The problem with this is the claim by VMWare that the way they do it (virtualizing a PC when the hardware doesnt' actually support it) is patented or at least patent pending, so there may simply not be a way to do this open source, at least not for the majority of the world that the patents would cover (presumably they're covering the EU, US, etc...).
This kind of system, in response to come people's comments about the license for Bosch, isn't nearly as complex. It'd take a lot less code to get a usable product (which Bosch still isn't...)
I think the best that could come from this effort is a rethinking by VMWare on their pricing, providing a more limited use version or non-commercial pricing for a more reasonable cost. The problem is, they've said they've been considering this, and maybe this effort is going to just stop them from really considering it.
I hope the opensource community isn't shooting itself in the foot on this one. I'd hate to see this effort anger VMWare into being more restrictive in the future with their product, all for the effort to create a product that may violate their patents anyway.
That's so funny, I mixed the two movies up in my head. I just assumed The Matrix was the movie I'd seen trailers for last year, but that's right, that was The 13th Floor.
Wierd.
I'm hating the thought of seeing The Matrix, because I've got a sinking feeling its plot may be very close to a book I've been writing... How much is that gonna suck?
Noooo.... I understood it to be like that at first, but what ICANN is doing is providing a framework that allows companies who become registrars to register domains in any of the root levels. NSI *isn't* going to be the only company that can register.com domains. Any company essentially with write access into ICANNs databases can register them. That's why NSI is pushing the "dot com people" thing, because they want the public to associate them with ".com" since there will soon be no reason people need to use them to register domains.
So my point is still valid. If NSI won't allow their records into the ICANN version of the whois database by saying they are their proprietary records, then ICANN wouldn't be able to validate requests for registrations or changes from other companies against what NSI says they register, and NSI might lose out.
I might be wrong about how that end of things works, but the new registration system isn't a simple case of NSI providing the.com zone file to ICANN. Otherwise people could register whatever.com on the same day on two different registries and who gets precidence?
Without a full disclosure of information, there'd be no way to track back to a domain. You'd have to run the equivalent of a whois against each of the half-dozen or more registries, without knowing which registry actually holds the record for the domain you're interested in. (Which you know now, since all domains in a given TLD come from a single database, NSI or otherwise...)
Here's a thought. Let NSI do whatever they damn well please. When ICANN takes over the root nameservers, simply give the other registration services that are freely providing their databases priority in the root servers. Ie, Microsoft registers microsoft.com with NSI, they better hope (and pressure) NSI to keep all the whois information completely and 100% free, or someone else will register microsoft.com with another registrar who is freely sharing information. (Linus website anyone?);)
In all seriousness though, if they want to claim the information is propriatary, then ICANN can't exactly integrate the information into the root servers, can they?
I don't know... listening experience... how can you describe it?
I don't expect much from the headphones. I can't hear the inaccuracies in the recording because of the inaccuracies in the headphones. Make any sense?
I'd rather hear lousy recordings on lousy equipment than lousy recordings on good equipment. Lousy equipment tents to muffle out and lose the detail in its own noise and inaccuracies. I'd rather not hear something because of lousy equipment than hear something that doesn't sound "right".
I can't get better sound anywhere than I can get in my car. Does that mean I don't listen to anything anywhere else? No, it means there some stuff I really don't like listening to in my car because of the low quality, and some stuff I don't like listening to outside of my car because it loses its impact. Most good classical falls into that latter category. Most '80s music falls in the former.:)
I'm no expert on human auditory perception, but personally I've noticed I can tolerate high volumes, for example, when the sound itself is very clean, and not distorted. A similarly high volume coming from a distorted or dirty source causes me headaches and makes my ears hurt.
That may have something to do with the whole headphone issue -- I don't listen to headphones as loud as I'd tend to listen to a stereo.
*shrug* I didn't say it was logical, I just said it was my observation.:)
FYI, that three to four dollars for a cd-club CD, not a dime goes to the artist. Part of the contracts they have with record labels is that CD's sold in certain circumstances don't pay royalties. The reason the CD's are so cheap is the CD-clubs pay a flat licensing fee to the label in exchange for access to its catalogs. They get masters -- and produce the copies themselves. That's why the printing is often inferior, and they usually say things like "Manufactured for BMG" etc on the case.
You can tell the artists with real clout in the industry -- they're the ones that you can't get in any CD catalog. They control their royalties, they decide how their art gets marketed and sold, thus they don't get bundled into flat-rate licensing from the labels.
A CD, with jewel case and printed material costs around a buck. The artist usually gets in the order of a dollar per sold, although that varies quite a bit based on contract.
They're sold to distributors in the 5-6 dollar range. Distributors (like WEA) resell them in the 9-10 dollar range. Stores mark them up from there.
The price from the distributors is why you don't see any places coming out and selling on razor thin margins and selling big-name CD's for eight bucks. (Witness prices at places like CD-Now and Walmart...)
So if you're spending $18 on the CD, the store itself is making 80% on that sale. In mall situations, they have to if they want to cover outrageous mall lease rates. If you think 80% is bad, you should see the markup on the clothing stores...
Big stores and indie stores in lower-rent locations are the ones that you'll find the $12 CD's at. But they simply don't get lower than that because of the cost from the manufacturer.
