As many people have pointed out, Microsoft's problem is that they don't seem to take the "big picture" approach to bug fixing often enough. I mean, how often have we known that buffer overflows are a problem? Microsoft itself even has a page on safe string handling functions to replace strcpy and its ilk. Switching to these functions is trivial.
Microsoft has harder problems facing it-- buffer overflows are only one class of problem. But it seems that Microsoft's highly compartmentalized development process prevents someone from saying, "You know what? We keep seeing the same kinds of bugs. We need to require that all our developers do X." Until someone at MS does this, we're going to see this patching go on indefinitely.
We buy MS licenses for 5000 seats at a time. We also have the "take home" deal. I looked into this option just last week when an informed employee asked about it. I, too, thought that the "take home" part of licensing had expired around the Office 97 era, but it turns out I was wrong. I got our current licensing terms directly from our Microsoft sales rep. So this program is still alive and well.
As somebody who works in IT for the publishing industry, I can say this: keep page layout features the HELL away from the word processor. There are a few reasons:
1) Advanced page-layout features add to the learning curve, and thus waste time.
2) Users are distracted by the fancy markup options they already have. When the copyeditors are done with their manuscripts and hand them over to production, you know what the first thing that production does with it? They strip out all formatting, and have the 'real' layout people do the layout: the interior book designers. So for us, even markup is a waste of time.
3) Yet another binary format that you can't use 20 years of accumulated text-processing tools with. People are floored when I show them what you can do with ed's regexp search-and-replace features. Or Perl. Or whatever. Binary files are teh suck, and stupid little assumptions like endianness or HFS+ metadata kills cross-platform compatibility.
I personally think that we should be using LaTEX around here, but speaking of learning curve... I'd never get that one very far...
We've firewalled that separate physical network, and all of the client connections go through proxies. Unfortunately, yes, at the moment we are using CIFS for filesharing-- Samba, actually-- through a VPN connection. That mitigates the share-hopping worm problem, but man, CIFS through VPN absolutely sucks. Dog slow.
We're currently looking at OpenAFS. It's built from the ground up to work remotely, deals with all (ok, most) file-locking the right way, and is quote a bit faster and more robust than Samba/Microsoft filesharing. Since OpenAFS relies heavily on Kerberos, I am learning the Kerb ropes before we start putting together a test network. The only hit we'll take is that our Macintosh users, who currently use AFP, will lose their HFS+ metadata unless they're careful to BinHex the sensitive files (mostly old fonts nowadays). And, if I'm not mistaken, OpenAFS doesn't support oplocks, so our Access databases will have to sit elsewhere.
It is beyond absurd that Microsoft can't get workable remote filesharing together. Sharepoint is not an option of course, since we're not a purely MS shop. I remember sitting at my father's Mac SE in 1990, for Christ's sake, doing remote filesharing through Apple Remote Access, on a 14.4 kilobaud Zoom modem-- and it was fast enough to work on. Anyway, OpenAFS looks to be very cool stuff. Clients for pretty much every OS you'd find at a workplace.
These old "analog" computers are really cool. The Communications of the ACM 60th anniversary retrospective issue from a few months ago talked about how the computing machinery field was once divided between people who wanted to build machines to do continuous computation and those who favored the discrete route. As computational machinery moved toward the discrete route, there were even "hybrid" machines where the digital side controlled an analog side. Of course, as TFA points out, these differential analyzers were not "programmable" in the modern sense. Interesting stuff-- makes you think about what kinds of tradeoffs we make to go with our current digital designs.
Where I work, that is not the case. Laptops aren't just on a separate LAN, they're on a separate physical network. That's not to say that we couldn't plug a laptop into that separate network in the server room-- which also happens to be where all of our cable runs terminate-- so that would be pretty easy.
Ingimp will not change this fact. The devs simply disagree whether it counts as a 'problem'. I don't think it's any secret that the majority of the people who come to The GIMP, especially those with a Photoshop background, simply can't grok the UI. And there's no denying that The GIMP's UI deviates from almost every other application's model out there. In fact, the only similar one I can think of is xv; hardly stuff for newbies.
This study perpetuates that denial: the people who give up on GIMP do so almost immediately. It's not just a Photoshop background that ruins it for you-- it's prior experience with Windows or the MacOS. The GIMP is simply too different. What useful information are you going to gain from people who give up immediately? Or if they spend all their time looking in menus having to do with color-- how will you know that they need a specific color model without them being able to tell you?
