I may be wrong, but I see Unix in general, and Linux specifically, as one huge development environment.
So do I, and many others. Indeed, that was one of the main goals for the OS back when Unix was being written in the early 1970s. However, there is no reason you cannot add more tools to that environment, such as Borland's offerings.
I don't know how much Borland can add to the C/C++ tools, and I doubt there's room for another scripting language. Is Delphi all they have?
Borland's development tools are very nice to work with. Their VCL (GUI framework) is the best I've ever used. Elegant and easy. You can use their visual form editor to drag-n-drop, but you can also write it in code -- or both. The IDE detects changes you make to the code it writes, and back-updates any data files it needs.
I would really like to see C++Builder on Linux. Delphi, too. I like Object Pascal for its clean design. It isn't a standard, nor is it portable, unfortunately, which limits its uses in "real world" code, but I find it fun to play with, if nothing else. And Borland has a number of other products as well, in the database, Java, and CORBA markets.
So, yes, I think this could work. Will it? I don't know. If I could predict the future, I wouldn't need to work on computers for a living.:-)
GTK is platform independent, too
on
Anarchy Online
·
· Score: 2
Hmm.. platform independencem, ease of poring... QT.. Another nail in the GTK casket.
Perhaps you're not aware, but GTK is platform independent as well. It has already been ported to the Win32 platform. And as long as we're flinging FUD around, I might point out that GTK is also language independent, while Qt is limited to C++.
I'm going to dust off one of my favorite soap-boxes, and try to introduce a few people here to a fond place in my memory. It's called FidoNet
FidoNet is a network of independent, dial-up bulletin board systems. You connect with a modem to a PC running in some guy's basement. Everything is text. "Graphics" means color text with line-draw characters. Almost none of these systems are connected to a permanent network other then the one run by AT&T. Every night, all these BBSes call the BBSes in the next towns over, and exchange mail and files. It isn't far wrong to say FidoNet was designed primarily to cheat the phone company.
For a long time, this was the only way for someone like myself, living in rural New Hampshire, USA, to get "online". The Internet was something to be found at universities. But FidoNet was everywhere. It had over 40,000 nodes when the Internet hit the big time and started killing it off. It's still alive, but slowly dying out as the 'net makes it obsolete.
What does this have to do with anything? Well, FidoNet had open discussion forms, like Usenet. Fido's "newsgroups" were called "echos". Like Usenet, FidoNet was largely self-regulated. The "coordinators" were a loosely-knit group of system operators who tried to keep everyone on the same sheet of music. But, unlike Usenet, Fido had controls in place to keep things like spam from getting out of hand.
Every Fido echo (discussion group) had a moderator. Not like a Usenet moderator, but more like an IRC "op". The moderator was responsible for keeping the echo in the echolist -- if noone cared about an echo, it would automatically remove itself in about six months. Moderators were free to implement whatever policy they chose. Anyone was free to start their own echo if they didn't like any of the current ones.
The big thing was: People were held accountable. If someone started making an ass out of themselves, the moderator would warn them. If the twit didn't listen, the moderator could mail the sysop of their board and get them removed from the echo. If the same guy kept getting kicked out of echos, the sysop would generally cut their echo access entirely.
But wait -- it gets better. If a particular board was a frequent source of twits, spammers, and the like, a coalition of moderators could contact the local network coordinator (NC). The NC would warn the board, but if it did not improve, it's FidoNet feed would be cut completely. There were ways for this to progress up the chain of command so that entire networks (local areas) could be cut.
If you are used to the loosely regulated anarchy of Usenet, this seems drastically different. But it did work, for the small population of Fido at the time. In effect, FidoNet has a cabal, and is better off for it.
Of course, we'll never know for sure if FidoNet would have scaled as well as UseNet has (and Usenet scaled, just not as well as we'd like). However, a system like FidoNet's might be something to consider for those looking to "replace" Usenet with something "better".
As for graphical browsers, the whole point of the Web is text, if you think about it (hypertext, remember?), so a "text-based broswer" is a pretty good tool for using it.
While I'm not disputing the usefulness of a text-based browser, I do disagree that the whole point of the web is text. It isn't. The point of the web is connected information. Information can be expressed in a number of ways: Written word, spoken word, images, etc. Ever hear the phrase "A picture is worth a thousand words?" It is often true. I object to sites that use images when text would do, but I also object to those who think the <IMG> tag should be banned when a diagram would obviously be much clearer then ten pages of text.
Well, apparently the current GNOME html code is based on KDE code.
Ya know, this single statement is actually rather insightful.
There are a lot of people (not pointing fingers here, just making an observation) that say GNOME and KDE are a duplicated effort and they they should merge into one project, yadda, yadda, yadda.
But here we see how Open Source Software makes that line of reasoning obsolete. Because all the code is open, both sides are free to borrow code from the other. Indeed, we have seen code being borrowed by one, improved, and then reincorporated back into the original. This not only reduces duplication of effort, but encourages the developers to make their software compatible whenever possible, because it may save them some work in the future.
