Make Dr. Fox explicitly completely insane. That way, you can have _two_ factions of people going back in time: one, the lunatics Dr. Fox sent back in time; two, the lunatics sent back in time by the authorities to stop the other lunatics who got sent back in time.
Should make things more interesting having the two "sides", IMO. I think I'll suggest this to these folks directly, too. ---
agreed. however, I view violation of a contract as immoral, at least, irrespective of legality, and vice versa. don't misunderstand me to be indicating that they always coincide. ---
Quite true. However, words still cary a lot of semantic baggage from their original meanings for a very long time. Of course the third meaning of "piracy" is accepted now, but that doesn't mean it doesnt inherit a significant amount of meaning from the older definitions.
People hear "piracy", they automatically think of things in terms of property crimes. Unauthorized copying, although wrong, is not a property crime.
However, thanks to this shift in usage, every time someone tries to make that point, when they propose a (more) neutral term be used instead, they're accused of trying to subvert the language.
The use of a neutral term in such a discussion is however necessary because the association between "piracy" (as in unauthorized copying) and "piracy" (as in material theft) is reinforced every time the word is used in that fashion. (It just so happens that the idea of unauthorized copying as "property crime" is extremely favorable to the arguments of organizations like the SPA, who initiated the use of the term in the first place.)
Capitalizing on semantic shifts is a great way to get people thinking in terms favorable to your case without them being aware of it, or even being able to rationally discount them. It's not always intentional, but it's something you have to be extremely careful about when choosing words, regardless.
Like it or not, your language has a substantial effect on the way you think. You really learn a new language, you learn a new way of thinking, quite literally. If you can effect a semantic shift, you CAN alter people's view of reality. ---
> I don't walk around telling carpenters that > they should build houses for free and I don't > expect software to be handed to me for free.
As long as we're using oversimplified analogies...
What if the physical laws of the universe were such that, once a house had been constructed by a carpenter (or by carpenters working for a contractor), any shmuck off the street could take a look at the house, wave his hands for a couple seconds, and magically get his own house just like it?
The two obvious solutions in such a case are:
often people have specific needs in a house that an existing house model just can't satisfy. Carpenters would be paid to take care of such things, making modifications to existing houses, and making bigger and better new houses. That way, everyone gets a house, and there's still an economic benefit (incentive) to being a carpenter (or a contractor, for that matter).
make it illegal for anyone but the original carpenter or contractor to copy a house without express permission. They'll be able to charge for the mere act of copying a house that way, too. Yes, you've created an artifical economy of scarcity, and increased the price of an individual house. So what? Yes, those people too poor to have one built or buy an existing one will just have to be faced with the choice of either going homeless or copying houses illegally, but hey, carpenters and contractors should be able to make a living, right?
> Most people do not think piracy is wrong because > most people do not identify it as > steeling (because there is no missing object)
If there's no missing property, it is by definition not stealing. That doesn't automatically make it right, however, and certainly doesn't automatically make it legal. ---
In economic parlance, software is "non-exclusionary", which means that you can have a piece of software without inherently denying it to someone else at the same time. (i.e. you can have your cake and someone else can eat it, too)
That is why it is licenced, instead of being sold. (If it were sold to you, you would pretty much have complete rights to do whatever you wanted with it, included sharing it with your friends, so instead, companies withold rights to their software and allow you to use/copy it under very restricted conditions.)
So, as a result, you're not actually taking a copy of software from a software licensor and giving them money in exchange, but rather getting the pretty cardboard packaging and media for free (modulo overhead), while paying them money for the privilege of agreeing to a contract (license agreement) that has the net effect of turning the software from a good into a service which you may have to pay for repeatedly?
Of course software makers have a right to be compensated for their work; just, shouldn't it be for their _work_, rather than for the software itself, which has no economic value? (look up the definition of a "public good" -- software, as a good, fits it to a tee). If you'll notice, most programmers don't get payed by the companies they work for for the software they write, but their labor. Why should the users pay for the software itself, instead of the packaging, labor, and any value added, if software is a good?
I suppose this is the question: is software a public good, or a service (in and of itself)?
