Why is it that all the videos I ever see on the internet are porn? I mean, is that all that anyone ever thinks about? What's the internet for anyway?
Downloading porn videos. Duh!
Thing is, porn videos have one thing going for them that a lot of the other video content on the net doesn't have. A very patient audience. Broadband is a LONG ways off from being the dominant player. Very few folks are going to set their modems aside for a 30 minute sitcom with a 15meg file size (horribly underestimating I realize). On the other hand, there's plenty of folks out there that'll leave the modem downloading all day to get a dirty video.
Despite the questionable product that porn sites provide, so far they've been a good indicator of what directions the Internet is or will be taking. We're just now seeing the same kind of power marketing the porn sites have been doing for a while getting into the main stream sites. They were also working out a lot of the eCommerce problems before a lot of other folks even considered it a viable option. Okay, so now they're causing the bulk of the problems here, but you can bet that you'll start seeing those same problems arise out of non-porn sites in the not too distant future.
Keep an eye on what them porn sites are doing, and not just the perty pictures either. They're battling out a highly competitive, multi-billion dollar market. The kinds of things that go on there eventually filter up to the rest of the web.
However, barring the emergence of a fundamentally different network structure, internet viewers cannot be identified by region.
An outstanding point, for the present. IPv6 on the other hand has built into it an IP address structure that may just make it possible to identify you by region. Although the marketing folks talk about it having more addresses than the ocean has water molecules, the actual spec calls for the first batch of numbers as an identifier of country and network. There's actually a lot of scary stuff built into IPv6, but I'll avoid going into all of that here.
The problem you refer to was caused by multiple exclusive licenses being granted to the same program, to different local markets (and, for ESPN etc., the complement of those markets).
Aren't programming licenses a civil matter? The FCC is in place to insure companies can share the radio spectrum without bouncing into eachother, not to haggle who sold what to whom. They've certainly gotten into that game now, despite having questionable jurisdiction.
I'll give ya this much, those were some very good items you pointed out. Heck, if ya dropped the porn story and logged in you'd be an interesting person to read more often. I almost didn't see the post because I cruise at +1 the bulk of the time.
Some regulations on issues such as leakage from cable lines were necessary. Some special cable channels had the same frequencies as the aircraft radio services. Those should remain, but will obviously become more and more irrelevant as cable systems gradually migrate to immense bandwidth digital fiber.
I've actually done calibration work for a major cable company for their test equipment. One of the pieces I had to do were these "sniffers" that all the field techs kept in their truck. They used these little buggers to check for RF leaks in the system. Thing is, way back when cable broadcasted on essentially the same frequencies as regular television. Not really that big a deal if a little RF leaked out in comparison to the kind of wattage that the broadcasters were pumping.
This all changed as the cable companies expanded their channel offerings, and begain using their own set of frequencies. With that in place they were putting RF out across the spectrum up to about 1GHz. Most of the older sites only really go up to about 500MHz, but that's still enough to trample on RF signals far more sensitive than neighboring TV stations.
Point is, I actually agree that the cable industry should have certain aspects of their infrastructure under the eye of the FCC. I wouldn't want some ambulance or police car's radio or digital transmissions interrupted by a rerun of "Who's The Boss".
I think the momentum is there to the point where it cannot be stopped even by the FCC
Not even in my wildest conspiracy fantasies would I think that either the FCC or the government would want to "stop" either the Internet, or broadband digital media. What I very well could imagine is a scenario that gets built up to allow some department into the regulating game. In this case, the FCC would be the most likely candidate to pass this on to. Keep in mind, we're not talking about destroying, we're talking about controlling.
In addition, I don't think we'll ever see congress actually do anything to regulate or slow down the Internet in any way. At least not directly. When you consider how the public just hates those evil folks at the IRS without consideration for who writes their rules you get a picture of what I'm talking about here. Be it the FCC, FTC, or some other entity, one of them is going to be put out there as the bad guy.
When (and if) streaming data starts getting into competing with the old style media television in the US, you can darn near guarantee congressional involvement.
Case in point, do you recall a few years back a scuffle between the cable industry and local broadcasters? Them local folks were griping that they were having to compete with other local broadcasters from outside their range. Nevermind the fact that they may have been sending a crappier signal, or had poorly chosen times for their program schedule. The whining and moaning kicked in about not being "fair".
This whole silliness got into the FCC's realm to correct. What in the hell the FCC has to do with a closed network system like cable programming is still a bit of a mystery to me. Anyhow, from this we now have shows blacked out from channels outside the local area if that same program is being shown on a local broadcaster. The consumer can go pound sand for all that the government cares.
What does this have to do with streaming data on the Internet? It's a closed system that doesn't involve broadcasting controlled frequencies. Thing is, as with the cable snafu you can just bet that we'll be seeing legislation down the road "protecting" the local broadcaster's from a situation that's not "fair" for them to compete in. It'll also be a fine place to have the government step in to starting to get a handle on all this free flow of info via the FCC.
I'm not talking about 2 weeks from now here. Nobody is seriously looking at impeding a still emerging market. That, and broadband hasn't yet penetrated deep enough to be a concern for the NAB. You can mark your calendars now though. At the present growth of broadband, you can damn sure bet we'll be seeing this coming at us within 2 to 3 years. It'll be a contest of a bunch of start up dot-coms who need to have Washington DC pointed out to them on the map versus the major broadcasters who have millions flowing into both major political parties on a very regular basis.
Aside from the money there will also be the issue of exposure on the minds of the politicians. They absolutely rely on traditional media to bring in their votes. How many folks in congress do you think are looking to bite the hand that feeds them?
In all honesty, I'm probably way off base here. What concerns me at this point is the stage is set for such a thing to occur. Here's hoping that I'm very wrong.
A lot of the DLL problems stem from not only allowing installations into the system folders, but also poor versioning. For example...
MS Windows 9x comes with foo.dll version 1.0 Vendor writes software to Ver. 2.0, then distributes updated DLL. MS releases a Ver 3.0 of foo.dll.
The problem here is that if you've updated your Windows with the latest patches from MS, then install this 3rd party software that thinks it's upgrading your system then conflicts arise.
On the plus side, I have noticed that some of the games I've installed on 98 now actually both to check to see what version of DirectX you have prior to trying to upgrade it. This kind of thing goes a long way to making the DLL concept actually work. The only problem is, there is no underlying system at the OS level that insures that 3rd party vendors (or even divisions of MS) play by the rules of checking before upgrading.
I haven't read anything about what Win2k is planning on doing to address this. I have heard Gate's address this directly in keynote speeches in the past as a problem with the Windows platform that MS needs to address. As with anything MS does with Windows, they have a large legacy issue they need to deal with at the same time.
What's interesting to me at this point is how similar this problem has become for Linux as well. At this point software either documents what libraries are required, leaving it to the user to go locate and upgrade, or it doesn't use external libraries at all. Package managers sort of deal with this, but they have plenty of their own problems as well.
Microsoft did not make Windows a market leader by having it handed to them by IBM. It had very little to do with their marketing. Everybody seems to have this extremely short memory of how this really went down.
What Microsoft focused on from day one with Windows was getting developer support. MS Excel for Windows was just as much a proof of concept as it was a product to be sold. MS did everything they could to attract developers, and in so doing built a software base on this new Windows platform of theirs.
No, your mom nor your other family members were concerned about "reusable code". They were concerned about what they could do with that computer when they got it home. What they could do with it is largely influenced by the software that can run on that platform. Reusable code, centralized libraries, built in driver support, top of the line development packages, and a whole host of other perfectly good reasons brought the developers to Windows. They in turn brought the users.
