Re:stick to e16 for a wm, but e17 has nice stuff
on
State of the E-nion
·
· Score: 1
the E folk are to be commended for their excellent modular development -- many of these components are already being used by other projects (imlib2 in particular), and many of the others either are or soon will be in shape to be used in other projects too.
Yes and no. Here's a quote from the article:
ecore: Currently it has basic IPC wrapping, X wrapping, Evas wrapping, job handling...
I think it's kind of a bad sign when you have to write a wrapper for your own library to be released with your library, so you can write your program which depends on... your library.
At the same time I commend them for the effort, and I'm glad they took the time to design it fully.
One thing I don't like is that Afterstep seems to have just disappeared. I remember going to their official homepage a few days ago (reflected by Google) but now is evidently the really broken homepage of some graphic design girl. A few years ago I was looking into Afterstep and noticed that their people had been working on a 2.0 release for quite some time, in the process creating libAfterImage which reportedly is blazing fast and produces beautiful results. I admit it, I kind of miss the crazy thousands-of-window-managers situation that Linux was in when I first got involved. And this whole GRFXGIRL thing is just weird. Especially since her copyright is last year and the Google page shows the afterstep page as being copyrighted this year.
There are milestone builds on a frequency on months wherein an attempt is made to level-set at a certain level of stability, and nightly builds that are expected to be fraught with bugs, but steadily progress towards the next milestone build.
It wasn't always like that though. I remember in the pre 0.9.x days, when they were still called Milestone X, the releases were multiple months apart. And with the last beta release of it coming out about a month ago, I don't really thing you have anything to complain about.
After all, Mozilla has thousands of developers, whereas Safari has what, a couple dozen if that? We're talking orders of magnitude difference here. The code doesn't write itself. You assume that every night there will be a cohesive program you can just download and install, but it might not be the case. Hell, they don't even have to guarantee that it will build every night.
I personally have had several bugs looked at in Mozilla/Chimera(Camino), and feel a much stronger involvement with those products as a direct result of this.
There's a concept in software engineering called egoless programming. The idea is that, you don't get too attached to the code you write. Your beautiful function might add bugs in other places, and you may have to remove it later, but it doesn't mean you should feel bad about it. By the same token, I'm not really sure why anyone should give a shit how much involvement you feel like you have with a project. Apple would rather release their browser on their terms, and fix the bugs on their terms. I'm content knowing that the engine of it is open source. Why isn't that enough? It's their code after all.
It's how Microsoft would do Open Source.
No, I'm pretty sure that's Shared Source, actually, with the premise being that if you want to see the source, you can (provided you don't modify it or use it any way).
There's nothing wrong with doing things for no other reason than fun.
This is true. But at the same time, the atmosphere in America (at least from where I'm sitting) is getting so passive that the thought of sitting down and actually doing something is really getting foreign to us. Why bother making a game, you can just wait for the next one to come out? Why bother writing? You can just wait for another novel to come out. Why bother learning an instrument? You'll never be as good as your heroes.
The bar has been raised so far it's effectively beyond the reach of your average person, unless they dedicate their lives to it.
Strangely enough, a lot of people like that wind up really good at games, because it's just something they find themselves in front of long enough to excel at. I think that's kind of the point of the parent poster. If instead of saying, "fuck it, I'll go play EQ" they said, "I'll spend an hour on my guitar tonight, and an hour writing, and an hour beating off" they are quite likely to eventually find themselves very good at guitar and writing. You can't help but improve if you do something enough.
And there are two good reasons to have your fun doing something "productive" as you put it.
1. You'll feel better about yourself. When you're laying in bed awake at night wondering what you're doing with yourself, it's easier to remember your skill with whatever you've been doing. It helps your memory ("remember when I was totally pathetic at Python? That was four years ago!"). You're not going to remember that you made it to level 30 in EQ after losing countless hours to the game.
2. You improve the world for other people. Commercialism pervades television, radio and is a visual nuissance in basically every direction you can look. Originality is unheard of on TV and the radio. Bringing some originality to the world is something community doesn't forget. You make friends, you make fans, you grow in vision and perspective. None of these things happen on EQ, except perhaps for making friends, and you'll be lucky if you can retain an EQ friend outside of EQ.
Also, to specifically knock EQ, I haven't met anyone yet who claims that EQ was a "pure joy." The players are confrontational, the company is disinterested, etc. At least with a pen-and-paper role playing game you're spending time with people you honestly enjoy and exercising your imagination.
All that said, if you spend all day being "productive" I understand if you don't want to do it at night. But in my experience, we put a little too much faith in the power of money to make us happy. It shouldn't be about that.
With the life we know, this indeed holds true. But I'm sure there could be life based on any number of weird building blocks, we just don't have them here.
With a little application of the anthropic principle, why should we expect other life-bearing planets to be wildly different? I agree with your point that we shouldn't be looking just for what we have here, but we have two reasons to do that: 1. we know our data is good, and 2. we really don't know what else to look for.
In fact, it reinforces my total agreement with you that [D|R]NA is not necessary for life. I believe that a good minimum for definining life is just adaptive behavior, i.e. evolution. Of course we aren't inclined to say things like evolutionary algorithms or simple adaptive chemical processes "are" life... but perhaps part of the problem with that is that we simply haven't let these things go on long enough to recognize them as life.
In a universe this vast, it seems impossible to me that we could be the only life. One thing which I expect we'll find if we explore the universe in greater detail is that it's full of weird things. The weirdness of life doesn't come across when we sit at home in ultra-introspective mode, categorizing the minute differences between insects as though they're legendary incredible differences. The weirdness will come across when we're confronted by complex interrelated chemical and physical processes on other worlds, and our biologists won't want to call it life, while the rest of us will (or vice versa).
For once a little manifest destiny would have been just fine. Instead, we're peering through expensive telescopes, while our ancestors are pointing at the leaning tower and asking us why we aren't dropping things from the top of it.
Probably because most of the truly hard-core lispers have transcended need for physical "containers" in this universe, and instead exist as pure energy...
The big lisper around here that I know once had a dream in which he was a lisp program, consing on additional limbs to do his tasks as it was going on.
Actually it works for all PIC compiled binaries, which on my system works out to everything except winex and svgalib (or something). It's essentially taking the binaries and libraries on your system and making them faster to load by doing part of the dynamic linker's work ahead of time.
I couldn't tell you why more people aren't using it, but it works for any binary on your system. I think it's just more noticable because KDE takes the most noticable amount of time to start.
I have actually turned my optimizations down to -march=athlon -O2 -pipe. -O3 turns on -finline-functions and -frename-registers. There is a school of thought that says that smaller binaries will execute faster because more of the code will be in the CPU cache at any given time. Of course, I guess if I really meant it, I'd also turn on -Os instead of -O2 to optimize for size.
In any case, the optimization that's most noticable for my computer is not a gcc option at all, but rather using prelink to make all of my programs start up faster. I read about it on the Gentoo page, it makes KDE start up in about 3 seconds instead of 8 or 10.
Don't take this as me sticking up for a pop star. She's cute in an odd sort of way, but I definitely don't like the music. I'm more of a Blue Oyster Cult/Led Zeppelin kind of guy, crossing over to power/progressive metal as of late. Actually now that I think about it, the current batch of pop girls (Avril, Michelle Branch, Vanessa Carlton) are the most attractive group I can remember.
