Actually, the Mac Genius isn't really for fixing the macs. Several things:
1. Macs are much simpler to configure for a novice user. More importantly, they're more RELIABLE to configure - it's much more rare to get some psychotic configuration behavior in OSX than in Windows, OSX better explains the dialog boxes, and OSX is more likely to prompt you with an appropriate correction. OSX has a better mix of manual and automatic and a better transition between them.
This doesn't mean they never break, but it does mean that you don't need a Geek Squad as much with OSX - in my experience a novice user will need less hand-holding help using OSX than Windows. (Not none, but less)
2. Last I checked, when you take a broken mac to an Apple Store, they only do like kind replacements or simple upgrades of a few of the most common components. It's much more like they do "installs" than real repairs. There are places that do real Apple Service around (I worked at one) but if your problem is actually complex and you bring it to the Apple Store they will overnight your computer to Texas where it will be fixed and overnighted back. To me this is a fine system for making sure you get a well-trained technician for your hard problems.
3. The Genius's real job is to be a consumer consultant. Their job is to talk and be helpful, not fix. (I'm not saying they _can't_ fix - the ones I've dealt with have seemed very competent) They need to be able to answer very open ended questions like "I want to learn video editing. I have no clue. Take my credit card." or "My computer is slow. How do I make it faster, or do I need a new one" and explain to the person what they need in minute detail. Their job is to get them what they actually need. Arguably figuring out what the customer wants is the part that requires being a Genius.
In short the Genius's job is much more about being able to answer a customer question of "what should I do" than about touching Macs - or at least it was last I checked.
We use it, but we're not a terrific data point. We haven't moved to 10 yet at all. But here are some basics:
I'm an Apple fan, and in my opinion SuSE is the Apple of Linux. (Of course, Apple is itself the Apple of *nix-like OSes.) They are not the earliest adopters of new technology, but they do a good job of integrating it. But moreover they make it easy to use and administer.*
Big business wants a Vendor, not a community. In the giant world that's pretty much RH and SuSE, or IBM selling someone else's.
Novell also makes a good business selling networking solutions for you whole office, not making you put them together youself.
SuSE will happily ship with the best available drivers and software, even if those are proprietary. For some people this is a reason not to use SuSE - zealots have their place, and I would not want the strictly OSS distros to go away - but if you are more interested in Linux-as-a-current-tool than Linux-as-a-political-statement to force vendors to open drivers, this is the right choice philosophy for you.
*Let me define "easy to use and administer" more: YaST puts a nice front end on whatever you're doing (package management and basically all other system administration) - with enough power to configure whatever you want however you want and enough guidance that you can do it even if it's not your speciality and you've been awake too long. It's the perfect kind of system that LETS you be knowledgeable but does not REQUIRE you to be knowledgeable. You can seamlessly escalate simple point and click management to advanced point and click management to tweaking files by hand that it then won't screw up.
So to me, "easy to use" means that I can use, in each instance, a system that is as automatic or as manual as I want, based on how much expertise I have in that area, and how much time and attention I have for the issue "right now"
Configuration entirely by manually touching files/registries/whatever is a little like walking through a minefield... get too tired, make a typo and all sorts of stuff might explode, and you've making a large number of changes. But I'd take it any day over a Windows GUI-only system where IF it explodes and, say, doesn't boot you have a much harder time putting it back together than you do with a text editor. But YaST is the ideal hybrid - it reduces your chances of stupid mistakes without limiting your power. You edit what you want, let YaST edit what you don't. It's not by far the only piece of software to do this, but I think it's a good example.
( I think much of OS X is similar. Can you enter complex firewall and packet forwarding configurations in their little GUI? No. Does their GUI work for most people? Yep. Does their GUI still use the standard BSD firewall, which you can configure however you want? Yep. )
Parent is completely correct about Rent-A-Coder etc (and there are many more) We get a lot of work from such places, but it's a lot of work GETTING the work - more than it is doing the work, usually. This isn't the kind of investment for someone who wants to program for fun.
Regarding the gp comment - I want to work less and feel less pressure to make money, so that I have more time - to work on the programming projects that I WANT to work on. I do plenty of other things with my life, but programming is wonderful and powerful and expressive and I'd still like to do more of it than I do. Furthermore, programming is broad - you can go home and do very significantly different things than you do at work.
If OP (or someone else) is good at what they do and interested, email me (use a subject like Slashdot Resume) or reply to this with some way to contact you. We have a steady stream of these smaller projects, we have many that would be appropriate for you to do in the evenings and we'd take care of the non-programming management for you.
Some sibling post mentioned programmers trying to find someone to do this business side - this is a big part of what we do.
(Furthermore we have at least one and possibly more projects that are definitely targetted at making the world a better place, if that adds to your preference for working on those projects.)
Instant Runoff Voting or Ranked Choice Voting are the most usual US names for this - which is definitely the best long term solution, and which is being used in some local elections already.
I will happily grant you that both major candidates may suck in any given election and that you might well want to protest by not voting for either one. (I do not agree with your idealistic "sullying my hands" position - I think if one of those candidates is less bad to you you should vote for them, and I think in most real cases one candidate is less bad to you if you bother to check. But that's not my major point, so I'll assume they're exactly even for now.)
But the _biggest_ consistent problem we have which makes the two candidates both suck is that the two incumbent parties have a strangehold on who we get to choose from. Voting for a third party candidates drives up the visibility of third parties existing and drives up the likelihood that OTHER people will vote for third parties.
As a bonus, if enough people do it for a presidental campaign then they get federal election money.
If you're going to publicly pick on someone's phrases you should really offer the correct examples, too. Especially since you don't know if English is the poster's first language. You can guarantee that it is not the first language of some of the readers and that some readers are trying to improve their skills.
"run the gamut" - 'to cover a whole range of variations' "run the gauntlet" (probably not what the poster meant) - 'to suffer severe criticism or tribulation' with background in this criticism coming from many sources. This usage would actually be funny if the poster meant that they were punished by a wide variety of technologies.
As a web application developer, I have to say that both the parent's experience (albeit with different tech) and the OP's experience ring true, but I love it.
As fundamentally a person who likes to create things, I love web development because there is no other domain where I can spend as little time to make something complex actually work and be available to many people. The first is the domain of all software and the second is specific to web development. (Obviously software projects can be huge and complex, but a physical invention of similar complexity would be much MORE work, often by more people, to get accomplished) The only potential challenge to this crown is being a popular blogger or flashmob organizer and actually getting zillions of actual people to do what you ask. And that usually pays worse.
Fundamentally web programmer pay is pretty good, I think. But it's highly variable, so you need to work on the system to make sure you get paid appropriately. I think many tech people think they'll get hired on their merits, when what matters are personal contacts/reputation and, at best, the appearance of merit (portfolio/resume) These are different skillsets, but they are important to getting paid enough.