Now it may be that for people with no prior experience with the above mentioned programs, or for those who are willing to spend the time to learn The GIMP's unique way of doing things, then maybe The GIMP really is the best way to edit images. But, as Eric Raymond has pointed out-- sometimes it's not enough to be "the best"; the cost of switching also has to outweigh the cost of learning a new way of doing things. Considering that Photoshop falls into the "extremely expensive software" category, what does this say about The GIMP when The GIMP is free of charge? If what the devs really want is for people to flock to The GIMP en masse, then they need to pull their heads out of the sand: The GIMP UI is a problem. Otherwise, we can safely put The GIMP into the category of extremely useful (and even revolutionary) but oddball software out there like Plan 9, TEX, emacs, and so on. It'll continue to be used rather successfully by a vocal minority, but it won't be the standard.
Oh, no kidding. But considering that Clarke himself didn't really write it, I basically think of it as a different story. Or fan fiction. Same deal with the Brian Herbert/Kevin J. Anderson crap. The only "collaboration" I've enjoyed so far was Asimov/Silverberg's novel version of Nightfall.
Wow, your sense of style interferes with your enjoyment of your computer. Do you trade in your house every couple of years, or do you just cringe through it?
If someone actually does do Rendezvous with Rama, they'd better do it right. The thing about that book-- and the thing that has always made me love Clarke's writing-- is that it captures the wonder and fear in an almost palpable way. But the fear part is hard for movie people to get right. It's the fear of the unknown. Not the fear of some big, drooling monster like Hollywood loves to put in the films. Rendezvous with Rama captured the weirdness of an alien species, and to my knowledge, Arthur C. Clarke is the only writer, next to Stanislaw Lem, who toys with the idea that actually communicating with aliens may not be possible.
Kubrick has made one of the only true sci-fi films in my mind-- 2001: A Space Odyssey. Rama would have to do something similar. Definitely a hard sell, but those kinds of films have staying power.
The notion that the use of regulation over competition is against "American business philosophy" is just wrong. Regulation does run counter to what CEOs say American business philosophy is all about, but it's all just rhetoric. They say these things to appease voters and stockholders; to keep them voting for the politicians and board members that keep the cash flow going. Look at all the corporate lobbyists out there. Look at all the corporate lawsuits. These things wouldn't exist if your claim were true. Modern corporations will use any legal method they have to win in the marketplace, and many of them will use illegal methods when the pros outweigh the cons. "Pretexing", backdating of stock options, cooking the books, bribery, SLAPPs, patent abuse-- I could go on.
It reminds me of this paper on evolved circuit design. On the surface, it works just the way you expect it to-- but underneath, there's a completely different story. Any physical characteristic that gives some advantage is utilized, even if the end result relies on components that are wildly out-of-spec or the blatant breakage of accepted design practices. Human behavior is the same deal in my mind. Laws, codes of ethics, yada yada, are certainly a part of the equation, but they are also just as certainly not the key determinants.
I first saw Dune when I was a kid. The movie blew me away. I've since watched it again many times, and it still blows me away. Not because of the special effects-- some of them look dated now-- but because David Lynch really understood the concept that we're discussing in this thread: make a movie, use special effects if you have to. Most of Dune is dialogue, and some weird costumes, and the best part is that you're thrown into this narrative with little discussion to prepare you for it. As a result, you really need to be engaged to understand the story.
Lynch deviated from Frank Herbert's book in some significant ways, but I still feel that Dune is an excellent movie. So that's one example, but there are many, many more great pre-CGI sci-fi movies out there. Another enjoyable film is Primer. There are no special effects at all in that one (and it's a modern film, too).
It's not true in my case. I work in an environment that has 8 men and roughly 150 women. All of us who work here are committed to getting our jobs done, and we work well together. True, being the only man in a meeting with a dozen women is certainly a different dynamic than being in a balanced or male-dominated environment. But I'd rather work with people who might go overboard to treat me like a human any day than work with a bunch of mouth-breathers who think calling each other 'cocksucker' and bragging about their drinking habits is funny. Thank God I don't work with the meatheads next door. If the 'bitchy' behavior happens here, I'm not seeing it.
Plus, about a quarter of our employees are girls straight out of college. I don't at all mind the eye candy.
Eric Raymond commented in his book that Plan 9 was the perfect example of why better software doesn't always just sweep away the current stuff (UNIX, in this case). It can't just be better, it has to be lots better. Plan 9 is some cool stuff, but UNIX does continue to get the job done.