But I have to stick up for her on a few points. Having heard two songs I think I could say that the lyrics don't seem to fit the genre of music. I read a slightly better interview of her in Newsweek a few months back, and I seem to recall she said something similar to what the previous poster had mentioned, that she focused on the lyrics and the label had used a lot of influence over the music on the first album. You can't really fault someone for saying "um" or "like" a lot in an interview because it is on the fly and most of us aren't practiced in rhetoric, don't take our time, and stutter all over the language in that situation. Also, when you're in a multiplatinum position, you probably don't want to talk about what the next album is going to be like if you don't know yet. I could forgive a lot saying she just came out of a huge album and tour, I don't blame her for being a bit exhausted and not really wanting to think about the next album right away. Particularly when you get asked it a lot and are going through the harrowing new star thing.
That said, my interpretation of the whole Avril thing is this: she was on the path to being one of their born 'n' bred country pop sensations. She for whatever reason came out differently than they expected. At 15, her "rebellion" probably didn't consist of walked into the CEO's office and terminating the contract over musical differences. Did anyone notice how mention of her parents is curiously missing from these interviews? My guess is that she told her parents she wasn't going to do it anymore. They went in and told the label, who came up with a "compromise." She could do "whatever she wanted"--as long as she followed along with what they wanted in the areas that didn't have to do with the music. Since she really didn't know a damn thing about music other than pop, it's what her first album sounds like. It's what she knew. Of course the rebellion didn't have anything to do with musical differences, she just didn't want to become a primped and preened mass media sex object.
Predictably, the label saw this as an excellent chance to make her a mass media sex object, only aiming her for the so-called angst-filled teenager market rather than the popular pop market. The uproar over this now seems really no different than the uproar that followed the release of American Pie or Something About Mary, except it's music and it's several orders of magnitude more benign. In actual fact, every teenager has angst, so her demographic is huge. Britney can't exactly convey the angst message, plus she suffered from over exposure. (I'd argue that pop is an inherently limited media that prevents more complex messages than simple teenage love/angst from being transmitted in the first place, but that's another rant.)
I bet they gave her all the freedom she could think of, and then just shuffled her off to do their photo shoots and various other PR without making a big deal about it. Being completely unworldly, she doesn't know 1) what she's rebelling against, or 2) what is intrinsic to the music business that she should be rebelling against, and isn't.
If my theory is correct, here's what I would expect to happen in the upcoming years:
1. Each successive album she creates is more of a departure from the first album until she finds her style (probably 2 albums from now).
2. Her fanbase grows smaller but more dedicated until she is taken seriously as a "real artist" in some circles. Along with that, it will be acknowledged that she has her own style, even if it's representative of some genre, but that genre will not be pop.
3. The label eventually drops her, inspite of which she continues to release albums on a smaller label and fill medium-sized venues well into her old age.
Is it likely? No. But the fact that her bass player left because he was tired of being a "marketing tool" might merely mean he is too talented to be wasted playing second fiddle for a clueless teenage girl who gets all the time in the spotlight. (It's not real likely he's talented either, but this is the music industry not the software industry). But here's what I expect would happen if she is nothing more than a marketing trick:
1. There are 2 additional albums from her, neither showing any marked improvement in skill in terms of songwriting or lyrics (or even any additional maturity or increase in vocabulary).
2. Each successive album cover shows her revealing more skin (in the other scenario, album covers are less likely to feature her prominently).
3. After the third album is a complete failure, the label drops her and she is never heard from again in any capacity. 20 years later, on VH1, we learn that she spent half her money moving to India where she teaches English and Computer Science in a middle school and is a devout Hindu. Or, alternately, she becomes an MTV anchorwench, which I think is at the same level of general interest and importance.
Unfortunately, I have to agree. I used to have a full 8 GB BeOS partition on my machine which I was happily neglecting until news of the BeOS MAX! "Distribution" fell on me, and I had to install it.
Well, right after I installed it was treated to booting into an OS that thought I was in France. Fine, not a big deal, but it would have been nice if the documentation (all in English) had mentioned this and maybe even described the process by which you fix it the language settings.
Then I got to experience the full glee of BeOS resending 16 email messages I had already sent from its mail program... about two years ago! I emailed the developers about this and got no response.
This is usually where the Be apologists will insert some vague praise of it's microkernel architecture. Yeah, it's great I can drag and drop a driver into the system folder and have it start working right then, but so what if it's totally impractical to actually *use.*
So I blew away the partition and said, "Goodbye BeOS, it had been fun." When an OS is new, you can forgive it for having no application support and no hardware drivers. Now, 2003, we're nearing 8 years with BeOS on the PC and almost all of the drivers we have today we had back then (exception: a non-accellerated nvidia driver). Application support continues to hover around the few commercial apps it had three years ago (though I believe Gobe has dropped BeOS support for Productive). When I last ran the OS, most of the software for it suffered from the same Windows delusion that every schmuck who downloads a shareware program is willing to chip in $10 for it. Consequently, actually achieving productivity with BeOS was difficult because you'd wind up paying hundreds in piss ant shareware fees to unlock the full features of whatever it was you wanted to use (see BeXL, all of the good code editors, SoundPlay).
So BeOS lost a fan in me. The only chance for redemption will be when OpenBeOS starts making releases, but even then it will be a long shot. If you doubt me, check my previous posts and you'll see, I used to be one of their supporters around here, but I give up. OS X certainly kicks BeOS in the nads, thanks in no small part to NEXTSTEP. I haven't used WinXP but wouldn't be surprised if much of what made BeOS advanced almost a decade ago had finally been integrated into Windows.
...more than 90% of the human genome is junk that has no expressive effect anyway (according to some theories it helps protect the rest of the genome.) Even point mutations in coding sections of the DNA often do not significantly alter the shape of the protein it codes for, and many proteins are coded for in several locations in the genome.
Do you think the DNA started off that way? Or is it possible it's more like RAID for organisms, that it evolved that way simply because it is more fault resistant that way?
To back up the parent of the previous post, I know someone who has been programming since the late sixties or early seventies. While that may not be quite 50 years ago, it certainly means he has had some experience with tape storage, even punch card FORTRAN initially, and probably worse though he doesn't talk about it very often.
What does he do now? He is still an application writer, his language of choice being Python and his file format of choice being XML. Frankly, I think this is quite telling: his opinion if I understand it correctly is that since we have the power, we shouldn't waste time writing things lower level than necessary. By using Python and XML he's far, far removed from the ordinary perils of yesteryear like memory management, pure procedural programming, even memory and disk size limitations.
And frankly, while those things are difficult to deal with, they're also very rote and don't leave much expression to the software engineer. People who favor C and to some extent C++ usually admit that there is some pleasure in the sheer amount of control in using the language; it's my opinion that people using Perl, Python and to some extent Java are the people reading books with "Practical" or "Design" in the title, and that's really a better way to do things.
In reply I would merely point you to the bloatware that exists today on all systems. You call that the work of efficient (read good) coders? I don't.
It's an easy attack to make, with some degree of merit. The qualifications for being a coder these days are certainly less strict than they were at one time. However, the observation of the post you were replying to was that the older systems had less to do than modern ones. When you resize your browser window you're doing an operation that, as far as a 386 would be concerned, is non-trivial. Add to that the sheer size of the parsed webpage which generates the pretty view you see, and you've got yourself a lot of graphical things to do, and a huge datastructure in RAM. This is not the kind of problem that can be solved simply by being able to manually manage memory from assembler. This is the kind of problem that requires an intelligent design from the get-go, so that optimizations can be placed in the places where they are required as needed.
Bloatware? Probably. People who needed computers for whatever reason seemed to be getting along with them just fine without GUIs, or multiprocessing, or realtime 3D games. All of these additions is going to consume resources both when written and when used. I won't argue with you that Windows would have been better if it were based on a clean design. Clearly it would have, and on Linux we now have many desktop systems based on (if not a good deal more forethought) at least the trial-and-error process that produced the early GUIs done with a faster turnaround. Unfortunately, the users have come to rely on GUIs, pretty widgets, and browsers that resize. If they were not, perhaps we could cut down on the code quite a bit.
