We need to add IP options that list allowed or disallowed countries for a packet to go
through or to.
My apologies if I missed something obvious, but
why not just do end-to-end encryption and go through any whichever country happens to be in the
route? That is assuming you are not as worried about espionage (eavesdropping, man-in-the-middle) as about legal issues with US residents accessing stuff.
Think of the words, "I promise", or in a wedding, "I do." These don't just describe a
physical situation, they actually make a concrete change in a social environment.
Apologies for snipping out just one small piece, but I think it might be enough to allow
me to refine my point.:-)
I think that the words "I do" still don't have any
intrinsic context-independent 'magical' powers.
You are right in that in right context they do
have more meaning than otherwise, thanks to traditions; semantics has been added as "I do" has become an idiom (?) in wedding ceremonies.
And depending on how people feel about the ceremony, political gathering, rally (whatever
context words are used in),
the words do get more weight, more power, from
their surroudings. And then there are people couldn't care less about the very same words,
even if being part of the ceremony (few people
would just ignore them when being wed though...)
What I mean is that words, sentences, even phrases, usually have a relatively small domain in which they have added connotations, and perhaps
additional powers of convincing people. I don't
think any words have "super-powers", to make
them taboo (or holy or whatever) independent of
context; independent of intentions / attitudes
of person who utters the words, independent of
social, political etc. situation/context of
discussion or speech.
I guess my rant against extreme-PC'ism doesn't really completely relate to gender-dependencies,
though, as it's a much larger topic. Not just
whether 'he' should always be used, whether it's
a "tainted" word (I don't think anyone claims
it is); more like whether it should always be used
as the 'neutral' or 'generic' pronoun just because that's what has traditionally been
used. I have no idea; not being a native english
speaker I can't really "feel" one way or the other. I do think, though, that it's good not to try to manipulate a living language in Orwellian NewSpeak way (something you mentioned too).
Then again, it's not like PC was the first movement that tries to "guide the language"; the trends are as old as english linguistic studies.:-)
Finally, I do agree that using (gender-) neutral terms makes perfect sense, when gender isn't really the point.
Even if "gender-coding" has just a slight influence, that influence is unlikely to be useful or positive. Just like I personally dislike colour coding baby clothes ("pink for girls, blue for boys"), just because that's what people have traditionally done. Or always buying dolls for girls and toy cars for boys (independent of what they might like to play with), "just because that's what they always play with". But I digress...:-)
Language does affect the way people think and for that reason...
Hmmmh. Perhaps it affects to small degree, but some people tend
to believe it even defines the way people think.
I think this is akin to superstition; the idea that
words have magical powers. And thus the 'extreme'
political correctionists (?) really look foolish;
condemning 'evil words' believing it's just bad
bad words (nigger etc), not the attitudes of the
people who use the words. And the result is that
new 'correct' words eventually get 'polluted'.
If you really want to present a derogatory racist comment, you can easily get phrase african-american (for example, replace with your
favourite ethnic euphenism here) sound like an insult, by varying tone and facial expressions.:-/
At any rate, a slightly on-topic remark; there are also languages that have no gender distinctions whatsoever. I'd even venture a guess that genders (for nouns) are more common in western (european) languages than in other ones (correct me if I'm wrong... could well be, I just know japanese doesn't have genders for nouns). I still don't believe it changes the way people think all that much. I don't feel like I'm more egaliatrian (or emansipated?) towards women than my american friends, just because my mother tongue doesn't have he/she distinction. If I do it's about culture, upraisinh and socio-economic things,
not the vocabulary, syntax or semantics of language(s) I speak.
it seems sorta backwards that they must open their network to rivals before they can offer data services. It's their network right?
It really depends. Usually telephone network was built by either state owned companies or companies heavily subsidized by state (practically everywhere in the world). And
that's because physical communication networks, like other basic
infrastructure (like roads, railroads, powerlines), are expensive to build and potentially costly to maintain, without actually
directly producing anything. They 'only' provide for actual services; they are not services per se.
Thus, to begin with, it's seldom the case that fully private companies completely funded "their" networks using just their money. But in addition, few people think it would be beneficial to have more than one (or two, perhaps three for redundancy, at most) phone line networks (or, more than one redundant cable network).
It just would usually be wasting lots and lots of money. Most of the time people need just one line, and want to choose the service provider (be it for cable or phone line... of course these may soon be combined, just need one physical highband network for all communication).
Instead, it's thought, it would make sense to share the physical network, and compete on services.