Seriously. My company, though large to me (couple thousand people), is still small compared to most of the people on this list. In our "branch office", we have around 160 employees. There are three people on the IT staff here. The atmosphere is relaxed, I have great benefits, three weeks of vacation time, flexible hours, and great coworkers who are open to new ideas. We'd never make a list like this, because we only hire someone once in a blue moon, but job satisfaction here is pretty high.
I'm working on a CS degree part time, and my company so far has reimbursed me 100% for all classes except the ones that have the word "math" in it (because it is not "directly related" to professional development). I can live with this.
A job with benefits like this really does give me the extra incentive to help keep the company in business.
The OpenBSD people have specifically stated that they will not pursue these kinds of certifications, because they take developer time away from actually making the operating system secure. IIRC, their opinion was that most of these certifications were based on a number of arbitrary tests that did not actually measure (nor accurately repsent) real-world security exposure. I don't know enough myself to comment on the subject, though. The subject may also be complicated by the fact that the OpenBSD's relationship with DARPA (which used to fund OpenSSH development), which if I understand correctly, is no longer friendly...
Your subject line made me think of something else-- A judge, ordering a defendent to start logging, without any relevance to the case at hand, sounds an awful lot like legislation. Which a judge is not entitled to do.
Which is silly, because email infrastructure has been around for a long time. SMTP/IMAP/POP works very well. If Blackberry's protocols gave you some advantage over the aforementioned methods (wrt SPAM, or reliability, etc), then Blackberry's way of doing things would be worth moving toward. But in my experience, all you get is one especially vulnerable single point of failure.
My Blackberry crashes frequently, using the standard apps. Total lockup. I end up doing a hard reset every week or two. I might just have a bad handset, but if the value of having email on the road weren't so high (oh, and having said email/voice plan being paid by my company), then I would be looking for another device.
In Blackberry's defense-- love the QWERTY keyboard. Couldn't live without it.
An interceptor missile that can only be used to shoot down an ICBM or a RV (reentry vehicle) is a defensive weapon by nature. Do you really think that a kinetic kill vehicle can wipe out a city?
As my uncle (a Lt. Col. and chopper pilot in the Army) pointed out to my father (an idealistic physics Ph.D. working on missle guidance systems at the time), the terms "defensive" and "offensive" are dubious distinctions. Anything that gives one side a tactical advantage can be considered offensive. Consider smokescreens-- a classic "defensive" technology. Only when he used it, in Vietnam, it was used to provide cover for troops on the offensive. Isn't it, then, an "offensive" weapon? He pointed out to my father that a missle defense system (which was what my father was working on) most certainly changes the power balance, which one side could exploit. Up until this point, my father considered working on defensive systems to be unquestionably moral, since it prevented deaths. But in reality, the scenario is far more complex-- it may actually enable one side to kill far more people than if they not had it.
Does anyone else here see the contradiction in an organization that identifies itself as a "market-oriented think tank" lobbying to pass laws that strengthen government-enforced monopolies on IP?
We believe that the technological change embodied in the digital revolution has created tremendous opportunities for enhanced individual liberty, as well as wealth creation and higher living standards. Those opportunities can only be realized if governments resist the temptation to regulate, tax and control. Government has important roles to play in society, including protecting property rights and individual liberties, but its tendency is to reach beyond its legitimate functions in ways that harm consumers, burden citizens and slow progress. Why, that's just pure BS! Actions speak louder than words.
Where do you live? In Boston, for instance, earning 30K a year puts you just above poverty, unless you want to commute an hour or more to the city. And then you need to consider the tradeoffs of transportation costs. I am intimately familiar with this subject.
In my experience, an inexpensive 2-bedroom apartment in Boston (and by Boston I mean "greater Boston") will run you $1500/mo, without utilities, generally in a less-than-desirable neighborhood. So, assuming that you have at least one roommate, your annual housing costs are going to be $9,000. That's nearly a third of your income, and you're not building equity of any kind. That leaves $21,000 for food, rent, utilities, and all your other expenses for the rest of the year, and whatever you have left, if you have anything at all, is "beer money". Did I mention that you can easily spend $1000 during the winter just for heat? At $30K, you'll worry about being able to make next month's rent. You'll worry about whether you can afford getting that filling at the dentist. It's not a happy situation.
Starting salary at my company for copy editors is around $26K/year. Most of these people are still relying on their parents a great deal. But if you want to work in publishing, you have to "pay your dues". Did I mention that interns often PAY US to get into the field? These poor suckers graduate from Emerson College with more debt than I can imagine, and then they PAY to get job experience. Thank God I'm in IT.