Incedentally, IANAL, but isn't unauthorized copying ("piracy") either a breach of contract or a tort, rather than a property crime?
It's one thing to take away (steal or destroy) property (or money) that a party already posesses -- that's a property crime. Isn't unauthorized copying, where the licensor is only denied _potential_ property (money), a matter of breach of contract (violating the software license) and/or tort (wronging them by denying them the money which they claim to have a natural "right" to) instead? ---
The best thing, IMO, would be to use MDI, and have the option of spawning additional MDI containers as needed. That serves to keep related browser windows nicely grouped together, while allowing you to play the workspace game, too.
Bonus if you can drag windows between MDI containers. (hrm... potential "killer feature" for the Qt version? I know that'd be pretty painful to implement in Windows, so I doubt we'd see it there... might be easier with a from-scratch MDI implementation) ---
> For one, seeing as how we are constantly > observing astonomical phenomenon, this would be > a good way for the aliens to improve our chances > of actually seeing the message, and two, we'd > probably notice any glaring anomalies. So, Could > this be the one?
What, by modulating the brightness of the star? I can just picture it...
SETI researcher #1: Hey, Dale, get over here... I think I've got the first character of the alien transmission...
SETI researcher #2: 'N'...
SETI researcher #1: I wonder what it means...
SETI researcher #2: Someday, in ages to come, our children shall know...
--- a couple hundred years later ---
SETI4 researcher #1: Enod, we've got another character of the transmission... five characters now...
SETI4 researcher #2: Someday, in ages to come, our children shall know...
--- a few thousand years pass ---
N'GANTHOK researcher #1: The alien message is complete, Miznok! Finally, we shall read together the cosmic truth sought by our anscestors for so many ages...
N'GANTHOK researcher #2: "NEED HELP SEND MORE BEER"...
> Are there any other stars by Eta Carinae they > can compare this to?
Some, yes. And they all seem normal for the most part.
> And how does this tie in to the way we usually > predict stars die, if that is what this thing is > indeed doing?
It doesn't... astronomers just can't figure out (pardon my french) what the fuck is going on with this thing.
In a way, it kind of sucks that it's already blown up or whatever 7500 years ago; we're just now getting to see what went on (and are just now going to be affected by it). At least it looks cool. ---
I don't have Windows, nor am I willing to pay money to license it. What is a person like me to do? Methinks it might have a bigger impact if everyone who wanted the service didn't play these games, and insisted on getting it installed in their Linux machine, or not at all. Right now, they don't, and so they aren't visible as Linux users who want the service. It's quite easy to interpret as a lack of demand, then. ---
> All these [hardware] companies would have to > drop their software divisions.
I'm not sure they'd miss them, either; in many cases, right now they're just playing loss-leader. Yes, they make a net profit, but often they spend more capital on the software development than they get back from it in revenue, and make it up from some of the revenue generated by their hardware.
There's a largely undocumented kiosk mode in Navigator. I forget the details, but you can find out about it deep in the bowels of the Netscape developer site.
> I was wondering if any SeaMonkeyUnix sources are > planned for integration with Berlin for DOM > stuff?
Maybe. Berlin isn't quite to the point yet where we can make those kinds of decisions. We're still a little further down the design/implementation ladder. Maybe in the next six months to a year.
It has already come up on the ML, though; you should really start following the archives if you're interested (better yet, subscribe and help development directly).
Wouldn't need to integrate the stuff in the actual codebase anyway; most of the stuff should be usable via CORBA, which is extremely pervasive in the Berlin architecture (read: everything is implemented using it). (just FYI, CORBA doesn't incur any extra function call overhead over normal C/C++ for method invocation between in-process objects, so performance should be quite decent)
> Was libggi/libgii port of SeaMonkey ever > attempted?
No, you'd need: a) a raw libggi windowing toolkit b) a widget set to use on top of it
I did write the first draft of libgwt (which was a step towards item a), but I never really had time to keep after it, and the other folks development fell to had the same problem.
Assuming that libgwt was finished at some point, all you'd probably need to do for item b is to port GDK to libgwt, and you've got GTK+, so the GTK FE could then be used on top of libggi directly.