I try to support the artists I actually think produce decent work -- their CD's come from stores. I have no problem buying CD's from hack bands that happen to have a song I find catchy from clubs... they're probably getting more $$$ than their music is worth anyway;)
I think you're probably right. The money I spent on the car stereo was excessive, to say the least. But I was young and had it.
The problem is that a lot of formats today (CD, DVD Audio, DAT) have better sound quality than the vast majority of equipment can reproduce. Its typically not, however, better than the studio equipment on newer recordings, so the quality capable in the recording is actually there in the recoding. As a result, you miss parts of the music.
So you spend more money to get better stereo equipment to handle the available quality in the format you choose. The problem you get from that comes when you listen to a source thats not as high a quality. MP3's are one example, but MD, and even old CD's copied from lower quality studio masters are all like that.
Its tough on good equipment to, for example, listen to CD's of stuff from the 80's. Interestingly I've found CD's of older material to sound better. During the 80's so much was being played on tape I guess they just pushed some parts of the recoding harder than they should have. They're all sort of piercing in the high frequencies. A Bangles CD is painful to listen to on real good equipment, but a Pink Floyd might sound fantastic. (Especially I've found the quality on the gold remasters to frequently be excellent)
My reason FWIW for spending all that $$$ on the equipment in my car was that the car was a convertible, and it takes a lot of power to get clear sound with the top down. (Most car stereo equipment takes advantage of the fact that you're essentially in a closed box to produce its sound -- without a top, you need to produce a higher volume, more power, etc...)
Its not a quality of hardware issue as much as an artifact of compression issue. Soundblaster 16's are very noisy. Some cards are better, and a lot are worse. Anything pushing the line out through a 1/8 in jack is going to sound lousy. If you spend some serious $$$ on a sound card you can get multichannel professional audio card with RCA outs, better S/N ratio, etc. Even on one of these, you'll get clearer sound but not more accurate sound.
My test I did with MP3's used 128kbit and 256kbit rips. I use a program under Linux (can't recall which) to generate 44khz 16 bit stereo WAV files. Those wav files were burned out to a CD, so I could test the quality of the format, not my computer's audio hardware. (Which is pretty good anyway, I do a lot of music production on it -- although I *never* record off the output, its always mixed down to a wav file and burned to a CD...)
I listened to the CD in three or four places. On my discman with the normal headphones it sounded fine. I didn't lose much quality compared to the original tracks.
With good ($100) headphones, it just struck me as being very... tinny? I'm not sure a good word for it. It was hollow sounding, but mostly in the high frequency ranges... it wasn't bad at vocal ranges, but electronic music (I had a Front Line Assembly track on the disk) sounded awful. Lost a *lot* of detail.
I also tried it on my home stereo. Not a great system, a Sony dolby digital setup, Bose 4001 speakers, custom built center channel (which is a lot higher quality than the Bose's -- I just haven't been able to afford to build left/rights yet...) The source was coming via a digital stream from my DVD player into the AC3 decoder, so it was digital straight to the amp. Sounded fine on here. (The lousy high and low end response on Bose speakers I'd attribute to that...)
The last place I tested it was in my car. My car's got about $5k of audio hardware in it. The speakers are custom enclosed Vifa components (some of the highest quality stuff you can get, car or home...) Very expensive wiring, 1000 watt amp, blah blah blah...
Results? Unlistenable. The Vifas are so accurate (especially with that much power able to take hold of them), every little inaccuracy of the source recording is audible. Everything sounded very hollow, there was some wierd side-to-side inaccuracies that I didn't recall noticing in the source material. I could go on...
That's just my results. A MP3 component plugged into most stereos will probably sound great. I'm not sure why I'd want to spend the $$$ for an MP3 component though, when I can get a 200 disc changer for $300, and not deal with the hassle of burning MP3 CDRs, and still being more limited in what I can listen to.
The idea of a portable player like the Rio is good, especially when I'm at the gym or biking or something. Its better sounding than a tape, and CDs are too prone to skipping.
I doubt MP3 has anything to do with declines in CD sales. I'd guess its just the general public getting bored with popular music these days. I mean seriously, among people on here who socialize with non-geeks, how many people do you know who have ever even heard of MP3, much less use it? Very few if any.
This is just a B.S. report from the RIAA to bolster their claims that MP3's are being used to pirate music and its severely hurting the industry. There's doubtfully any truth to it at all. Most people that have a few MP3's probably have them of something they wouldn't have payed any money for otherwise.
Its also silly given the way that CD clubs license music from record labels -- CD clubs hurt profits and particularly artist royalties more than MP3 ever will.
For the same reason the idea of TV tuners in PCs never really took off, MP3s playing through PCs for mainstream users will never take off. And I don't know a single person with a stereo component that plays them. No one wants to fire up their PC and sit in that room to listen to music or watch TV. I know a dozen people who've bought Toshiba or other "name" PC's with TV tuners, radios, etc... and none of them ever use them. These are virtually computer illiterate users. (ie, most of the general public)
On top of that, you've got the audiophiles -- people who tend to spend a lot of $$$ on audio-related hardware. No true audiophile would want to listen to anything on MP3. On anything but lousy headphones or low end computer speakers or stereo equipment they just sound lousy. (And that comment isn't flame bait -- people may disagree, but most people don't have even reasonable quality audio equipment...)
In a nutshell, I think its a growing lack of innovative and creative music in the mainstream popular music area thats responsible for the drop in sales. The big companies (and the companies that are RIAA members) aren't the ones selling the new and interesting music... The RIAA is an association of the big boys, and the big boys are growing old and tired.