I understand the feeling "it's mine all mine, why should I play with others". It's familiar feeling from childhood. But putting things in perspective, it really would be foolish to force competitors to build their own duplicate networks. That would be really really really inefficient; and as such would enforce de facto monopolies (that telcos now have... the weak competition between cable and telcos is just an excuse to prevent real competition)
The problem left is, of course, what is the 'fair cost' of a physical line. It would be much easier if the lines were taken care of by non-profict organizations, perhaps utility companies, perhaps counties. But unfortunately that's not the case. And that's why legislation tries to figure out a somewhat fair way to allow for actual competition.
The goal is to create images that reflect what small segments of the
population want. Life becomes less of a search and more of a pick and choose. Employees become
employees around the clock, walking viral marketers.
That is one way to view it, but I think (having skimmed the review) the idea was not to FORCE (or coerce in any way) people to MARKET their company or its product(s). Instead it's to allow employees to do what they would do otherwise; when/if they are happy working there, they will usually talk about the company to their friends, relatives. Just normal conversation. And that's likely to be (at the microlevel at least) more effective PR, perhaps even marketing, than high-level mass market advertising.
I mean, think about it. You have a job, at company X, working on product Y. You don't walk around asking your friends to go buy Y when it comes available. That's a bad idea. However, if you like what you are doing, and are interested in creating the product, you are likely to mention the job, company, products. If someone happens to be looking for a product in category Y is in, you are highly likely to mention existence (and features) of Y, without marketing. Or should you try to keep your mouth shut, never mention it? Same thing about the company in general. You like it there? Does it have nice benefits? Sane managers? Good products? People do ask these questions, just small talk. But the same thing applies to web too; home pages people have usually do contain stuff about their job (as well as family, photos, whatever). No one is asking (or should ask for that matter) employees to put company's logo in the page, or to add direct marketing stuff. And yet people usually do add links; perhaps mention something more about actual work (projects etc) they are doing.
Of course you don't need to talk about micro-marketing here. It's just good old "let's keep employees happy as that's the best PR we can have" approach. There are a few high-tech companies that have good reputation, partly because of the way they treat their employees (AFAIK companies like Cisco, Sun, even Microsoft have pretty decent reputation are where they are thanks to being good employers).
And based on what I read about the Clue Train Manifesto, fundamental idea is to really empower (sic! buzz-word, I know) employees, literally, to let them be creative, use their energy. Let employees have their own web pages, at work, at home. Let those contain whatever people want to put there. More restrictions, less useful things people do on their free time voluntarily. No one wants to be a corporate peon, much less 24/7.
But interestingly enough, with freedom and equipment, oftentimes resulting web-pages are surprisingly useful; especially spontaneously created intranet pages. Even if no one "authorizes" or "plans" or "orders" help-pages, how-to - pages, people are likely to create such, as long as they are not explicitly prevented to. Read the Cluetrain Manifesto for more about these ideas if interested. It's good reading.
I think RMS and Larry are native US citizens, but
Linus is a finn, so even though he probably wouldn't really want to, he could easily move
back to Finland. Or to any other EU country, basically, as he wouldn't need any visas to be
able to work (unlike in US... though he probably already has a green card by now).
Just curious; is Java-compiler actually allowed to do that? I'd guess a JIT/native-compiler would be allowed to do that (if it can figure out there are no side-effects, going through constructor and append-methods of StringBuffer), but not javac?
Then again, I remember String and StringBuffer having some interesting smugs way of avoiding unnecessary copies (ie. sharing array for as long as possible), and perhaps even that could prevent excessive array allocations/copies.
Hmmh. Time to check out java decompilers again.:-)
However, since the linux operating system is
being used, that negates these factors.
Well, Linux was
originally created as a unix-clone that runs
efficiently on low-end (usually single user)
systems. Thus, performance and efficiency were
important when doing architectural / design
choices; scalability and scheduling fairness
(for example) were of lesser importance (and
that bit when trying to make kernel SMP-friendly)
Now,
obviously OSes designed specifically for ultra low-end embedded market (QNX?) are likely to
be even more efficient, but really, StrongArm
with ~200 mhz clock frequence has plenty of
power to spare, especially when it's only feeding
tiny display (like you said). OS overhead is
likely to be negligible.
...it won't be interesting for big
corporations (like Sun, IBM, HP) to invest in Amaricans Universities. So the college
education will lower quality, which will take a step back for US economy in a long term.
Don't take this as US bashing (as it is not written as such), but US education system has been unable to "supply" enough people with post grad degrees for years now. Perhaps even more importantly, many of those graduated have been non-native, so "production" has been based on
imported raw materials...
There's nothing wrong with the latter, of course, if you consider it a new form of immigration.