Thankfully it seems that the folks working on KDE and Gnome understand this concept. On the downside, there's a huge task ahead for folks working on Linux to make this actually possible. It's unfortunate that so many folks posting here either don't understand this, or simply find it more convenient to strap on to their more than slightly skewed beliefs.
Since we're trying to get facts straight here, at the time it was known as "Galactic Digital Research". As is a common excuse for a LOT of things from that era, it was the 70's.
If Linux fails as a desktop platform, you can bet it'll have something to do with this attitude.
Until there is a flavor of Linux that has the droolproofing and sneezeguard that the more popular OS's come with, it will not begin to push Microsoft out of the home market.
Droolproofing? So, someone who doesn't spend 12 plus hours researching the intricacies of LISP to get their menu buttons to turn blue is just some drooling idiot to you? It wouldn't be so bad if this were just a stupid comment, but this kind of attitude towards the users today, and those of tomorrow is down right dangerous. Are you so blind as to not see how this goes to undermine the very concept of open source software?
This may be old news to some of you, but I just recently discovered this one. Had one of my users bring me his laptop with a variety of problems on it. Had the usual glitches that form up after a while on Win98, but one of them was especially interesting.
His Netscape kept loading up this GoHip web site as it's default home page. Even going into the preferences in NS would only change this until the next re-boot. Had me poking around all over his system trying to figure out how his default home page kept getting changed. I couldn't find anything in the registry or.ini files that looked to be starting up that was out of the ordinary.
I then popped on over to this GoHip web site to have a look. Right on their front page is a link that states something like "Make GoHip your default home page". The clever bit was that this was not a link to some how-to about preferences. It linked directly to a.reg file. This site was able to tweak registry entries directly from the web!
Once I managed to download this.reg file to my local PC I was then able to trace back what all it had changed and get this thing off his system. I knew Windows had some security problems, but I had no idea it was THAT open to an attack.
Now just imagine sending someone an E-Mail with an embedded meta tag that redirected you to some.reg file you've got mirrored on a number of free web hosts. Heck, all I'd have to do at that point is delete the file association to.exe and.com files, which is just two lines of the registry, and I'd have your system rendered useless.
Mind you, I strongly disagree with this monopoly case that is presently going on. The details of this I'll save for later. On the other hand, I would have no problems at all with Microsoft being held criminally liable for gross negligience. None of what I'm talking about here is a secret to Microsoft, and still they continue to put out a known faulty product. How long do you think folks would put up with flaws like this from Ford, Honda, or any other car maker?
As cool as seeing a Woz interview is, I actually found the follow up interview to be more interesting. Need to go shopping around for this guy's book.
Anyhow I felt I just had to repeat this paragraph from that second interview here. This is singuarly the finest analysis of the whole MS monopoly issue I have seen to date.
Any thoughts about the Microsoft case? I don't think the whole story was about Microsoft as a monopoly. The Justice department coming in had to do with power. Orrin Hatch let the cat out of the bag when he said, "Microsoft should have thought about working in Washington a lot sooner." Spend more money on campaigns, spend more money on lobbyists, get back here and schmooze us. Tug on your forelock and show respect. Look what happened after the Justice department announced its case. Microsoft gave away millions to campaigns, got a whole army of lobbyists back there, and Bill Gates went around from door to door to all the congressmen. That's why Intel did such a better job. Intel, the moment they announced the Justice department case, the whole executive team flew back there and said, "How can we settle this in an amicable way?" No one even remembers there was an antitrust case against Intel. They got out of that sucker in a week and a half. Microsoft was too arrogant. They said we don't need a federal government. They should have remembered Al Capone; he said the same thing. Federal judges and the United States Army, they'll get you. Gates' mistake was he didn't immediately apologize, settle, and say, "How can we make this thing better?" and by the way, we're donating a million dollars to the Democratic national committee and 2 million dollars to the Republican national committee. He's paying for being Bill Gates. Part of Gates' personality is to never, ever give up an inch of ground. But I think what they're scared of now is that they've now been categorized as evil. Everybody knew that in the industry anyway. Where it's going to hurt them is recruiting. The key to all these companies is what kind of talent they can recruit for the next generation of products. Do you really want to go work for the Evil Empire? People are embarrassed to say they work for Microsoft now. In the long run that's what kills you because the quality of your talent starts falling and you can't fix it. And so your products don't get out on time and they're not as good as they used to be. They should have just caved the first day. Big mistake.
RANT:ON: Nobody, and I mean not a single soul, in congress gives a squat about whether or not Microsoft bruised other companies around them. Fact is, they didn't pay their "protection" money like Intel, IBM, Netscape, Sun, Apple, Oracle, and the rest of the industry testifying against Microsoft has done. MS has not been fighting a battle to innovate, but rather to not have to pay the "protection" money.
When all this plays out, no court ruling is going to replace the desktop OS's of the world. Only a better product will do that.
High ease of use
Low consumer cost
Heavy support by developers
At this point, Linux is perhaps the best bet for actually going after the MS market share. The point is, nobody in the open source world should be looking to this monkey trial as any kind of victory no matter how it plays out. Instead, it should be waking folks up to the cold hard fact of the payola system that's really on trial here.
RANT:OFF: The rest of the interview was definitely worth a read as well.
At least we're not filled to the brim with shotgun-toting inbreds. Lots of stupid people, yes.. but not many Anonymous Inbreds.
Wow, guess Kyle's mom really was right.
You can always tell when you've got a REALLY good Slashdot article when you get the bigots and zealots identifying themselves. Hey, got some hardcore socialists into the mix too. Now this is the good stuff!
First off, it's just plain silly trying to push a political agenda here on Slashdot. For you, or the orginal poster.
Oh my, I just can't resist.... What do you suppose the statistical odds of a repeat offence following a death penalty being carried out? No, you can't quote any Stephen King books either. For what it's worth, there's probably already been a $100mil study looking into this very thing.
I've got a very similar problem concering a VPN solution. My company has a Checkpoint firewall with VPN support that we've been quite happy with running on NT. The main problem that I've been running into is the software for client side only supports Windows platforms and the client must have a routable IP address.
For the most part, this isn't a problem since Windows is the dominant desktop platform around our office. With home networking kicking in with a lot of my folks, they're finding a need to have a single routable IP solution at home for multiple boxes AND having VPN support for them. I also have one remote office that presently has to have unique routable IP's for each client. To further complicate matters, that remote office has a couple of Macs tossed into the mix.
I've been looking about for a reasonable server side solution that I can deploy to a number of locations to handle the chit chat between it and this Checkpoint firewall. If I can get either Linux or a flavor of BSD to act as a proxy and VPN solution, freeware will get migrated into my office setup for the first time.
Aside from getting this to work at all, I do have support concerns. Between Linux and BSD, I've at least spent some time using Linux but there's apparently stability concerns with S/Wan. OpenBSD looks interesting, but I have zero BSD experience at this point in time. All the support and configuration falls squarely into my lap to implement.
I had rather hoped to find something that was a Windows based solution, mostly since that's what all my remote users are using. Not too many folks want to go out and purchase a seperate PC just to handle network proxying. Granted, this isn't nearly as much of an issue as the remote office is, as I can easily get another PC to deal with this there.
Bottom line: I need a solution that proxies non-routable IP addresses to the internet while providing for VPN support to a Checkpoint firewall.
Why would you symlink to things, instead of just putting them in the structure you want them in in the first place?