> Should Berlin make sense for a SeaMonkey hack to > get a smaller memory footprint?
I don't understand the question. email me if you like, rather than continuing this offtopic thread.
> Would it be that hard for them to port it to > Linux if they already have IE 4.01 and IE 5.0 > for Solaris and HP-UX? (They do have it > available for download.) I mean, would there be > a lot of code to rewrite If they already got the > thing working on other POSIX compliant OS's?
It would really depend on how careful/competent the porting team had been. It's easy to write to the POSIX spec and expect the software to build and work on any POSIX-compliant OS, but it's another matter to write it to a specific POSIX-compliant OS (or any small set of them) and then expect it to build and work on any other POSIX-compliant OS.
This is especially true when you throw architectural differences into the mix; SPARC and PA-RISC are both 64-bit big-endian architecures (well, PA-RISC might be little-endian; I can't remember). It may well be that their IE for Unix codebase (hopefully they didn't do two individual ports) contains a lot of 64-bitisms. I have learned not to expect good software engineering practices from them.
> Even though I LOVE Linux, I do welcome IE, even > though it's from M$, It is the superior browser, > that wasn't the case when they were both like > 2.0, but the more they progress, the more bugs I > see in netscape, and the the more I see IE > handle browsing the web better and better.
IMO, the existing (old) Netscape codebase is slowly collapsing under its own weight. It's high time a Mozilla-derived product got into Alpha...
> The biggest reason I welcome IE to Linux is > because I'd like to see a Linux browser that > properly handles Cascading Style Sheets. I have > yet to see that on Linux browsers.
I have yet to see that on Microsoft browsers. Try some of the CSS conformance tests in IE sometime.
Personally, I'd rather wait for Mozilla on Linux than for IE. Yes... I know... Mozilla might not not be all there yet, but IE for Linux sure as hell isn't.
> I guess this puts to rest all the arguments of > those people who were saying that it's > slashdot's setup, not the Linux kernel, that was > causing the problem. Apparently it was a bug in > the kernel itself. A known bug, but a bug > nonetheless, and a serious one that caused the > site to lose uptime at that.
If you're talking about the latest batch of downtimes, yes.
However, I should point out that most of the problems in the past (and Rob has announced the causes in a similar fashion to this in the past) have been due to mySQL crapping out, or more occasionally bugs in the particular patchlevel of mod_perl that Rob was running. Keep in mind, kernel 2.2.8 hasn't even existed for most of/.'s lifetime.
The bug was stomped pretty quickly in 2.2.9 as I understand, too; the only reason it's not fixed for/. is because Rob hasn't upgraded from 2.2.8 yet
It is dissapointing to have a stable kernel that's this flaky (although it is just the specific driver). I'm sure we'll have our share of 2.0.34s in the future, too.
I still think that _on the balance_ (and I think history bears out my point here), Linux does very well in terms of low bug rates and fast bug fixes.
I should have persisted in trying to find actual documentation before posting. JOOC, I seem to remember there being some controversy with NeXT, though -- was that just a matter of caving-under-bad-pr, or further misinformation on my part?
> I don't know enough about the itty-bitty details of graphics-device > interfaces to take a particular stand on whether they should go into > the kernel or user space or a little of both (the last seems most > likely).
That is in fact what GGI actually does. KGI drivers are typically pretty thin (generally enough to successfully arbitrate hardware access among a number of userland processes) and then most of the stuff goes on in LibGGI. LibGGI, incdentally, can work on top of a lot of stuff besides KGI (i.e. X), so it's compatible with other Unixes (even those without a KGI layer) too.
GPL has held up in court
on
BSD vs GPL
·
· Score: 1
Ever notice the GNU Objective C compiler? The NeXT folks took gcc and adapted it for their Objective C compiler. Only problem is, they tried to make the resulting compiler proprietary. The FSF took them to court. The FSF won, and now GCC compiles Objective C too. There have been a handfull of other cases too.
sounds like a stupid compiler thing
on
Linux 2.3.0
·
· Score: 1
the variables probably need to be marked 'volatile' to keep the compiler from optimizing them away.