Also, one thing about my friend I mentioned earlier: while his code is extremely well-designed, he seems to have a fundamental lack of understanding of ideas such as UI design and concurrency. None of his programs as far as I have seen have used threading, even the GUI ones, and the few GUI programs I have seen were beyond the ugliness I expect from TK. He wrote an abstraction layer for a database that implemented foreign key constraints, and was at a bit of a loss when I first tried to explain to him that it wouldn't carry over necessarily if multiple copies of his application were running simultaneously. So we all have these problems, and we all try to get better.
If you want to see well designed and implemented code, I recommend you pick up a copy of BeOS. By sacrificing backwards compatibility, they managed to create an operating system from scratch based on object-oriented principals. It's quite amazing when you realize the things that you could do with it that you couldn't do with Windows, yet it was a tiny fraction of the size of Windows when fully installed. For example:
Active queries. Linux acquired something similar via FAM but you need application support for it. Basically, you could search for files based on their attributes, and as files were removed or added to the system they would disappear or appear in the query. The query could be used like a directory for all programs that could access one (AFAICT).
Device drivers took effect immediately upon placing them in the appropriate system directory (except display drivers).
Applications were tiny - the HTML 3.0 compliant browser came in at under a meg for the whole binary. I never saw an app larger that 5 MB.
A full install came in at about 300 MB (comparable to OpenBSD) IIRC.
Now I'm going to get some sleep and try to forget about the sorry state of computing we're in right now.
Most gamers would get real tired of having to shell out tons of money just to have a powerful character, and having to put up with little shits with too much money and attitude ruining it for everyone else.
It worked for Wizards of the Coast.
I've had friends that have spent upwards of $500 on Magic cards, and that probably is a small figure compared to what some people here have spent. One of them said, "If I saw a $500 game in the store, I would back away slowly and make sure not to touch it." But since it was just $8 here, $8 there, a couple expansions for $3 or $4... it added up quickly. Also, there was always a new "killer deck" that could be made by combining certain rare cards in interesting ways, at great cost.
I haven't seen their scheme (the website is down) but if their smart about it, they can make a deal no power-addicted gamer can refuse. It's happened before and it will happen again. People are always willing to pay a premium for their entertainment.
I'd be willing to be this is an effect of your desire to see it. A few loud people can seem like a lot of people; as we all know it's that misperception that special interest groups thrive on.
At my school, most of my friends argue for little to no restriction of either guns or information. After all, one wants to be free, and the other guarantees our freedom. I have an EFF sticker on my car, and most of my friends own guns. I might be "the liberal," but in saying that they really mean I'm the one who doesn't own a gun or a membership card to the NRA. I may not be a hardcore gun proponent, but I certainly try defend the 2nd amendment, with the little bit of 1st amendment rights I still have.
Of course, this might just be my perception too.:)
I've been looking for someone to ask this question to, because I'm interested in setting up a RAID at some point in the near future. I'm going to need between 480 and 640 GB of actual storage in the end. Cost is an issue, but I'm willing to spend ~$1000, so I'm thinking IDE is what I can afford.
The purpose of the array is to serve high quality audio files to myself and my friends. We're talking about 30MB - 150MB sized files probably averaging around 50 MB in size, and it should be capable of streaming about four of them at a time at a minimum of 1 or 2 MB/s.
My first thought was to use RAID 5, since it traditionally is the most economical variety. But then understanding the difference between 3 and 5 seems to indicate that 3 would stream faster and deal with large files better. But, on the other hand, 5 handles multiple simultaneous reads or writes more gracefully than 3, which essentially has all of the drives doing the same thing at the same time.
So I'm wondering which RAID level I should use for this application. These aren't video editing sized files if I understand the size they can get, but they are definitely not what I would normally consider small files either. Is the performance difference between 3 and 5 truly all that great? Does it go to the toilet when multiple people are using the same drive? What problems will I notice with 3 or 5?
Since I would be sharing this via AFS, it is safe to say that these files are going to be copied around the network essentially as soon as they are opened, which means each file read will probably be reading the whole file, by the way (if you don't use AFS, not intending to sound condescending).
Maybe it's not for Mp3s. I'm interested in converting my audio collection to FLAC format so I don't have to tolerate loss of quality. If I do that, I computed that my puny 15 GB colletion would take up about 100 GB, maybe more. And so I have room to grow, I'd want to have more like 300 or 400 GB of free space to do this.
I've been planning on buying a RAID set up to accomplish this. RAID, as you all know, uses more than one disk. You all know, apparantly, that the R in RAID is for redundancy. I'm not being redundant, I hope. RAID would give me the room for this, as you all know, by using one disk.
Apparantly.
So like, I need a lot of space. And this looks cool (though it might not be obvious).:)
1. Do they need to know how to install the OS first, or should I let them look that up on their own while I make them power-users?
I learned how to use Linux on a coworker's computer at an internship I had. I didn't learn how to install Linux until I was comfortable using it at work, and I found the installation to be pretty simple.
You'll run into a bit of a chicken-and-egg problem here if you think about it too hard. In order to learn how to install the OS they will have to deal with a bunch of things all at once: partitioning, filesystems, mountpoints, and thousands of names of programs that sound cool. Installation is pretty easy if you have a handle on these concepts, but if you don't you'll find it very exasperating--little documentation and a lot at stake. But if you make then a power user first, then they get hands-on experience with the OS, they have a reason to want to install it, and they'll be more capable at installing the OS itself on their own when it comes time to do that.
2. What distributions of Linux and BSD should they be first introduced to? (I'm only familiar with Debian, and I know virtually nil about *BSD.)
If you focus on Unix, and what the Unix philosophy is, then it won't matter much. Being able to cast around sed & awk, make command lines with for loops, and configure networking manually is a lot better than knowing how to use a graphical system configuration/installation utility like linuxconf or dselect. Plus, focusing on the tools is more likely to carry over to other operating systems. Teach them regular expressions. Teach them loops and conditionals. Don't teach them how to run RPM just because they might have to someday; it's rote, they won't remember it when they have to use it, and they'll be falling asleep as you explain it.
3. Initially, do they need to be more adept at the GUI, or do they first need to know how to use the shell?
No question: the shell. Even if you use the GUI all day (I run KDE!), lots of low-level things are written in bash/ksh/csh. If you're going to be able to sit down at any Unix computer and use it, you're going to have to be competant with at least sh/bash. The first thing I do when I sit down is fire up a console. They should be doing that automatically by the end of your class: the console is where the work is done.
There's a lot of variety with GUIs, and it can be great fun to mess around with window managers and so forth. Give them the ability to change these things (teach them how to edit their.xinitrc), and point them to the Window Managers for X webpage, let them tinker with it. But definitely teach the shell, because that's where the real power is.
4. Should I give away Debian CDs no-questions-asked, or should I talk with the almighty Parents so little Daniel doesn't install Linux over Dad's 'work computer.
Make the CDs available, but do talk to the parents. Make sure the parents realize that if the kid is going to do anything with "partitioning" they should go out and buy a copy of Partition Magic so they don't permanently screw everything up. That software was the best $50 I spent in high school, when I was a big OS installing weirdo. My parents never lost any data.:)
Clearly, some of the kids are going to want to install it at home, and it's not necessary that their parents understand the whole idea behind partitioning (though it would be nice). It's also not necessary to scare the parents into preventing the kid from doing what they want to do out of fear. I geniunely believe PM is the right answer to this problem: it's got undo until the last possible moment, and it's very user-friendly, which helps a lot when you're just getting started. Once that's taken care of there is very little that can be screwed up (most installers either ignore other drives or assume that if it's type "vfat" it shouldn't be formatted).