For NT Workstation I'm kinda doing the same kind of hack that the original poster is talking about. Unlike the 9x flavors of Windows, NT does a pretty fair job of providing a multi-user OS. Perhaps the worst aspect of this is how deep into the Winnt directory you need to go before you actually get to a specific profile.
To get around this, I keep a number of short cuts on my desktop that point to those specific user folders. For example, I've got one called "My Documents" that actually points to the document directory within my profile. When setting up an NT Server, I'll set up short cuts to the short cuts that are in the start menu for the various admin tools, so that service packs and upgrades know where to find the proper path to add their short cuts.
Thing is, even in NT there are certain "expected" directories that just make life easier to leave them be. I have no idea exactly how much havoc you'd cause if you went in and just changed the "System32" directory's name to something else. I sure know it wouldn't be pretty.
I do have to wonder what the true value of having a mutli-user OS is for everyone out there using computers. Heck, even on NT I pretty much use it 100% with "Administrator" (aka root) rights all the time. The multi-user thing starts making a lot more sense to me when you get into talking about servers, and different admins with different rights to it. For workstation use it feels more like a barrier than a feature. The significant difference here being between local and network rights and permissions.
On my Linux set up I pretty much went and did it like I was supposed to. I created a user account with limited rights where I spend the bulk of my time. Thing is, hardly a day goes by where I don't have some reason to "su" my way into control. Not sure if this is a good or bad thing at this point, just an observation. If this were an actual production box I'd probably be a good bit more concerned with the specifics of the setup. As it is, this thing is on a non-routable IP behind a proxy that's on my NT.
I think if we're going to package managers, we should standardize what they do
Whole heartedly agree. When I think "standardizing" I tend to be thinking more along the lines of how HTML compares to Word Docs. By standardizing HTML it allows for a wide variety of applications to view and edit these files, where as Doc files can't. To carry this analogy further, I never meant to suggest that standardizing meant that only Netscape should be used as a browser. When standards are done properly and adopted, like with HTML, it actually expands the number of applications that can be built around it rather than stifling them.
WinZip isn't really a good analogy
The only thing I was attempting to illustrate was WinZip's ability to track an application's installation process. You are quite correct in that it makes for a poor example to directly compare this to what a package manager does. I don't think I illustrated my meaning well enough there.
You know, for the most part, I think the issue is file interoperability between distributions and/or similar software... if dpkg and rpm could use the same files (successfully!) or if KDE and gnome could use the same files (for example), we'd be in business.
Maybe this conversation could draw more bad analogies to WinZip. For example, what WinZip actually does is front end several types of compression routines. Heck, it's only the more recent versions that don't rely on having PKZIP around on the system. It'll recognize.tar,.gz, .arc,.zip, as well as a whole stack of others. As previously stated, a package manager has a lot more to do than just uncompress, but WinZip is a fair example of a front end for several back end processes bringing it all together in a pretty decent UI. Unfortunately, I think the problem is far more complicated by the fact that each package manager is maintaining their own db without regard to what the other is doing.
As far as the Gnome & KDE issue, I'm pretty much able to run KDE apps under Gnome without too much of a glitch. My biggest complaint has to do with file associations. Since RH installed Gnome by default, the KDE installation I did later didn't pick up any file associations. I was amazed that something this basic wasn't handled well below the window manager level. Until such a time comes about where I want to spend many an hour manually creating associations, I can't really use KDE full time whether I like it or not.
Anyway, thanks for sharing. I do agree that there is a need for some semblance of standards to not push distributions too far away from each other, but I think we have to be *very* careful about it.
Your quite welcome. And a return thanks for a reply that reads like a well thought out letter. Just goes to show, folks coming from Linux and NT backgrounds can have civil conversations. Sometimes easy to forget that point after reading a bit too much/.
It seems like a lot of the windows-y people coming to linux want to see linux look/feel/act like windows without looking into what linux has for itself, and where it's going on it's own.
Had a phrase where I worked a long while ago I rather liked. "The customer is not always right, but the customer is always the customer." The notion here is that us folks coming from a Windows GUI perspective have certain expectations from an OS. Sometimes these are good things, other times not so. Despite the rhetoric, Microsoft really does have some very cool UI ideas built into Windows. To date, they are still the only company to bring an OS to the masses. There is considerable value to Linux in recognizing this as projects such as Gnome and KDE mature.
Huge Disclaimer: Nothing of what I stated above is meant to suggest that Linux should look to clone it's likeness into Windows. Get the good stuff out of it, then move on.
The only other part of the article(s) that I couldn't stand was the "even your parents, sister, or girlfriend" could use it gem. Excuse me, but my sister is pretty damn smart, and it is I who teaches my husband about linux. *snort*
Yeah, I work with my ex on various programming projects still yet. She still runs circles around me when it comes to SQL and XML. She's still very heavily in the realms of MS, as she's making a very nice living knowing how to muck around with Site Server and ASP.
If I had moderator points that I could use, I'd give you a +1 for your contribution.
That bash for win32 might be fun to play with. I'll give a look around for it as time permits. I should note that I meant my post to illustrate strengths and weaknesses in both CLI's. For the most part, I still feel more comfortable on NT's CLI. Yes, I fully realize this probably has far more to do with prior experience then which one is actually "better".
One thing that just came to mind concerning a comparison of the two CLI's has to do with the help system involved. After playing with the two, I have to say that NT's help at the CLI level is still better than Linux's.
I realize that Linux tends to provide a LOT of depth about each and every command through the use of "man" or "--help". The problem is, it's a LOT of depth. One example here is a paragraph from the "cp --help" command.
"By default, sparse SOURCE files are detected by a crude heuristic and the corresponding DEST file is made sparse as well. That is the behavior selected by --sparse=auto. Specify --sparse=always to create a sparse DEST file whenever the SOURCE file contains a long enough sequence of zero bytes. Use --sparse=never to inhibit creation of sparse files."
Here I am, the idiot user who wants a file to go from over here to over there. After looking at this I get to thinking that maybe that file is doing hunky dory right where it is. I've been using a variety of computer platforms for over a decade now, and I still have no idea what a "crude heuristic" is or why I should care. I know Linux is international and all that, but where do I go to pick up the version in English?
NT's help system on the other hand tends to be a lot more straight forward to getting the info I need. DOS used to have a help system that resembled the "man" system quite a bit for it's commands. It was actually pretty cool, and it may still be around in Win9x, but I haven't looked for it.
Just to get the disclaimers out of the way here, I know that a *nix CLI is always going to be at some disadvantage to me due to many years of working with DOS (back to before 2.0) and all its variants. Furthermore, I also realize that many of the *nix commands have a lot more capability built into them. Thing is, I still find it far easier to pick up a DOS style command than a *nix style command. This may be due to the above reasoning, but I also believe that the help system plays a very large factor.
Maybe some of the more basic CLI tools need a couple levels of help. "--helpwalk" and "--helprun" kind of approach. Just a thought.
Last bit of a rant here, as a few weeks ago I located what might have been a really useful "How-To" which got into explaining to someone with a DOS background the equivalent actions on Unix. The thing was written like a propaganda sheet for Unix rather than a tutorial. There didn't seem to be a paragraph in the whole thing that didn't include something along the lines of "...and that is why Unix is so much better than DOS...". To make matters worse, the author had several mistakes in his interpertation of DOS commands and what they could do. Mind you, I don't have a great love for DOS beyond the notion that I do understand how to use it pretty well. I just can't stand reading the work of a zealot, even if I agree with them. How can you concentrate on material if the voice in your head keeps screaming, "Bullshit!"