I just wanted to get that off my chest. Yes, I feel better now.:-)
Hmmmm, it appears The Great Taco has disable ALL HTML in "Extrans" mode. No doubt related to the CERT alert. However, "HTML" mode still works. How odd. I liked Extrans; it saved the effort of marking up all my paragraphs. Oh well.
A lot of people find Netscape Communicator on Linux to be unstable. And they are right. However, there are some things you can do to dramatically improve stability.
First and foremost, download Netscape Navigator (the stand-alone browser version), and NOT Communicator. The mail, news, and HTML editor components of Communicator seem to significantly reduce stability. This alone has cut my Netscape crashes to only occasionally.
Next, make sure you have all the proper fonts installed. Netscape expects certain fonts in a few places, and gets rather confused if it doesn't get them. The Java VM in particular has this problem.
Notably, some versions of Red Hat Linux don't configure all the fonts properly. Check the/etc/X11/fs/config file's catalogue section to be sure all of the following are included:
It is okay to have more, just make sure the above are included, both unscaled and regular.
If you can do without it, turn off Java support (not JavaScript -- the two are completely separate things). Netscape's JVM is remarkably unstable.
Consider turning off JavaScript, too. Not only can it be abused (CERT advisory, blah, blah), poorly designed code can make Netscape screwy.
Keep an eye on the memory usage of Navigator. Navigator has some severe memory leaks in it. If it starts to grow larger then 50 MB or so Virtual Segment Size, exit and restart it.
No, none of these things are acceptable behavior for a browser, but they will get you by until Mozilla, Konqueror, Opera, or whatever GUI browser replacement you favor is ready.
As an aside, if you are using Communicator, ALWAYS turn off "JavaScript in mail and news" -- apparently Netscape wants to be like Microsoft and allow people to send you emails that take actions.
If I'm not mistaken, this is part of the Cluster Satellite project, investigating "the Earth?s magnetic field and its interaction with the solar wind."
I think you mean Cluster II. The first one ended up in a swamp when the French Ariane 5 rocket blew up on its maiden flight due to a software bug. Parts of the instrument were actually recovered and are being used in the new mission under the name "Phoenix", a particularly apt use of the name. I say this only because I was working as a sysadmin at the UNH Space Science Center at the time. The SCC was building one of the instruments for Cluster. I remember a lot of people's dreams went down with that rocket.
Success and failure come about because of the nature of the person. The heroes are not (obviously) born into their fate and position. As for the hobbits, do you remember who they were. ALL were the children of the great families of the hobbits.
I really think you're reading something into the story that isn't there. One of the major themes of The Lord of the Rings is the struggle of a few persons against an overwhelmingly powerful evil. They weren't born into it; they fell into it, and proved their worth by their actions. Not by fate or a birthright.
Indeed, I don't remember fate being a significant part of LotR, other then for the elves (the choice of staying in Middle-earth, blah, blah). There were instances of people trying to see the future, but that isn't fate. There was Aragorn, but he wasn't fated to be king; he was the rightful king of Gondor, by line of decent. If that's fate, then everything is.
Is there something specific you're thinking about and I'm forgetting?
Try Patricia McKillip. The Cygnet and the Sorceress and The Cygnet and the Firebird are, to my mind, great.
Added to my virtual to-read queue. No guarantees on when I'll get around to it, tho. It's a big queue.:-)
Does it bother anyone else that in Tolkeins world, you have to be born to greatness?
Huh? What about Bilbo? Frodo? The other hobbits of the Fellowship? Did I miss something here? Or did you miss one of the main themes of the story?
And what's more, all the great people are tall and fair. The only evil humans are short and swarthy.
Oh, come on. First, again: The Hobbits. Second, Tolkien is neither the first nor the last to equate good with light and evil with darkness. It is a classic metaphor, and one that works very well.
The Hobbit is a marvelous story. I really like The Silmarillion. After trying to reread Lord of the Rings, I gave up.
They're really all part of the same story, you know.
There is better fantasy out there. Much, much better.
You're entitled to your opinion. Personally, I think The Lord of the Rings is the best fantasy ever written.
Anyway, some things are not easy or enjoyable, but are good to read.
Yes, but that does not mean easy and enjoyable stuff is necessarily bad. Tolkien's stuff is enjoyable, IMO, but not easy to read. Another of my personal favorites, Anne McCaffrey's works, I find both easy and enjoyable. Personal preference.
Everyone else on the Internet has chimed in; I might as well, too.
Must Reads
Asimov, Heinlein, Clarke, Bradbury. They still pack more ideas into a small space then any truck-full of cyberpunk.:-)
J.R.R. Tolkien's The Hobbit should be read as young as possible, as kids can enjoy that sort of story much better then adults. Later on, they can try out The Lord of the Rings, but that requires some work to appreciate it fully, so go easy early on.
Ender's Game by Orson Scott Card. Quite simply one of the best books I've ever read.
The Hitchhiker's Guide to the Galaxy by Douglas Adams. Follow up with the sequels if you like (you will).
Personal Favorites
The Dragonriders of Pern series, by Anne McCaffrey. A pleasant mix of sci-fi and fantasy, with excellent characterization in relatively light reading. And you'll fall in love with the dragons, too.
