimate clues that players will find satisfying, and that range can be written up. For example, there are extensive spoiler files for nethack, angband, and similar games, which explain useful strategies for dealing with the various random elements. In the murder-mystery game you describe, I suspect that the guide would list a variety of different possible forms of clues, and what each one tells you, and how trustworthy it is. Either there's some method which works for solving the mystery, and the guide can tell you that strategy, or there is no method, and the player will have some chance of being unable to solve the game.
The fundamental part of making a game interesting is to make determining the correct strategy or following it enjoyable. I think, actually, that the rise of strategy guides is due to the increase in arbitrary elements of games. A strategy guide is a lot less tempting if it is reasonable for players to figure out the tricks themselves, rather than having to guess something unlikely or try a lot of possibilities.