Maybe some day down the road when I actually get to understanding Unix CLI a bit better I'll get into writing my own How-To for this topic. In the mean time, I'll just keep plugging along.
stty erase to define the backspace key to be some other key... Use with caution!
I'm sure there's some good advice in there, but unfortunately you went zipping past me. I'll be out and about looking to figure out what you're talking about later on to be sure.
rm "This is a single filename with a lot of spaces in it."
Doh! Ya know something, I actually knew that but had forgotten. I mentioned it in my post because I recall having problems with that. Sortta like I STILL can't seem to remember to add that "./" before a file when I'm in the same dir.
In time you will try to highlight right click, click center click even in NT. and it won't work there:(
There's already tons of things I get messed up on going between OS's these days. Sometimes it's little things, like trying to use Ctrl-C in Netscape under Linux, or trying to drag a window to a virtual desktop that ain't there on NT. I could write up a very lengthy post with all the little oddities that arise when you constantly bounce back and forth between NT and Linux, with just a smattering of Mac once in a while. I'll spare the viewing audience that rant for now.
Here is a good reason. I hate trying to debug scripts where the same program is referred to as Wipeout, WIPEOUT, wipeout, WipeOut and wipeOut interchangeable. It makes it nearly impossible to find every reference to the command wipeout when I am replacing that script with something else.
I can't tell if you're agreeing with me or not here. It seems that this would be an outstanding argument against case sensitivity, as most Windows based editors allow for search and replace without regard to case. This is probably a bit trickier with a regular expression style of s&r, but I never did get a handle on them darn things anyway.
But it doesn't have too! UNIX has had job control for 25 years now. Instead of typing "xview stars.jpg" type "stars.jpg &" and you will run the GUI command in the background. Quite handy and far beyond the capability of Windows9x. To find out more about job control run the command "man bash" and look for a section titled JOB CONTROL.
Linux can institute a number of standards and still allow for a ton of customization. There are a number of areas that should be centralized, to at least agree on what kinds of things are stored where.
1) I mentioned in an earlier post my strong feelings towards standardizing package management. This is a key component to getting software on and off a system. Standardizing this should be able to include compiling from source as well as dealing with binaries. Heck, if WinZip can track where a group of files go why can't Linux?
2) I happen to like KDE for some things, but I end up mostly using Gnome because of how it handles virtual desktops. Bouncing between the two of them there's a couple of things I just don't get. Why can't these two managers actually see eachothers "Start" (for lack of a better term) menu? Why are the file associations mutually exclusive between the two? There's quite a bit that could be done here to standardize things without impacting the unique features each bring to the game.
3) The article did get into specifying some of the problems that lack of certain standards has had. Stuff like having to figure out which version of X, glibc, Gnome, KDE, kernel, etc. you happen to be running. Unless you did the building Linux From Scratch route, you're most likely going to have to scramble about your system to figure this stuff out. For a newbie, their going to have to spend a couple of hours researching how to go about scrambling through their system. I don't know if simply stating what components are to be standard is the solution, but there is definitely a problem here that needs addressed.
4) I remember the issue of distros coming together or fading away from eachother talked about over a year ago. On one hand, if they start coming together too closely they can no longer differentiate themselves from the rest of the pack. Going the other way, we end up down the road that split the Unix world in the past. It's a delicate balancing act that all the distros are playing.
5) I fully agree, "let's be like windows" is a poor way to go about things. Linux should be looking to grab the good stuff that is in Windows and move beyond. I think we're still in the grabbing stage for the most part (when talking about UI) and now is definitely the time to encourage setting the stage for moving beyond.
From an NT user here who actually knows how to get around in NT from the command prompt. Actually pretty good at getting around the entire NT GUI without a mouse as well, but that's another story.
My Impressions of the Linux CLI 1) Love the tab complete!! 2) Nice roll back history. Between sessions even! 3) Why can't the Delete key NOT act like a backspace? ACK! 4) Still haven't figured out how to deal with spaces in a file name or directory. 5) Rather like the hi-light with left, paste with middle mouse button thing. Not 100% sure I prefer it to NT's clipboard. 6) I hate case sensitivity. I have yet to hear a single reasonable argument as to why this is a "feature". pizza != Pizza != PiZZa. Has a logic train like this ever done anything but cause errors? 7) Don't much care for launching GUI apps from the CLI. Locks out that CLI from being used for anything else.
Impressions of NT's CLI 1) Nice roll back history, but it doesn't remember between CLI sessions. 2) "Insert", "Delete", "Home", and "End" do exactly what they are supposed to do here. 3) Copying is very easy (hi-light, press Enter), but pasting is a pain. 4) Spaces in file names and directories not a problem. Supports either "PROGRA~1" or "cd c:\Program Files\". 5) No case sensitivity issues!! 6) Launching GUI apps gives the user control of the CLI after the app has launched.
Windows 98 CLI (aka Win95SR-3, DOS 7.0 Shell, etc.) 1) Run screaming. 2) Don't turn around.
One major strike against Windows in my mind is that f'ing registry. For some apparently morbid reason I keep finding myself missing some of the stuff in Win 3.1, like being able to edit the.ini files from a DOS prompt. Definite plus for Linux, but now I just need to figure out where they're all at!
CLI programs for NT are all but dead when it comes to development. About the only new stuff I see there is some cheesy driver config. It is nice to see that Linux is still actively working on CLI apps. Now if I could just get a port of "edit.com" on over to Linux!:) Yeah, I know somebody out there has probably already done it, I've just been too lazy to go look. I think I may have 5 too many text editors on here now as it is.
What is pif you ask? I haven't got the foggiest idea.
Sounds like you just asked. I forget what the letters stand for, but this is a carry over from Win 3.1 that is still in use. Probably in Win2k, but not certain. What this file does is store the attributes of a shortcut. Stuff like where the file is stored, what kind of memory to allocate to it, how it will run, and other stuff like that. Under 3.1 you used to have to go into a seperate application to edit this, or tweak it directly with a text editor. With every version after 95 you edit those files by going into the properties of the shortcut. Roughly the equivalent of setting up symlinks in Unix, but with a bunch of extra stuff tossed in so you can actually get them old DOS apps to run under Windows.
Wow, have to go back 6-7 years to appreciate why those files exist. Now imagine how tough it is for folks like myself just now getting accustomed to Unix having to go back better than 20 years to comprehend why certain things are done in the way they are.
Along these lines, I can appreciate where the original poster is coming from. Being brought up from the world of DOS I'm pretty much used to defining whatever darn directory structure I want. I'm just now beginning to get a feel for where key items are in Linux, but I have no idea WHY they need to be there. I am slowly learning it, but I can fully appreciate exactly how intimidating that directory structure looks to a newbie.
The user shouldn't care how the filesystem is laid out so long as he/she knows how to find that document he wrote.
On this point we are in disagreement. I truly believe that the user should be able to understand what the heck is going on with their directory structure. The often mentioned "Program Files" is actually a great idea, though there are devils in them details to be sure. It sure makes a lot more sense than "/usr/bin" and all it's related cousins. Both Windows and Unix have a lot of legacy they need to bring along with them, so I don't expect either platform to get this truly friendly and/or organized any time soon.
Establish some standards that programmers can use to know where to place files.
Believe it or not, Bill Gates is in agreement with you. The problem here goes back to legacy support. Windows to this day has DOS apps to consider, not to mention how they set up Win 3.1 originally. I believe that if MS had the chance to start over from scratch there would be certain folders set aside that would not allow 3rd party files to be added. As I understand it (never played with it myself) BeOS does something very similar to this. BeOS presently enjoys not having to support a huge legacy of apps like MS and Unix presently have to deal with, so they can define these things without breaking anything.