Larry Niven. One of my favorite authors, his stories pack an education in the human condition and physics into the same space. I recommend his two short story collections, N-Space and Playgrounds of the Mind. If you prefer something longer, Ringworld is great. I also recommend The Mote in God's Eye, an excellent First Contact novel, and as Robert A. Heinlein (yes, him) said, "Possibly the best science fiction novel I have ever read."
C.S. Lewis's The Chronicles of Narnia is an excellent series of fantasy, in the classic tradition of the fairy-tale. If the reader hasn't lost that childlike sense of wonder, they are excellent books. (There is also some Christian allegory if you like that sort of thing, but you don't have to get that part if you don't want to.)
I'm going to stop now, before I list my entire library.:-) Cheers!
Sci Fi is a great way to get childern into reading, however, they may fill up only on it, ignoring the classic readings that makes up her/your culture.
"Our myths, the myths of Tolkien and Homer, of Heinlein and Mallory are eternal; they exchange one name for another, cast off one mask and assume the next." -- J. Michael Straczynski, creator of Babylon 5
I'm not so sure I agree with your assertion that what academia says are "classics" are "must reads" just because they are classics. Even ignoring the circular logic, you have to understand that as JMS has so astutely observed, stories repeat themselves; it is simply the dressing, the flavor, that changes.
I don't think the "our culture" argument holds, either. I know I didn't relate to anything in any of those works you listed as well as I related to, say, Ender's Game. Niven, Heinlein, McCaffrey and Tolkien had far more influence on my life then anything pushed on me in school.
If I can be so bold as to speak for those other then myself, I suspect a lot of the Slashdot readership's "culture" is science fiction and fantasy before "traditional classics".
Now, keep one thing clear: I am not saying one should not read those works. Frankly, I think they are all very good. But the reason for reading them should be "Because they are good reads", not because they define some arbitrary cultural tradition that may or may not fit anyone personally.
Remember that the Classics are the foundation that most of todays readings are build on.
True, but you don't have to sleep on the foundation to enjoy the benefits of a well-built house. If someone wants to study the "classics" because literary history interests them or whatever, then fine. But it should be the reader's choice, as only they know what they like. Far too many teachers I've had in the past assume that they can like a book for me. It simply isn't true.
Cold, highs in the low 20s. Expect lows in the single digits tonight, and up to 40 below with the wind chill factor. Snow tomorrow, four to eight inches.
Good words for OSS from Gartner? I don't believe it.
Wow! That Server51 is one cool site. Personally, I think the "alien conspiracy" thing is bit overused, but they've done it well, and made excellent use of a few small graphics to create a very cool look.
I say that you DO have to regulate this trash. If you can't keep off-shore folfs from doing it then just keep off-shore sites off our internet. Place a great big firewall up and regualte what goes through it.
You really want the government telling you who you can and cannot connect to? Personally, that idea scares me. For one, it is just about the same thing as censorship. Sorry, but I don't want the government in that business.
You can setup your own proxy server or firewall to prevent such privacy invasions, however. I recommend The Internet Junkbuster; I set it up in like ten minutes the other day and it works great.
This is all well and good, but don't the Pentium IIIs have a "thumbprint" that allows for them to see what we're doing?
You got a network card in your system? That has a "thumbprint" too. The MAC address.
You got any commercial software (e.g., Windows) on your system that you had to enter a software key to use? There's another "thumbprint" for you.
How about a static IP address? Ever time you send a network request, you're identifying yourself.
You think you're safe because you have a dynamic address? Do you at least always call the same ISP at the same phone number? You'll always be getting the same range of IP numbers, then. You and maybe a few dozen or hundred more people. That is almost as good as a unique personal ID, as far as demographics go.
Fact of the matter is, tracking a computer is not that hard to do. If you ever give out any personal information at all (name, email, phone number, ZIP code), that can be combined with any of the above to nail down exactly who you are.
I think Scott McNeally's right on this one. Privacy on the Internet is dead.
The only way to improve things would be for the government to step in and make such unauthorized tracking illegal, with hefty fines for violators. You could even do some good by donating said fines to the EFF.
... what I really want to do is remove any and all info about me from their database. I'm sure they've used other methods to collect info on me, and I want it removed. What are my options?
Unfortunately, that information was likely collected using perfectly legal means, and is thus their property. You can control how they use it (e.g., stopping them from calling you to sell you things), but not the fact that they have it. You can usually tell them not to rent or sell your name, but I believe the law isn't clear on your rights in such cases.
Yah, I had to deal with Arcserve on a Netware server when I was working at Unnamed U. It didn't help matters much that the server had some flaky hardware in it to begin with, but that is what makes a good backup all the more important.
ArcServe was the wrong choice for the job.
The UI was horrible, consisting of a scattering of programs, some loaded on the server, some on the client. Only backup software I know of that forced you to load a program on another machine just to eject a tape.
It used to go off the trolly somewhere into a la-la land that only ArcServe could find. It would keep spitting out useless console messages and eating up CPU time. The only way to fix it was to forcibly kill the program, which is not good for the health of your Netware server.