It is risky however; if the flow of foreign-born students slows down (or dries up) hi-tech sector will get serious shortage of skilled researchers and engineers. Finally, the flow of people who already have all their education (from their home country or country they first immigrated to), is
what makes up of shortage of people graduating
from US universities (insert your favourite H1B bashing stuff here).
So far the reasons for foreigners to come to USA have been higher salaries (compared to almost any other country in the world) and high-level of post graduate university degrees. It has been a decent deal, too, since those people have already gotten their basic education (possibly including under grad studies, or perhaps the whole education),
and thus the costs to US educational system is minimal (considering they'll be paying taxes from high salaries pretty soon -- even though some people think H1B 'peons' are only paid peanuts, DoL restricts wages to reasonable levels, and those levels are sure to be higher than the average US salaries... even if not equal to salaries US born equals make)
Whether this will change in near future is anyone's guess. It might be wise for politicians to try to bolster education of "natives", however, just to make sure system is not too dependant on motivated bright people "imported" from other countries (I don't like using words like "import", "production" etc., as it's almost like admitting people are consumable products... but interestingly the flow of people has so many similarities to the trade of, say, crude oil, that the words are useful here for making analogies).
Whatever word processor they use, it'll have to
work well and have everything they need and want
it to.
And, perhaps even more importantly, has to coexist with/be able to talk to _other_ 'competing' (open, closed, whatever) applications.
If Free/Open s/w history is any indication, there'll be n+1 applications for any task you can
name. Having n+1 alternatives doesn't have to be a problem in itself, as long as they can talk to each other.
And now... having just read an article about Open/Star Office saving its documents in Open format (its 'own' XML dtd... but perhaps that grows to be a de facto Open Office Document standard?), which in turn uses other open XML-based languages (SVG, Dublin Core RDF vocabulary), I'm kind of optimistic about the future of open office applications. If all goes well, this may be the classic case of "whoever implements it first well enough creates a new de facto standard".
And the weird thing is that Open Office is heavily sponsored by a big good-old proprietary company (although one traditionally getting most of its income from h/w).:-)
But then again, Mozilla is (was?) funded by Netscape. Perhaps it is the way of future.
I don't really have experience to give 'real' answer, but I think that (like someone else already suggested) in many cases original founders / owners of the small company are:
In need of many (why sell otherwise)
Grown tired of the thing company does, or infighting, or too-long work days, or all of the above, meaning sell-out makes even more sense.
After selling possibly have sour grapes towards each other (how many is shared, was the price right, etc. etc. etc.)
So, in some cases perhaps they aren't in speaking terms, or at least not interested in getting back together to start a new business.:-)
And of course it may be that people get interested in other things to do; even though they know a great deal about certain subject they may well want to try out something new. That's (from my experience) what many talented people do; want to learn new things, tricks, try their wings in different areas. After all, after sell-out most people really have to take a deep breath and think about what to do next.
Oh and also the terms of sale matter; there may be non-competition clauses in there, just because it has happened often that people start a competing business. 8-)
...or from which lessons are not being learned. Multics, TOPS-10, and TOPS-20 come to mind. Any others?
And all this time I thought lessons from Multics were in fact well learnt and understood? You do know where name 'Unix' (originally, pun intended)
comes from, right?:-)
Yes, I guess YMMV. Thinking back, a sub-project that was completely written in C++ that did use STL was much slower to compile. Its size was in the same order of magnitude as yours (100 - 200kloc), but even that didn't take (AFAIR) more than something like 5 - 10 minutes. Of course that's (considering code size) much slower than the main application's compilation time.
One more thing about VC++ is that it really makes difference how you design your precompiled headers; usage is not quite as automatic as one would hope, so it's (too) easy not to get anything precompiled/cached.
... and of course Mac-heads could start talking about Metrowerks CodeWarrior. It supposedly has even faster compiler (on MacOS, anyways), and it certainly seemed to be able to do partial recompilation much better than VC++ (which sometimes partially compiled project and produced broken binary... ie. failed to really re-compile all it needs to, but nothing more).
Well, "right tool for the job". If you create emails, use text editors, if you need simple formatting, use simple word processors. If you need more features, use full-size office suites.
Not saying that your concern wasn't valid at all (it still kind of is, office suites are hogs), but...
Thing is, OpenOffice being based on open sourced Star Office code base (after Sun acquired the company that created it) was aimed at "full-featured" Office Suite market. Kind of like SUVs of "productivity" applications (ie. bloated, powerful, ugly). Thus, it wasn't started out from scratch. There are more light-weight word processors (and office suites) around, such as AbiWord, but they might (still) not be as mature as, say, Star/OpenOffice. So, having all the bloat already built-in it's much more difficult to trim the fat, than building a leaner application from scratch. But on the other hand, you do have a usable finished application to work with.