Its good to see a drop in CD sales. Maybe one day the big labels will start to take risks again and promote quality music rather than generic clone bands. If the RIAA wants to twist that to their political agenda against MP3, who can stop them?
Yes, and the low memory boot disk copies itself to the harddrive, and boots from there, hense getting it to work when installing from a local harddrive.:)
Going back to the topic of this whole posting, maybe that's a good machine to try out Slackware 4.0 on.:)
I don't have any experience with that model, but Debian is probably easier to get installed, it seems to give a lot more flexibility in the install process.
Another option you might have is to boot the system off something else -- if the kernel loads (I assume it does, since that's a BIOS thing, not a Linux thing), then you could always boot the kernel off the floppy and provide some other device to boot from. I assume the system doesn't have a CD-Rom, but if you can boot to DOS on it, you could install the packages you need on a DOS/umsdos partition and try booting it from loadlin.
I'm very impressed with the way the LinuxPPC demo thats in MacWorld works -- it contains all the filesystems and stuff in a single MacOS file, and the system can boot and run from that file itself. That strikes me as a good installation method that is not supported by most (all?) distributions... Put the install system (3-5 meg probably) into a file, and boot linux using loadlin and mounting that file as root. It'd run fine obviously off a CD, and in cases of wierd installs you could in a worst case scenario, split the root file into 1.44 meg chunks, copy them into a small DOS partition on the harddrive, cat them back together and boot from the harddrive into it.
But the idea that RedHat can't install onto laptops is just silly. I'm sure there are thousands of/. readers who've done it and are doing it. I've got it on three, one of them that doesn't even have a floppy drive... (yes, it did have one the *first* time I installed Linux on it three years ago, but I've since had to remove it because it broke...)
Its not that RedHat is bad, IMHO, its just that RedHat strangely requires you to run through their partitioning/mounting system before you can mount the swap. A significant upgrade where memory is concerned with RedHat is if they'd give the user in "Advanced" mode the ability to scan the drives on the local system for swap partitions and mount them BEFORE anything else happens during the install.
The RedHat installer seems to load itself into a ramdisk, and then load any additional modules into a second ramdisk mounted into the filesystem. That plus the kernel tends to eat 5-6 meg of RAM, and with the libraries and such their Disk Druid program and fdisk both can't load, which are prerequisites for mounting the swap space.
As an alternative, documenting how it can be done from the command line in the shells thats opened would help. As I said in another post, I wasn't able to find what program it was using, since the usual utilities don't seem to be there.
I have 176 meg of ram in my system at work, and those applications do tend to swell in size (particularly Netscape). I actually try to close Netscape once every hour or two (figuring it doesn't crash on its own) to keep memory bloat down. I don't know if its a "feature" or just a bad memory leak, but it is bad.
Now, where 8 meg is concerned, 8 meg most certainly isn't a joke. Back in the days of the pre-1.0 kernel, I ran a rather useful system for several years in 8 meg of RAM. When I was in school, virtually all of my papers were written on that with vi and nroff/groff. It handled e-mail serving and reading, usenet reading and posting, and I was running a mailing list getting almost 30 postings a day to 200 people on it. I also gave user accounts to friends who needed a machine to work on from terminals around the campus.
Never had a problem with it, at one point had nearly 9 months uptime on it.
I currently have three systems here with only 8 meg RAM. One's being used as a development platform for embedded linux POS applications. One handles my internet dialup, masquerading, routing, firewalling, fax sending and receiving, and voicemail. The third is the system I mentioned above, which is an old notebook computer. Slow, low in RAM, small screen, but the battery lasts almost five hours, and its great for writing when I want to be outside. The router box used to handle printing too, but ghostscript eats too much RAM, so I moved it off to the system I've been experimenting with Oracle on.
All three of those applications don't need more than eight meg of RAM. They're providing important functions for me without costing me any excessive amount of money, using old parts I've scraped up.
Its a lot of functionality in not a lot of hardware. If you want to know why the ability to run in 8 meg is important, that's the exact reason. There's lots of very inexpensive hardware that people can buy or have laying around that can be made useful as print servers, or any of a dozen other functions, and 8 meg is enough for many of them.
Unfortunately that doesn't help with RedHat. RedHat doesn't give you the ability to mount a swap partition until after you have partitioned the drive -- regardless whether or not the drive is already partitioned. The real problem in the RedHat install with 8 meg seems to be the amount of memory it takes to fork Disk Druid or even fdisk. The very next step is mounting the swap space, but in one attempt I made to install it onto a 486/80 (AMD) with 8 meg ram, I let it run swapping for two hours just to see if it'd ever actually manage to get fdisk or Disk Druid to load up. It never happened.
Debian is better -- it lets you mount a swap parition if you've already created one. In the previously mentioned install, I was able to boot off a RedHat install disk, load no drivers for anything, and just squeak by after killing the installer to be able to load fdisk and partition the drive. My hope had been to mount the swap before getting to the partitioning step, but I was unable to find the program that's actually mounting the swap space. I'm assuming RedHat isn't using swapon / mkswap for it, but I might've just missed it.
I ended up ordering Debian CD's and got it installed on that system with no troubles. I remember I had to install the barest minimum I could, then reboot to get it to handle the full install. I never did figure out why that was the case.