> Anyone else tried ttyquake? Just get the version > of svgalib that is a hack over the aalib (?) > char cell graphics libraries, then use > LD_PRELOAD to grab them... Weird.
Alternatively, use LibGGI via the svgalib wrapper with GGI_DISPLAY set to "aa".
That way, depending on what GGI_DISPLAY is set to (if unset, it autodetects), you can use the same rig without further modification to display on X or on a monitor wall or on console with fbcon or all kinds of other really freaky shit (see http://www.ggi-project.org/screenshots.html) -- these people are EVIL&#*@^$*#&^
Just judging by the majority (75% currently) of the replies, it would seem that my experience is atypical. Some of these replies (which are all quite good) should be scored up somewhere closer to the original post.
> If "they" don't teach generalized skills at your > school, you need to get a better school.
I think so too. However, _so far_ none of the CS programs that I've looked at in other schools look better enough to bother switching. This may just be a regional phenomena.
> In these parts, they teach classes on: > Databases (including all sorts of theory, including some subsets of logic) > Operating systems (which is to say, general principles underlying operating systems, not "How to configure NT") > Computer languages (overview of different types of languages, including Prolog, Lisp, and others.) > Assembly language > Data structures > Algorithms > Circuit design > And a whole lot more
> And, oh, they teach all of this using Unix systems
This is all refreshing to hear.
> and require that CS students take twice as many non-CS classes as CS ones.
Well, that much is a given, even if you're stuck in some crappy school like myself.
> And I don't think the U. of Wisconsin is particularly unusual in all these regards.
It'd be interesting to do a general survey of various colleges' curriculae and see exactly what is the most typical case. I suspect our individual perceptions are both pretty colored by regional variations.
Regardless, I think my comments about high school in particular are still applicable -- certainly, the majority of people from around the country I've talked to have indicated the same kind of deficiencies I've described. I haven't talked to as wide a variety of people about college in the same way -- at least not yet.
Slashdot Mirror
Make Dr. Fox explicitly completely insane. That way, you can have _two_ factions of people going back in time: one, the lunatics Dr. Fox sent back in time; two, the lunatics sent back in time by the authorities to stop the other lunatics who got sent back in time.
Should make things more interesting having the two "sides", IMO. I think I'll suggest this to these folks directly, too.
---
Nomad.
*grins*
---
> legal != right
agreed. however, I view violation of a contract
as immoral, at least, irrespective of legality,
and vice versa. don't misunderstand me to be
indicating that they always coincide.
---
> Definitions of words CHANGE.
Quite true. However, words still cary a lot of semantic baggage from their original meanings for a very long time. Of course the third meaning of "piracy" is accepted now, but that doesn't mean it doesnt inherit a significant amount of meaning from the older definitions.
People hear "piracy", they automatically think of things in terms of property crimes. Unauthorized copying, although wrong, is not a property crime.
However, thanks to this shift in usage, every time someone tries to make that point, when they propose a (more) neutral term be used instead, they're accused of trying to subvert the language.
The use of a neutral term in such a discussion is however necessary because the association between "piracy" (as in unauthorized copying) and "piracy" (as in material theft) is reinforced every time the word is used in that fashion. (It just so happens that the idea of unauthorized copying as "property crime" is extremely favorable to the arguments of organizations like the SPA, who initiated the use of the term in the first place.)
Capitalizing on semantic shifts is a great way to get people thinking in terms favorable to your case without them being aware of it, or even being able to rationally discount them. It's not always intentional, but it's something you have to be extremely careful about when choosing words, regardless.
Like it or not, your language has a substantial effect on the way you think. You really learn a new language, you learn a new way of thinking, quite literally. If you can effect a semantic shift, you CAN alter people's view of reality.
---
> they should build houses for free and I don't
> expect software to be handed to me for free.
As long as we're using oversimplified analogies...
What if the physical laws of the universe were such that, once a house had been constructed by a carpenter (or by carpenters working for a contractor), any shmuck off the street could take a look at the house, wave his hands for a couple seconds, and magically get his own house just like it?