These are not simple changes, but they are necessary to gaming development on the Linux platform.
For all the good points in this post, this one misses the mark. These kinds of things are going to be needed for ANY use of Linux on the desktop. It's these kinds of innovations that Linux should be pointing at as why it is so much better than Windows. Thing is, at this point it isn't.
1) I'm running RH 6.1 here myself, and at first I really loved the updating agent it had. Oh boy, then did that love ever go sour. Their FTP server they provide without cost is never available, and I have yet to get any of the supposed mirrors to work. To make matters worse, as I got more comfortable where to find the actual updates I discover that the stuff from the RH updater is usually more than a couple of versions back.
<rant val="on"> Just a stupid newbie question here, but couldn't an agent be written that utilizes FreshMeat as a backend? FreshMeat has all the latest stuff, summaries of the changes, usually functional links even. Instead of having a system that relies upon the distributor to keep things up to date, it seems that a 3rd party like FreshMeat would be perfect. Wonder what Andover would think of it?
The biggest problem I see with this is that Linux has just too many package managers. I like variety as much as the next fella, but not in this case. I fully realize the religous aspects to this debate, but this is a huge obstacle on the road to Linux ease of use. Why can't the Linux community bring in the best of RedHat's, Debian's, and even BSD's package management and label it a standard?
In the wild world of MS there's 3 or 4 major installer apps that most folks use. On occasion I do see a few that appear to be custom written, but they are the exception. Thing is, they all pretty much function the same. Boring as that may be, it provides a level of consistancy that is presently lacking on Linux. Oh god, then we could go on a while about the fun of uninstalling apps. Where the heck did that "make install" put everything? ACK! This is the kind of stuff that an OS should be keeping track of for me. Of course, it won't ever be able to until there is some level of standardization to application installation. <rant val="off">
2) From what little I understand about the architecture X it may need to be replaced to support everything you suggest. There are a couple of projects in their infancy now that are attempting to address this. Thing is, they are VERY much in their infancy, and there is also a huge legacy that comes along with X.
3) To my original point, OpenGL at the OS level would have implications that go beyond gaming. I could envision new GUI technologies being developed around this, not to mention advanced support of graphics editing packages. We would certainly see the impact on the games before anything else though.
Re:shocking may be the easy configuration...
on
XFree86 4.0.1 Review
·
· Score: 1
I've been running a dual monitor set up now under NT for a number of years. Been doing it for a long time at work, and more recently got set up at home to do the same. I've got to say, it is something I definitely miss when trying to work on someone else's PC.
Thing of it is, when I upgraded my NT to this Matrox G400, I moved my old Voodoo3000 into my Linux box. I'm finding that I really love having all these virtual desktops under Linux, even with just one screen.
The dual screen comes in really handy for apps like Photoshop, Imageready, Dreamweaver, Homesite, or darn near anything else that has a lot of floating tools. On the other hand, I've gotten quite used to dragging lots of opened browser windows to different desktops on Linux. Gnome and Sawfish handle this really nicely.
I don't know if there's much of a point to this post, other than I believe that each kind of environment has it's pros and cons. I haven't seen a dual monitor set up for Linux yet. I'd have to imagine that dual head with virtual desktops would get pretty confusing real quick. Sounds like it'd be fun to play with if such a thing were possible though.
Okay, here's a point of view that's been rarely expressed here, but here goes: font antialiasing is a crufty outdated process which isn't really very useful on today's monitors.
Not even close to truth. Whatever performance hit you might take from anti-aliasing fonts, it is absolutely worth it. I'm running an NT Workstation next to this Linux box I'm typing this on now, and there is a marked difference in the font quality that is quite visible. Both are on 17" monitors running 1152x864. And yes, I can quite easily tell when AA is turned off on the NT side, even with true type fonts being used.
In trying to see if Linux could act as my primary desktop I've tried using Abiword and StarOffice to simply write up a document. Every time I try I get about a paragraph into it, then realizing I can't read what I'm writing! I'd rather use a CLI based word processor than one in X, because at least there the text is readable.
Even with TrueType fonts there is a marked difference in visual quality as the size of the font grows. Might explain why the Mac only supports font smoothing for large type fonts. Even still, NT actually presents far clearer fonts than Mac as well, primarily for support for font smoothing down to small fonts.
If Microsoft ever figures out how to move that new font rendering technology of theirs to CRT support, it will raise the bar on visual quality quite a bit. If you haven't seen what they've done with this, you might want to go check it out. Smooth, crisp, clean fonts with no anti-aliasing involved. Only working on LCD based screens at this point though.
Bottom line, some day down the road Linux may work out to be my primary desktop. Aside from missing a few key apps, one of the biggest show stoppers for me is the readability of the fonts. I hope that either a future version of X corrects this, or X gets replaced by something that can. Heck, I'd love to see Linux one up Windows by working in things like kerning and tracking of fonts at the OS level. That'd be an awsome display, and a compelling reason for folks doing print work to seriously consider Linux.
Slashdot Mirror
Why is it that all the videos I ever see on the internet are porn? I mean, is that all that anyone ever thinks about? What's the internet for anyway?
Downloading porn videos. Duh!
Thing is, porn videos have one thing going for them that a lot of the other video content on the net doesn't have. A very patient audience. Broadband is a LONG ways off from being the dominant player. Very few folks are going to set their modems aside for a 30 minute sitcom with a 15meg file size (horribly underestimating I realize). On the other hand, there's plenty of folks out there that'll leave the modem downloading all day to get a dirty video.
Despite the questionable product that porn sites provide, so far they've been a good indicator of what directions the Internet is or will be taking. We're just now seeing the same kind of power marketing the porn sites have been doing for a while getting into the main stream sites. They were also working out a lot of the eCommerce problems before a lot of other folks even considered it a viable option. Okay, so now they're causing the bulk of the problems here, but you can bet that you'll start seeing those same problems arise out of non-porn sites in the not too distant future.
Keep an eye on what them porn sites are doing, and not just the perty pictures either. They're battling out a highly competitive, multi-billion dollar market. The kinds of things that go on there eventually filter up to the rest of the web.
However, barring the emergence of a fundamentally different network structure, internet viewers cannot be identified by region.
An outstanding point, for the present. IPv6 on the other hand has built into it an IP address structure that may just make it possible to identify you by region. Although the marketing folks talk about it having more addresses than the ocean has water molecules, the actual spec calls for the first batch of numbers as an identifier of country and network. There's actually a lot of scary stuff built into IPv6, but I'll avoid going into all of that here.
The problem you refer to was caused by multiple exclusive licenses being granted to the same program, to different local markets (and, for ESPN etc., the complement of those markets).
Aren't programming licenses a civil matter? The FCC is in place to insure companies can share the radio spectrum without bouncing into eachother, not to haggle who sold what to whom. They've certainly gotten into that game now, despite having questionable jurisdiction.
I'll give ya this much, those were some very good items you pointed out. Heck, if ya dropped the porn story and logged in you'd be an interesting person to read more often. I almost didn't see the post because I cruise at +1 the bulk of the time.
Some regulations on issues such as leakage from cable lines were necessary. Some special cable channels had the same frequencies as the aircraft radio services. Those should remain, but will obviously become more and more irrelevant as cable systems gradually migrate to immense bandwidth digital fiber.