Part of the autoLinux stuff I've been working on is getting a good mid-size distribution together. (Bigger than the various router projects out there, but smaller than what I remember the Slackware "A" series being...)
Anyone know what the requirements for a system are going to be for Slackware 4? I've been disappointed lately to see they keep growing with Redhat (in particular) and to some extent with Debian.
I know Linux on a 4 meg system may not be reasonable any more, but it seems 8 should be doable, but 8's too small to install RedHat, and I 've only gotten Debian to install in 8 meg successfully when it was installing from a local harddrive.
Is the minimum (A) set still fairly small? It'd be nice to have a standard distribution that can get a core system installed in 15-20 meg.
I can't speak to its stability compared to Solaris's version -- never had a speck of trouble with either, but I've never pushed either very hard.
I can tell you, on a 586/133 (AMD) with 64 meg of ram, and a five year old slow-as-molassis IDE harddrive, it takes a good four hours to install. Most of it the process of setting up the initial database, AFAIK, and not needing any interaction. But it did work.
A nice spankin' new Pentium II or III might work better.;)
Huh? There is a core for Mozilla, what do you think new-layout was? I'm not comparing M3 to Navigator 4.5... I'm comparing M3 to builds from a few weeks ago. It certainly is backtracking if something that's been working for weeks or months isn't after a codefreeze.
Given the constraints of HTML tables (setting borders to 0, cellspacing to 0 and cellpadding to 0), the interior size of the table should match the exterior size.
It seems whats changed in Mozilla, after some other testing, is settings on tables are no longer inhereted from tables they might be within. Up until now (and this might be a bug in Mozilla, I'm not sure...) nested tables (or for that matter any nested tags) took on as defaults the setting from the element they're nested within. Setting "cellspacing=0" on a table passed that setting onto tables it contains. Same with font tags, or other tags.
Only some settings on elements seems to be inhereted... border, for example, seems to be inhereted, but cellspacing and cellpadding don't. Or at least don't consistantly get inherited.
Adding them manually to the table fixes the problem. I looked that the HTML spec, and its not really clear IMHO on this issue.
Maybe I'll submit this as a bug at the risk of looking dumb when its not.:)
Sorry, this seems to be a step back. Sure, the browsing window has some (ugly, nonfunctioning) UI elements wrapped around it, but the core of the browser -- the renderer -- seems to have taken a significant step back. The nightly build from mid-february I was running layed out most pages I looked at fairly well, 90% of the time they were correct as far as I can see. Now half the sites that worked before no longer lay out properly.
Some things I've noticed that are broken now:
Tables inside of tables where the width of the inside table is the same as the width of the cell its in -- no longer fit. The inside table for whatever reason starts to wrap cells. Very strange...
Frames. (I can't honestly remember if the other one had frame support at all, but I get a lot of frame content showing up in the wrong frame...)
Although it *says* its got editor support built into it, the editor doesn't work at all. That's a step back. Opening editor pages just opens up new browser windows (which cause it to blow core when you close them...)
Javascript -- working just enough not to work.:) Unfortunately it blows core when I try to open the console window, so I can't see exactly whats not working...
The widgets... I thought this was GTK? What version of GTK is it using? Themes don't work -- and the widgets themselves don't look like any of the GTK ones I've got installed. Where's it getting them from? Either way, I get toolbars showing up in the scrollbars, although once a page is loaded they usually go away.
Resizing the window doesn't work...
Forms don't seem to work, whereas they did before.
I couldn't get any background images to display. Some JPG's were not displaying either, and the ones that don't display seem to badly mess up table rendering. (IE, a 200 pixel wide image aligned left in a table cell with text around it will push the table 200 pixels wider...)
Of course, most of the people getting all worked up about the release think "M3" is some official release or something, its just a development milestone. No promise or guarantee it'd be better than what was there before, its just a point to freeze the code and take stock on whats done and what needs to be done.
I'm just suprised that the core of it seems to have backtracked so far...
Ever since freshmeat changed their design, the homepage causes an instant browser crash on two of my PC's, this one (Netscape 4.5, RedHat 5.2) and mine at work (Netscape 4.5, mutilated RedHat 5.0...)
At least Mozilla's getting better. One of these days I might get to look at freshmeat again.:)
There are lots of these things
on
Mini Board PC
·
· Score: 1
Hard to come by? What makes you think that? Twenty seconds on yahoo and I can find dozens of them...
Perfect for webcams... if cheaper
on
Mini Board PC
·
· Score: 1
Do some poking around the net. I can't remember the company, but there's a place that sells webcams that simply plug right into the ethernet. They're essentially a SBC PC, camera, and embedded OS (Linux maybe?). I seem to remember they were $700 or something like that.
There are probably a hundred companies that make these, and most are a *LOT* cheaper than this. I've seen systems that size at that power for $400, not $800.
Just do a search in yahoo for embedded PC. You'll find lots of them. Or you can look at some of the links on my autoLinux site at http://www.bangsplat.org/autolinux.
Not once... not a single time, EVER, has an application I've been using under Linux caused the OS to crash where I had to reboot.
Netscape crashes a lot itself... but its lousy software, not a lousy OS. Once in a while that spawning window problem happens and X might hang -- it doesn't crash, it just gets hung up. Telnetting into my box from another machine and killing Netscape always fixes the problem.