One thing I'm wondering though is the compilation time. The company I used to work for had a similarly-sized (ie. couple of millions of lines of C++/C-code) application, and it compiled in 5 - 10 minutes on Visual C++ (back then on 350 mhz machines). Much of the code was straight-forward C (not C++), and even C++-code didn't make heavy use of many of C++'s slow-compilable features (templates)... And VC++ has a good compiler plus pre-compiles headers nicely. Still, more than an order of magnitude slower compilation sounds weird; it shouldn't take hours to compile that thing. Fortunately end users need not worry about that. The reason I would worry (as a developer) is that slow compilation is often caused by too many dependencies between classes that shouldn't be dependant on each other, which is usually a sign of problems at architectural level. Encapsulation and insulation should be used to reduce physical dependencies, not just logical ones (book "Large-scale C++ - projects" is a good one for reading more about the problems and solutions).
Just a small comment; even though US administration certainly has lots of indirect influence on other countries' jurisdictions, it's not all that straight-forward to shut down foreign companies/websites. If I'm not mistaken, ZK is a canadian company, and like IP-name says, anon.penet.fi was a finnish site (a hobby of a finnish Internet pioneer who got lots of unfair shit because of a few uninforment news articles back then).
And no, it's not a co-incidence that practically all anonymity-enhancing services have been located outside US of A for years now.
Yeah, that's why IBM PC compatible clone market died in 80s; nobody wanted non-proprietary hardware platform. Same for the TCP/IP and WWW on 80s and 90s; neither had a chance against cool stuff like decnet, netware and various cool hypertext protocols that rule nowadays.
As long as it can be muted before I ever hear a single syllable, it's fine with me.:-)
I guess I would pay (if I had to) for a PC that doesn't talk back to me. Hmmh... but think about this; if Windows tried to soothe you when BSOD arrives, what would you do? Feel relaxed and smile?
My preference for output method might be a simple led-display (with just one or two digits, referring to error codes), but I'm not sure how that could easily be pre-configured to be fail-safe like simple aural devices (as they need not be connected to case). Or perhaps a set of red/green leds, indicating progress of self-diagnostics (although, once again, to see the leds, they'd need to connect to the case or you have to open the case).
If someone puts Linux on a toaster, he doesn't
have to supply you with source!
Note that GPL doesn't force you to supply source with binaries, but it does force you to give that option, ie. if asked, to provide the source. There's a big difference here.
But aside from that point, why is toaster / embedded device so different from software-only product? Are they not shipping GPLed program code in binary form; be it embedded or stand-alone? Isn't this exactly what GPL covers? The only difference is the visibility; it's easier to obfuscate / hide origins of software when software is embedded, ie. not directly accessible. Aside from that, how is this so different?
Most people don't really care if you could build stuff yourself. If a company sells a decent product -- even if you could build such a thing bit cheaper/for free yourself -- it may well be selling tons of the product. Especially in this case, since it's an embedded product, which means that h/w (small embedded device?) is an integral part of the success (?) of the product.
In short; you can't download the physical system they are selling.
Unless I'm mistaken, the exact problem is just that; they are not giving away the source code... (and more than that not even acknowledging they have any obligation to do so).
If they meant "optical or magnetic storage media," they
should have said that.
I must admit I don't remember exact wording of the whole GPL by heart, but I would be surprised if it did not explicitly and exactly specify what constitutes acceptable provision of source code. And that definition is probably slightly more verbose than simple "physical media".
And as to cost of electrons; they may be free, but how about shipping them? (which original poster mentioned in addition to the cost of the media itself)?:-)
One note; I understand PU companies would offer more infrastructure (roads, physical DSL/cable/whatever lines) than actual services (ISP). It may be that due to opposition from big players they have to offer some bare-bones ISP too, but in the perfect world, they wouldn't have to.
I know car/road analogies are used too often and seldom fit nicely, but let's try this one; think of getting paved roads for your tax money, but paying yourself for gas (and indirectly, thus, gas stations), car, maps and all other service that relies on existence of roads. Similarly, having option to get your bare-bones wiring (and routing at least up to ISP) as low-level infrastructure thing, not as a service.
The sad fact is there is no money to be made in providing cabling and basic connectivity.
At least no more than in providing for paved roads; there are toll roads, but for the most part roads are in fact public. Money is supposed to come from services; few people are disputing that basic assumption. So far telephone co's and others have tried milking (big) companies, knowing end users won't be as profitable (due to cust. support etc). Now that companies are tightening their belts, this income source has mostly dried up.