If you take a regular job and it doesn't pay enough, you need to look for another one. Or you need to get more skills at getting hired so you can.
I am self employed and love that part too - but you have to be good at making clients want to hire you and pay you - the hiring process over and over again - and you have to throw a wide enough net. Being self employed at anything involves building up a clientele, and that's a lot of work. I certainly went through a period of intermittently not enough employment. Realize that your volume and price are intertwined. Essentially there's a minimarket for your skills, so your volume will drop as your price rises. I generally think this means you should start with a fairly low price, get as much work as you can to build reputation, and as you get too much work raise your price and realize that some clients will no longer be able to afford you. Reliable repeat clients are often worth having even if you have to charge them less, because there's no overhead.
If you're good at what you do and still interested in freelancing, feel free to email me with what you know and make sure to put "Slashdot resume" in the subject.
You have to be good, and I still do not guarantee to pay well; there's a wide open market of developers out there. But we'll do the work of finding clients, selling clients and having a client-base, we'll add to your volume at the low end, and you'll still be free to take on other clients at the high end simultaneously and you can build up your own clientbase.
I believe in the long term quality of pressed DVDs - but not so much of burned ones, because insufficient evidence about such has really been developed about that.
I nonetheless agree that a burned DVD is likely to last longer than a HD. If the theory is that I'm going to go into a coma for 10 years and then want my data, I'd absolutely think DVD is a pretty good solution, although tape might well be better.
But I would rather have important data on 2 HDs in different locations than only on 1 set of DVDs. Especially if I'm checking the HDs with any kind of regularity and replacing the one when it goes bad. Either of these is insufficient, of course, for any kind of reasonable security.
Furthermore, it is much easier for me to integrity check my HDs than your set of DVDs (because there are more DVDs to check and a limited number of IDE connections) which we should definitely be doing in either case...
My basic belief is that you should play this game until you have a number of harddrives you feel comfortable with, you should automatically test them on a regular basis, and you should replace them as they go bad. Chances of them going bad? 100%. Chances of them ALL going bad before you can replace them? Very low even for a handful of identical copies.
(Obviously you need a scheme not only for backup but also for versioning on them)
We had an experience with a very light (because it was very small).22 LR semiautomatic pistol where it would repeatedly jam - but only for certain users. Same gun, different shooter, no jams. We eventually discovered that the since the reloading action depends on the slide moving compared to the base, but the base didn't weigh anything, it would jam if you didn't hold it firmly enough in place. So it jammed for anybody that didn't hold it firmly enough.
Which was funny, because it didn't really kick either way, being a.22LR.
Harness is extremely complex to measure in any theoretically sensible units. (That is, units derived from other units, like MPa.) Instead we use standardized _practical_ tests, like "how big is the hole if we press X hard with a diamond bit"
The first link you list specificially has hardness discussed just below modulus - and you're quoting us modulus numbers which are of only slight relevance. (That is, the same material in various alloys is usually harder and has a harder modulus as you make it less ductile.)
Ti is commonly harder than mild carbon steel. So is stainless steel, which is more common in jewelry than mild carbon steel. But, you simply can't figure out which is harder without knowing the alloys and heat treatments involved because this makes a huge difference.
However, precious jewelry metals (silver, gold) are supposed to be very pure and are relatively very soft. So if your old ring cutter is only for those metals, Ti OR Steel is going to be bad news.
Certainly you are still _vulnerable_ to social engineering, but you are not AS vulnerable to this kind of social engineering attack.
1. With Windows you are apparently vulnerable just BROWSING the flash drive. Or so says many posts on here, at last. With Linux you must run an executable in there, and you are less likely to run a nonexecutable file in an exploit happy ActiveX environment (eg, IE) or similar.
2. To root your machine with Linux you ALSO need a privilege escalation attack. With Windows if you don't run as Administrator on a normal desktop you face an uphill battle of getting applications to work, because Microsoft has made little attempt to force the body of applications - even major commercial ones - to run not-as-root and even install not-as-root unless they really need root privs. Linux and OS X both do this nicely because the OS demanded it of the application developers. Without escalation the payload is always limited in what it can do to the overall OS... at a minimum you'll be able to clean the computer, which is often not feasible with Windows.
This is perhaps a narrow point, but you can't REALLY build a Mac out of common PC parts.
To run stock OSX without hackish tweaks you need an Apple. Part of this is certainly software, but part of it is EFI: PC components expect to get initialized by BIOS which doesn't exist even on an Intel Mac, it only has EFI. Because Windows will never support 32 bit EFI, common PC motherboards don't support EFI. So to use OTHER Mac parts, you need to use a Mac motherboard, and then you can't use regular PC parts.
(This does not mean they are VERY different - the core chips in the different component versions are usually identical. Clearly someone could manufacturer crossover components, but only if they do it on purpose. Historically you've had to flash PCI components to go from Mac to PC. One presumes that you wouldn't have to flash EFI-compatible PC server components to go to Mac, but that's narrow)
Not to mention that to really be a current Mac you'd need Rosetta to support the earlier apps.
I generally disagree with you about HD vs DVD. Pricewatch seems to say a cheap DVD-R is about 22c in bulk, for a price of about 4.7c/GB and HDs @ about 26.5c/GB. Of course, especially with a scheme where you're going to routinely send remote backups, you've likely to write data to the HD more than 5.6 times. Oh, and mailing them doubles the cost. BUT, that's not the important part at all.
The important part is that to weekly burn and mail a bunch of disks takes a bunch of time you're not accounting for, time that increases linearly as your data exceeds 4.7 GB. Whereas if you add another HD to several other people's computers (with different kinds of HDs), install OpenVPN and some kind of backup manager you get redundant multisite backups which you can easily have run nightly - and completely automatically. And if you want to transfer ALL of your data from older disks onto newer ones, that is relatively quick to do.
If you really only have a few gigs of data total and a complete backup fits on a DVD, then your way is probably fine. But it doesn't scale even a little bit; being able to make complete backups without a huge struggle is a very important consideration.
I have two points, some based on 'uncle' posts, and I'm starting with the minor one:
1. I don't believe in a 'centerless' repository. That is, FOR sharing, I believe it is better to have a designated central server that you know should have high uptime instead of somebody's laptop. If someone might want to poke at the version on MY laptop, it should be copied to the server appropriately so they can still get to it while I'm on an airplane. This is not to imply that distributed redundancy isn't useful, awesome, or worth automating and improving. But there is a difference between replicating as much as possible into the relatively self-sufficient nodes and actually cutting off the head and saying "there should not be a primary server", which I think some in this thread mean.
2. Your other point - essentially about 'experimental' versions - is VERY well taken. And I do not believe that tons of little mini-forks is a good solution to this, although that seems to be a common solution. I definitely believe you do need a system that allows multiple "tiers" of commits - where you can commit to your "personal" repository (which I agree should be local), or decide you're ready for your "team" repository, the project repository, etc.