The two obvious solutions in such a case are:
Yes, you've created an artifical economy of scarcity, and increased the price of an individual house. So what? Yes, those people too poor to have one built or buy an existing one will just have to be faced with the choice of either going homeless or copying houses illegally, but hey, carpenters and contractors should be able to make a living, right?
Of course they should.
---
> Most people do not think piracy is wrong because
> most people do not identify it as
> steeling (because there is no missing object)
If there's no missing property, it is by definition not stealing. That doesn't automatically make it right, however, and certainly doesn't automatically make it legal.
---
In economic parlance, software is "non-exclusionary", which means that you can have a piece of software without inherently denying it to someone else at the same time. (i.e. you can have your cake and someone else can eat it, too)
That is why it is licenced, instead of being sold. (If it were sold to you, you would pretty much have complete rights to do whatever you wanted with it, included sharing it with your friends, so instead, companies withold rights to their software and allow you to use/copy it under very restricted conditions.)
So, as a result, you're not actually taking a copy of software from a software licensor and giving them money in exchange, but rather getting the pretty cardboard packaging and media for free (modulo overhead), while paying them money for the privilege of agreeing to a contract (license agreement) that has the net effect of turning the software from a good into a service which you may have to pay for repeatedly?
Of course software makers have a right to be compensated for their work; just, shouldn't it be for their _work_, rather than for the software itself, which has no economic value? (look up the definition of a "public good" -- software, as a good, fits it to a tee). If you'll notice, most programmers don't get payed by the companies they work for for the software they write, but their labor. Why should the users pay for the software itself, instead of the packaging, labor, and any value added, if software is a good?
I suppose this is the question: is software a public good, or a service (in and of itself)?
Incedentally, IANAL, but isn't unauthorized copying ("piracy") either a breach of contract or a tort, rather than a property crime?
It's one thing to take away (steal or destroy) property (or money) that a party already posesses -- that's a property crime. Isn't unauthorized copying, where the licensor is only denied _potential_ property (money), a matter of breach of contract (violating the software license) and/or tort (wronging them by denying them the money which they claim to have a natural "right" to) instead?
---
The best thing, IMO, would be to use MDI, and have the option of spawning additional MDI containers as needed. That serves to keep related browser windows nicely grouped together, while allowing you to play the workspace game, too.
Bonus if you can drag windows between MDI containers. (hrm... potential "killer feature" for the Qt version? I know that'd be pretty painful to implement in Windows, so I doubt we'd see it there... might be easier with a from-scratch MDI implementation)
---
> For one, seeing as how we are constantly
...
... I think I've got the first character of the alien transmission...
...
... 'E' ... 'E' ... 'D' ... 'H'?
... but, what's "needh"?
> observing astonomical phenomenon, this would be
> a good way for the aliens to improve our chances
> of actually seeing the message, and two, we'd
> probably notice any glaring anomalies. So, Could
> this be the one?
What, by modulating the brightness of the star? I can just picture it
SETI researcher #1: Hey, Dale, get over here
SETI researcher #2: 'N'
SETI researcher #1: I wonder what it means...
SETI researcher #2: Someday, in ages to come, our children shall know...
--- a couple hundred years later ---
SETI4 researcher #1: Enod, we've got another character of the transmission... five characters now...
SETI4 researcher #2: 'N'
SETI4 researcher #1: Yeah
SETI4 researcher #2: Someday, in ages to come, our children shall know...
--- a few thousand years pass ---
N'GANTHOK researcher #1: The alien message is complete, Miznok! Finally, we shall read together the cosmic truth sought by our anscestors for so many ages...
N'GANTHOK researcher #2: "NEED HELP SEND MORE BEER"...
---
> Are there any other stars by Eta Carinae they > can compare this to?
... astronomers just can't figure out (pardon my french) what the fuck is going on with this thing.
Some, yes. And they all seem normal for the most part.
> And how does this tie in to the way we usually
> predict stars die, if that is what this thing is
> indeed doing?