And it was always spewing error messages like "Unexpected error nnnn" or "Tape server error nnnn" where "nnnn" was some number not found in the manual. You'd call tech support, sit on hold for an hour, and that get a tech who'd look it up and say something like "It means your backup is toast, try it again" or "Oh, we don't know what that means, either. It wasn't documented by the engineering team, and they haven't had a chance to go back and find it yet."
Geez, am I glad I don't have to deal with that POS anymore.
Perhaps another malloc comes along wanting the same size (roughly) as the thing just freed. We just give it the recently freed chunk. We then don't have to build a new one.
IIRC from what I have read about the Linux kernel, it does something very like this. Blocks of free memory are kept track of in a hash table, where the hash key is the size of the block. When you free a block, it gets merged with any adjacent blocks and put in the appropriate spot in the hash table.
Or it might have been an array of lists of blocks, where each element in the array represents a list of blocks twice as big as the previous element's list. It's been awhile; I forget.:)
The point is that the average application has inflated enormeous over the years.
So has the average machine. Which is cause and which is effect? Do we have bigger applications because with today's faster processors and bigger memories, throwing more hardware at the problem is cheaper then trying to bum five bytes out of a 50 KB program?
... why is it not possible to use GTK widgets to render inside forms?
Basically, they need to be able to do things like apply CSS styles, layer widgets, composite them, and do other things that many toolkits/platforms don't support.
This also results in having the same UI regardless of what platform you're running Mozilla on. This goes back to the Netscape goal of turning Navigator into a development platform. There are pros and cons to this approach, but I guess they figured they were already most of the way there, so they might as well finish the job.
It also reduces platform-specific code, which any developer will tell you makes your life easier.
Given that both Mozilla and GTK are Open Source Software, I'm sure someone will eventually do the work to merge the two, even if it requires rewritting GTK.:-)
Netscape is badly written junk (oh, sorry, I have a knack for being redundant).
I wouldn't go that far. Netscape Navigator is a case of code pressed beyond the entropy point. Any non-trivial code base which is being modified will experience a case of gradually diminishing returns. Eventually, the system becomes unmaintainable and you have to start over.
Navigator 3 was pretty good (for its day), but Navigator 4 really shows the age of the codebase. And, as the Mozilla project found out, trying to do anything major with the V4 codebase resulted in a total collapse.
This doesn't make the software "badly written junk"; it makes it software which should have been thrown away after the V3 series. Unfortunately, management so rarely recognizes such cases.
Read Fred Brooks's timeless work on software engineering, The Mythical Man-Month, for more about this and other software principles.
More seriously, Netscape on Linux needs 64 megs of memory.
Um, I'm not sure what version of Netscape you're running, but on my system, right now, Navigator 4.7 has a virtual segment size of just under 15 MB and a resident size of just under 11 MB.
If Linux can't swap Netscape in and out fast enough, Netscape gets impatient and dies.
If a process needs more memory, it will block until the kernel satisfies the request or fails it. It cannot "get impatient"; it will not be run by the kernel while memory is being managed.
Netscape seems to have a HUGE memory leak that causes it to eat up memory, especially on pages with lots of images (ahem!).
Here you have found the real problem. There is a memory leak in Navigator V4 related to pages with complex layouts and/or lots of images. This causes Navigator to gradually leak memory over time. It isn't using this memory, but it eats up swap space and slows down the system. If you don't exit and restart it, Navigator will eventually exhaust virtual memory, resulting in failed memory allocation requests and a program abort.
This is, of course, a Bad Thing. My workaround (until Mozilla V5 becomes viable) is to keep an eye on the memory gauge in the GNOME Panel, and exit and restart Navigator if it grows beyond a reasonable size.
Netscape can bring the entire system down if a) your CPU overheats or b) your video card overheats.
That's a very misleading statement. The cause of such a crash is a heat problem. Anything could cause that. Blaming Navigator because it was the forground application at the time is silly. If your system is overheating, you need to fix your system cooling, not blame the software!
Netscape can also bring the entire system down if by some random chance it happens to eat memory just as another process (syslog, etc.) tries to grab it, but for some reason Netscape wins the race. In this case, I've seen INIT panic and Ctrl-Alt-Del get disabled.
Again, here you're blaming Navigator for a system problem. It is a known problem that Linux handles out-of-memory situations poorly. I believe the 2.3 kernel addresses this, by making sure system processes have a proper reserve and doing a better job of killing processes to free memory in an emergency.
Anything that eats memory will cause this problem. while(1) malloc(10000); will do it. Are you going to blame GCC for that, next?
What I recommend doing is instituing user resource limits (using ulimit) to prevent any single process from exhausting your system's memory.
In the end, if an application crashes your system, there is something wrong with the entire system. It doesn't matter which application did it.
Slashdot Mirror
I may be wrong, but I see Unix in general, and Linux specifically, as one huge development environment.
So do I, and many others. Indeed, that was one of the main goals for the OS back when Unix was being written in the early 1970s. However, there is no reason you cannot add more tools to that environment, such as Borland's offerings.