Which means you definitely need a way of saying "take my current personal repository and merge just the changed parts into the team repository so I can test them there"
However, this doesn't seem like it needs any server changes at all - it just seems like it needs a client to be 'overlapping repository aware'.
This kind of repository tiering of the same filespace is something I would find very useful, but "distributed" isn't my first choice of name. Is this what those distributed systems do? Are these systems mature?
parent, I wish I could mod you "Funny/Informative Satire" For those people who can't reverse your satire:
It is relatively difficult to add more database servers. It is relatively easy to add more webservers. Any caching structure you could possibly use in AJAX you could definitely use on your webserver - and many more, because you have power over those servers.
You MIGHT use AJAX to reduce the load on your webserver in some cases, but if it reduces load on your DB server you didn't have caching setup right. Generally you use AJAX for a better user experience.
I want to take this opportunity to mention Flash as an alternative. Certainly it has been abused, but over the last few years Actionscript (used in Flash and Flex) has become a full fledged OOP language. It is ideal for these sorts of applications, precisely because the results you get on one platform or browser are identical to the ones you get on another one, in Windows, OSX or Linux (i386) It has the highest installed userbase of any software in use by web users (higher than IE, because it's also installed on non-IE browsers) Applications without a lot of GUI can be quite small. A significant part of my living is creating database driven Flash/Flex apps.
The AJAX alternative would seem to be to use some kind of AJAX framework that abstracted the differences away for you. Offhand I can't recall a specific one, but I've seen a couple.
I don't know how other SVN clients work, but we use an SVN server and Eclipse clients with the Subclipse plugin. The effect is that you login to Eclipse and sync whatever files, directories, or parts you want, and it copies them to your local filesystem. You can then open them in Eclipse (online or offline) but you can ALSO open them using your favorite filesystem tools, including grepping them. The current tree is completely distributed, and you never ever need to sync up unless you WANT to share something or get something new...
Sharing is an inherently centralized process anyway (Whether it happens on a central server or through some peer mechanism is irrelevant, it's centralized on you putting it the same kind of place someone else looks for it)
If being able to destroy the server and rebuild the current version from any client isn't good enough setup some darned backups of the server onto the clients. (I imagine you could install an Eclipse client on the server and have it sync the Subversion directory into ANOTHER repository if you wanted to be psychotic, but since you wouldn't be able to merge changes you wouldn't really gain anything over regular backups.)
Microsoft is the only guiding entity of the Windows world, and they are therefore responsible for guiding it. They have not done a good job of this. If there were merely a FEW bad actors, certainly they are not responsible. But they are responsible for the PREVALENCE of such applications even from major vendors.
Clearly you didn't understand my system. Under my proposal: 1. Old apps all run, run as administrator, and run EXACTLY like they do now with a simple startup warning after "x" date/version which can be completely and safely ignored by the user. Akin to the "unsigned software" warning you see now sometimes. 2. Developers "opt-in" to the new low-priv system, which completely defeats the "chicken-egg" problem they have now where they can't get developers to work like that without starting with lower privs and can't start with lower privs because the apps don't work like that.
*****
I probably shouldn't bother answering this, because someone who thinks priv separation is pointless is probably either too misguided for a post to fix or a troll, but I'll try anyway.
Users will always be vulnerable to social attacks like phishing, and will be vulnerable to trojans in programs they want to install, like screensavers, or will click on anything that tells them to. It clearly doesn't fix any of those problems which are basically social and educational problems and technologically unfixable.
But there is an entire additional class of problem - a very COMMON class of problem - that you CAN fix, and priv separation goes a long way to fixing it.
For a single home user, having an Administrator password that is different from your user password is probably essentially useless except to remind you which account you're using. But having distinct ACCOUNTS is very useful.
For instance - simply BROWSING to the wrong site in IE with ActiveX enabled can compromise your browser. Design problem of Microsoft's. BUT without priv separation, this ActiveX weakness means that your whole MACHINE is compromised, not just IE. Without escalated privs, the recent Word trojan wouldn't be able to compromise your whole machine. Obviously there are plenty of others...
And in all these cases because it is compromised at the system level many users can never manage to get, install, or run software that might be able to cleanup the problem. For a rootkit to be able to hide itself from even simple security checks it necessarily must have admin privs - and therefore can make itself invisible to detection software.
With priv separation, any attack must have BOTH an initial vector AND a priv escalation attack.
I personally think the OS X solution is a good one - enter the SAME password again, but a dialog that clearly separates program execution from administrative control. Mac security is far from perfect, but this basic user ARCHITECTURE is a very good idea.
The first solution is to avoid - and get everyone you can to avoid - voting for major-party candidates unless they know something about the actual candidates.
But the second is to support Instant Runoff Voting whereever possible.
The two party system can only be strong when the parties can leverage: 1. The feeling that voting for a non-major candidate is "wasting" your vote.
2. the fact that two people running from their party (or even with similar positions) will inevitably give the race to the opponent.
These are true in our current simple general elections, but it isn't true with IRV at all. With IRV a third-party candidate has no artificial barriers to people who just LIKE them voting for them first and their major-party-of-choice second without impacting their major candidate at all. This takes out all the _inherent_ advantage of the major parties. (But not the _inertial_ advantage; they still have lots of money and power.)
Any major change like this can only happen locally - so get your local politicians to adopt IRV!
I really want the reverse: VGA - IN for laptops
on
Portables as Servers?
·
· Score: 1
What I really want is VGA _in_ for laptops. I want to take any old laptop - with or without a functioning OS, and with or without it _running_ that OS - and plug IN the video from another PC or laptop. In short, I want to be able to retask a laptop with a functioning LCD as an LCD screen.
Furthermore, I want to be able to do this on the fly. I want to lay two laptops on the desk and have one span both displays. Most laptops these days support spanning into an external monitor, but most don't accept input.
(Synergy2 is cool, but you can't drag windows across it.)
(I'd be all about other retasking uses, too. The only other one that comes immediately to mind is that the last 7 years of mac laptops can be booted holding "T" to make them act as an external harddrive for another machine, and I have to give props to that.)
I do NOT recommend RunAs. Because for every program that can't run as non-admin there's two that can't run as a different user than they were installed in.
So to me there are _3_ "kinds" of users from a program requirements point of view: Admin, Normal/PU, and "Normal/PU but must have admin to install AND must run as the same user they were installed by"
Our standard operating procedure for the installation of new software on a nonprivileged single user machine with 2k or XP is to: logout, login as administrator as admin, elevate the local user's privs, logout, login as user install software reboot if necessary run software to ensure it does its "first run" stuff reduce user's privs back to what they should have been logout and log back in.
But MakeMeAdmin looks awesome, I hadn't seen that before.
Microsoft does share the blame, and in fact they have _most_ of the blame.