Blanket statements about $30K being "quite enough money" are just plain wrong. It depends on your costs.
Slashdot Mirror
As many people have pointed out, Microsoft's problem is that they don't seem to take the "big picture" approach to bug fixing often enough. I mean, how often have we known that buffer overflows are a problem? Microsoft itself even has a page on safe string handling functions to replace strcpy and its ilk. Switching to these functions is trivial.
Microsoft has harder problems facing it-- buffer overflows are only one class of problem. But it seems that Microsoft's highly compartmentalized development process prevents someone from saying, "You know what? We keep seeing the same kinds of bugs. We need to require that all our developers do X." Until someone at MS does this, we're going to see this patching go on indefinitely.
We buy MS licenses for 5000 seats at a time. We also have the "take home" deal. I looked into this option just last week when an informed employee asked about it. I, too, thought that the "take home" part of licensing had expired around the Office 97 era, but it turns out I was wrong. I got our current licensing terms directly from our Microsoft sales rep. So this program is still alive and well.
As somebody who works in IT for the publishing industry, I can say this: keep page layout features the HELL away from the word processor. There are a few reasons:
1) Advanced page-layout features add to the learning curve, and thus waste time.
2) Users are distracted by the fancy markup options they already have. When the copyeditors are done with their manuscripts and hand them over to production, you know what the first thing that production does with it? They strip out all formatting, and have the 'real' layout people do the layout: the interior book designers. So for us, even markup is a waste of time.
3) Yet another binary format that you can't use 20 years of accumulated text-processing tools with. People are floored when I show them what you can do with ed's regexp search-and-replace features. Or Perl. Or whatever. Binary files are teh suck, and stupid little assumptions like endianness or HFS+ metadata kills cross-platform compatibility.
I personally think that we should be using LaTEX around here, but speaking of learning curve... I'd never get that one very far...
We've firewalled that separate physical network, and all of the client connections go through proxies. Unfortunately, yes, at the moment we are using CIFS for filesharing-- Samba, actually-- through a VPN connection. That mitigates the share-hopping worm problem, but man, CIFS through VPN absolutely sucks. Dog slow.
We're currently looking at OpenAFS. It's built from the ground up to work remotely, deals with all (ok, most) file-locking the right way, and is quote a bit faster and more robust than Samba/Microsoft filesharing. Since OpenAFS relies heavily on Kerberos, I am learning the Kerb ropes before we start putting together a test network. The only hit we'll take is that our Macintosh users, who currently use AFP, will lose their HFS+ metadata unless they're careful to BinHex the sensitive files (mostly old fonts nowadays). And, if I'm not mistaken, OpenAFS doesn't support oplocks, so our Access databases will have to sit elsewhere.
It is beyond absurd that Microsoft can't get workable remote filesharing together. Sharepoint is not an option of course, since we're not a purely MS shop. I remember sitting at my father's Mac SE in 1990, for Christ's sake, doing remote filesharing through Apple Remote Access, on a 14.4 kilobaud Zoom modem-- and it was fast enough to work on. Anyway, OpenAFS looks to be very cool stuff. Clients for pretty much every OS you'd find at a workplace.
These old "analog" computers are really cool. The Communications of the ACM 60th anniversary retrospective issue from a few months ago talked about how the computing machinery field was once divided between people who wanted to build machines to do continuous computation and those who favored the discrete route. As computational machinery moved toward the discrete route, there were even "hybrid" machines where the digital side controlled an analog side. Of course, as TFA points out, these differential analyzers were not "programmable" in the modern sense. Interesting stuff-- makes you think about what kinds of tradeoffs we make to go with our current digital designs.
Where I work, that is not the case. Laptops aren't just on a separate LAN, they're on a separate physical network. That's not to say that we couldn't plug a laptop into that separate network in the server room-- which also happens to be where all of our cable runs terminate-- so that would be pretty easy.
Ingimp will not change this fact. The devs simply disagree whether it counts as a 'problem'. I don't think it's any secret that the majority of the people who come to The GIMP, especially those with a Photoshop background, simply can't grok the UI. And there's no denying that The GIMP's UI deviates from almost every other application's model out there. In fact, the only similar one I can think of is xv; hardly stuff for newbies.
This study perpetuates that denial: the people who give up on GIMP do so almost immediately. It's not just a Photoshop background that ruins it for you-- it's prior experience with Windows or the MacOS. The GIMP is simply too different. What useful information are you going to gain from people who give up immediately? Or if they spend all their time looking in menus having to do with color-- how will you know that they need a specific color model without them being able to tell you?