5. Are there any other key issue I need to think about?
Yes, definitely try and think of a curriculum. Kids who are interested can absorb tons of material, but you've got to keep it interesting. The best way to do that is to not focus on the rote doing of things, but to talk about the possibilities of things. Don't just say, here's regular expressions, instead say with regular expressions you can strip out everything in a file you don't want, get a list of all the email addresses on this webpage, etc. Keep them interested in the possibilities that you're opening up for them by teaching them the software.
Be aware of your philosophy and how it affects your decisions. If you believe in focus-follows-mouse, set up all the machines that way. Let them change it but make them try it out first: expose them to as many possibilities as possible. Have vi and emacs on the machines, and let them pick. You've probably got a whole set of quasi-religious beliefs about Linux. This is your chance to instill those beliefs in other people, but make sure that you point out your biases. Most of the kids who take your class are going to take Debian home to install, which is fine as long as you don't focus on the "Debianness of it all."
Show them the Linux that newbies don't usually get to see, because of the religious wars that go on. Regular expressions are another great example, because vi and emacs both support them. You may not be able to extend vi in lisp (and you might with vim), so don't focus on that. Keep it open.
Here's my advice on the actual software to talk about:
Shell: bash
Distribution: Debian or Redhat
Window Manager: Blackbox or WindowMaker
Browser: Galeon
I suggest bash because of it's ubiquity. Everybody uses it and knows it, and it's very expressive in its weird little way. Debian or Redhat because they are also everywhere, but don't focus on rpm or apt. I suggest Blackbox or WindowMaker because they don't have a bazillion settings that you can spend all day on. If they fire up GNOME or KDE they'll be so absorbed in the coolness of it all that they'll spend the whole day configuring it and not listening to you. Galeon is simpler than Mozilla but uses the same rendering engine, and makes webpages look pretty decent. If you have them firing up Netscape 4.7 or Mozilla, they'll either be disappointed or distracted (my guess). Past these options I don't think the software matters much if they have choice.
Good luck with your project! It sounds like a step in the right direction to me.
GNU-Darwin is Apple's Darwin. Or at least a binary compatible re-distribution of it. At least a fork. Frankly, their website isn't completely informative on this issue, but there seem to be three Darwins:
Frankly, I'm a little unclear on the differences but either way calling it a "shitty distro with ripped off GUI graphics" is a stretch. GNU Darwin seems to me to be a GNU operating system built on an Apple-modified BSD kernel. Which sounds kind of perverted, but not necessarily "shit." Hey, they've ported it to x86! It's got to be at least important to x86 as NetBSD.:)
Apple's lawyers are going to have a field day with this one.
The source is open. Read all about it at Apple's Darwin page. There's nothing to sue anyone over, although Apple can via their license simply "revoke" the source and keep all of the outside changes.
Actually, according to the license, when you take any source covered by the APSL, you're required to register with Apple. If the developers didn't do that, Apple would have a valid case to sue them over. If they did (and I'm positive they did, since they link to the damn license off their page), then Apple really doesn't have anything to get them on, unless they're keeping changes private. If they were doing that it wouldn't be GNU either.
I think your reaction is a little uninformed. A simple websearch turned up quite a bit of information on this topic, even a nice rant from the FSF about the APSL.
I completely agree. In fact, that's why Python is my language of choice--I get more done faster, and unless I write a shitty algorithm I really don't notice a speed problem.
However, when it comes to this particular benchmark, it has more to do with performance on a server somewhere handling thousands of simultaneous connections. Web applications, if you will. And in this particular case, a performance difference of 10% could mean a hundred users get locked out, and responsiveness to the others will be very bad.
If you ask me, it's one of the great mysteries of the computer industry why desktop apps are written in C/C++ and web apps are written in Java, Perl, or.Net. After all, we've been admitting for years that scripting languages are OK unless you need the extra performance, and these guys do need that extra drop. So why aren't they writing these programs in C or C++?:) Clearly it isn't the libraries or types involved, since these are almost always available to C or C++ also.
That's the $15 million dollar question, now isn't it? We seem to optimizing for RAD and performance, rather than just performance. Perhaps we should rethink our priorities, move scripting languages onto the desktop, and compiled languages onto the server...
Well, they're about slashdotted now. They lost my last request, and it says they have almost 2000 anonymous users. I sometimes think the reason I like reading Slashdot isn't because of the great links and articles, but instead because I like being a part of the goddamned Slashdot effect.:)
Which brings me to the point. Ya know, about the only site that can handle the Slashdot effect is Slashdot. So maybe Taco should write an article like this (or maybe he has?). The Slashdot guys know what they're doing, we should pay attention. Although I find it interesting that when slashdot does "go down," the only way I know is because for some reason it's telling me I have to log in (which is a lot nicer than Squid telling me the server's gone).
It's "suite," asshole. You could at least make sure you got the word right before you use it 6 times.
No one asked you what was noble and what wasn't. If you know the slightest thing about programming you'd be aware that you don't start writing something beat what's already been made, you start by writing something, and then you add to it, and as your knowledge increases you can take on more and more complex or difficult tasks. At this moment, there is no open source 3D modeller/renderer/kitchen sink which these guys can go to and learn how to do these complex things. The algorithms used by these complicated applications are all but unknown to the community; the proprietary ones certainly aren't published in books, and even if they were it would take a great deal of skill to "get inside" them, figure out how they work, etc.
We have to start at the beginning.
Did you use the Gimp three years ago? I tell you it wasn't a "photoshop-alike" then. We have to start somewhere and then move up. There are programs we have in free software that are so advanced, there's no analog for them in closed-source. For example, the RADIANCE renderer is the only one that does light accurately, and it has been free for ages and will become open source in December. Aspell uses a new algorithm that beats every other spell checker. At its inception, would you have said we shouldn't invest this time and effort into another spell checker?
The problem is that people like you look at open source development like closed source development. As though when a project is formed, some segment of the total number of developers have to be allocated to it. Fortunately, the way it really works is developers work on whatever they are attracted to. Even huge projects like Mozilla have at the core less than 20 developers. I haven't checked, but I would guess that these 3D projects will have 2-4 core developers. Everyone else will contribute from time to time, or possibly even just once. But the people who work on Moonlight are not people who necessarily would have worked on Blender. They are not even necessarily people who would have developed for anything at all.
What does a professional setting have to do with whether or not something should be developed? Take enlightenment for example. I bet many people said, this is a window manager which will never be used in a professional setting. And yet I know several sysadmins who have used enlightenment as their window manager! It's like science, we don't research the things we think are going to bring about "useful" discoveries, because that's counter-productive and we never know which research will result in useful discoveries. Instead, we just generate all of the software we'd like to generate, and some of it will get used, and some will not. AWK is a good counter example; it was developed for a particular purpose for which it is extremely useful (parsing text files with very uniform structure). And yet, it has fallen by the wayside because Perl can do essentially the same things, but is a more powerful programming language in general.
To summarize:
Yes, we do need another <insert software-type here>.
No, it doesn't matter what the software is. You think every window manager should be like TWM?
People will develop what they want to develop, regardless of whether or not YOU think it is prudent or a good use of "resources" (ie. people besides you).
Start at the beginning, and work your way up. RMS stated that the GNU system would encompass everything from a shell to a spreadsheet. Linus just wanted a kernel. Which one has been useful the longest? Don't bite off more than you can chew.