Slashdot Mirror
The problem with this is the claim by VMWare that the way they do it (virtualizing a PC when the hardware doesnt' actually support it) is patented or at least patent pending, so there may simply not be a way to do this open source, at least not for the majority of the world that the patents would cover (presumably they're covering the EU, US, etc...).
This kind of system, in response to come people's comments about the license for Bosch, isn't nearly as complex. It'd take a lot less code to get a usable product (which Bosch still isn't...)
I think the best that could come from this effort is a rethinking by VMWare on their pricing, providing a more limited use version or non-commercial pricing for a more reasonable cost. The problem is, they've said they've been considering this, and maybe this effort is going to just stop them from really considering it.
I hope the opensource community isn't shooting itself in the foot on this one. I'd hate to see this effort anger VMWare into being more restrictive in the future with their product, all for the effort to create a product that may violate their patents anyway.
Okay, bad me. Read the spoiler comment down below. Its nothin' like my book. Yay! Now I can go see it without having that to worry about. :)
That's so funny, I mixed the two movies up in my head. I just assumed The Matrix was the movie I'd seen trailers for last year, but that's right, that was The 13th Floor.
Wierd.
I'm hating the thought of seeing The Matrix, because I've got a sinking feeling its plot may be very close to a book I've been writing... How much is that gonna suck?
Noooo.... I understood it to be like that at first, but what ICANN is doing is providing a framework that allows companies who become registrars to register domains in any of the root levels. NSI *isn't* going to be the only company that can register .com domains. Any company essentially with write access into ICANNs databases can register them. That's why NSI is pushing the "dot com people" thing, because they want the public to associate them with ".com" since there will soon be no reason people need to use them to register domains.
.com zone file to ICANN. Otherwise people could register whatever.com on the same day on two different registries and who gets precidence?
So my point is still valid. If NSI won't allow their records into the ICANN version of the whois database by saying they are their proprietary records, then ICANN wouldn't be able to validate requests for registrations or changes from other companies against what NSI says they register, and NSI might lose out.
I might be wrong about how that end of things works, but the new registration system isn't a simple case of NSI providing the
Without a full disclosure of information, there'd be no way to track back to a domain. You'd have to run the equivalent of a whois against each of the half-dozen or more registries, without knowing which registry actually holds the record for the domain you're interested in. (Which you know now, since all domains in a given TLD come from a single database, NSI or otherwise...)
Errrr.... Linux website I meant... Guess I should've clicked Preview. ;)
Here's a thought. Let NSI do whatever they damn well please. When ICANN takes over the root nameservers, simply give the other registration services that are freely providing their databases priority in the root servers. Ie, Microsoft registers microsoft.com with NSI, they better hope (and pressure) NSI to keep all the whois information completely and 100% free, or someone else will register microsoft.com with another registrar who is freely sharing information. (Linus website anyone?) ;)
In all seriousness though, if they want to claim the information is propriatary, then ICANN can't exactly integrate the information into the root servers, can they?
I don't know... listening experience... how can you describe it?
:)
:)
I don't expect much from the headphones. I can't hear the inaccuracies in the recording because of the inaccuracies in the headphones. Make any sense?
I'd rather hear lousy recordings on lousy equipment than lousy recordings on good equipment. Lousy equipment tents to muffle out and lose the detail in its own noise and inaccuracies. I'd rather not hear something because of lousy equipment than hear something that doesn't sound "right".
I can't get better sound anywhere than I can get in my car. Does that mean I don't listen to anything anywhere else? No, it means there some stuff I really don't like listening to in my car because of the low quality, and some stuff I don't like listening to outside of my car because it loses its impact. Most good classical falls into that latter category. Most '80s music falls in the former.
I'm no expert on human auditory perception, but personally I've noticed I can tolerate high volumes, for example, when the sound itself is very clean, and not distorted. A similarly high volume coming from a distorted or dirty source causes me headaches and makes my ears hurt.
That may have something to do with the whole headphone issue -- I don't listen to headphones as loud as I'd tend to listen to a stereo.
*shrug* I didn't say it was logical, I just said it was my observation.
FYI, that three to four dollars for a cd-club CD, not a dime goes to the artist. Part of the contracts they have with record labels is that CD's sold in certain circumstances don't pay royalties. The reason the CD's are so cheap is the CD-clubs pay a flat licensing fee to the label in exchange for access to its catalogs. They get masters -- and produce the copies themselves. That's why the printing is often inferior, and they usually say things like "Manufactured for BMG" etc on the case.
;)
You can tell the artists with real clout in the industry -- they're the ones that you can't get in any CD catalog. They control their royalties, they decide how their art gets marketed and sold, thus they don't get bundled into flat-rate licensing from the labels.
A CD, with jewel case and printed material costs around a buck. The artist usually gets in the order of a dollar per sold, although that varies quite a bit based on contract.
They're sold to distributors in the 5-6 dollar range. Distributors (like WEA) resell them in the 9-10 dollar range. Stores mark them up from there.
The price from the distributors is why you don't see any places coming out and selling on razor thin margins and selling big-name CD's for eight bucks. (Witness prices at places like CD-Now and Walmart...)
So if you're spending $18 on the CD, the store itself is making 80% on that sale. In mall situations, they have to if they want to cover outrageous mall lease rates. If you think 80% is bad, you should see the markup on the clothing stores...
Big stores and indie stores in lower-rent locations are the ones that you'll find the $12 CD's at. But they simply don't get lower than that because of the cost from the manufacturer.