Now it may be that for people with no prior experience with the above mentioned programs, or for those who are willing to spend the time to learn The GIMP's unique way of doing things, then maybe The GIMP really is the best way to edit images. But, as Eric Raymond has pointed out-- sometimes it's not enough to be "the best"; the cost of switching also has to outweigh the cost of learning a new way of doing things. Considering that Photoshop falls into the "extremely expensive software" category, what does this say about The GIMP when The GIMP is free of charge? If what the devs really want is for people to flock to The GIMP en masse, then they need to pull their heads out of the sand: The GIMP UI is a problem. Otherwise, we can safely put The GIMP into the category of extremely useful (and even revolutionary) but oddball software out there like Plan 9, TEX, emacs, and so on. It'll continue to be used rather successfully by a vocal minority, but it won't be the standard.
Who wrote your BIOS?
Oh, no kidding. But considering that Clarke himself didn't really write it, I basically think of it as a different story. Or fan fiction. Same deal with the Brian Herbert/Kevin J. Anderson crap. The only "collaboration" I've enjoyed so far was Asimov/Silverberg's novel version of Nightfall.
Wow, your sense of style interferes with your enjoyment of your computer. Do you trade in your house every couple of years, or do you just cringe through it?
If someone actually does do Rendezvous with Rama, they'd better do it right. The thing about that book-- and the thing that has always made me love Clarke's writing-- is that it captures the wonder and fear in an almost palpable way. But the fear part is hard for movie people to get right. It's the fear of the unknown. Not the fear of some big, drooling monster like Hollywood loves to put in the films. Rendezvous with Rama captured the weirdness of an alien species, and to my knowledge, Arthur C. Clarke is the only writer, next to Stanislaw Lem, who toys with the idea that actually communicating with aliens may not be possible.
Kubrick has made one of the only true sci-fi films in my mind-- 2001: A Space Odyssey. Rama would have to do something similar. Definitely a hard sell, but those kinds of films have staying power.
Mel Gibson is Australian. Fortunately, we still have Clayton Bigsby.
I heard about this the other day on the news, and wondered if it was indeed The Yes Men. Ha ha. Now I know.
The notion that the use of regulation over competition is against "American business philosophy" is just wrong. Regulation does run counter to what CEOs say American business philosophy is all about, but it's all just rhetoric. They say these things to appease voters and stockholders; to keep them voting for the politicians and board members that keep the cash flow going. Look at all the corporate lobbyists out there. Look at all the corporate lawsuits. These things wouldn't exist if your claim were true. Modern corporations will use any legal method they have to win in the marketplace, and many of them will use illegal methods when the pros outweigh the cons. "Pretexing", backdating of stock options, cooking the books, bribery, SLAPPs, patent abuse-- I could go on.
It reminds me of this paper on evolved circuit design. On the surface, it works just the way you expect it to-- but underneath, there's a completely different story. Any physical characteristic that gives some advantage is utilized, even if the end result relies on components that are wildly out-of-spec or the blatant breakage of accepted design practices. Human behavior is the same deal in my mind. Laws, codes of ethics, yada yada, are certainly a part of the equation, but they are also just as certainly not the key determinants.
I first saw Dune when I was a kid. The movie blew me away. I've since watched it again many times, and it still blows me away. Not because of the special effects-- some of them look dated now-- but because David Lynch really understood the concept that we're discussing in this thread: make a movie, use special effects if you have to. Most of Dune is dialogue, and some weird costumes, and the best part is that you're thrown into this narrative with little discussion to prepare you for it. As a result, you really need to be engaged to understand the story.
Lynch deviated from Frank Herbert's book in some significant ways, but I still feel that Dune is an excellent movie. So that's one example, but there are many, many more great pre-CGI sci-fi movies out there. Another enjoyable film is Primer. There are no special effects at all in that one (and it's a modern film, too).
It's not true in my case. I work in an environment that has 8 men and roughly 150 women. All of us who work here are committed to getting our jobs done, and we work well together. True, being the only man in a meeting with a dozen women is certainly a different dynamic than being in a balanced or male-dominated environment. But I'd rather work with people who might go overboard to treat me like a human any day than work with a bunch of mouth-breathers who think calling each other 'cocksucker' and bragging about their drinking habits is funny. Thank God I don't work with the meatheads next door. If the 'bitchy' behavior happens here, I'm not seeing it.