While I was on a road trip over the summer we stopped in Oklahoma somewhere for gas. The ATM was "down," but I went over and played with it for a few minutes, found out that it was in fact running OS/2.
It's freaking weird to see the OS/2 window decoration on an error message in amber on an ATM though! Especially when you know it's the kind of ATM that normally looks like a telnet connection (characters only).
BTW: if you read about the server components, you should have noticed that most of these things exist already (postfix, cyrus, apache, inetd, proftpd, openldap2). Kolab is just going to tie them together and slap a pretty administative UI on top.
Slashdot Mirror
the E folk are to be commended for their excellent modular development -- many of these components are already being used by other projects (imlib2 in particular), and many of the others either are or soon will be in shape to be used in other projects too.
...
Yes and no. Here's a quote from the article:
ecore: Currently it has basic IPC wrapping, X wrapping, Evas wrapping, job handling
I think it's kind of a bad sign when you have to write a wrapper for your own library to be released with your library, so you can write your program which depends on... your library.
At the same time I commend them for the effort, and I'm glad they took the time to design it fully.
One thing I don't like is that Afterstep seems to have just disappeared. I remember going to their official homepage a few days ago (reflected by Google) but now is evidently the really broken homepage of some graphic design girl. A few years ago I was looking into Afterstep and noticed that their people had been working on a 2.0 release for quite some time, in the process creating libAfterImage which reportedly is blazing fast and produces beautiful results. I admit it, I kind of miss the crazy thousands-of-window-managers situation that Linux was in when I first got involved. And this whole GRFXGIRL thing is just weird. Especially since her copyright is last year and the Google page shows the afterstep page as being copyrighted this year.
--
Daniel
There are milestone builds on a frequency on months wherein an attempt is made to level-set at a certain level of stability, and nightly builds that are expected to be fraught with bugs, but steadily progress towards the next milestone build.
It wasn't always like that though. I remember in the pre 0.9.x days, when they were still called Milestone X, the releases were multiple months apart. And with the last beta release of it coming out about a month ago, I don't really thing you have anything to complain about.
After all, Mozilla has thousands of developers, whereas Safari has what, a couple dozen if that? We're talking orders of magnitude difference here. The code doesn't write itself. You assume that every night there will be a cohesive program you can just download and install, but it might not be the case. Hell, they don't even have to guarantee that it will build every night.
I personally have had several bugs looked at in Mozilla/Chimera(Camino), and feel a much stronger involvement with those products as a direct result of this.
There's a concept in software engineering called egoless programming. The idea is that, you don't get too attached to the code you write. Your beautiful function might add bugs in other places, and you may have to remove it later, but it doesn't mean you should feel bad about it. By the same token, I'm not really sure why anyone should give a shit how much involvement you feel like you have with a project. Apple would rather release their browser on their terms, and fix the bugs on their terms. I'm content knowing that the engine of it is open source. Why isn't that enough? It's their code after all.
It's how Microsoft would do Open Source.
No, I'm pretty sure that's Shared Source, actually, with the premise being that if you want to see the source, you can (provided you don't modify it or use it any way).
--
Daniel
at least it's a community. :)
--
Daniel
There's nothing wrong with doing things for no other reason than fun.
This is true. But at the same time, the atmosphere in America (at least from where I'm sitting) is getting so passive that the thought of sitting down and actually doing something is really getting foreign to us. Why bother making a game, you can just wait for the next one to come out? Why bother writing? You can just wait for another novel to come out. Why bother learning an instrument? You'll never be as good as your heroes.
The bar has been raised so far it's effectively beyond the reach of your average person, unless they dedicate their lives to it.
Strangely enough, a lot of people like that wind up really good at games, because it's just something they find themselves in front of long enough to excel at. I think that's kind of the point of the parent poster. If instead of saying, "fuck it, I'll go play EQ" they said, "I'll spend an hour on my guitar tonight, and an hour writing, and an hour beating off" they are quite likely to eventually find themselves very good at guitar and writing. You can't help but improve if you do something enough.
And there are two good reasons to have your fun doing something "productive" as you put it.
1. You'll feel better about yourself. When you're laying in bed awake at night wondering what you're doing with yourself, it's easier to remember your skill with whatever you've been doing. It helps your memory ("remember when I was totally pathetic at Python? That was four years ago!"). You're not going to remember that you made it to level 30 in EQ after losing countless hours to the game.
2. You improve the world for other people. Commercialism pervades television, radio and is a visual nuissance in basically every direction you can look. Originality is unheard of on TV and the radio. Bringing some originality to the world is something community doesn't forget. You make friends, you make fans, you grow in vision and perspective. None of these things happen on EQ, except perhaps for making friends, and you'll be lucky if you can retain an EQ friend outside of EQ.
Also, to specifically knock EQ, I haven't met anyone yet who claims that EQ was a "pure joy." The players are confrontational, the company is disinterested, etc. At least with a pen-and-paper role playing game you're spending time with people you honestly enjoy and exercising your imagination.
All that said, if you spend all day being "productive" I understand if you don't want to do it at night. But in my experience, we put a little too much faith in the power of money to make us happy. It shouldn't be about that.
--
Daniel
With the life we know, this indeed holds true. But I'm sure there could be life based on any number of weird building blocks, we just don't have them here.
With a little application of the anthropic principle, why should we expect other life-bearing planets to be wildly different? I agree with your point that we shouldn't be looking just for what we have here, but we have two reasons to do that: 1. we know our data is good, and 2. we really don't know what else to look for.
In fact, it reinforces my total agreement with you that [D|R]NA is not necessary for life. I believe that a good minimum for definining life is just adaptive behavior, i.e. evolution. Of course we aren't inclined to say things like evolutionary algorithms or simple adaptive chemical processes "are" life... but perhaps part of the problem with that is that we simply haven't let these things go on long enough to recognize them as life.
In a universe this vast, it seems impossible to me that we could be the only life. One thing which I expect we'll find if we explore the universe in greater detail is that it's full of weird things. The weirdness of life doesn't come across when we sit at home in ultra-introspective mode, categorizing the minute differences between insects as though they're legendary incredible differences. The weirdness will come across when we're confronted by complex interrelated chemical and physical processes on other worlds, and our biologists won't want to call it life, while the rest of us will (or vice versa).
For once a little manifest destiny would have been just fine. Instead, we're peering through expensive telescopes, while our ancestors are pointing at the leaning tower and asking us why we aren't dropping things from the top of it.
--
Daniel
Probably because most of the truly hard-core lispers have transcended need for physical "containers" in this universe, and instead exist as pure energy...
The big lisper around here that I know once had a dream in which he was a lisp program, consing on additional limbs to do his tasks as it was going on.
--
Daniel
It would be nice if the second link contained any data, or linked to anything with any data.
--
Daniel
See the Freshmeat project page for some of the answers here.
Actually it works for all PIC compiled binaries, which on my system works out to everything except winex and svgalib (or something). It's essentially taking the binaries and libraries on your system and making them faster to load by doing part of the dynamic linker's work ahead of time.
I couldn't tell you why more people aren't using it, but it works for any binary on your system. I think it's just more noticable because KDE takes the most noticable amount of time to start.
--
Daniel
I have actually turned my optimizations down to -march=athlon -O2 -pipe. -O3 turns on -finline-functions and -frename-registers. There is a school of thought that says that smaller binaries will execute faster because more of the code will be in the CPU cache at any given time. Of course, I guess if I really meant it, I'd also turn on -Os instead of -O2 to optimize for size.
In any case, the optimization that's most noticable for my computer is not a gcc option at all, but rather using prelink to make all of my programs start up faster. I read about it on the Gentoo page, it makes KDE start up in about 3 seconds instead of 8 or 10.
