I know it depends on what you are doing... but if you aren't pretty sure about the safety of the access point you are connecting to, you are opening yourself to potentially serious security problems. The person running the open network can easily redirect any address he wants to one of his own choosing - i.e. one under his control. You enter "www.passwordsite.com" but are connected to his web-server - presenting you with a login that looks exactly like what you are expecting. You try to login, can't because it isn't the real site, but the fake site just displays some kind of plausible error message about excessive load or something so you ignore it. You try again later, but since he's seen you mac address before and knows that he's already harvested your info for that site, this time he just passes you to the real site. He now has whatever information you entered and you're none the wiser. (This is why some sites now present you with some "secret information" of your choice, which you should verify, before you enter your password.)
I know that this has been said over and over, but names mater - and GIMP puts people off. Strongly. I don't know why geeks don't get this. I have worked on products where the marketing teem has spent, literally, millions of dollars in market research and consulting fees to come up with a product name. This happens all the time. Product names evoke images/moods/whatever in potential users. A bad name can tarnish a product, even an excellent product... sometimes fatally. The name is actually attached to the product in users' minds just like th UI. It really matters.
Contrary to the previous Slashdot stroy, the name need not be descriptive to be effective (e.g. Firefox is a good name), but it doesn't hurt (Photoshop). The name should make people feel good about the product, and feel good about using the product - if it makes them feel uncomfortable, or worse, creepy, they aren't going to use it. They just aren't. They will actually avoid it. Clever, geeky, inside joke names rarely work. You and I may know what GIMP stands for (but is GNU Image Manipulation Program really much better?) but the rest of the world doesn't... and they probably do have some sense about "gimp" - and it's bad.
Geeks: please, please, think more about product names. If you want to move beyond just other geeks (in the case of GIMP photographers and graphic designers) you have to come up with names (and logos/splash screens) that appeal to more than just other geeks. You simply have to accept the fact that what geeks think is cool is not necessarily what the rest of the (potential) user community does. And these people are not "lusers" for not "getting it."
I was almost a "switcher." I was quite literally within two days of ordering a pair or G5's. Up until this point I was *not* a Mac user and now I probably will continue to *not* be a Mac user. First, understand that I couldn't care less about x86 vs. PPC - but when you sell hardware as expensive as Apple's, you need to pick something and stay with it. I have no interest in buying $5000 paperweights.
I know, I know... I can hear it as I type: "But the hardware you buy today will work just as well when they stop making it as it does now." True - if you buy a computer as it is now and never expect more from it. TiVo could start making new hardware with an abacus in it for all I care as long as mine keeps working as it does now - but for $4-5K you had better be able to install new software on a thing, and I do not believe that third parties will continue to support PPC after the x86 switch. Trust me, it's more than just a "simple recompile." I don't care whether the new machines can emulate PPC or not (although I am supremely doubtful about that as well - I just can not see how an x86, as the archecture stands now, could possibly hope to reasonably be used to emulate the PPC - nothing against the x86, it's simply a matter of register count), I care whether the machine I buy today will continue to be supported, by both Apple and crucial third parties, for an amount of time to make paying Apple prices worth it. I don't think that they will - the first time an app that you must have requires x86 for a necessary upgrade, you've got a paperweight. I for one, am not willing to take that risk.
At my university, a full degree takes 8 semesters, or approximately 4300 hours of coursework
How did you come up with this number? When I was in school the general rule was 4-to-5 hours out of class for every hour in class. A BSEE degree required 140 semester hours at my school with 15 week semesters. Assuming the low number I get:
(140 semester hours)*(15 weeks/semester)*(5 reals hours/semester hour/week) = 10,500 real hours of work for a BSEE
This works out to about 70-80 hours/week for 4-5 years - which *is* about what I and my classmates put in per week back then.
Granted this was in EE... and I'm old (BSEE in '85)... but is a BSCS really that easy? Or have times just changed so much that an engineering degree is that easy?
I am literally beside myself. I have been a (paid) Redhat user since 5.0 (and unix for a decade before that) and currently run 8.0. The huge problem here is what now? The systems I run are quite customized with local init-scripts, a patched kernel, and many custom or home-grown applications. Up2date and all are nice, but not the key issue... that being the *unbelievable* migration headache of moving to another distribution. Weeks of work at the very least... and how do I even approach that? I can't take the systems I have now off-line for the time necessary to move everything to Debian or Suse - and where do I do this since it can't be done in-place as it were? And what about the learning-curve? I'm not talking about learning a new desktop or something silly like that, I'm talking about how the system boots, where it keeps stuff, how its package management works, etc., etc., etc.
I am not an "Enterprise" as Redhat seems to conceives of the term and cannot afford what Redhat Enterprise costs - nor do I need that level of support. I am also not a hobbiest who just wants to "Play around with linux" - it runs 24/7 here. I guess what I am is a guy who made a mistake back in 1996 or 1997 when he decided to rely on Redhat.
What am I supposed to do Mr. Szulik? I trusted your company and now I'm screwed (and Fedora *is* *not* the answer.) Isn't this exactly the kind of thing that certain companies have said would happen if we went with open source? Doesn't this give the entire linux community a black eye?