I've actually done calibration work for a major cable company for their test equipment. One of the pieces I had to do were these "sniffers" that all the field techs kept in their truck. They used these little buggers to check for RF leaks in the system. Thing is, way back when cable broadcasted on essentially the same frequencies as regular television. Not really that big a deal if a little RF leaked out in comparison to the kind of wattage that the broadcasters were pumping.
This all changed as the cable companies expanded their channel offerings, and begain using their own set of frequencies. With that in place they were putting RF out across the spectrum up to about 1GHz. Most of the older sites only really go up to about 500MHz, but that's still enough to trample on RF signals far more sensitive than neighboring TV stations.
Point is, I actually agree that the cable industry should have certain aspects of their infrastructure under the eye of the FCC. I wouldn't want some ambulance or police car's radio or digital transmissions interrupted by a rerun of "Who's The Boss".
I think the momentum is there to the point where it cannot be stopped even by the FCC
Not even in my wildest conspiracy fantasies would I think that either the FCC or the government would want to "stop" either the Internet, or broadband digital media. What I very well could imagine is a scenario that gets built up to allow some department into the regulating game. In this case, the FCC would be the most likely candidate to pass this on to. Keep in mind, we're not talking about destroying, we're talking about controlling.
In addition, I don't think we'll ever see congress actually do anything to regulate or slow down the Internet in any way. At least not directly. When you consider how the public just hates those evil folks at the IRS without consideration for who writes their rules you get a picture of what I'm talking about here. Be it the FCC, FTC, or some other entity, one of them is going to be put out there as the bad guy.
When (and if) streaming data starts getting into competing with the old style media television in the US, you can darn near guarantee congressional involvement.
Case in point, do you recall a few years back a scuffle between the cable industry and local broadcasters? Them local folks were griping that they were having to compete with other local broadcasters from outside their range. Nevermind the fact that they may have been sending a crappier signal, or had poorly chosen times for their program schedule. The whining and moaning kicked in about not being "fair".
This whole silliness got into the FCC's realm to correct. What in the hell the FCC has to do with a closed network system like cable programming is still a bit of a mystery to me. Anyhow, from this we now have shows blacked out from channels outside the local area if that same program is being shown on a local broadcaster. The consumer can go pound sand for all that the government cares.
What does this have to do with streaming data on the Internet? It's a closed system that doesn't involve broadcasting controlled frequencies. Thing is, as with the cable snafu you can just bet that we'll be seeing legislation down the road "protecting" the local broadcaster's from a situation that's not "fair" for them to compete in. It'll also be a fine place to have the government step in to starting to get a handle on all this free flow of info via the FCC.
I'm not talking about 2 weeks from now here. Nobody is seriously looking at impeding a still emerging market. That, and broadband hasn't yet penetrated deep enough to be a concern for the NAB. You can mark your calendars now though. At the present growth of broadband, you can damn sure bet we'll be seeing this coming at us within 2 to 3 years. It'll be a contest of a bunch of start up dot-coms who need to have Washington DC pointed out to them on the map versus the major broadcasters who have millions flowing into both major political parties on a very regular basis.
Aside from the money there will also be the issue of exposure on the minds of the politicians. They absolutely rely on traditional media to bring in their votes. How many folks in congress do you think are looking to bite the hand that feeds them?
In all honesty, I'm probably way off base here. What concerns me at this point is the stage is set for such a thing to occur. Here's hoping that I'm very wrong.
K-LOC's
'nuff said about OS/2 right there.
A lot of the DLL problems stem from not only allowing installations into the system folders, but also poor versioning. For example...
MS Windows 9x comes with foo.dll version 1.0
Vendor writes software to Ver. 2.0, then distributes updated DLL.
MS releases a Ver 3.0 of foo.dll.
The problem here is that if you've updated your Windows with the latest patches from MS, then install this 3rd party software that thinks it's upgrading your system then conflicts arise.
On the plus side, I have noticed that some of the games I've installed on 98 now actually both to check to see what version of DirectX you have prior to trying to upgrade it. This kind of thing goes a long way to making the DLL concept actually work. The only problem is, there is no underlying system at the OS level that insures that 3rd party vendors (or even divisions of MS) play by the rules of checking before upgrading.
I haven't read anything about what Win2k is planning on doing to address this. I have heard Gate's address this directly in keynote speeches in the past as a problem with the Windows platform that MS needs to address. As with anything MS does with Windows, they have a large legacy issue they need to deal with at the same time.
What's interesting to me at this point is how similar this problem has become for Linux as well. At this point software either documents what libraries are required, leaving it to the user to go locate and upgrade, or it doesn't use external libraries at all. Package managers sort of deal with this, but they have plenty of their own problems as well.
Microsoft did not make Windows a market leader by having it handed to them by IBM. It had very little to do with their marketing. Everybody seems to have this extremely short memory of how this really went down.
What Microsoft focused on from day one with Windows was getting developer support. MS Excel for Windows was just as much a proof of concept as it was a product to be sold. MS did everything they could to attract developers, and in so doing built a software base on this new Windows platform of theirs.
No, your mom nor your other family members were concerned about "reusable code". They were concerned about what they could do with that computer when they got it home. What they could do with it is largely influenced by the software that can run on that platform. Reusable code, centralized libraries, built in driver support, top of the line development packages, and a whole host of other perfectly good reasons brought the developers to Windows. They in turn brought the users.
Thankfully it seems that the folks working on KDE and Gnome understand this concept. On the downside, there's a huge task ahead for folks working on Linux to make this actually possible. It's unfortunate that so many folks posting here either don't understand this, or simply find it more convenient to strap on to their more than slightly skewed beliefs.
Since we're trying to get facts straight here, at the time it was known as "Galactic Digital Research". As is a common excuse for a LOT of things from that era, it was the 70's.
If Linux fails as a desktop platform, you can bet it'll have something to do with this attitude.
Until there is a flavor of Linux that has the droolproofing and sneezeguard that the more popular OS's come with, it will not begin to push Microsoft out of the home market.
Droolproofing? So, someone who doesn't spend 12 plus hours researching the intricacies of LISP to get their menu buttons to turn blue is just some drooling idiot to you? It wouldn't be so bad if this were just a stupid comment, but this kind of attitude towards the users today, and those of tomorrow is down right dangerous. Are you so blind as to not see how this goes to undermine the very concept of open source software?
This may be old news to some of you, but I just recently discovered this one. Had one of my users bring me his laptop with a variety of problems on it. Had the usual glitches that form up after a while on Win98, but one of them was especially interesting.
.ini files that looked to be starting up that was out of the ordinary.
.reg file. This site was able to tweak registry entries directly from the web!
.reg file to my local PC I was then able to trace back what all it had changed and get this thing off his system. I knew Windows had some security problems, but I had no idea it was THAT open to an attack.
.reg file you've got mirrored on a number of free web hosts. Heck, all I'd have to do at that point is delete the file association to .exe and .com files, which is just two lines of the registry, and I'd have your system rendered useless.
His Netscape kept loading up this GoHip web site as it's default home page. Even going into the preferences in NS would only change this until the next re-boot. Had me poking around all over his system trying to figure out how his default home page kept getting changed. I couldn't find anything in the registry or
I then popped on over to this GoHip web site to have a look. Right on their front page is a link that states something like "Make GoHip your default home page". The clever bit was that this was not a link to some how-to about preferences. It linked directly to a
Once I managed to download this
Now just imagine sending someone an E-Mail with an embedded meta tag that redirected you to some
Mind you, I strongly disagree with this monopoly case that is presently going on. The details of this I'll save for later. On the other hand, I would have no problems at all with Microsoft being held criminally liable for gross negligience. None of what I'm talking about here is a secret to Microsoft, and still they continue to put out a known faulty product. How long do you think folks would put up with flaws like this from Ford, Honda, or any other car maker?