I don't know how much Borland can add to the C/C++ tools, and I doubt there's room for another scripting language. Is Delphi all they have?
Borland's development tools are very nice to work with. Their VCL (GUI framework) is the best I've ever used. Elegant and easy. You can use their visual form editor to drag-n-drop, but you can also write it in code -- or both. The IDE detects changes you make to the code it writes, and back-updates any data files it needs.
I would really like to see C++Builder on Linux. Delphi, too. I like Object Pascal for its clean design. It isn't a standard, nor is it portable, unfortunately, which limits its uses in "real world" code, but I find it fun to play with, if nothing else. And Borland has a number of other products as well, in the database, Java, and CORBA markets.
So, yes, I think this could work. Will it? I don't know. If I could predict the future, I wouldn't need to work on computers for a living. :-)
Hmm.. platform independencem, ease of poring... QT.. Another nail in the GTK casket.
Perhaps you're not aware, but GTK is platform independent as well. It has already been ported to the Win32 platform. And as long as we're flinging FUD around, I might point out that GTK is also language independent, while Qt is limited to C++.
I'm going to dust off one of my favorite soap-boxes, and try to introduce a few people here to a fond place in my memory. It's called FidoNet
FidoNet is a network of independent, dial-up bulletin board systems. You connect with a modem to a PC running in some guy's basement. Everything is text. "Graphics" means color text with line-draw characters. Almost none of these systems are connected to a permanent network other then the one run by AT&T. Every night, all these BBSes call the BBSes in the next towns over, and exchange mail and files. It isn't far wrong to say FidoNet was designed primarily to cheat the phone company.
For a long time, this was the only way for someone like myself, living in rural New Hampshire, USA, to get "online". The Internet was something to be found at universities. But FidoNet was everywhere. It had over 40,000 nodes when the Internet hit the big time and started killing it off. It's still alive, but slowly dying out as the 'net makes it obsolete.
What does this have to do with anything? Well, FidoNet had open discussion forms, like Usenet. Fido's "newsgroups" were called "echos". Like Usenet, FidoNet was largely self-regulated. The "coordinators" were a loosely-knit group of system operators who tried to keep everyone on the same sheet of music. But, unlike Usenet, Fido had controls in place to keep things like spam from getting out of hand.
Every Fido echo (discussion group) had a moderator. Not like a Usenet moderator, but more like an IRC "op". The moderator was responsible for keeping the echo in the echolist -- if noone cared about an echo, it would automatically remove itself in about six months. Moderators were free to implement whatever policy they chose. Anyone was free to start their own echo if they didn't like any of the current ones.
The big thing was: People were held accountable. If someone started making an ass out of themselves, the moderator would warn them. If the twit didn't listen, the moderator could mail the sysop of their board and get them removed from the echo. If the same guy kept getting kicked out of echos, the sysop would generally cut their echo access entirely.
But wait -- it gets better. If a particular board was a frequent source of twits, spammers, and the like, a coalition of moderators could contact the local network coordinator (NC). The NC would warn the board, but if it did not improve, it's FidoNet feed would be cut completely. There were ways for this to progress up the chain of command so that entire networks (local areas) could be cut.
If you are used to the loosely regulated anarchy of Usenet, this seems drastically different. But it did work, for the small population of Fido at the time. In effect, FidoNet has a cabal, and is better off for it.
Of course, we'll never know for sure if FidoNet would have scaled as well as UseNet has (and Usenet scaled, just not as well as we'd like). However, a system like FidoNet's might be something to consider for those looking to "replace" Usenet with something "better".
*climbs back down off soap-box*
Just my 1/4 of a byte. ;-)
No, no, no. *REAL* surfers don't even do that. They use their superior mental powers to read the minds of those on the other end of the link. :)
As for graphical browsers, the whole point of the Web is text, if you think about it (hypertext, remember?), so a "text-based broswer" is a pretty good tool for using it.
While I'm not disputing the usefulness of a text-based browser, I do disagree that the whole point of the web is text. It isn't. The point of the web is connected information. Information can be expressed in a number of ways: Written word, spoken word, images, etc. Ever hear the phrase "A picture is worth a thousand words?" It is often true. I object to sites that use images when text would do, but I also object to those who think the <IMG> tag should be banned when a diagram would obviously be much clearer then ten pages of text.
End of rant. :-)
(Drifting off topic here, sorry...)
Well, apparently the current GNOME html code is based on KDE code.
Ya know, this single statement is actually rather insightful.
There are a lot of people (not pointing fingers here, just making an observation) that say GNOME and KDE are a duplicated effort and they they should merge into one project, yadda, yadda, yadda.
But here we see how Open Source Software makes that line of reasoning obsolete. Because all the code is open, both sides are free to borrow code from the other. Indeed, we have seen code being borrowed by one, improved, and then reincorporated back into the original. This not only reduces duplication of effort, but encourages the developers to make their software compatible whenever possible, because it may save them some work in the future.