Slashdot Mirror
Tea Party favorite, Governor Rick Scott actually has a similar plan for Florida. http://www.gainesville.com/article/20111011/ARTICLES/111019928
I know it depends on what you are doing... but if you aren't pretty sure about the safety of the access point you are connecting to, you are opening yourself to potentially serious security problems. The person running the open network can easily redirect any address he wants to one of his own choosing - i.e. one under his control. You enter "www.passwordsite.com" but are connected to his web-server - presenting you with a login that looks exactly like what you are expecting. You try to login, can't because it isn't the real site, but the fake site just displays some kind of plausible error message about excessive load or something so you ignore it. You try again later, but since he's seen you mac address before and knows that he's already harvested your info for that site, this time he just passes you to the real site. He now has whatever information you entered and you're none the wiser. (This is why some sites now present you with some "secret information" of your choice, which you should verify, before you enter your password.)
I know that this has been said over and over, but names mater - and GIMP puts people off. Strongly. I don't know why geeks don't get this. I have worked on products where the marketing teem has spent, literally, millions of dollars in market research and consulting fees to come up with a product name. This happens all the time. Product names evoke images/moods/whatever in potential users. A bad name can tarnish a product, even an excellent product... sometimes fatally. The name is actually attached to the product in users' minds just like th UI. It really matters.
Contrary to the previous Slashdot stroy, the name need not be descriptive to be effective (e.g. Firefox is a good name), but it doesn't hurt (Photoshop). The name should make people feel good about the product, and feel good about using the product - if it makes them feel uncomfortable, or worse, creepy, they aren't going to use it. They just aren't. They will actually avoid it. Clever, geeky, inside joke names rarely work. You and I may know what GIMP stands for (but is GNU Image Manipulation Program really much better?) but the rest of the world doesn't... and they probably do have some sense about "gimp" - and it's bad.
Geeks: please, please, think more about product names. If you want to move beyond just other geeks (in the case of GIMP photographers and graphic designers) you have to come up with names (and logos/splash screens) that appeal to more than just other geeks. You simply have to accept the fact that what geeks think is cool is not necessarily what the rest of the (potential) user community does. And these people are not "lusers" for not "getting it."
On the flips side...
I was almost a "switcher." I was quite literally within two days of ordering a pair or G5's. Up until this point I was *not* a Mac user and now I probably will continue to *not* be a Mac user. First, understand that I couldn't care less about x86 vs. PPC - but when you sell hardware as expensive as Apple's, you need to pick something and stay with it. I have no interest in buying $5000 paperweights.
I know, I know... I can hear it as I type: "But the hardware you buy today will work just as well when they stop making it as it does now." True - if you buy a computer as it is now and never expect more from it. TiVo could start making new hardware with an abacus in it for all I care as long as mine keeps working as it does now - but for $4-5K you had better be able to install new software on a thing, and I do not believe that third parties will continue to support PPC after the x86 switch. Trust me, it's more than just a "simple recompile." I don't care whether the new machines can emulate PPC or not (although I am supremely doubtful about that as well - I just can not see how an x86, as the archecture stands now, could possibly hope to reasonably be used to emulate the PPC - nothing against the x86, it's simply a matter of register count), I care whether the machine I buy today will continue to be supported, by both Apple and crucial third parties, for an amount of time to make paying Apple prices worth it. I don't think that they will - the first time an app that you must have requires x86 for a necessary upgrade, you've got a paperweight. I for one, am not willing to take that risk.
Sorry Steve, you just lost a "Switcher."
At my university, a full degree takes 8 semesters, or approximately 4300 hours of coursework
How did you come up with this number? When I was in school the general rule was 4-to-5 hours out of class for every hour in class. A BSEE degree required 140 semester hours at my school with 15 week semesters. Assuming the low number I get:
(140 semester hours)*(15 weeks/semester)*(5 reals hours/semester hour/week) = 10,500 real hours of work for a BSEE
This works out to about 70-80 hours/week for 4-5 years - which *is* about what I and my classmates put in per week back then.
Granted this was in EE... and I'm old (BSEE in '85)... but is a BSCS really that easy? Or have times just changed so much that an engineering degree is that easy?
I find these responses extremely unhelpful.
I am literally beside myself. I have been a (paid) Redhat user since 5.0 (and unix for a decade before that) and currently run 8.0. The huge problem here is what now? The systems I run are quite customized with local init-scripts, a patched kernel, and many custom or home-grown applications. Up2date and all are nice, but not the key issue... that being the *unbelievable* migration headache of moving to another distribution. Weeks of work at the very least... and how do I even approach that? I can't take the systems I have now off-line for the time necessary to move everything to Debian or Suse - and where do I do this since it can't be done in-place as it were? And what about the learning-curve? I'm not talking about learning a new desktop or something silly like that, I'm talking about how the system boots, where it keeps stuff, how its package management works, etc., etc., etc.
I am not an "Enterprise" as Redhat seems to conceives of the term and cannot afford what Redhat Enterprise costs - nor do I need that level of support. I am also not a hobbiest who just wants to "Play around with linux" - it runs 24/7 here. I guess what I am is a guy who made a mistake back in 1996 or 1997 when he decided to rely on Redhat. What am I supposed to do Mr. Szulik? I trusted your company and now I'm screwed (and Fedora *is* *not* the answer.) Isn't this exactly the kind of thing that certain companies have said would happen if we went with open source? Doesn't this give the entire linux community a black eye?