It doesn't
In a way, it kind of sucks that it's already blown up or whatever 7500 years ago; we're just now getting to see what went on (and are just now going to be affected by it). At least it looks cool.
---
I don't have Windows, nor am I willing to pay
money to license it. What is a person like me
to do?
Methinks it might have a bigger impact if everyone
who wanted the service didn't play these games,
and insisted on getting it installed in their
Linux machine, or not at all. Right now, they
don't, and so they aren't visible as Linux users
who want the service. It's quite easy to interpret
as a lack of demand, then.
---
> All these [hardware] companies would have to
> drop their software divisions.
I'm not sure they'd miss them, either; in many cases, right now they're just playing loss-leader. Yes, they make a net profit, but often they spend more capital on the software development than they get back from it in revenue, and make it up from some of the revenue generated by their hardware.
There's a largely undocumented kiosk mode in Navigator. I forget the details, but you can find out about it deep in the bowels of the Netscape developer site.
> I was wondering if any SeaMonkeyUnix sources are
> planned for integration with Berlin for DOM
> stuff?
Maybe. Berlin isn't quite to the point yet where we can make those kinds of decisions. We're still a little further down the design/implementation ladder. Maybe in the next six months to a year.
It has already come up on the ML, though; you should really start following the archives if you're interested (better yet, subscribe and help development directly).
Wouldn't need to integrate the stuff in the actual codebase anyway; most of the stuff should be usable via CORBA, which is extremely pervasive in the Berlin architecture (read: everything is implemented using it). (just FYI, CORBA doesn't incur any extra function call overhead over normal C/C++ for method invocation between in-process objects, so performance should be quite decent)
> Was libggi/libgii port of SeaMonkey ever
> attempted?
No, you'd need:
a) a raw libggi windowing toolkit
b) a widget set to use on top of it
I did write the first draft of libgwt (which was a step towards item a), but I never really had time to keep after it, and the other folks development fell to had the same problem.
Assuming that libgwt was finished at some point, all you'd probably need to do for item b is to port GDK to libgwt, and you've got GTK+, so the GTK FE could then be used on top of libggi directly.
> Should Berlin make sense for a SeaMonkey hack to
> get a smaller memory footprint?
I don't understand the question. email me if you like, rather than continuing this offtopic thread.
> Would it be that hard for them to port it to
> Linux if they already have IE 4.01 and IE 5.0
> for Solaris and HP-UX? (They do have it
> available for download.) I mean, would there be
> a lot of code to rewrite If they already got the
> thing working on other POSIX compliant OS's?
It would really depend on how careful/competent the porting team had been. It's easy to write to the POSIX spec and expect the software to build and work on any POSIX-compliant OS, but it's another matter to write it to a specific POSIX-compliant OS (or any small set of them) and then expect it to build and work on any other POSIX-compliant OS.
This is especially true when you throw architectural differences into the mix; SPARC and PA-RISC are both 64-bit big-endian architecures (well, PA-RISC might be little-endian; I can't remember). It may well be that their IE for Unix codebase (hopefully they didn't do two individual ports) contains a lot of 64-bitisms. I have learned not to expect good software engineering practices from them.
> Even though I LOVE Linux, I do welcome IE, even
> though it's from M$, It is the superior browser,
> that wasn't the case when they were both like
> 2.0, but the more they progress, the more bugs I
> see in netscape, and the the more I see IE
> handle browsing the web better and better.
IMO, the existing (old) Netscape codebase is slowly collapsing under its own weight. It's high time a Mozilla-derived product got into Alpha...
> The biggest reason I welcome IE to Linux is
> because I'd like to see a Linux browser that
> properly handles Cascading Style Sheets. I have
> yet to see that on Linux browsers.
I have yet to see that on Microsoft browsers. Try some of the CSS conformance tests in IE sometime.
Personally, I'd rather wait for Mozilla on Linux than for IE. Yes... I know... Mozilla might not not be all there yet, but IE for Linux sure as hell isn't.
ssia. it's not exactly been easy to find.
> I guess this puts to rest all the arguments of
/.'s lifetime.