--
Daniel
Don't take this as me sticking up for a pop star. She's cute in an odd sort of way, but I definitely don't like the music. I'm more of a Blue Oyster Cult/Led Zeppelin kind of guy, crossing over to power/progressive metal as of late. Actually now that I think about it, the current batch of pop girls (Avril, Michelle Branch, Vanessa Carlton) are the most attractive group I can remember.
:)
But I have to stick up for her on a few points. Having heard two songs I think I could say that the lyrics don't seem to fit the genre of music. I read a slightly better interview of her in Newsweek a few months back, and I seem to recall she said something similar to what the previous poster had mentioned, that she focused on the lyrics and the label had used a lot of influence over the music on the first album. You can't really fault someone for saying "um" or "like" a lot in an interview because it is on the fly and most of us aren't practiced in rhetoric, don't take our time, and stutter all over the language in that situation. Also, when you're in a multiplatinum position, you probably don't want to talk about what the next album is going to be like if you don't know yet. I could forgive a lot saying she just came out of a huge album and tour, I don't blame her for being a bit exhausted and not really wanting to think about the next album right away. Particularly when you get asked it a lot and are going through the harrowing new star thing.
That said, my interpretation of the whole Avril thing is this: she was on the path to being one of their born 'n' bred country pop sensations. She for whatever reason came out differently than they expected. At 15, her "rebellion" probably didn't consist of walked into the CEO's office and terminating the contract over musical differences. Did anyone notice how mention of her parents is curiously missing from these interviews? My guess is that she told her parents she wasn't going to do it anymore. They went in and told the label, who came up with a "compromise." She could do "whatever she wanted"--as long as she followed along with what they wanted in the areas that didn't have to do with the music. Since she really didn't know a damn thing about music other than pop, it's what her first album sounds like. It's what she knew. Of course the rebellion didn't have anything to do with musical differences, she just didn't want to become a primped and preened mass media sex object.
Predictably, the label saw this as an excellent chance to make her a mass media sex object, only aiming her for the so-called angst-filled teenager market rather than the popular pop market. The uproar over this now seems really no different than the uproar that followed the release of American Pie or Something About Mary, except it's music and it's several orders of magnitude more benign. In actual fact, every teenager has angst, so her demographic is huge. Britney can't exactly convey the angst message, plus she suffered from over exposure. (I'd argue that pop is an inherently limited media that prevents more complex messages than simple teenage love/angst from being transmitted in the first place, but that's another rant.)
I bet they gave her all the freedom she could think of, and then just shuffled her off to do their photo shoots and various other PR without making a big deal about it. Being completely unworldly, she doesn't know 1) what she's rebelling against, or 2) what is intrinsic to the music business that she should be rebelling against, and isn't.
If my theory is correct, here's what I would expect to happen in the upcoming years:
1. Each successive album she creates is more of a departure from the first album until she finds her style (probably 2 albums from now).
2. Her fanbase grows smaller but more dedicated until she is taken seriously as a "real artist" in some circles. Along with that, it will be acknowledged that she has her own style, even if it's representative of some genre, but that genre will not be pop.
3. The label eventually drops her, inspite of which she continues to release albums on a smaller label and fill medium-sized venues well into her old age.
Is it likely? No. But the fact that her bass player left because he was tired of being a "marketing tool" might merely mean he is too talented to be wasted playing second fiddle for a clueless teenage girl who gets all the time in the spotlight. (It's not real likely he's talented either, but this is the music industry not the software industry). But here's what I expect would happen if she is nothing more than a marketing trick:
1. There are 2 additional albums from her, neither showing any marked improvement in skill in terms of songwriting or lyrics (or even any additional maturity or increase in vocabulary).
2. Each successive album cover shows her revealing more skin (in the other scenario, album covers are less likely to feature her prominently).
3. After the third album is a complete failure, the label drops her and she is never heard from again in any capacity. 20 years later, on VH1, we learn that she spent half her money moving to India where she teaches English and Computer Science in a middle school and is a devout Hindu. Or, alternately, she becomes an MTV anchorwench, which I think is at the same level of general interest and importance.
Of course, I could be wrong.
--
Daniel
forgot the additional -1 for saying "anywho". :)
--
Daniel
Unfortunately, I have to agree. I used to have a full 8 GB BeOS partition on my machine which I was happily neglecting until news of the BeOS MAX! "Distribution" fell on me, and I had to install it.
Well, right after I installed it was treated to booting into an OS that thought I was in France. Fine, not a big deal, but it would have been nice if the documentation (all in English) had mentioned this and maybe even described the process by which you fix it the language settings.
Then I got to experience the full glee of BeOS resending 16 email messages I had already sent from its mail program... about two years ago! I emailed the developers about this and got no response.
This is usually where the Be apologists will insert some vague praise of it's microkernel architecture. Yeah, it's great I can drag and drop a driver into the system folder and have it start working right then, but so what if it's totally impractical to actually *use.*
So I blew away the partition and said, "Goodbye BeOS, it had been fun." When an OS is new, you can forgive it for having no application support and no hardware drivers. Now, 2003, we're nearing 8 years with BeOS on the PC and almost all of the drivers we have today we had back then (exception: a non-accellerated nvidia driver). Application support continues to hover around the few commercial apps it had three years ago (though I believe Gobe has dropped BeOS support for Productive). When I last ran the OS, most of the software for it suffered from the same Windows delusion that every schmuck who downloads a shareware program is willing to chip in $10 for it. Consequently, actually achieving productivity with BeOS was difficult because you'd wind up paying hundreds in piss ant shareware fees to unlock the full features of whatever it was you wanted to use (see BeXL, all of the good code editors, SoundPlay).
So BeOS lost a fan in me. The only chance for redemption will be when OpenBeOS starts making releases, but even then it will be a long shot. If you doubt me, check my previous posts and you'll see, I used to be one of their supporters around here, but I give up. OS X certainly kicks BeOS in the nads, thanks in no small part to NEXTSTEP. I haven't used WinXP but wouldn't be surprised if much of what made BeOS advanced almost a decade ago had finally been integrated into Windows.
--
Daniel
...more than 90% of the human genome is junk that has no expressive effect anyway (according to some theories it helps protect the rest of the genome.) Even point mutations in coding sections of the DNA often do not significantly alter the shape of the protein it codes for, and many proteins are coded for in several locations in the genome.
Do you think the DNA started off that way? Or is it possible it's more like RAID for organisms, that it evolved that way simply because it is more fault resistant that way?
--
Daniel
What does he do now? He is still an application writer, his language of choice being Python and his file format of choice being XML. Frankly, I think this is quite telling: his opinion if I understand it correctly is that since we have the power, we shouldn't waste time writing things lower level than necessary. By using Python and XML he's far, far removed from the ordinary perils of yesteryear like memory management, pure procedural programming, even memory and disk size limitations.
And frankly, while those things are difficult to deal with, they're also very rote and don't leave much expression to the software engineer. People who favor C and to some extent C++ usually admit that there is some pleasure in the sheer amount of control in using the language; it's my opinion that people using Perl, Python and to some extent Java are the people reading books with "Practical" or "Design" in the title, and that's really a better way to do things.
In reply I would merely point you to the bloatware that exists today on all systems. You call that the work of efficient (read good) coders? I don't.
It's an easy attack to make, with some degree of merit. The qualifications for being a coder these days are certainly less strict than they were at one time. However, the observation of the post you were replying to was that the older systems had less to do than modern ones. When you resize your browser window you're doing an operation that, as far as a 386 would be concerned, is non-trivial. Add to that the sheer size of the parsed webpage which generates the pretty view you see, and you've got yourself a lot of graphical things to do, and a huge datastructure in RAM. This is not the kind of problem that can be solved simply by being able to manually manage memory from assembler. This is the kind of problem that requires an intelligent design from the get-go, so that optimizations can be placed in the places where they are required as needed.