Certainly this problem happens because of lazy developers, but the market forces at work imply that developers would do the least work necessary to market their program to the majority of users.
Furthermore, no developer has central control over "all programs" - the direction of development of "all programs" rests squarely with MS. We aren't talking about a minimal set of back actors here.
So the fault lies with M$ for shipping an OS that expects to be installed by default as an admin. If the majority of XP machines were NOT shipping this way - if these applications would break for MOST users - then these people would stop shipping these apps.
Here's a random trivial solution they could've implemented: Step 1: Add a "modernapp" flag that software can have/set that says "I'm approved for any user" Step 2: Create a "super-power-user" priv, where everything runs as admin, except things setting the modernapp flag which run as PU. Again, ONLY the apps that ASK for Windows to enforce strict checking get strict checking therefore you have no legacy software problem. Step 3: Popup a warning whenever you run apps that _don't_ have modernapp set. Don't even prevent it, don't even ask for comfirmation - just a warning popup would be sufficient to make those apps fixed.
This would keep it from RUNNING at all. But it would create a reason why Joe Consumer would call the app manufacturer and complain, and after a point it would be cheaper to fix it than answer their questions.
This parallels the way the signed software stuff works, and that would've been a good time to implement it.
Allergens can be tricky. I don't have eczema, but I've done a lot to fight airborne allergies, and here's some stuff I've learned along the way:
The first thing I have to say is that unless you have a very effective (perhaps HEPA) vaccum, vaccumming can be MUCH worse than not vaccumming at all in terms of getting rid of allergens. Vaccums as a rule pick up big particles but just throw little particles in the air - and the little particles are usually the problem.
Second, some people have reactions to outgassed chemicals from solids that you buy. The prime example is when you buy a new computer or monitor, for at least years it spits out a bunch of chemicals whenever it's on. (Primarily but not solely from power heating the plastics) But there plenty of smaller examples where you're buying something that you wouldn't assume gives off chemicals. (Microwaving plastics gives off pseudoestrogens too.) There is plenty of argument about the EFFECT of these emmissions, but the emissions are there. And if your child is having symptoms you can't track down, it's something to think about.
I personally know that many cosmetics here outgas; I'm not sensitive in an allergic way but I can smell the volatiles in them, even in ones that aren't particularly supposed to have a scent.
The only broad-spectrum solution I know to combat volatile airborne chemicals is the extensive use of activated charcoal. I believe it is sometimes possible to buy this in big bags instead of tiny filters - as I understand it it's really just evenly crushed charcoal that's been baked very high.
Outdoor air is actually better than indoor air in many circumstances. Some new construction now blows in a certain amount of outside air to help this. (And to counteract the lower inside pressure of running the heat or AC) Oh and a conventional house furnance filter is useless for allergens. You have to install an electronic one or get a specific high efficiency one - 3M makes one.
I have to bring up laundry detergent because I THINK you're including that in "household chemicals" but I couldn't be sure. Extra rinsing helps but isn't perfect... and of course many clothes come with preexisting chemicals in them from the manufacturing process.
And of course you can have a pretty frightening amount of mold in a house in the walls and such without knowing. Fake wood paneling and carpet padding are the most common offendors, but it can be anywhere.
Generally the "problems" are because you've only known the MS way. Most (but not all) are behavior I slightly prefer...
The printing is definitely different, and we had training problems getting people to convert. Calc printing had some issues with printing a ton of blank pages, but so does Excel. Generally I think it was superior to MS's tendency to reformat everything...
The problem that literally got it uninstalled and MS purchased on one machine was the lack of a comma button. That is, a button on the toolbar that automatically added or removed the comma formatting from numbers in Calc. (The formatting options were there, but not the button. I was tempted to try to write one, but I didn't have time at the time...)
Quality of Service is great. BUT THE ISPs ARE NOT TALKING ABOUT QOS, so it's just a distraction.
VoIP with poor bandwidth will behave worse than many other things with poor bandwidth.
For an ISP to say "Vonage, will you pay $X if we guarantee exceptional QoS to your packets" I would not object to that. In fact, I'm sure this already goes on and partner providers are already faster in some cases. Nor would I object if they said "customer, for $X you can upgrade to a high QoS service. This has the same peak bandwidth, but your VoIP will work better"
This is NOT what they're saying. What Net Neutrality is fighting against is very different. It is the ISP saying: "we realize Vonage packets are competing with services we would like to offer or might like to offer. So we will institute QoS rules to specifically make Vonage packets much WORSE - or totally blocked - than your typical internet packet, unless they pay us a fee not to worsen their connection."
The reason this is a big difference is that there is no way they are going to have agreements with the vast majority of internet sites. So the ability to selectively block things amounts to them being able to send a bill to anyone they want and break up the internet into little fiefdoms again.
This makes a big difference to customers: If their connection just sux, then almost all high-speed stuff is going to be slow. Either they'll be ok with that or not (and leave) But if everywhere but _1_ is slow, they'll think it's that one place...
Personally, I hope that no website ever pays these fees out of pocket. A sane position for a content provider would be to tell their customers "your ISP is blocking us; you may pay $X to have access. Try these other great ISPs instead." This strategy might have short-term problems, but it would result in a significant flight away from the bad ISPs.
Somebody clearly needs to start a coalition of broadband ISPs that agree to support net neutrality - and the content providers need to start giving them props.
I'm not trying to suggest that an appropriate model is to always add every feature anybody thinks of. There are very valid reasons not to bloat your code like that, to keep it easier and simpler to maintain, test, audit, secure, etc.
But there is still a difference there. Perhaps I could best explain it as: You can never take your existing software, spend time/money optimizing it, and somehow make it cost you less to produce a million copies of the software.
(More USERS, maybe, but not more copies of the software - in a situation where you make it handle bigger load on smaller hardware. More PROFIT maybe, if more people want to buy the new version.)
Very rarely is the reason a feature shouldn't be present that the cost-to-make-copies is too high. In cars, for example, a lot of work goes into taking existing components and making the production cheaper. There are plenty of other things you can optimize code for, but you can only rarely spend significant time optimizing existing features for "cheaper" You never go back, take your existing really well-written function and replace it with a less-effective one because the less-effective one because the worse one is cheaper to copy. (Unless you've licensed that great code from someone else, but that's somewhat artificial)
You might choose a faster algorithm over a more accurate one - or a simple and therefore easier to audit algorithm over a complex but more accurate one. But you are still missing the entire optimization about the "produce copies" part of the development of _anything_ else that people produce. To me this is the dominant reason why it's so easy to make complex software and the dominant reason why software is massively more complex than anything else for a given number of hours of development.