Bolting a GUI on top of an existing OS may look like a losing idea, but what happened to noble goal of modularity? Separation to achieve clarity, layering functionality to allow for more parallelism on development? I'm not claiming Unix with X-Windows is necessarily a good example of this, but it really isn't as bad as some people think. The real beef is of course whether low-level windows management (X-windows) and GUI itself (widgets, toolkits) should be integrated or not.
At one point he said something along the lines of "tying OS more tightly to GUI". I think he miss-phrased it (rather, "modifying GUI to make more use of the nifty features OS provides for"); if he didn't it seems like an ass-backwards way... Windows has been widely criticized for mixing up the separation between OS, GUI and apps. There is such thing as 'too much integration'; for embedded devices it's more acceptable, but on work stations...
Re:The correct name for these bricks is LEGOS
on
When Lego Meet Rubik
·
· Score: 2
I'm a European enough, and in my native language we most certainly call them "legot" (plural of "lego"). "Play with legoS" instead of "play with lego", and so forth. And it doesn't sound stupid (in my ears) in any of the languages I know of. YMMV I s'pose. I could even say I've never met anyone before that thinks "lego" is an adjective, or a purely singular noun (like some words in english, "luggage", "furniture" etc are, always used in singular form, even if meaning plural things).
Any danes around who could give "definite" answer to this dilemma?!? Is "Legoer" (or something like that) a valid term in the land of Hamlet?
Could be as I haven't seen any mention of that (nor about BeOS version). For OS/2 there seemed to be a few enthusiasts (polls for getting Opera port got most support from there), so of course it's possible some brave coder might step up but...
Slashdot Mirror
My apologies if I missed something obvious, but why not just do end-to-end encryption and go through any whichever country happens to be in the route? That is assuming you are not as worried about espionage (eavesdropping, man-in-the-middle) as about legal issues with US residents accessing stuff.
Apologies for snipping out just one small piece, but I think it might be enough to allow me to refine my point. :-)
I think that the words "I do" still don't have any intrinsic context-independent 'magical' powers. You are right in that in right context they do have more meaning than otherwise, thanks to traditions; semantics has been added as "I do" has become an idiom (?) in wedding ceremonies. And depending on how people feel about the ceremony, political gathering, rally (whatever context words are used in), the words do get more weight, more power, from their surroudings. And then there are people couldn't care less about the very same words, even if being part of the ceremony (few people would just ignore them when being wed though...)
What I mean is that words, sentences, even phrases, usually have a relatively small domain in which they have added connotations, and perhaps additional powers of convincing people. I don't think any words have "super-powers", to make them taboo (or holy or whatever) independent of context; independent of intentions / attitudes of person who utters the words, independent of social, political etc. situation/context of discussion or speech.
I guess my rant against extreme-PC'ism doesn't really completely relate to gender-dependencies, though, as it's a much larger topic. Not just whether 'he' should always be used, whether it's a "tainted" word (I don't think anyone claims it is); more like whether it should always be used as the 'neutral' or 'generic' pronoun just because that's what has traditionally been used. I have no idea; not being a native english speaker I can't really "feel" one way or the other. I do think, though, that it's good not to try to manipulate a living language in Orwellian NewSpeak way (something you mentioned too). Then again, it's not like PC was the first movement that tries to "guide the language"; the trends are as old as english linguistic studies. :-)
Finally, I do agree that using (gender-) neutral terms makes perfect sense, when gender isn't really the point. Even if "gender-coding" has just a slight influence, that influence is unlikely to be useful or positive. Just like I personally dislike colour coding baby clothes ("pink for girls, blue for boys"), just because that's what people have traditionally done. Or always buying dolls for girls and toy cars for boys (independent of what they might like to play with), "just because that's what they always play with". But I digress... :-)
Hmmmh. Perhaps it affects to small degree, but some people tend to believe it even defines the way people think. I think this is akin to superstition; the idea that words have magical powers. And thus the 'extreme' political correctionists (?) really look foolish; condemning 'evil words' believing it's just bad bad words (nigger etc), not the attitudes of the people who use the words. And the result is that new 'correct' words eventually get 'polluted'. If you really want to present a derogatory racist comment, you can easily get phrase african-american (for example, replace with your favourite ethnic euphenism here) sound like an insult, by varying tone and facial expressions. :-/
At any rate, a slightly on-topic remark; there are also languages that have no gender distinctions whatsoever. I'd even venture a guess that genders (for nouns) are more common in western (european) languages than in other ones (correct me if I'm wrong... could well be, I just know japanese doesn't have genders for nouns). I still don't believe it changes the way people think all that much. I don't feel like I'm more egaliatrian (or emansipated?) towards women than my american friends, just because my mother tongue doesn't have he/she distinction. If I do it's about culture, upraisinh and socio-economic things, not the vocabulary, syntax or semantics of language(s) I speak.