Bloatware? Probably. People who needed computers for whatever reason seemed to be getting along with them just fine without GUIs, or multiprocessing, or realtime 3D games. All of these additions is going to consume resources both when written and when used. I won't argue with you that Windows would have been better if it were based on a clean design. Clearly it would have, and on Linux we now have many desktop systems based on (if not a good deal more forethought) at least the trial-and-error process that produced the early GUIs done with a faster turnaround. Unfortunately, the users have come to rely on GUIs, pretty widgets, and browsers that resize. If they were not, perhaps we could cut down on the code quite a bit.
Also, one thing about my friend I mentioned earlier: while his code is extremely well-designed, he seems to have a fundamental lack of understanding of ideas such as UI design and concurrency. None of his programs as far as I have seen have used threading, even the GUI ones, and the few GUI programs I have seen were beyond the ugliness I expect from TK. He wrote an abstraction layer for a database that implemented foreign key constraints, and was at a bit of a loss when I first tried to explain to him that it wouldn't carry over necessarily if multiple copies of his application were running simultaneously. So we all have these problems, and we all try to get better.
If you want to see well designed and implemented code, I recommend you pick up a copy of BeOS. By sacrificing backwards compatibility, they managed to create an operating system from scratch based on object-oriented principals. It's quite amazing when you realize the things that you could do with it that you couldn't do with Windows, yet it was a tiny fraction of the size of Windows when fully installed. For example:
Now I'm going to get some sleep and try to forget about the sorry state of computing we're in right now.
--
Daniel
Most gamers would get real tired of having to shell out tons of money just to have a powerful character, and having to put up with little shits with too much money and attitude ruining it for everyone else.
It worked for Wizards of the Coast.
I've had friends that have spent upwards of $500 on Magic cards, and that probably is a small figure compared to what some people here have spent. One of them said, "If I saw a $500 game in the store, I would back away slowly and make sure not to touch it." But since it was just $8 here, $8 there, a couple expansions for $3 or $4... it added up quickly. Also, there was always a new "killer deck" that could be made by combining certain rare cards in interesting ways, at great cost.
I haven't seen their scheme (the website is down) but if their smart about it, they can make a deal no power-addicted gamer can refuse. It's happened before and it will happen again. People are always willing to pay a premium for their entertainment.
--
Daniel
I'd be willing to be this is an effect of your desire to see it. A few loud people can seem like a lot of people; as we all know it's that misperception that special interest groups thrive on.
:)
At my school, most of my friends argue for little to no restriction of either guns or information. After all, one wants to be free, and the other guarantees our freedom. I have an EFF sticker on my car, and most of my friends own guns. I might be "the liberal," but in saying that they really mean I'm the one who doesn't own a gun or a membership card to the NRA. I may not be a hardcore gun proponent, but I certainly try defend the 2nd amendment, with the little bit of 1st amendment rights I still have.
Of course, this might just be my perception too.
--
Daniel
I've been looking for someone to ask this question to, because I'm interested in setting up a RAID at some point in the near future. I'm going to need between 480 and 640 GB of actual storage in the end. Cost is an issue, but I'm willing to spend ~$1000, so I'm thinking IDE is what I can afford.
The purpose of the array is to serve high quality audio files to myself and my friends. We're talking about 30MB - 150MB sized files probably averaging around 50 MB in size, and it should be capable of streaming about four of them at a time at a minimum of 1 or 2 MB/s.
My first thought was to use RAID 5, since it traditionally is the most economical variety. But then understanding the difference between 3 and 5 seems to indicate that 3 would stream faster and deal with large files better. But, on the other hand, 5 handles multiple simultaneous reads or writes more gracefully than 3, which essentially has all of the drives doing the same thing at the same time.
So I'm wondering which RAID level I should use for this application. These aren't video editing sized files if I understand the size they can get, but they are definitely not what I would normally consider small files either. Is the performance difference between 3 and 5 truly all that great? Does it go to the toilet when multiple people are using the same drive? What problems will I notice with 3 or 5?
Since I would be sharing this via AFS, it is safe to say that these files are going to be copied around the network essentially as soon as they are opened, which means each file read will probably be reading the whole file, by the way (if you don't use AFS, not intending to sound condescending).
Thanks,
--
Daniel
Maybe it's not for Mp3s. I'm interested in converting my audio collection to FLAC format so I don't have to tolerate loss of quality. If I do that, I computed that my puny 15 GB colletion would take up about 100 GB, maybe more. And so I have room to grow, I'd want to have more like 300 or 400 GB of free space to do this.
:)
I've been planning on buying a RAID set up to accomplish this. RAID, as you all know, uses more than one disk. You all know, apparantly, that the R in RAID is for redundancy. I'm not being redundant, I hope. RAID would give me the room for this, as you all know, by using one disk.
Apparantly.
So like, I need a lot of space. And this looks cool (though it might not be obvious).
--
Daniel
I learned how to use Linux on a coworker's computer at an internship I had. I didn't learn how to install Linux until I was comfortable using it at work, and I found the installation to be pretty simple.
You'll run into a bit of a chicken-and-egg problem here if you think about it too hard. In order to learn how to install the OS they will have to deal with a bunch of things all at once: partitioning, filesystems, mountpoints, and thousands of names of programs that sound cool. Installation is pretty easy if you have a handle on these concepts, but if you don't you'll find it very exasperating--little documentation and a lot at stake. But if you make then a power user first, then they get hands-on experience with the OS, they have a reason to want to install it, and they'll be more capable at installing the OS itself on their own when it comes time to do that.
2. What distributions of Linux and BSD should they be first introduced to? (I'm only familiar with Debian, and I know virtually nil about *BSD.)
If you focus on Unix, and what the Unix philosophy is, then it won't matter much. Being able to cast around sed & awk, make command lines with for loops, and configure networking manually is a lot better than knowing how to use a graphical system configuration/installation utility like linuxconf or dselect. Plus, focusing on the tools is more likely to carry over to other operating systems. Teach them regular expressions. Teach them loops and conditionals. Don't teach them how to run RPM just because they might have to someday; it's rote, they won't remember it when they have to use it, and they'll be falling asleep as you explain it.
3. Initially, do they need to be more adept at the GUI, or do they first need to know how to use the shell?
No question: the shell. Even if you use the GUI all day (I run KDE!), lots of low-level things are written in bash/ksh/csh. If you're going to be able to sit down at any Unix computer and use it, you're going to have to be competant with at least sh/bash. The first thing I do when I sit down is fire up a console. They should be doing that automatically by the end of your class: the console is where the work is done.
There's a lot of variety with GUIs, and it can be great fun to mess around with window managers and so forth. Give them the ability to change these things (teach them how to edit their
4. Should I give away Debian CDs no-questions-asked, or should I talk with the almighty Parents so little Daniel doesn't install Linux over Dad's 'work computer.
Make the CDs available, but do talk to the parents. Make sure the parents realize that if the kid is going to do anything with "partitioning" they should go out and buy a copy of Partition Magic so they don't permanently screw everything up. That software was the best $50 I spent in high school, when I was a big OS installing weirdo. My parents never lost any data.
Clearly, some of the kids are going to want to install it at home, and it's not necessary that their parents understand the whole idea behind partitioning (though it would be nice). It's also not necessary to scare the parents into preventing the kid from doing what they want to do out of fear. I geniunely believe PM is the right answer to this problem: it's got undo until the last possible moment, and it's very user-friendly, which helps a lot when you're just getting started. Once that's taken care of there is very little that can be screwed up (most installers either ignore other drives or assume that if it's type "vfat" it shouldn't be formatted).