(Which is not to say that, say, the CPU hardware isn't very complex. But a LOT of people have spent a LOT of time on that. )
Slashdot Mirror
Actually, the Mac Genius isn't really for fixing the macs. Several things:
1. Macs are much simpler to configure for a novice user. More importantly, they're more RELIABLE to configure - it's much more rare to get some psychotic configuration behavior in OSX than in Windows, OSX better explains the dialog boxes, and OSX is more likely to prompt you with an appropriate correction. OSX has a better mix of manual and automatic and a better transition between them.
This doesn't mean they never break, but it does mean that you don't need a Geek Squad as much with OSX - in my experience a novice user will need less hand-holding help using OSX than Windows. (Not none, but less)
2. Last I checked, when you take a broken mac to an Apple Store, they only do like kind replacements or simple upgrades of a few of the most common components. It's much more like they do "installs" than real repairs. There are places that do real Apple Service around (I worked at one) but if your problem is actually complex and you bring it to the Apple Store they will overnight your computer to Texas where it will be fixed and overnighted back. To me this is a fine system for making sure you get a well-trained technician for your hard problems.
3. The Genius's real job is to be a consumer consultant. Their job is to talk and be helpful, not fix. (I'm not saying they _can't_ fix - the ones I've dealt with have seemed very competent) They need to be able to answer very open ended questions like "I want to learn video editing. I have no clue. Take my credit card." or "My computer is slow. How do I make it faster, or do I need a new one" and explain to the person what they need in minute detail. Their job is to get them what they actually need. Arguably figuring out what the customer wants is the part that requires being a Genius.
In short the Genius's job is much more about being able to answer a customer question of "what should I do" than about touching Macs - or at least it was last I checked.
We use it, but we're not a terrific data point. We haven't moved to 10 yet at all. But here are some basics:
I'm an Apple fan, and in my opinion SuSE is the Apple of Linux. (Of course, Apple is itself the Apple of *nix-like OSes.) They are not the earliest adopters of new technology, but they do a good job of integrating it. But moreover they make it easy to use and administer.*
Big business wants a Vendor, not a community. In the giant world that's pretty much RH and SuSE, or IBM selling someone else's.
Novell also makes a good business selling networking solutions for you whole office, not making you put them together youself.
SuSE will happily ship with the best available drivers and software, even if those are proprietary. For some people this is a reason not to use SuSE - zealots have their place, and I would not want the strictly OSS distros to go away - but if you are more interested in Linux-as-a-current-tool than Linux-as-a-political-statement to force vendors to open drivers, this is the right choice philosophy for you.
*Let me define "easy to use and administer" more: YaST puts a nice front end on whatever you're doing (package management and basically all other system administration) - with enough power to configure whatever you want however you want and enough guidance that you can do it even if it's not your speciality and you've been awake too long. It's the perfect kind of system that LETS you be knowledgeable but does not REQUIRE you to be knowledgeable. You can seamlessly escalate simple point and click management to advanced point and click management to tweaking files by hand that it then won't screw up.
So to me, "easy to use" means that I can use, in each instance, a system that is as automatic or as manual as I want, based on how much expertise I have in that area, and how much time and attention I have for the issue "right now"
Configuration entirely by manually touching files/registries/whatever is a little like walking through a minefield... get too tired, make a typo and all sorts of stuff might explode, and you've making a large number of changes. But I'd take it any day over a Windows GUI-only system where IF it explodes and, say, doesn't boot you have a much harder time putting it back together than you do with a text editor. But YaST is the ideal hybrid - it reduces your chances of stupid mistakes without limiting your power. You edit what you want, let YaST edit what you don't. It's not by far the only piece of software to do this, but I think it's a good example.
( I think much of OS X is similar. Can you enter complex firewall and packet forwarding configurations in their little GUI? No. Does their GUI work for most people? Yep. Does their GUI still use the standard BSD firewall, which you can configure however you want? Yep. )
Parent is completely correct about Rent-A-Coder etc (and there are many more) We get a lot of work from such places, but it's a lot of work GETTING the work - more than it is doing the work, usually. This isn't the kind of investment for someone who wants to program for fun.
Regarding the gp comment - I want to work less and feel less pressure to make money, so that I have more time - to work on the programming projects that I WANT to work on. I do plenty of other things with my life, but programming is wonderful and powerful and expressive and I'd still like to do more of it than I do. Furthermore, programming is broad - you can go home and do very significantly different things than you do at work.
If OP (or someone else) is good at what they do and interested, email me (use a subject like Slashdot Resume) or reply to this with some way to contact you. We have a steady stream of these smaller projects, we have many that would be appropriate for you to do in the evenings and we'd take care of the non-programming management for you.
Some sibling post mentioned programmers trying to find someone to do this business side - this is a big part of what we do.
(Furthermore we have at least one and possibly more projects that are definitely targetted at making the world a better place, if that adds to your preference for working on those projects.)
Instant Runoff Voting or Ranked Choice Voting are the most usual US names for this - which is definitely the best long term solution, and which is being used in some local elections already.
Not voting is no good.
I will happily grant you that both major candidates may suck in any given election and that you might well want to protest by not voting for either one. (I do not agree with your idealistic "sullying my hands" position - I think if one of those candidates is less bad to you you should vote for them, and I think in most real cases one candidate is less bad to you if you bother to check. But that's not my major point, so I'll assume they're exactly even for now.)
But the _biggest_ consistent problem we have which makes the two candidates both suck is that the two incumbent parties have a strangehold on who we get to choose from. Voting for a third party candidates drives up the visibility of third parties existing and drives up the likelihood that OTHER people will vote for third parties.
As a bonus, if enough people do it for a presidental campaign then they get federal election money.
Hate to taunt a girl on /., but
If you're going to publicly pick on someone's phrases you should really offer the correct examples, too. Especially since you don't know if English is the poster's first language. You can guarantee that it is not the first language of some of the readers and that some readers are trying to improve their skills.
"run the gamut" - 'to cover a whole range of variations'
"run the gauntlet" (probably not what the poster meant) - 'to suffer severe criticism or tribulation' with background in this criticism coming from many sources. This usage would actually be funny if the poster meant that they were punished by a wide variety of technologies.
"for all intents and purposes"
"fringe benefits"
And no, I didn't spellcheck this.
As a web application developer, I have to say that both the parent's experience (albeit with different tech) and the OP's experience ring true, but I love it.
As fundamentally a person who likes to create things, I love web development because there is no other domain where I can spend as little time to make something complex actually work and be available to many people. The first is the domain of all software and the second is specific to web development. (Obviously software projects can be huge and complex, but a physical invention of similar complexity would be much MORE work, often by more people, to get accomplished) The only potential challenge to this crown is being a popular blogger or flashmob organizer and actually getting zillions of actual people to do what you ask. And that usually pays worse.
Fundamentally web programmer pay is pretty good, I think. But it's highly variable, so you need to work on the system to make sure you get paid appropriately. I think many tech people think they'll get hired on their merits, when what matters are personal contacts/reputation and, at best, the appearance of merit (portfolio/resume) These are different skillsets, but they are important to getting paid enough.