It really depends. Usually telephone network was built by either state owned companies or companies heavily subsidized by state (practically everywhere in the world). And that's because physical communication networks, like other basic infrastructure (like roads, railroads, powerlines), are expensive to build and potentially costly to maintain, without actually directly producing anything. They 'only' provide for actual services; they are not services per se.
Thus, to begin with, it's seldom the case that fully private companies completely funded "their" networks using just their money. But in addition, few people think it would be beneficial to have more than one (or two, perhaps three for redundancy, at most) phone line networks (or, more than one redundant cable network). It just would usually be wasting lots and lots of money. Most of the time people need just one line, and want to choose the service provider (be it for cable or phone line... of course these may soon be combined, just need one physical highband network for all communication). Instead, it's thought, it would make sense to share the physical network, and compete on services.
I understand the feeling "it's mine all mine, why should I play with others". It's familiar feeling from childhood. But putting things in perspective, it really would be foolish to force competitors to build their own duplicate networks. That would be really really really inefficient; and as such would enforce de facto monopolies (that telcos now have... the weak competition between cable and telcos is just an excuse to prevent real competition)
The problem left is, of course, what is the 'fair cost' of a physical line. It would be much easier if the lines were taken care of by non-profict organizations, perhaps utility companies, perhaps counties. But unfortunately that's not the case. And that's why legislation tries to figure out a somewhat fair way to allow for actual competition.
The goal is to create images that reflect what small segments of the
population want. Life becomes less of a search and more of a pick and choose. Employees become
employees around the clock, walking viral marketers.
That is one way to view it, but I think (having skimmed the review) the idea was not to FORCE (or coerce in any way) people to MARKET their company or its product(s). Instead it's to allow employees to do what they would do otherwise; when/if they are happy working there, they will usually talk about the company to their friends, relatives. Just normal conversation. And that's likely to be (at the microlevel at least) more effective PR, perhaps even marketing, than high-level mass market advertising.
I mean, think about it. You have a job, at company X, working on product Y. You don't walk around asking your friends to go buy Y when it comes available. That's a bad idea. However, if you like what you are doing, and are interested in creating the product, you are likely to mention the job, company, products. If someone happens to be looking for a product in category Y is in, you are highly likely to mention existence (and features) of Y, without marketing. Or should you try to keep your mouth shut, never mention it? Same thing about the company in general. You like it there? Does it have nice benefits? Sane managers? Good products? People do ask these questions, just small talk. But the same thing applies to web too; home pages people have usually do contain stuff about their job (as well as family, photos, whatever). No one is asking (or should ask for that matter) employees to put company's logo in the page, or to add direct marketing stuff. And yet people usually do add links; perhaps mention something more about actual work (projects etc) they are doing.
Of course you don't need to talk about micro-marketing here. It's just good old "let's keep employees happy as that's the best PR we can have" approach. There are a few high-tech companies that have good reputation, partly because of the way they treat their employees (AFAIK companies like Cisco, Sun, even Microsoft have pretty decent reputation are where they are thanks to being good employers).
And based on what I read about the Clue Train Manifesto, fundamental idea is to really empower (sic! buzz-word, I know) employees, literally, to let them be creative, use their energy. Let employees have their own web pages, at work, at home. Let those contain whatever people want to put there. More restrictions, less useful things people do on their free time voluntarily. No one wants to be a corporate peon, much less 24/7.
But interestingly enough, with freedom and equipment, oftentimes resulting web-pages are surprisingly useful; especially spontaneously created intranet pages. Even if no one "authorizes" or "plans" or "orders" help-pages, how-to - pages, people are likely to create such, as long as they are not explicitly prevented to. Read the Cluetrain Manifesto for more about these ideas if interested. It's good reading.
I think RMS and Larry are native US citizens, but
Linus is a finn, so even though he probably wouldn't really want to, he could easily move
back to Finland. Or to any other EU country, basically, as he wouldn't need any visas to be
able to work (unlike in US... though he probably already has a green card by now).
Then again, I remember String and StringBuffer having some interesting smugs way of avoiding unnecessary copies (ie. sharing array for as long as possible), and perhaps even that could prevent excessive array allocations/copies.
Hmmh. Time to check out java decompilers again.
However, since the linux operating system is
being used, that negates these factors.