As cool as seeing a Woz interview is, I actually found the follow up interview to be more interesting. Need to go shopping around for this guy's book.
Anyhow I felt I just had to repeat this paragraph from that second interview here. This is singuarly the finest analysis of the whole MS monopoly issue I have seen to date.
Any thoughts about the Microsoft case?
I don't think the whole story was about Microsoft as a monopoly. The Justice department coming in had to do with power. Orrin Hatch let the cat out of the bag when he said, "Microsoft should have thought about working in Washington a lot sooner." Spend more money on campaigns, spend more money on lobbyists, get back here and schmooze us. Tug on your forelock and show respect. Look what happened after the Justice department announced its case. Microsoft gave away millions to campaigns, got a whole army of lobbyists back there, and Bill Gates went around from door to door to all the congressmen. That's why Intel did such a better job. Intel, the moment they announced the Justice department case, the whole executive team flew back there and said, "How can we settle this in an amicable way?" No one even remembers there was an antitrust case against Intel. They got out of that sucker in a week and a half. Microsoft was too arrogant. They said we don't need a federal government. They should have remembered Al Capone; he said the same thing. Federal judges and the United States Army, they'll get you. Gates' mistake was he didn't immediately apologize, settle, and say, "How can we make this thing better?" and by the way, we're donating a million dollars to the Democratic national committee and 2 million dollars to the Republican national committee. He's paying for being Bill Gates. Part of Gates' personality is to never, ever give up an inch of ground. But I think what they're scared of now is that they've now been categorized as evil. Everybody knew that in the industry anyway. Where it's going to hurt them is recruiting. The key to all these companies is what kind of talent they can recruit for the next generation of products. Do you really want to go work for the Evil Empire? People are embarrassed to say they work for Microsoft now. In the long run that's what kills you because the quality of your talent starts falling and you can't fix it. And so your products don't get out on time and they're not as good as they used to be. They should have just caved the first day. Big mistake.
RANT :ON:
Nobody, and I mean not a single soul, in congress gives a squat about whether or not Microsoft bruised other companies around them. Fact is, they didn't pay their "protection" money like Intel, IBM, Netscape, Sun, Apple, Oracle, and the rest of the industry testifying against Microsoft has done. MS has not been fighting a battle to innovate, but rather to not have to pay the "protection" money.
When all this plays out, no court ruling is going to replace the desktop OS's of the world. Only a better product will do that.
- High ease of use
- Low consumer cost
- Heavy support by developers
At this point, Linux is perhaps the best bet for actually going after the MS market share. The point is, nobody in the open source world should be looking to this monkey trial as any kind of victory no matter how it plays out. Instead, it should be waking folks up to the cold hard fact of the payola system that's really on trial here.RANT :OFF:
The rest of the interview was definitely worth a read as well.
At least we're not filled to the brim with shotgun-toting inbreds. Lots of stupid people, yes.. but not many Anonymous Inbreds.
Wow, guess Kyle's mom really was right.
You can always tell when you've got a REALLY good Slashdot article when you get the bigots and zealots identifying themselves. Hey, got some hardcore socialists into the mix too. Now this is the good stuff!
First off, it's just plain silly trying to push a political agenda here on Slashdot. For you, or the orginal poster.
Oh my, I just can't resist.... What do you suppose the statistical odds of a repeat offence following a death penalty being carried out? No, you can't quote any Stephen King books either. For what it's worth, there's probably already been a $100mil study looking into this very thing.
I've got a very similar problem concering a VPN solution. My company has a Checkpoint firewall with VPN support that we've been quite happy with running on NT. The main problem that I've been running into is the software for client side only supports Windows platforms and the client must have a routable IP address.
For the most part, this isn't a problem since Windows is the dominant desktop platform around our office. With home networking kicking in with a lot of my folks, they're finding a need to have a single routable IP solution at home for multiple boxes AND having VPN support for them. I also have one remote office that presently has to have unique routable IP's for each client. To further complicate matters, that remote office has a couple of Macs tossed into the mix.
I've been looking about for a reasonable server side solution that I can deploy to a number of locations to handle the chit chat between it and this Checkpoint firewall. If I can get either Linux or a flavor of BSD to act as a proxy and VPN solution, freeware will get migrated into my office setup for the first time.
Aside from getting this to work at all, I do have support concerns. Between Linux and BSD, I've at least spent some time using Linux but there's apparently stability concerns with S/Wan. OpenBSD looks interesting, but I have zero BSD experience at this point in time. All the support and configuration falls squarely into my lap to implement.
I had rather hoped to find something that was a Windows based solution, mostly since that's what all my remote users are using. Not too many folks want to go out and purchase a seperate PC just to handle network proxying. Granted, this isn't nearly as much of an issue as the remote office is, as I can easily get another PC to deal with this there.
Bottom line: I need a solution that proxies non-routable IP addresses to the internet while providing for VPN support to a Checkpoint firewall.
Why would you symlink to things, instead of just putting them in the structure you want them in in the first place?
For NT Workstation I'm kinda doing the same kind of hack that the original poster is talking about. Unlike the 9x flavors of Windows, NT does a pretty fair job of providing a multi-user OS. Perhaps the worst aspect of this is how deep into the Winnt directory you need to go before you actually get to a specific profile.
To get around this, I keep a number of short cuts on my desktop that point to those specific user folders. For example, I've got one called "My Documents" that actually points to the document directory within my profile. When setting up an NT Server, I'll set up short cuts to the short cuts that are in the start menu for the various admin tools, so that service packs and upgrades know where to find the proper path to add their short cuts.
Thing is, even in NT there are certain "expected" directories that just make life easier to leave them be. I have no idea exactly how much havoc you'd cause if you went in and just changed the "System32" directory's name to something else. I sure know it wouldn't be pretty.
I do have to wonder what the true value of having a mutli-user OS is for everyone out there using computers. Heck, even on NT I pretty much use it 100% with "Administrator" (aka root) rights all the time. The multi-user thing starts making a lot more sense to me when you get into talking about servers, and different admins with different rights to it. For workstation use it feels more like a barrier than a feature. The significant difference here being between local and network rights and permissions.
On my Linux set up I pretty much went and did it like I was supposed to. I created a user account with limited rights where I spend the bulk of my time. Thing is, hardly a day goes by where I don't have some reason to "su" my way into control. Not sure if this is a good or bad thing at this point, just an observation. If this were an actual production box I'd probably be a good bit more concerned with the specifics of the setup. As it is, this thing is on a non-routable IP behind a proxy that's on my NT.
I think if we're going to package managers, we should standardize what they do
.tar, .gz, .zip, as well as a whole stack of others. As previously stated, a package manager has a lot more to do than just uncompress, but WinZip is
/.
:)
Whole heartedly agree. When I think "standardizing" I tend to be thinking more along the lines of how HTML compares to Word Docs. By
standardizing HTML it allows for a wide variety of applications to view and edit these files, where as Doc files can't. To carry this analogy
further, I never meant to suggest that standardizing meant that only Netscape should be used as a browser. When standards are done properly
and adopted, like with HTML, it actually expands the number of applications that can be built around it rather than stifling them.