If you take a regular job and it doesn't pay enough, you need to look for another one. Or you need to get more skills at getting hired so you can.
I am self employed and love that part too - but you have to be good at making clients want to hire you and pay you - the hiring process over and over again - and you have to throw a wide enough net. Being self employed at anything involves building up a clientele, and that's a lot of work. I certainly went through a period of intermittently not enough employment. Realize that your volume and price are intertwined. Essentially there's a minimarket for your skills, so your volume will drop as your price rises. I generally think this means you should start with a fairly low price, get as much work as you can to build reputation, and as you get too much work raise your price and realize that some clients will no longer be able to afford you. Reliable repeat clients are often worth having even if you have to charge them less, because there's no overhead.
If you're good at what you do and still interested in freelancing, feel free to email me with what you know and make sure to put "Slashdot resume" in the subject.
You have to be good, and I still do not guarantee to pay well; there's a wide open market of developers out there. But we'll do the work of finding clients, selling clients and having a client-base, we'll add to your volume at the low end, and you'll still be free to take on other clients at the high end simultaneously and you can build up your own clientbase.
I believe in the long term quality of pressed DVDs - but not so much of burned ones, because insufficient evidence about such has really been developed about that.
I nonetheless agree that a burned DVD is likely to last longer than a HD. If the theory is that I'm going to go into a coma for 10 years and then want my data, I'd absolutely think DVD is a pretty good solution, although tape might well be better.
But I would rather have important data on 2 HDs in different locations than only on 1 set of DVDs. Especially if I'm checking the HDs with any kind of regularity and replacing the one when it goes bad. Either of these is insufficient, of course, for any kind of reasonable security.
Furthermore, it is much easier for me to integrity check my HDs than your set of DVDs (because there are more DVDs to check and a limited number of IDE connections) which we should definitely be doing in either case...
My basic belief is that you should play this game until you have a number of harddrives you feel comfortable with, you should automatically test them on a regular basis, and you should replace them as they go bad. Chances of them going bad? 100%. Chances of them ALL going bad before you can replace them? Very low even for a handful of identical copies.
(Obviously you need a scheme not only for backup but also for versioning on them)
We had an experience with a very light (because it was very small) .22 LR semiautomatic pistol where it would repeatedly jam - but only for certain users. Same gun, different shooter, no jams. We eventually discovered that the since the reloading action depends on the slide moving compared to the base, but the base didn't weigh anything, it would jam if you didn't hold it firmly enough in place. So it jammed for anybody that didn't hold it firmly enough.
.22LR.
Which was funny, because it didn't really kick either way, being a
Harness is extremely complex to measure in any theoretically sensible units. (That is, units derived from other units, like MPa.) Instead we use standardized _practical_ tests, like "how big is the hole if we press X hard with a diamond bit"
The first link you list specificially has hardness discussed just below modulus - and you're quoting us modulus numbers which are of only slight relevance. (That is, the same material in various alloys is usually harder and has a harder modulus as you make it less ductile.)
Ti is commonly harder than mild carbon steel. So is stainless steel, which is more common in jewelry than mild carbon steel. But, you simply can't figure out which is harder without knowing the alloys and heat treatments involved because this makes a huge difference.
However, precious jewelry metals (silver, gold) are supposed to be very pure and are relatively very soft. So if your old ring cutter is only for those metals, Ti OR Steel is going to be bad news.
Certainly you are still _vulnerable_ to social engineering, but you are not AS vulnerable to this kind of social engineering attack.
1. With Windows you are apparently vulnerable just BROWSING the flash drive. Or so says many posts on here, at last. With Linux you must run an executable in there, and you are less likely to run a nonexecutable file in an exploit happy ActiveX environment (eg, IE) or similar.
2. To root your machine with Linux you ALSO need a privilege escalation attack. With Windows if you don't run as Administrator on a normal desktop you face an uphill battle of getting applications to work, because Microsoft has made little attempt to force the body of applications - even major commercial ones - to run not-as-root and even install not-as-root unless they really need root privs. Linux and OS X both do this nicely because the OS demanded it of the application developers. Without escalation the payload is always limited in what it can do to the overall OS... at a minimum you'll be able to clean the computer, which is often not feasible with Windows.
This is perhaps a narrow point, but you can't REALLY build a Mac out of common PC parts.
To run stock OSX without hackish tweaks you need an Apple. Part of this is certainly software, but part of it is EFI: PC components expect to get initialized by BIOS which doesn't exist even on an Intel Mac, it only has EFI. Because Windows will never support 32 bit EFI, common PC motherboards don't support EFI. So to use OTHER Mac parts, you need to use a Mac motherboard, and then you can't use regular PC parts.
(This does not mean they are VERY different - the core chips in the different component versions are usually identical. Clearly someone could manufacturer crossover components, but only if they do it on purpose. Historically you've had to flash PCI components to go from Mac to PC. One presumes that you wouldn't have to flash EFI-compatible PC server components to go to Mac, but that's narrow)
Not to mention that to really be a current Mac you'd need Rosetta to support the earlier apps.
I generally disagree with you about HD vs DVD. Pricewatch seems to say a cheap DVD-R is about 22c in bulk, for a price of about 4.7c/GB and HDs @ about 26.5c/GB. Of course, especially with a scheme where you're going to routinely send remote backups, you've likely to write data to the HD more than 5.6 times. Oh, and mailing them doubles the cost. BUT, that's not the important part at all.
The important part is that to weekly burn and mail a bunch of disks takes a bunch of time you're not accounting for, time that increases linearly as your data exceeds 4.7 GB. Whereas if you add another HD to several other people's computers (with different kinds of HDs), install OpenVPN and some kind of backup manager you get redundant multisite backups which you can easily have run nightly - and completely automatically. And if you want to transfer ALL of your data from older disks onto newer ones, that is relatively quick to do.
If you really only have a few gigs of data total and a complete backup fits on a DVD, then your way is probably fine. But it doesn't scale even a little bit; being able to make complete backups without a huge struggle is a very important consideration.
I have two points, some based on 'uncle' posts, and I'm starting with the minor one:
1. I don't believe in a 'centerless' repository. That is, FOR sharing, I believe it is better to have a designated central server that you know should have high uptime instead of somebody's laptop. If someone might want to poke at the version on MY laptop, it should be copied to the server appropriately so they can still get to it while I'm on an airplane. This is not to imply that distributed redundancy isn't useful, awesome, or worth automating and improving. But there is a difference between replicating as much as possible into the relatively self-sufficient nodes and actually cutting off the head and saying "there should not be a primary server", which I think some in this thread mean.
2. Your other point - essentially about 'experimental' versions - is VERY well taken. And I do not believe that tons of little mini-forks is a good solution to this, although that seems to be a common solution. I definitely believe you do need a system that allows multiple "tiers" of commits - where you can commit to your "personal" repository (which I agree should be local), or decide you're ready for your "team" repository, the project repository, etc.