I just wanted to get that off my chest. Yes, I feel better now. :-)
Hmmmm, it appears The Great Taco has disable ALL HTML in "Extrans" mode. No doubt related to the CERT alert. However, "HTML" mode still works. How odd. I liked Extrans; it saved the effort of marking up all my paragraphs. Oh well.
A lot of people find Netscape Communicator on Linux to be unstable. And they are right. However, there are some things you can do to dramatically improve stability.
First and foremost, download Netscape Navigator (the stand-alone browser version), and NOT Communicator. The mail, news, and HTML editor components of Communicator seem to significantly reduce stability. This alone has cut my Netscape crashes to only occasionally.
Next, make sure you have all the proper fonts installed. Netscape expects certain fonts in a few places, and gets rather confused if it doesn't get them. The Java VM in particular has this problem.
Notably, some versions of Red Hat Linux don't configure all the fonts properly. Check the /etc/X11/fs/config file's catalogue section to be sure all of the following are included:
It is okay to have more, just make sure the above are included, both unscaled and regular.
If you can do without it, turn off Java support (not JavaScript -- the two are completely separate things). Netscape's JVM is remarkably unstable.
Consider turning off JavaScript, too. Not only can it be abused (CERT advisory, blah, blah), poorly designed code can make Netscape screwy.
Keep an eye on the memory usage of Navigator. Navigator has some severe memory leaks in it. If it starts to grow larger then 50 MB or so Virtual Segment Size, exit and restart it.
No, none of these things are acceptable behavior for a browser, but they will get you by until Mozilla, Konqueror, Opera, or whatever GUI browser replacement you favor is ready.
As an aside, if you are using Communicator, ALWAYS turn off "JavaScript in mail and news" -- apparently Netscape wants to be like Microsoft and allow people to send you emails that take actions.
If I'm not mistaken, this is part of the Cluster Satellite project, investigating "the Earth?s magnetic field and its interaction with the solar wind."
I think you mean Cluster II. The first one ended up in a swamp when the French Ariane 5 rocket blew up on its maiden flight due to a software bug. Parts of the instrument were actually recovered and are being used in the new mission under the name "Phoenix", a particularly apt use of the name. I say this only because I was working as a sysadmin at the UNH Space Science Center at the time. The SCC was building one of the instruments for Cluster. I remember a lot of people's dreams went down with that rocket.
Success and failure come about because of the nature of the person. The heroes are not (obviously) born into their fate and position. As for the hobbits, do you remember who they were. ALL were the children of the great families of the hobbits.
:-)
I really think you're reading something into the story that isn't there. One of the major themes of The Lord of the Rings is the struggle of a few persons against an overwhelmingly powerful evil. They weren't born into it; they fell into it, and proved their worth by their actions. Not by fate or a birthright.
Indeed, I don't remember fate being a significant part of LotR, other then for the elves (the choice of staying in Middle-earth, blah, blah). There were instances of people trying to see the future, but that isn't fate. There was Aragorn, but he wasn't fated to be king; he was the rightful king of Gondor, by line of decent. If that's fate, then everything is.
Is there something specific you're thinking about and I'm forgetting?
Try Patricia McKillip. The Cygnet and the Sorceress and The Cygnet and the Firebird are, to my mind, great.
Added to my virtual to-read queue. No guarantees on when I'll get around to it, tho. It's a big queue.
Does it bother anyone else that in Tolkeins world, you have to be born to greatness?
Huh? What about Bilbo? Frodo? The other hobbits of the Fellowship? Did I miss something here? Or did you miss one of the main themes of the story?
And what's more, all the great people are tall and fair. The only evil humans are short and swarthy.
Oh, come on. First, again: The Hobbits. Second, Tolkien is neither the first nor the last to equate good with light and evil with darkness. It is a classic metaphor, and one that works very well.
The Hobbit is a marvelous story. I really like The Silmarillion. After trying to reread Lord of the Rings, I gave up.
They're really all part of the same story, you know.
There is better fantasy out there. Much, much better.
You're entitled to your opinion. Personally, I think The Lord of the Rings is the best fantasy ever written.
Anyway, some things are not easy or enjoyable, but are good to read.
Yes, but that does not mean easy and enjoyable stuff is necessarily bad. Tolkien's stuff is enjoyable, IMO, but not easy to read. Another of my personal favorites, Anne McCaffrey's works, I find both easy and enjoyable. Personal preference.
Everyone else on the Internet has chimed in; I might as well, too.
:-)
:-) Cheers!
Must Reads
Asimov, Heinlein, Clarke, Bradbury. They still pack more ideas into a small space then any truck-full of cyberpunk.
J.R.R. Tolkien's The Hobbit should be read as young as possible, as kids can enjoy that sort of story much better then adults. Later on, they can try out The Lord of the Rings, but that requires some work to appreciate it fully, so go easy early on.
Ender's Game by Orson Scott Card. Quite simply one of the best books I've ever read.
The Hitchhiker's Guide to the Galaxy by Douglas Adams. Follow up with the sequels if you like (you will).
Personal Favorites
The Dragonriders of Pern series, by Anne McCaffrey. A pleasant mix of sci-fi and fantasy, with excellent characterization in relatively light reading. And you'll fall in love with the dragons, too.