/. is because Rob hasn't upgraded from 2.2.8 yet
> those people who were saying that it's
> slashdot's setup, not the Linux kernel, that was > causing the problem. Apparently it was a bug in
> the kernel itself. A known bug, but a bug
> nonetheless, and a serious one that caused the
> site to lose uptime at that.
If you're talking about the latest batch of downtimes, yes.
However, I should point out that most of the problems in the past (and Rob has announced the causes in a similar fashion to this in the past) have been due to mySQL crapping out, or more occasionally bugs in the particular patchlevel of mod_perl that Rob was running. Keep in mind, kernel 2.2.8 hasn't even existed for most of
The bug was stomped pretty quickly in 2.2.9 as I understand, too; the only reason it's not fixed for
It is dissapointing to have a stable kernel that's this flaky (although it is just the specific driver). I'm sure we'll have our share of 2.0.34s in the future, too.
I still think that _on the balance_ (and I think history bears out my point here), Linux does very well in terms of low bug rates and fast bug fixes.
I should have persisted in trying to find
actual documentation before posting.
JOOC, I seem to remember there being some
controversy with NeXT, though -- was that just
a matter of caving-under-bad-pr, or further
misinformation on my part?
as someone who is not likely to have time to
see the movie, how did they do it?
> I don't know enough about the itty-bitty details of graphics-device
> interfaces to take a particular stand on whether they should go into
> the kernel or user space or a little of both (the last seems most
> likely).
That is in fact what GGI actually does. KGI
drivers are typically pretty thin (generally enough to successfully
arbitrate hardware access among a number of userland processes)
and then most of the stuff goes on in LibGGI. LibGGI,
incdentally, can work on top of a lot of stuff besides KGI (i.e. X), so it's compatible with other Unixes (even those without a KGI layer) too.
Ever notice the GNU Objective C compiler?
The NeXT folks took gcc and adapted it for their
Objective C compiler. Only problem is, they
tried to make the resulting compiler proprietary.
The FSF took them to court. The FSF won, and
now GCC compiles Objective C too.
There have been a handfull of other cases too.
the variables probably need to be marked 'volatile' to keep the compiler from optimizing them away.
> Anyone else tried ttyquake? Just get the version
> of svgalib that is a hack over the aalib (?)
> char cell graphics libraries, then use
> LD_PRELOAD to grab them... Weird.
Alternatively, use LibGGI via the svgalib wrapper with GGI_DISPLAY set to "aa".
That way, depending on what GGI_DISPLAY is set to (if unset, it autodetects), you can use the same rig without further modification to display on X or on a monitor wall or on console with fbcon or all kinds of other really freaky shit (see http://www.ggi-project.org/screenshots.html) -- these people are EVIL&#*@^$*#&^
*grin*
Just judging by the majority (75% currently) of the replies, it would seem that my experience is atypical. Some of these replies (which are all quite good) should be scored up somewhere closer to the original post.
> If "they" don't teach generalized skills at your
> school, you need to get a better school.
I think so too. However, _so far_ none of the CS programs that I've looked at in other schools look better enough to bother switching. This may just be a regional phenomena.
> In these parts, they teach classes on:
> Databases (including all sorts of theory, including some subsets of logic)
> Operating systems (which is to say, general principles underlying operating systems, not "How to configure NT")
> Computer languages (overview of different types of languages, including Prolog, Lisp, and others.)
> Assembly language
> Data structures
> Algorithms
> Circuit design
> And a whole lot more
> And, oh, they teach all of this using Unix systems
This is all refreshing to hear.
> and require that CS students take twice as many non-CS classes as CS ones.
Well, that much is a given, even if you're stuck in some crappy school like myself.
> And I don't think the U. of Wisconsin is particularly unusual in all these regards.
It'd be interesting to do a general survey of various colleges' curriculae and see exactly what is the most typical case. I suspect our individual perceptions are both pretty colored by regional variations.
Regardless, I think my comments about high school in particular are still applicable -- certainly, the majority of people from around the country I've talked to have indicated the same kind of deficiencies I've described. I haven't talked to as wide a variety of people about college in the same way -- at least not yet.