I try to support the artists I actually think produce decent work -- their CD's come from stores. I have no problem buying CD's from hack bands that happen to have a song I find catchy from clubs... they're probably getting more $$$ than their music is worth anyway
I think you're probably right. The money I spent on the car stereo was excessive, to say the least. But I was young and had it.
The problem is that a lot of formats today (CD, DVD Audio, DAT) have better sound quality than the vast majority of equipment can reproduce. Its typically not, however, better than the studio equipment on newer recordings, so the quality capable in the recording is actually there in the recoding. As a result, you miss parts of the music.
So you spend more money to get better stereo equipment to handle the available quality in the format you choose. The problem you get from that comes when you listen to a source thats not as high a quality. MP3's are one example, but MD, and even old CD's copied from lower quality studio masters are all like that.
Its tough on good equipment to, for example, listen to CD's of stuff from the 80's. Interestingly I've found CD's of older material to sound better. During the 80's so much was being played on tape I guess they just pushed some parts of the recoding harder than they should have. They're all sort of piercing in the high frequencies. A Bangles CD is painful to listen to on real good equipment, but a Pink Floyd might sound fantastic. (Especially I've found the quality on the gold remasters to frequently be excellent)
My reason FWIW for spending all that $$$ on the equipment in my car was that the car was a convertible, and it takes a lot of power to get clear sound with the top down. (Most car stereo equipment takes advantage of the fact that you're essentially in a closed box to produce its sound -- without a top, you need to produce a higher volume, more power, etc...)
Its not a quality of hardware issue as much as an artifact of compression issue. Soundblaster 16's are very noisy. Some cards are better, and a lot are worse. Anything pushing the line out through a 1/8 in jack is going to sound lousy. If you spend some serious $$$ on a sound card you can get multichannel professional audio card with RCA outs, better S/N ratio, etc. Even on one of these, you'll get clearer sound but not more accurate sound.
My test I did with MP3's used 128kbit and 256kbit rips. I use a program under Linux (can't recall which) to generate 44khz 16 bit stereo WAV files. Those wav files were burned out to a CD, so I could test the quality of the format, not my computer's audio hardware. (Which is pretty good anyway, I do a lot of music production on it -- although I *never* record off the output, its always mixed down to a wav file and burned to a CD...)
I listened to the CD in three or four places. On my discman with the normal headphones it sounded fine. I didn't lose much quality compared to the original tracks.
With good ($100) headphones, it just struck me as being very... tinny? I'm not sure a good word for it. It was hollow sounding, but mostly in the high frequency ranges... it wasn't bad at vocal ranges, but electronic music (I had a Front Line Assembly track on the disk) sounded awful. Lost a *lot* of detail.
I also tried it on my home stereo. Not a great system, a Sony dolby digital setup, Bose 4001 speakers, custom built center channel (which is a lot higher quality than the Bose's -- I just haven't been able to afford to build left/rights yet...) The source was coming via a digital stream from my DVD player into the AC3 decoder, so it was digital straight to the amp. Sounded fine on here. (The lousy high and low end response on Bose speakers I'd attribute to that...)
The last place I tested it was in my car. My car's got about $5k of audio hardware in it. The speakers are custom enclosed Vifa components (some of the highest quality stuff you can get, car or home...) Very expensive wiring, 1000 watt amp, blah blah blah...
Results? Unlistenable. The Vifas are so accurate (especially with that much power able to take hold of them), every little inaccuracy of the source recording is audible. Everything sounded very hollow, there was some wierd side-to-side inaccuracies that I didn't recall noticing in the source material. I could go on...
That's just my results. A MP3 component plugged into most stereos will probably sound great. I'm not sure why I'd want to spend the $$$ for an MP3 component though, when I can get a 200 disc changer for $300, and not deal with the hassle of burning MP3 CDRs, and still being more limited in what I can listen to.
The idea of a portable player like the Rio is good, especially when I'm at the gym or biking or something. Its better sounding than a tape, and CDs are too prone to skipping.
I doubt MP3 has anything to do with declines in CD sales. I'd guess its just the general public getting bored with popular music these days. I mean seriously, among people on here who socialize with non-geeks, how many people do you know who have ever even heard of MP3, much less use it? Very few if any.
This is just a B.S. report from the RIAA to bolster their claims that MP3's are being used to pirate music and its severely hurting the industry. There's doubtfully any truth to it at all. Most people that have a few MP3's probably have them of something they wouldn't have payed any money for otherwise.
Its also silly given the way that CD clubs license music from record labels -- CD clubs hurt profits and particularly artist royalties more than MP3 ever will.
For the same reason the idea of TV tuners in PCs never really took off, MP3s playing through PCs for mainstream users will never take off. And I don't know a single person with a stereo component that plays them. No one wants to fire up their PC and sit in that room to listen to music or watch TV. I know a dozen people who've bought Toshiba or other "name" PC's with TV tuners, radios, etc... and none of them ever use them. These are virtually computer illiterate users. (ie, most of the general public)
On top of that, you've got the audiophiles -- people who tend to spend a lot of $$$ on audio-related hardware. No true audiophile would want to listen to anything on MP3. On anything but lousy headphones or low end computer speakers or stereo equipment they just sound lousy. (And that comment isn't flame bait -- people may disagree, but most people don't have even reasonable quality audio equipment...)