Which means you definitely need a way of saying "take my current personal repository and merge just the changed parts into the team repository so I can test them there"
However, this doesn't seem like it needs any server changes at all - it just seems like it needs a client to be 'overlapping repository aware'.
This kind of repository tiering of the same filespace is something I would find very useful, but "distributed" isn't my first choice of name. Is this what those distributed systems do? Are these systems mature?
parent, I wish I could mod you "Funny/Informative Satire" For those people who can't reverse your satire:
It is relatively difficult to add more database servers. It is relatively easy to add more webservers. Any caching structure you could possibly use in AJAX you could definitely use on your webserver - and many more, because you have power over those servers.
You MIGHT use AJAX to reduce the load on your webserver in some cases, but if it reduces load on your DB server you didn't have caching setup right. Generally you use AJAX for a better user experience.
I want to take this opportunity to mention Flash as an alternative. Certainly it has been abused, but over the last few years Actionscript (used in Flash and Flex) has become a full fledged OOP language. It is ideal for these sorts of applications, precisely because the results you get on one platform or browser are identical to the ones you get on another one, in Windows, OSX or Linux (i386) It has the highest installed userbase of any software in use by web users (higher than IE, because it's also installed on non-IE browsers) Applications without a lot of GUI can be quite small. A significant part of my living is creating database driven Flash/Flex apps.
The AJAX alternative would seem to be to use some kind of AJAX framework that abstracted the differences away for you. Offhand I can't recall a specific one, but I've seen a couple.
Explain what you mean be "distributed"
I don't know how other SVN clients work, but we use an SVN server and Eclipse clients with the Subclipse plugin. The effect is that you login to Eclipse and sync whatever files, directories, or parts you want, and it copies them to your local filesystem. You can then open them in Eclipse (online or offline) but you can ALSO open them using your favorite filesystem tools, including grepping them. The current tree is completely distributed, and you never ever need to sync up unless you WANT to share something or get something new...
Sharing is an inherently centralized process anyway (Whether it happens on a central server or through some peer mechanism is irrelevant, it's centralized on you putting it the same kind of place someone else looks for it)
If being able to destroy the server and rebuild the current version from any client isn't good enough setup some darned backups of the server onto the clients. (I imagine you could install an Eclipse client on the server and have it sync the Subversion directory into ANOTHER repository if you wanted to be psychotic, but since you wouldn't be able to merge changes you wouldn't really gain anything over regular backups.)
Microsoft is the only guiding entity of the Windows world, and they are therefore responsible for guiding it. They have not done a good job of this. If there were merely a FEW bad actors, certainly they are not responsible. But they are responsible for the PREVALENCE of such applications even from major vendors.
Clearly you didn't understand my system. Under my proposal: 1. Old apps all run, run as administrator, and run EXACTLY like they do now with a simple startup warning after "x" date/version which can be completely and safely ignored by the user. Akin to the "unsigned software" warning you see now sometimes. 2. Developers "opt-in" to the new low-priv system, which completely defeats the "chicken-egg" problem they have now where they can't get developers to work like that without starting with lower privs and can't start with lower privs because the apps don't work like that.
*****
I probably shouldn't bother answering this, because someone who thinks priv separation is pointless is probably either too misguided for a post to fix or a troll, but I'll try anyway.
Users will always be vulnerable to social attacks like phishing, and will be vulnerable to trojans in programs they want to install, like screensavers, or will click on anything that tells them to. It clearly doesn't fix any of those problems which are basically social and educational problems and technologically unfixable.
But there is an entire additional class of problem - a very COMMON class of problem - that you CAN fix, and priv separation goes a long way to fixing it.
For a single home user, having an Administrator password that is different from your user password is probably essentially useless except to remind you which account you're using. But having distinct ACCOUNTS is very useful.
For instance - simply BROWSING to the wrong site in IE with ActiveX enabled can compromise your browser. Design problem of Microsoft's. BUT without priv separation, this ActiveX weakness means that your whole MACHINE is compromised, not just IE. Without escalated privs, the recent Word trojan wouldn't be able to compromise your whole machine. Obviously there are plenty of others...
And in all these cases because it is compromised at the system level many users can never manage to get, install, or run software that might be able to cleanup the problem. For a rootkit to be able to hide itself from even simple security checks it necessarily must have admin privs - and therefore can make itself invisible to detection software.
With priv separation, any attack must have BOTH an initial vector AND a priv escalation attack.
I personally think the OS X solution is a good one - enter the SAME password again, but a dialog that clearly separates program execution from administrative control. Mac security is far from perfect, but this basic user ARCHITECTURE is a very good idea.
I agree, and would mod you if I had points.
The first solution is to avoid - and get everyone you can to avoid - voting for major-party candidates unless they know something about the actual candidates.
But the second is to support Instant Runoff Voting whereever possible.
The two party system can only be strong when the parties can leverage:
1. The feeling that voting for a non-major candidate is "wasting" your vote.
2. the fact that two people running from their party (or even with similar positions) will inevitably give the race to the opponent.
These are true in our current simple general elections, but it isn't true with IRV at all. With IRV a third-party candidate has no artificial barriers to people who just LIKE them voting for them first and their major-party-of-choice second without impacting their major candidate at all. This takes out all the _inherent_ advantage of the major parties. (But not the _inertial_ advantage; they still have lots of money and power.)
Any major change like this can only happen locally - so get your local politicians to adopt IRV!
What I really want is VGA _in_ for laptops. I want to take any old laptop - with or without a functioning OS, and with or without it _running_ that OS - and plug IN the video from another PC or laptop. In short, I want to be able to retask a laptop with a functioning LCD as an LCD screen.
Furthermore, I want to be able to do this on the fly. I want to lay two laptops on the desk and have one span both displays. Most laptops these days support spanning into an external monitor, but most don't accept input.
(Synergy2 is cool, but you can't drag windows across it.)
(I'd be all about other retasking uses, too. The only other one that comes immediately to mind is that the last 7 years of mac laptops can be booted holding "T" to make them act as an external harddrive for another machine, and I have to give props to that.)
I do NOT recommend RunAs. Because for every program that can't run as non-admin there's two that can't run as a different user than they were installed in.
So to me there are _3_ "kinds" of users from a program requirements point of view:
Admin, Normal/PU, and "Normal/PU but must have admin to install AND must run as the same user they were installed by"
Our standard operating procedure for the installation of new software on a nonprivileged single user machine with 2k or XP is to:
logout, login as administrator
as admin, elevate the local user's privs,
logout, login as user
install software
reboot if necessary
run software to ensure it does its "first run" stuff
reduce user's privs back to what they should have been
logout and log back in.
But MakeMeAdmin looks awesome, I hadn't seen that before.