Well, Linux was
originally created as a unix-clone that runs
efficiently on low-end (usually single user)
systems. Thus, performance and efficiency were
important when doing architectural / design
choices; scalability and scheduling fairness
(for example) were of lesser importance (and
that bit when trying to make kernel SMP-friendly)
Now,
obviously OSes designed specifically for ultra low-end embedded market (QNX?) are likely to
be even more efficient, but really, StrongArm
with ~200 mhz clock frequence has plenty of
power to spare, especially when it's only feeding
tiny display (like you said). OS overhead is
likely to be negligible.
...it won't be interesting for big
corporations (like Sun, IBM, HP) to invest in Amaricans Universities. So the college
education will lower quality, which will take a step back for US economy in a long term.
Don't take this as US bashing (as it is not written as such), but US education system has been unable to "supply" enough people with post grad degrees for years now. Perhaps even more importantly, many of those graduated have been non-native, so "production" has been based on
imported raw materials...
There's nothing wrong with the latter, of course, if you consider it a new form of immigration.
It is risky however; if the flow of foreign-born students slows down (or dries up) hi-tech sector will get serious shortage of skilled researchers and engineers. Finally, the flow of people who already have all their education (from their home country or country they first immigrated to), is
what makes up of shortage of people graduating
from US universities (insert your favourite H1B bashing stuff here).
So far the reasons for foreigners to come to USA have been higher salaries (compared to almost any other country in the world) and high-level of post graduate university degrees. It has been a decent deal, too, since those people have already gotten their basic education (possibly including under grad studies, or perhaps the whole education),
and thus the costs to US educational system is minimal (considering they'll be paying taxes from high salaries pretty soon -- even though some people think H1B 'peons' are only paid peanuts, DoL restricts wages to reasonable levels, and those levels are sure to be higher than the average US salaries... even if not equal to salaries US born equals make)
Whether this will change in near future is anyone's guess. It might be wise for politicians to try to bolster education of "natives", however, just to make sure system is not too dependant on motivated bright people "imported" from other countries (I don't like using words like "import", "production" etc., as it's almost like admitting people are consumable products... but interestingly the flow of people has so many similarities to the trade of, say, crude oil, that the words are useful here for making analogies).
Whatever word processor they use, it'll have to
work well and have everything they need and want
it to.
And, perhaps even more importantly, has to coexist with/be able to talk to _other_ 'competing' (open, closed, whatever) applications.
If Free/Open s/w history is any indication, there'll be n+1 applications for any task you can
name. Having n+1 alternatives doesn't have to be a problem in itself, as long as they can talk to each other.
And now... having just read an article about Open/Star Office saving its documents in Open format (its 'own' XML dtd... but perhaps that grows to be a de facto Open Office Document standard?), which in turn uses other open XML-based languages (SVG, Dublin Core RDF vocabulary), I'm kind of optimistic about the future of open office applications. If all goes well, this may be the classic case of "whoever implements it first well enough creates a new de facto standard".
And the weird thing is that Open Office is heavily sponsored by a big good-old proprietary company (although one traditionally getting most of its income from h/w).
But then again, Mozilla is (was?) funded by Netscape. Perhaps it is the way of future.
So, in some cases perhaps they aren't in speaking terms, or at least not interested in getting back together to start a new business.
And of course it may be that people get interested in other things to do; even though they know a great deal about certain subject they may well want to try out something new. That's (from my experience) what many talented people do; want to learn new things, tricks, try their wings in different areas. After all, after sell-out most people really have to take a deep breath and think about what to do next.
Oh and also the terms of sale matter; there may be non-competition clauses in there, just because it has happened often that people start a competing business. 8-)
...or from which lessons are not being learned. Multics, TOPS-10, and TOPS-20 come to mind. Any others?
And all this time I thought lessons from Multics were in fact well learnt and understood? You do know where name 'Unix' (originally, pun intended)
comes from, right?
One more thing about VC++ is that it really makes difference how you design your precompiled headers; usage is not quite as automatic as one would hope, so it's (too) easy not to get anything precompiled/cached.
... and of course Mac-heads could start talking about Metrowerks CodeWarrior. It supposedly has even faster compiler (on MacOS, anyways), and it certainly seemed to be able to do partial recompilation much better than VC++ (which sometimes partially compiled project and produced broken binary... ie. failed to really re-compile all it needs to, but nothing more).
Not saying that your concern wasn't valid at all (it still kind of is, office suites are hogs), but...