In a nutshell, I think its a growing lack of innovative and creative music in the mainstream popular music area thats responsible for the drop in sales. The big companies (and the companies that are RIAA members) aren't the ones selling the new and interesting music... The RIAA is an association of the big boys, and the big boys are growing old and tired.
Its good to see a drop in CD sales. Maybe one day the big labels will start to take risks again and promote quality music rather than generic clone bands. If the RIAA wants to twist that to their political agenda against MP3, who can stop them?
Yes, and the low memory boot disk copies itself to the harddrive, and boots from there, hense getting it to work when installing from a local harddrive. :)
Going back to the topic of this whole posting, maybe that's a good machine to try out Slackware 4.0 on. :)
/. readers who've done it and are doing it. I've got it on three, one of them that doesn't even have a floppy drive... (yes, it did have one the *first* time I installed Linux on it three years ago, but I've since had to remove it because it broke...)
I don't have any experience with that model, but Debian is probably easier to get installed, it seems to give a lot more flexibility in the install process.
Another option you might have is to boot the system off something else -- if the kernel loads (I assume it does, since that's a BIOS thing, not a Linux thing), then you could always boot the kernel off the floppy and provide some other device to boot from. I assume the system doesn't have a CD-Rom, but if you can boot to DOS on it, you could install the packages you need on a DOS/umsdos partition and try booting it from loadlin.
I'm very impressed with the way the LinuxPPC demo thats in MacWorld works -- it contains all the filesystems and stuff in a single MacOS file, and the system can boot and run from that file itself. That strikes me as a good installation method that is not supported by most (all?) distributions... Put the install system (3-5 meg probably) into a file, and boot linux using loadlin and mounting that file as root. It'd run fine obviously off a CD, and in cases of wierd installs you could in a worst case scenario, split the root file into 1.44 meg chunks, copy them into a small DOS partition on the harddrive, cat them back together and boot from the harddrive into it.
But the idea that RedHat can't install onto laptops is just silly. I'm sure there are thousands of
Its not that RedHat is bad, IMHO, its just that RedHat strangely requires you to run through their partitioning/mounting system before you can mount the swap. A significant upgrade where memory is concerned with RedHat is if they'd give the user in "Advanced" mode the ability to scan the drives on the local system for swap partitions and mount them BEFORE anything else happens during the install.
The RedHat installer seems to load itself into a ramdisk, and then load any additional modules into a second ramdisk mounted into the filesystem. That plus the kernel tends to eat 5-6 meg of RAM, and with the libraries and such their Disk Druid program and fdisk both can't load, which are prerequisites for mounting the swap space.
As an alternative, documenting how it can be done from the command line in the shells thats opened would help. As I said in another post, I wasn't able to find what program it was using, since the usual utilities don't seem to be there.
I have 176 meg of ram in my system at work, and those applications do tend to swell in size (particularly Netscape). I actually try to close Netscape once every hour or two (figuring it doesn't crash on its own) to keep memory bloat down. I don't know if its a "feature" or just a bad memory leak, but it is bad.
Now, where 8 meg is concerned, 8 meg most certainly isn't a joke. Back in the days of the pre-1.0 kernel, I ran a rather useful system for several years in 8 meg of RAM. When I was in school, virtually all of my papers were written on that with vi and nroff/groff. It handled e-mail serving and reading, usenet reading and posting, and I was running a mailing list getting almost 30 postings a day to 200 people on it. I also gave user accounts to friends who needed a machine to work on from terminals around the campus.
Never had a problem with it, at one point had nearly 9 months uptime on it.
I currently have three systems here with only 8 meg RAM. One's being used as a development platform for embedded linux POS applications. One handles my internet dialup, masquerading, routing, firewalling, fax sending and receiving, and voicemail. The third is the system I mentioned above, which is an old notebook computer. Slow, low in RAM, small screen, but the battery lasts almost five hours, and its great for writing when I want to be outside. The router box used to handle printing too, but ghostscript eats too much RAM, so I moved it off to the system I've been experimenting with Oracle on.
All three of those applications don't need more than eight meg of RAM. They're providing important functions for me without costing me any excessive amount of money, using old parts I've scraped up.
Its a lot of functionality in not a lot of hardware. If you want to know why the ability to run in 8 meg is important, that's the exact reason. There's lots of very inexpensive hardware that people can buy or have laying around that can be made useful as print servers, or any of a dozen other functions, and 8 meg is enough for many of them.
Unfortunately that doesn't help with RedHat. RedHat doesn't give you the ability to mount a swap partition until after you have partitioned the drive -- regardless whether or not the drive is already partitioned. The real problem in the RedHat install with 8 meg seems to be the amount of memory it takes to fork Disk Druid or even fdisk. The very next step is mounting the swap space, but in one attempt I made to install it onto a 486/80 (AMD) with 8 meg ram, I let it run swapping for two hours just to see if it'd ever actually manage to get fdisk or Disk Druid to load up. It never happened.
Debian is better -- it lets you mount a swap parition if you've already created one. In the previously mentioned install, I was able to boot off a RedHat install disk, load no drivers for anything, and just squeak by after killing the installer to be able to load fdisk and partition the drive. My hope had been to mount the swap before getting to the partitioning step, but I was unable to find the program that's actually mounting the swap space. I'm assuming RedHat isn't using swapon / mkswap for it, but I might've just missed it.