Microsoft does share the blame, and in fact they have _most_ of the blame.
Certainly this problem happens because of lazy developers, but the market forces at work imply that developers would do the least work necessary to market their program to the majority of users.
Furthermore, no developer has central control over "all programs" - the direction of development of "all programs" rests squarely with MS. We aren't talking about a minimal set of back actors here.
So the fault lies with M$ for shipping an OS that expects to be installed by default as an admin. If the majority of XP machines were NOT shipping this way - if these applications would break for MOST users - then these people would stop shipping these apps.
Here's a random trivial solution they could've implemented:
Step 1: Add a "modernapp" flag that software can have/set that says "I'm approved for any user"
Step 2: Create a "super-power-user" priv, where everything runs as admin, except things setting the modernapp flag which run as PU. Again, ONLY the apps that ASK for Windows to enforce strict checking get strict checking therefore you have no legacy software problem.
Step 3: Popup a warning whenever you run apps that _don't_ have modernapp set. Don't even prevent it, don't even ask for comfirmation - just a warning popup would be sufficient to make those apps fixed.
This would keep it from RUNNING at all. But it would create a reason why Joe Consumer would call the app manufacturer and complain, and after a point it would be cheaper to fix it than answer their questions.
This parallels the way the signed software stuff works, and that would've been a good time to implement it.
Allergens can be tricky. I don't have eczema, but I've done a lot to fight airborne allergies, and here's some stuff I've learned along the way:
The first thing I have to say is that unless you have a very effective (perhaps HEPA) vaccum, vaccumming can be MUCH worse than not vaccumming at all in terms of getting rid of allergens. Vaccums as a rule pick up big particles but just throw little particles in the air - and the little particles are usually the problem.
Second, some people have reactions to outgassed chemicals from solids that you buy. The prime example is when you buy a new computer or monitor, for at least years it spits out a bunch of chemicals whenever it's on. (Primarily but not solely from power heating the plastics) But there plenty of smaller examples where you're buying something that you wouldn't assume gives off chemicals. (Microwaving plastics gives off pseudoestrogens too.) There is plenty of argument about the EFFECT of these emmissions, but the emissions are there. And if your child is having symptoms you can't track down, it's something to think about.
I personally know that many cosmetics here outgas; I'm not sensitive in an allergic way but I can smell the volatiles in them, even in ones that aren't particularly supposed to have a scent.
The only broad-spectrum solution I know to combat volatile airborne chemicals is the extensive use of activated charcoal. I believe it is sometimes possible to buy this in big bags instead of tiny filters - as I understand it it's really just evenly crushed charcoal that's been baked very high.
Outdoor air is actually better than indoor air in many circumstances. Some new construction now blows in a certain amount of outside air to help this. (And to counteract the lower inside pressure of running the heat or AC) Oh and a conventional house furnance filter is useless for allergens. You have to install an electronic one or get a specific high efficiency one - 3M makes one.
I have to bring up laundry detergent because I THINK you're including that in "household chemicals" but I couldn't be sure. Extra rinsing helps but isn't perfect... and of course many clothes come with preexisting chemicals in them from the manufacturing process.
And of course you can have a pretty frightening amount of mold in a house in the walls and such without knowing. Fake wood paneling and carpet padding are the most common offendors, but it can be anywhere.
Based admittedly on OOo v1:
Generally the "problems" are because you've only known the MS way. Most (but not all) are behavior I slightly prefer...
The printing is definitely different, and we had training problems getting people to convert. Calc printing had some issues with printing a ton of blank pages, but so does Excel. Generally I think it was superior to MS's tendency to reformat everything...
The problem that literally got it uninstalled and MS purchased on one machine was the lack of a comma button. That is, a button on the toolbar that automatically added or removed the comma formatting from numbers in Calc. (The formatting options were there, but not the button. I was tempted to try to write one, but I didn't have time at the time...)
Quality of Service is great. BUT THE ISPs ARE NOT TALKING ABOUT QOS, so it's just a distraction.
VoIP with poor bandwidth will behave worse than many other things with poor bandwidth.
For an ISP to say "Vonage, will you pay $X if we guarantee exceptional QoS to your packets" I would not object to that. In fact, I'm sure this already goes on and partner providers are already faster in some cases. Nor would I object if they said "customer, for $X you can upgrade to a high QoS service. This has the same peak bandwidth, but your VoIP will work better"
This is NOT what they're saying. What Net Neutrality is fighting against is very different. It is the ISP saying: "we realize Vonage packets are competing with services we would like to offer or might like to offer. So we will institute QoS rules to specifically make Vonage packets much WORSE - or totally blocked - than your typical internet packet, unless they pay us a fee not to worsen their connection."
The reason this is a big difference is that there is no way they are going to have agreements with the vast majority of internet sites. So the ability to selectively block things amounts to them being able to send a bill to anyone they want and break up the internet into little fiefdoms again.
This makes a big difference to customers: If their connection just sux, then almost all high-speed stuff is going to be slow. Either they'll be ok with that or not (and leave) But if everywhere but _1_ is slow, they'll think it's that one place...
Personally, I hope that no website ever pays these fees out of pocket. A sane position for a content provider would be to tell their customers "your ISP is blocking us; you may pay $X to have access. Try these other great ISPs instead." This strategy might have short-term problems, but it would result in a significant flight away from the bad ISPs.
Somebody clearly needs to start a coalition of broadband ISPs that agree to support net neutrality - and the content providers need to start giving them props.
I'm not trying to suggest that an appropriate model is to always add every feature anybody thinks of. There are very valid reasons not to bloat your code like that, to keep it easier and simpler to maintain, test, audit, secure, etc.
But there is still a difference there. Perhaps I could best explain it as: You can never take your existing software, spend time/money optimizing it, and somehow make it cost you less to produce a million copies of the software.
(More USERS, maybe, but not more copies of the software - in a situation where you make it handle bigger load on smaller hardware. More PROFIT maybe, if more people want to buy the new version.)
Very rarely is the reason a feature shouldn't be present that the cost-to-make-copies is too high. In cars, for example, a lot of work goes into taking existing components and making the production cheaper. There are plenty of other things you can optimize code for, but you can only rarely spend significant time optimizing existing features for "cheaper" You never go back, take your existing really well-written function and replace it with a less-effective one because the less-effective one because the worse one is cheaper to copy. (Unless you've licensed that great code from someone else, but that's somewhat artificial)
You might choose a faster algorithm over a more accurate one - or a simple and therefore easier to audit algorithm over a complex but more accurate one. But you are still missing the entire optimization about the "produce copies" part of the development of _anything_ else that people produce. To me this is the dominant reason why it's so easy to make complex software and the dominant reason why software is massively more complex than anything else for a given number of hours of development.
(Which is not to say that, say, the CPU hardware isn't very complex. But a LOT of people have spent a LOT of time on that. )