5. Are there any other key issue I need to think about?
Yes, definitely try and think of a curriculum. Kids who are interested can absorb tons of material, but you've got to keep it interesting. The best way to do that is to not focus on the rote doing of things, but to talk about the possibilities of things. Don't just say, here's regular expressions, instead say with regular expressions you can strip out everything in a file you don't want, get a list of all the email addresses on this webpage, etc. Keep them interested in the possibilities that you're opening up for them by teaching them the software.
Be aware of your philosophy and how it affects your decisions. If you believe in focus-follows-mouse, set up all the machines that way. Let them change it but make them try it out first: expose them to as many possibilities as possible. Have vi and emacs on the machines, and let them pick. You've probably got a whole set of quasi-religious beliefs about Linux. This is your chance to instill those beliefs in other people, but make sure that you point out your biases. Most of the kids who take your class are going to take Debian home to install, which is fine as long as you don't focus on the "Debianness of it all."
Show them the Linux that newbies don't usually get to see, because of the religious wars that go on. Regular expressions are another great example, because vi and emacs both support them. You may not be able to extend vi in lisp (and you might with vim), so don't focus on that. Keep it open.
Here's my advice on the actual software to talk about:
I suggest bash because of it's ubiquity. Everybody uses it and knows it, and it's very expressive in its weird little way. Debian or Redhat because they are also everywhere, but don't focus on rpm or apt. I suggest Blackbox or WindowMaker because they don't have a bazillion settings that you can spend all day on. If they fire up GNOME or KDE they'll be so absorbed in the coolness of it all that they'll spend the whole day configuring it and not listening to you. Galeon is simpler than Mozilla but uses the same rendering engine, and makes webpages look pretty decent. If you have them firing up Netscape 4.7 or Mozilla, they'll either be disappointed or distracted (my guess). Past these options I don't think the software matters much if they have choice.
Good luck with your project! It sounds like a step in the right direction to me.
--
Daniel
GNU-Darwin is Apple's Darwin. Or at least a binary compatible re-distribution of it. At least a fork. Frankly, their website isn't completely informative on this issue, but there seem to be three Darwins:
Frankly, I'm a little unclear on the differences but either way calling it a "shitty distro with ripped off GUI graphics" is a stretch. GNU Darwin seems to me to be a GNU operating system built on an Apple-modified BSD kernel. Which sounds kind of perverted, but not necessarily "shit." Hey, they've ported it to x86! It's got to be at least important to x86 as NetBSD.
Apple's lawyers are going to have a field day with this one.
The source is open. Read all about it at Apple's Darwin page. There's nothing to sue anyone over, although Apple can via their license simply "revoke" the source and keep all of the outside changes.
Actually, according to the license, when you take any source covered by the APSL, you're required to register with Apple. If the developers didn't do that, Apple would have a valid case to sue them over. If they did (and I'm positive they did, since they link to the damn license off their page), then Apple really doesn't have anything to get them on, unless they're keeping changes private. If they were doing that it wouldn't be GNU either.
I think your reaction is a little uninformed. A simple websearch turned up quite a bit of information on this topic, even a nice rant from the FSF about the APSL.
--
Daniel
I completely agree. In fact, that's why Python is my language of choice--I get more done faster, and unless I write a shitty algorithm I really don't notice a speed problem.
.Net. After all, we've been admitting for years that scripting languages are OK unless you need the extra performance, and these guys do need that extra drop. So why aren't they writing these programs in C or C++? :) Clearly it isn't the libraries or types involved, since these are almost always available to C or C++ also.
However, when it comes to this particular benchmark, it has more to do with performance on a server somewhere handling thousands of simultaneous connections. Web applications, if you will. And in this particular case, a performance difference of 10% could mean a hundred users get locked out, and responsiveness to the others will be very bad.
If you ask me, it's one of the great mysteries of the computer industry why desktop apps are written in C/C++ and web apps are written in Java, Perl, or
That's the $15 million dollar question, now isn't it? We seem to optimizing for RAD and performance, rather than just performance. Perhaps we should rethink our priorities, move scripting languages onto the desktop, and compiled languages onto the server...
--
Daniel
Well, they're about slashdotted now. They lost my last request, and it says they have almost 2000 anonymous users. I sometimes think the reason I like reading Slashdot isn't because of the great links and articles, but instead because I like being a part of the goddamned Slashdot effect. :)
Which brings me to the point. Ya know, about the only site that can handle the Slashdot effect is Slashdot. So maybe Taco should write an article like this (or maybe he has?). The Slashdot guys know what they're doing, we should pay attention. Although I find it interesting that when slashdot does "go down," the only way I know is because for some reason it's telling me I have to log in (which is a lot nicer than Squid telling me the server's gone).
--
Daniel
No one asked you what was noble and what wasn't. If you know the slightest thing about programming you'd be aware that you don't start writing something beat what's already been made, you start by writing something, and then you add to it, and as your knowledge increases you can take on more and more complex or difficult tasks. At this moment, there is no open source 3D modeller/renderer/kitchen sink which these guys can go to and learn how to do these complex things. The algorithms used by these complicated applications are all but unknown to the community; the proprietary ones certainly aren't published in books, and even if they were it would take a great deal of skill to "get inside" them, figure out how they work, etc.
We have to start at the beginning.
Did you use the Gimp three years ago? I tell you it wasn't a "photoshop-alike" then. We have to start somewhere and then move up. There are programs we have in free software that are so advanced, there's no analog for them in closed-source. For example, the RADIANCE renderer is the only one that does light accurately, and it has been free for ages and will become open source in December. Aspell uses a new algorithm that beats every other spell checker. At its inception, would you have said we shouldn't invest this time and effort into another spell checker?
The problem is that people like you look at open source development like closed source development. As though when a project is formed, some segment of the total number of developers have to be allocated to it. Fortunately, the way it really works is developers work on whatever they are attracted to. Even huge projects like Mozilla have at the core less than 20 developers. I haven't checked, but I would guess that these 3D projects will have 2-4 core developers. Everyone else will contribute from time to time, or possibly even just once. But the people who work on Moonlight are not people who necessarily would have worked on Blender. They are not even necessarily people who would have developed for anything at all.
What does a professional setting have to do with whether or not something should be developed? Take enlightenment for example. I bet many people said, this is a window manager which will never be used in a professional setting. And yet I know several sysadmins who have used enlightenment as their window manager! It's like science, we don't research the things we think are going to bring about "useful" discoveries, because that's counter-productive and we never know which research will result in useful discoveries. Instead, we just generate all of the software we'd like to generate, and some of it will get used, and some will not. AWK is a good counter example; it was developed for a particular purpose for which it is extremely useful (parsing text files with very uniform structure). And yet, it has fallen by the wayside because Perl can do essentially the same things, but is a more powerful programming language in general.
To summarize:
--
Daniel
While I was on a road trip over the summer we stopped in Oklahoma somewhere for gas. The ATM was "down," but I went over and played with it for a few minutes, found out that it was in fact running OS/2.
It's freaking weird to see the OS/2 window decoration on an error message in amber on an ATM though! Especially when you know it's the kind of ATM that normally looks like a telnet connection (characters only).
--
Daniel
How do you take a screenshot of a server?
BTW: if you read about the server components, you should have noticed that most of these things exist already (postfix, cyrus, apache, inetd, proftpd, openldap2). Kolab is just going to tie them together and slap a pretty administative UI on top.
--
Daniel