Plus, about a quarter of our employees are girls straight out of college. I don't at all mind the eye candy.
Eric Raymond commented in his book that Plan 9 was the perfect example of why better software doesn't always just sweep away the current stuff (UNIX, in this case). It can't just be better, it has to be lots better. Plan 9 is some cool stuff, but UNIX does continue to get the job done.
Seriously. My company, though large to me (couple thousand people), is still small compared to most of the people on this list. In our "branch office", we have around 160 employees. There are three people on the IT staff here. The atmosphere is relaxed, I have great benefits, three weeks of vacation time, flexible hours, and great coworkers who are open to new ideas. We'd never make a list like this, because we only hire someone once in a blue moon, but job satisfaction here is pretty high.
I'm working on a CS degree part time, and my company so far has reimbursed me 100% for all classes except the ones that have the word "math" in it (because it is not "directly related" to professional development). I can live with this.
A job with benefits like this really does give me the extra incentive to help keep the company in business.
The OpenBSD people have specifically stated that they will not pursue these kinds of certifications, because they take developer time away from actually making the operating system secure. IIRC, their opinion was that most of these certifications were based on a number of arbitrary tests that did not actually measure (nor accurately repsent) real-world security exposure. I don't know enough myself to comment on the subject, though. The subject may also be complicated by the fact that the OpenBSD's relationship with DARPA (which used to fund OpenSSH development), which if I understand correctly, is no longer friendly...
Your subject line made me think of something else-- A judge, ordering a defendent to start logging, without any relevance to the case at hand, sounds an awful lot like legislation. Which a judge is not entitled to do.
Which is silly, because email infrastructure has been around for a long time. SMTP/IMAP/POP works very well. If Blackberry's protocols gave you some advantage over the aforementioned methods (wrt SPAM, or reliability, etc), then Blackberry's way of doing things would be worth moving toward. But in my experience, all you get is one especially vulnerable single point of failure.
My Blackberry crashes frequently, using the standard apps. Total lockup. I end up doing a hard reset every week or two. I might just have a bad handset, but if the value of having email on the road weren't so high (oh, and having said email/voice plan being paid by my company), then I would be looking for another device.
In Blackberry's defense-- love the QWERTY keyboard. Couldn't live without it.
Einstein didn't exactly do a lot of experimentation. But I understand your point.
An interceptor missile that can only be used to shoot down an ICBM or a RV (reentry vehicle) is a defensive weapon by nature. Do you really think that a kinetic kill vehicle can wipe out a city?
As my uncle (a Lt. Col. and chopper pilot in the Army) pointed out to my father (an idealistic physics Ph.D. working on missle guidance systems at the time), the terms "defensive" and "offensive" are dubious distinctions. Anything that gives one side a tactical advantage can be considered offensive. Consider smokescreens-- a classic "defensive" technology. Only when he used it, in Vietnam, it was used to provide cover for troops on the offensive. Isn't it, then, an "offensive" weapon? He pointed out to my father that a missle defense system (which was what my father was working on) most certainly changes the power balance, which one side could exploit. Up until this point, my father considered working on defensive systems to be unquestionably moral, since it prevented deaths. But in reality, the scenario is far more complex-- it may actually enable one side to kill far more people than if they not had it.
Or maybe they don't, apparently...
Where do you live? In Boston, for instance, earning 30K a year puts you just above poverty, unless you want to commute an hour or more to the city. And then you need to consider the tradeoffs of transportation costs. I am intimately familiar with this subject.
In my experience, an inexpensive 2-bedroom apartment in Boston (and by Boston I mean "greater Boston") will run you $1500/mo, without utilities, generally in a less-than-desirable neighborhood. So, assuming that you have at least one roommate, your annual housing costs are going to be $9,000. That's nearly a third of your income, and you're not building equity of any kind. That leaves $21,000 for food, rent, utilities, and all your other expenses for the rest of the year, and whatever you have left, if you have anything at all, is "beer money". Did I mention that you can easily spend $1000 during the winter just for heat? At $30K, you'll worry about being able to make next month's rent. You'll worry about whether you can afford getting that filling at the dentist. It's not a happy situation.
Starting salary at my company for copy editors is around $26K/year. Most of these people are still relying on their parents a great deal. But if you want to work in publishing, you have to "pay your dues". Did I mention that interns often PAY US to get into the field? These poor suckers graduate from Emerson College with more debt than I can imagine, and then they PAY to get job experience. Thank God I'm in IT.
Blanket statements about $30K being "quite enough money" are just plain wrong. It depends on your costs.