Larry Niven. One of my favorite authors, his stories pack an education in the human condition and physics into the same space. I recommend his two short story collections, N-Space and Playgrounds of the Mind. If you prefer something longer, Ringworld is great. I also recommend The Mote in God's Eye, an excellent First Contact novel, and as Robert A. Heinlein (yes, him) said, "Possibly the best science fiction novel I have ever read."
C.S. Lewis's The Chronicles of Narnia is an excellent series of fantasy, in the classic tradition of the fairy-tale. If the reader hasn't lost that childlike sense of wonder, they are excellent books. (There is also some Christian allegory if you like that sort of thing, but you don't have to get that part if you don't want to.)
I'm going to stop now, before I list my entire library.
Sci Fi is a great way to get childern into reading, however, they may fill up only on it, ignoring the classic readings that makes up her/your culture.
"Our myths, the myths of Tolkien and Homer, of Heinlein and Mallory are eternal; they exchange one name for another, cast off one mask and assume the next."
-- J. Michael Straczynski, creator of Babylon 5
I'm not so sure I agree with your assertion that what academia says are "classics" are "must reads" just because they are classics. Even ignoring the circular logic, you have to understand that as JMS has so astutely observed, stories repeat themselves; it is simply the dressing, the flavor, that changes.
I don't think the "our culture" argument holds, either. I know I didn't relate to anything in any of those works you listed as well as I related to, say, Ender's Game. Niven, Heinlein, McCaffrey and Tolkien had far more influence on my life then anything pushed on me in school.
If I can be so bold as to speak for those other then myself, I suspect a lot of the Slashdot readership's "culture" is science fiction and fantasy before "traditional classics".
Now, keep one thing clear: I am not saying one should not read those works. Frankly, I think they are all very good. But the reason for reading them should be "Because they are good reads", not because they define some arbitrary cultural tradition that may or may not fit anyone personally.
Remember that the Classics are the foundation that most of todays readings are build on.
True, but you don't have to sleep on the foundation to enjoy the benefits of a well-built house. If someone wants to study the "classics" because literary history interests them or whatever, then fine. But it should be the reader's choice, as only they know what they like. Far too many teachers I've had in the past assume that they can like a book for me. It simply isn't true.
Cold, highs in the low 20s. Expect lows in the single digits tonight, and up to 40 below with the wind chill factor. Snow tomorrow, four to eight inches.
Good words for OSS from Gartner? I don't believe it.
I mean the American Association Against Acronym And Alliteration Abuse, of course. ;-)
Wow! That Server51 is one cool site. Personally, I think the "alien conspiracy" thing is bit overused, but they've done it well, and made excellent use of a few small graphics to create a very cool look.
"Me too!"
I say that you DO have to regulate this trash. If you can't keep off-shore folfs from doing it then just keep off-shore sites off our internet. Place a great big firewall up and regualte what goes through it.
You really want the government telling you who you can and cannot connect to? Personally, that idea scares me. For one, it is just about the same thing as censorship. Sorry, but I don't want the government in that business.
You can setup your own proxy server or firewall to prevent such privacy invasions, however. I recommend The Internet Junkbuster; I set it up in like ten minutes the other day and it works great.
This is all well and good, but don't the Pentium IIIs have a "thumbprint" that allows for them to see what we're doing?
You got a network card in your system? That has a "thumbprint" too. The MAC address.
You got any commercial software (e.g., Windows) on your system that you had to enter a software key to use? There's another "thumbprint" for you.
How about a static IP address? Ever time you send a network request, you're identifying yourself.
You think you're safe because you have a dynamic address? Do you at least always call the same ISP at the same phone number? You'll always be getting the same range of IP numbers, then. You and maybe a few dozen or hundred more people. That is almost as good as a unique personal ID, as far as demographics go.
Fact of the matter is, tracking a computer is not that hard to do. If you ever give out any personal information at all (name, email, phone number, ZIP code), that can be combined with any of the above to nail down exactly who you are.
I think Scott McNeally's right on this one. Privacy on the Internet is dead.
The only way to improve things would be for the government to step in and make such unauthorized tracking illegal, with hefty fines for violators. You could even do some good by donating said fines to the EFF.
... what I really want to do is remove any and all info about me from their database. I'm sure they've used other methods to collect info on me, and I want it removed. What are my options?
Unfortunately, that information was likely collected using perfectly legal means, and is thus their property. You can control how they use it (e.g., stopping them from calling you to sell you things), but not the fact that they have it. You can usually tell them not to rent or sell your name, but I believe the law isn't clear on your rights in such cases.
Check out the Data About You page at JunkBusters.com for more information about this sort of thing.
When I went to click on that opt-out link, I got a message saying the Internet JunkBuster had blocked that URL.
;-)
Aw, darn.
Yah, I had to deal with Arcserve on a Netware server when I was working at Unnamed U. It didn't help matters much that the server had some flaky hardware in it to begin with, but that is what makes a good backup all the more important.
ArcServe was the wrong choice for the job.