Thing is, OpenOffice being based on open sourced Star Office code base (after Sun acquired the company that created it) was aimed at "full-featured" Office Suite market. Kind of like SUVs of "productivity" applications (ie. bloated, powerful, ugly). Thus, it wasn't started out from scratch. There are more light-weight word processors (and office suites) around, such as AbiWord, but they might (still) not be as mature as, say, Star/OpenOffice. So, having all the bloat already built-in it's much more difficult to trim the fat, than building a leaner application from scratch. But on the other hand, you do have a usable finished application to work with.
One thing I'm wondering though is the compilation time. The company I used to work for had a similarly-sized (ie. couple of millions of lines of C++/C-code) application, and it compiled in 5 - 10 minutes on Visual C++ (back then on 350 mhz machines). Much of the code was straight-forward C (not C++), and even C++-code didn't make heavy use of many of C++'s slow-compilable features (templates)... And VC++ has a good compiler plus pre-compiles headers nicely. Still, more than an order of magnitude slower compilation sounds weird; it shouldn't take hours to compile that thing. Fortunately end users need not worry about that. The reason I would worry (as a developer) is that slow compilation is often caused by too many dependencies between classes that shouldn't be dependant on each other, which is usually a sign of problems at architectural level. Encapsulation and insulation should be used to reduce physical dependencies, not just logical ones (book "Large-scale C++ - projects" is a good one for reading more about the problems and solutions).
And no, it's not a co-incidence that practically all anonymity-enhancing services have been located outside US of A for years now.
Yeah, that's why IBM PC compatible clone market died in 80s; nobody wanted non-proprietary hardware platform. Same for the TCP/IP and WWW on 80s and 90s; neither had a chance against cool stuff like decnet, netware and various cool hypertext protocols that rule nowadays.
I guess I would pay (if I had to) for a PC that doesn't talk back to me. Hmmh... but think about this; if Windows tried to soothe you when BSOD arrives, what would you do? Feel relaxed and smile?
My preference for output method might be a simple led-display (with just one or two digits, referring to error codes), but I'm not sure how that could easily be pre-configured to be fail-safe like simple aural devices (as they need not be connected to case). Or perhaps a set of red/green leds, indicating progress of self-diagnostics (although, once again, to see the leds, they'd need to connect to the case or you have to open the case).
If someone puts Linux on a toaster, he doesn't
have to supply you with source!
Note that GPL doesn't force you to supply source with binaries, but it does force you to give that option, ie. if asked, to provide the source. There's a big difference here.
But aside from that point, why is toaster / embedded device so different from software-only product? Are they not shipping GPLed program code in binary form; be it embedded or stand-alone? Isn't this exactly what GPL covers? The only difference is the visibility; it's easier to obfuscate / hide origins of software when software is embedded, ie. not directly accessible. Aside from that, how is this so different?
In short; you can't download the physical system they are selling.
they'll probably just give away the source, too.
Unless I'm mistaken, the exact problem is just that; they are not giving away the source code... (and more than that not even acknowledging they have any obligation to do so).
If they meant "optical or magnetic storage media," they
should have said that.
I must admit I don't remember exact wording of the whole GPL by heart, but I would be surprised if it did not explicitly and exactly specify what constitutes acceptable provision of source code. And that definition is probably slightly more verbose than simple "physical media".
And as to cost of electrons; they may be free, but how about shipping them? (which original poster mentioned in addition to the cost of the media itself)?
I know car/road analogies are used too often and seldom fit nicely, but let's try this one; think of getting paved roads for your tax money, but paying yourself for gas (and indirectly, thus, gas stations), car, maps and all other service that relies on existence of roads. Similarly, having option to get your bare-bones wiring (and routing at least up to ISP) as low-level infrastructure thing, not as a service.
The sad fact is there is no money to be made in providing cabling and basic connectivity.
At least no more than in providing for paved roads; there are toll roads, but for the most part roads are in fact public. Money is supposed to come from services; few people are disputing that basic assumption. So far telephone co's and others have tried milking (big) companies, knowing end users won't be as profitable (due to cust. support etc). Now that companies are tightening their belts, this income source has mostly dried up.
At one point he said something along the lines of "tying OS more tightly to GUI". I think he miss-phrased it (rather, "modifying GUI to make more use of the nifty features OS provides for"); if he didn't it seems like an ass-backwards way... Windows has been widely criticized for mixing up the separation between OS, GUI and apps. There is such thing as 'too much integration'; for embedded devices it's more acceptable, but on work stations...
Any danes around who could give "definite" answer to this dilemma?!? Is "Legoer" (or something like that) a valid term in the land of Hamlet?
Could be as I haven't seen any mention of that (nor about BeOS version). For OS/2 there seemed to be a few enthusiasts (polls for getting Opera port got most support from there), so of course it's possible some brave coder might step up but...