I ended up ordering Debian CD's and got it installed on that system with no troubles. I remember I had to install the barest minimum I could, then reboot to get it to handle the full install. I never did figure out why that was the case.
Part of the autoLinux stuff I've been working on is getting a good mid-size distribution together. (Bigger than the various router projects out there, but smaller than what I remember the Slackware "A" series being...)
Anyone know what the requirements for a system are going to be for Slackware 4? I've been disappointed lately to see they keep growing with Redhat (in particular) and to some extent with Debian.
I know Linux on a 4 meg system may not be reasonable any more, but it seems 8 should be doable, but 8's too small to install RedHat, and I 've only gotten Debian to install in 8 meg successfully when it was installing from a local harddrive.
Is the minimum (A) set still fairly small? It'd be nice to have a standard distribution that can get a core system installed in 15-20 meg.
I can't speak to its stability compared to Solaris's version -- never had a speck of trouble with either, but I've never pushed either very hard.
;)
I can tell you, on a 586/133 (AMD) with 64 meg of ram, and a five year old slow-as-molassis IDE harddrive, it takes a good four hours to install. Most of it the process of setting up the initial database, AFAIK, and not needing any interaction. But it did work.
A nice spankin' new Pentium II or III might work better.
Huh? There is a core for Mozilla, what do you think new-layout was? I'm not comparing M3 to Navigator 4.5... I'm comparing M3 to builds from a few weeks ago. It certainly is backtracking if something that's been working for weeks or months isn't after a codefreeze.
:)
Given the constraints of HTML tables (setting borders to 0, cellspacing to 0 and cellpadding to 0), the interior size of the table should match the exterior size.
It seems whats changed in Mozilla, after some other testing, is settings on tables are no longer inhereted from tables they might be within. Up until now (and this might be a bug in Mozilla, I'm not sure...) nested tables (or for that matter any nested tags) took on as defaults the setting from the element they're nested within. Setting "cellspacing=0" on a table passed that setting onto tables it contains. Same with font tags, or other tags.
Only some settings on elements seems to be inhereted... border, for example, seems to be inhereted, but cellspacing and cellpadding don't. Or at least don't consistantly get inherited.
Adding them manually to the table fixes the problem. I looked that the HTML spec, and its not really clear IMHO on this issue.
Maybe I'll submit this as a bug at the risk of looking dumb when its not.
Sorry, this seems to be a step back. Sure, the browsing window has some (ugly, nonfunctioning) UI elements wrapped around it, but the core of the browser -- the renderer -- seems to have taken a significant step back. The nightly build from mid-february I was running layed out most pages I looked at fairly well, 90% of the time they were correct as far as I can see. Now half the sites that worked before no longer lay out properly.
:) Unfortunately it blows core when I try to open the console window, so I can't see exactly whats not working...
Some things I've noticed that are broken now:
Tables inside of tables where the width of the inside table is the same as the width of the cell its in -- no longer fit. The inside table for whatever reason starts to wrap cells. Very strange...
Frames. (I can't honestly remember if the other one had frame support at all, but I get a lot of frame content showing up in the wrong frame...)
Although it *says* its got editor support built into it, the editor doesn't work at all. That's a step back. Opening editor pages just opens up new browser windows (which cause it to blow core when you close them...)
Javascript -- working just enough not to work.
The widgets... I thought this was GTK? What version of GTK is it using? Themes don't work -- and the widgets themselves don't look like any of the GTK ones I've got installed. Where's it getting them from? Either way, I get toolbars showing up in the scrollbars, although once a page is loaded they usually go away.
Resizing the window doesn't work...
Forms don't seem to work, whereas they did before.
I couldn't get any background images to display. Some JPG's were not displaying either, and the ones that don't display seem to badly mess up table rendering. (IE, a 200 pixel wide image aligned left in a table cell with text around it will push the table 200 pixels wider...)
Of course, most of the people getting all worked up about the release think "M3" is some official release or something, its just a development milestone. No promise or guarantee it'd be better than what was there before, its just a point to freeze the code and take stock on whats done and what needs to be done.
I'm just suprised that the core of it seems to have backtracked so far...
Ever since freshmeat changed their design, the homepage causes an instant browser crash on two of my PC's, this one (Netscape 4.5, RedHat 5.2) and mine at work (Netscape 4.5, mutilated RedHat 5.0...)
:)
At least Mozilla's getting better. One of these days I might get to look at freshmeat again.
Hard to come by? What makes you think that? Twenty seconds on yahoo and I can find dozens of them...
Do some poking around the net. I can't remember the company, but there's a place that sells webcams that simply plug right into the ethernet. They're essentially a SBC PC, camera, and embedded OS (Linux maybe?). I seem to remember they were $700 or something like that.
There are probably a hundred companies that make these, and most are a *LOT* cheaper than this. I've seen systems that size at that power for $400, not $800.
Just do a search in yahoo for embedded PC. You'll find lots of them. Or you can look at some of the links on my autoLinux site at http://www.bangsplat.org/autolinux.
I seem to remember I linked to some there...
Not once... not a single time, EVER, has an application I've been using under Linux caused the OS to crash where I had to reboot.
Netscape crashes a lot itself... but its lousy software, not a lousy OS. Once in a while that spawning window problem happens and X might hang -- it doesn't crash, it just gets hung up. Telnetting into my box from another machine and killing Netscape always fixes the problem.