The UI was horrible, consisting of a scattering of programs, some loaded on the server, some on the client. Only backup software I know of that forced you to load a program on another machine just to eject a tape.
It used to go off the trolly somewhere into a la-la land that only ArcServe could find. It would keep spitting out useless console messages and eating up CPU time. The only way to fix it was to forcibly kill the program, which is not good for the health of your Netware server.
And it was always spewing error messages like "Unexpected error nnnn" or "Tape server error nnnn" where "nnnn" was some number not found in the manual. You'd call tech support, sit on hold for an hour, and that get a tech who'd look it up and say something like "It means your backup is toast, try it again" or "Oh, we don't know what that means, either. It wasn't documented by the engineering team, and they haven't had a chance to go back and find it yet."
Geez, am I glad I don't have to deal with that POS anymore.
Perhaps another malloc comes along wanting the same size (roughly) as the thing just freed. We just give it the recently freed chunk. We then don't have to build a new one.
:)
IIRC from what I have read about the Linux kernel, it does something very like this. Blocks of free memory are kept track of in a hash table, where the hash key is the size of the block. When you free a block, it gets merged with any adjacent blocks and put in the appropriate spot in the hash table.
Or it might have been an array of lists of blocks, where each element in the array represents a list of blocks twice as big as the previous element's list. It's been awhile; I forget.
The point is that the average application has inflated enormeous over the years.
So has the average machine. Which is cause and which is effect? Do we have bigger applications because with today's faster processors and bigger memories, throwing more hardware at the problem is cheaper then trying to bum five bytes out of a 50 KB program?
Something to think about, anyway.
... why is it not possible to use GTK widgets to render inside forms?
:-)
Basically, they need to be able to do things like apply CSS styles, layer widgets, composite them, and do other things that many toolkits/platforms don't support.
This also results in having the same UI regardless of what platform you're running Mozilla on. This goes back to the Netscape goal of turning Navigator into a development platform. There are pros and cons to this approach, but I guess they figured they were already most of the way there, so they might as well finish the job.
It also reduces platform-specific code, which any developer will tell you makes your life easier.
Given that both Mozilla and GTK are Open Source Software, I'm sure someone will eventually do the work to merge the two, even if it requires rewritting GTK.
Netscape is badly written junk (oh, sorry, I have a knack for being redundant).
I wouldn't go that far. Netscape Navigator is a case of code pressed beyond the entropy point. Any non-trivial code base which is being modified will experience a case of gradually diminishing returns. Eventually, the system becomes unmaintainable and you have to start over.
Navigator 3 was pretty good (for its day), but Navigator 4 really shows the age of the codebase. And, as the Mozilla project found out, trying to do anything major with the V4 codebase resulted in a total collapse.
This doesn't make the software "badly written junk"; it makes it software which should have been thrown away after the V3 series. Unfortunately, management so rarely recognizes such cases.
Read Fred Brooks's timeless work on software engineering, The Mythical Man-Month, for more about this and other software principles.
More seriously, Netscape on Linux needs 64 megs of memory.
Um, I'm not sure what version of Netscape you're running, but on my system, right now, Navigator 4.7 has a virtual segment size of just under 15 MB and a resident size of just under 11 MB.
If Linux can't swap Netscape in and out fast enough, Netscape gets impatient and dies.
If a process needs more memory, it will block until the kernel satisfies the request or fails it. It cannot "get impatient"; it will not be run by the kernel while memory is being managed.
Netscape seems to have a HUGE memory leak that causes it to eat up memory, especially on pages with lots of images (ahem!).
Here you have found the real problem. There is a memory leak in Navigator V4 related to pages with complex layouts and/or lots of images. This causes Navigator to gradually leak memory over time. It isn't using this memory, but it eats up swap space and slows down the system. If you don't exit and restart it, Navigator will eventually exhaust virtual memory, resulting in failed memory allocation requests and a program abort.
This is, of course, a Bad Thing. My workaround (until Mozilla V5 becomes viable) is to keep an eye on the memory gauge in the GNOME Panel, and exit and restart Navigator if it grows beyond a reasonable size.
Netscape can bring the entire system down if a) your CPU overheats or b) your video card overheats.
That's a very misleading statement. The cause of such a crash is a heat problem. Anything could cause that. Blaming Navigator because it was the forground application at the time is silly. If your system is overheating, you need to fix your system cooling, not blame the software!
Netscape can also bring the entire system down if by some random chance it happens to eat memory just as another process (syslog, etc.) tries to grab it, but for some reason Netscape wins the race. In this case, I've seen INIT panic and Ctrl-Alt-Del get disabled.
Again, here you're blaming Navigator for a system problem. It is a known problem that Linux handles out-of-memory situations poorly. I believe the 2.3 kernel addresses this, by making sure system processes have a proper reserve and doing a better job of killing processes to free memory in an emergency.
Anything that eats memory will cause this problem. while(1) malloc(10000); will do it. Are you going to blame GCC for that, next?
What I recommend doing is instituing user resource limits (using ulimit) to prevent any single process from exhausting your system's memory.
In the end, if an application crashes your system, there is something wrong with the entire system. It doesn't matter which application did it.