WinZip isn't really a good analogy
The only thing I was attempting to illustrate was WinZip's ability to track an application's installation process. You are quite correct in that it
makes for a poor example to directly compare this to what a package manager does. I don't think I illustrated my meaning well enough there.
You know, for the most part, I think the issue is file interoperability between distributions and/or similar software... if dpkg and rpm could use the
same files (successfully!) or if KDE and gnome could use the same files (for example), we'd be in business.
Maybe this conversation could draw more bad analogies to WinZip. For example, what WinZip actually does is front end several types of
compression routines. Heck, it's only the more recent versions that don't rely on having PKZIP around on the system. It'll recognize
.arc,
a fair example of a front end for several back end processes bringing it all together in a pretty decent UI. Unfortunately, I think the problem is far
more complicated by the fact that each package manager is maintaining their own db without regard to what the other is doing.
As far as the Gnome & KDE issue, I'm pretty much able to run KDE apps under Gnome without too much of a glitch. My biggest complaint has to
do with file associations. Since RH installed Gnome by default, the KDE installation I did later didn't pick up any file associations. I was amazed
that something this basic wasn't handled well below the window manager level. Until such a time comes about where I want to spend many an
hour manually creating associations, I can't really use KDE full time whether I like it or not.
Anyway, thanks for sharing. I do agree that there is a need for some semblance of standards to not push distributions too far away from each
other, but I think we have to be *very* careful about it.
Your quite welcome. And a return thanks for a reply that reads like a well thought out letter. Just goes to show, folks coming from Linux and NT
backgrounds can have civil conversations. Sometimes easy to forget that point after reading a bit too much
It seems like a lot of the windows-y people coming to linux want to see linux look/feel/act like windows without looking into what linux has for
itself, and where it's going on it's own.
Had a phrase where I worked a long while ago I rather liked. "The customer is not always right, but the customer is always the customer." The
notion here is that us folks coming from a Windows GUI perspective have certain expectations from an OS. Sometimes these are good things,
other times not so. Despite the rhetoric, Microsoft really does have some very cool UI ideas built into Windows. To date, they are still the only
company to bring an OS to the masses. There is considerable value to Linux in recognizing this as projects such as Gnome and KDE mature.
Huge Disclaimer: Nothing of what I stated above is meant to suggest that Linux should look to clone it's likeness into Windows. Get the good stuff
out of it, then move on.
The only other part of the article(s) that I couldn't stand was the "even your parents, sister, or girlfriend" could use it gem. Excuse me, but my
sister is pretty damn smart, and it is I who teaches my husband about linux. *snort*
Yeah, I work with my ex on various programming projects still yet. She still runs circles around me when it comes to SQL and XML. She's still
very heavily in the realms of MS, as she's making a very nice living knowing how to muck around with Site Server and ASP.
If I had moderator points that I could use, I'd give you a +1 for your contribution.
Okay, I'm a sucker for flattery
That bash for win32 might be fun to play with. I'll give a look around for it as time permits. I should note that I meant my post to illustrate strengths and weaknesses in both CLI's. For the most part, I still feel more comfortable on NT's CLI. Yes, I fully realize this probably has far more to do with prior experience then which one is actually "better".
One thing that just came to mind concerning a comparison of the two CLI's has to do with the help system involved. After playing with the two, I have to say that NT's help at the CLI level is still better than Linux's.
I realize that Linux tends to provide a LOT of depth about each and every command through the use of "man" or "--help". The problem is, it's a LOT of depth. One example here is a paragraph from the "cp --help" command.
"By default, sparse SOURCE files are detected by a crude heuristic and the
corresponding DEST file is made sparse as well. That is the behavior
selected by --sparse=auto. Specify --sparse=always to create a sparse DEST
file whenever the SOURCE file contains a long enough sequence of zero bytes.
Use --sparse=never to inhibit creation of sparse files."
Here I am, the idiot user who wants a file to go from over here to over there. After looking at this I get to thinking that maybe that file is doing hunky dory right where it is. I've been using a variety of computer platforms for over a decade now, and I still have no idea what a "crude heuristic" is or why I should care. I know Linux is international and all that, but where do I go to pick up the version in English?
NT's help system on the other hand tends to be a lot more straight forward to getting the info I need. DOS used to have a help system that resembled the "man" system quite a bit for it's commands. It was actually pretty cool, and it may still be around in Win9x, but I haven't looked for it.
Just to get the disclaimers out of the way here, I know that a *nix CLI is always going to be at some disadvantage to me due to many years of working with DOS (back to before 2.0) and all its variants. Furthermore, I also realize that many of the *nix commands have a lot more capability built into them. Thing is, I still find it far easier to pick up a DOS style command than a *nix style command. This may be due to the above reasoning, but I also believe that the help system plays a very large factor.
Maybe some of the more basic CLI tools need a couple levels of help. "--helpwalk" and "--helprun" kind of approach. Just a thought.
Last bit of a rant here, as a few weeks ago I located what might have been a really useful "How-To" which got into explaining to someone with a DOS background the equivalent actions on Unix. The thing was written like a propaganda sheet for Unix rather than a tutorial. There didn't seem to be a paragraph in the whole thing that didn't include something along the lines of "...and that is why Unix is so much better than DOS...". To make matters worse, the author had several mistakes in his interpertation of DOS commands and what they could do. Mind you, I don't have a great love for DOS beyond the notion that I do understand how to use it pretty well. I just can't stand reading the work of a zealot, even if I agree with them. How can you concentrate on material if the voice in your head keeps screaming, "Bullshit!"
Maybe some day down the road when I actually get to understanding Unix CLI a bit better I'll get into writing my own How-To for this topic. In the mean time, I'll just keep plugging along.
stty erase to define the backspace key to be some other key... Use with caution!
:(
I'm sure there's some good advice in there, but unfortunately you went zipping past me. I'll be out and about looking to figure out what you're talking about later on to be sure.
rm "This is a single filename with a lot of spaces in it."
Doh! Ya know something, I actually knew that but had forgotten. I mentioned it in my post because I recall having problems with that. Sortta like I STILL can't seem to remember to add that "./" before a file when I'm in the same dir.
In time you will try to highlight right click, click center click even in NT. and it won't work there
There's already tons of things I get messed up on going between OS's these days. Sometimes it's little things, like trying to use Ctrl-C in Netscape under Linux, or trying to drag a window to a virtual desktop that ain't there on NT. I could write up a very lengthy post with all the little oddities that arise when you constantly bounce back and forth between NT and Linux, with just a smattering of Mac once in a while. I'll spare the viewing audience that rant for now.
Here is a good reason. I hate trying to debug scripts where the same program is referred to as Wipeout, WIPEOUT, wipeout, WipeOut and wipeOut interchangeable. It makes it nearly impossible to find every reference to the command wipeout when I am replacing that script with something else.
I can't tell if you're agreeing with me or not here. It seems that this would be an outstanding argument against case sensitivity, as most Windows based editors allow for search and replace without regard to case. This is probably a bit trickier with a regular expression style of s&r, but I never did get a handle on them darn things anyway.
But it doesn't have too! UNIX has had job control for 25 years now. Instead of typing "xview stars.jpg" type "stars.jpg &" and you will run the GUI command in the background. Quite handy and far beyond the capability of Windows9x.
To find out more about job control run the command "man bash" and look for a section titled JOB CONTROL.
Cool! I'll have to play with that later on here.
Linux can institute a number of standards and still allow for a ton of customization. There are a number of areas that should be centralized, to at least agree on what kinds of things are stored where.
1) I mentioned in an earlier post my strong feelings towards standardizing package management. This is a key component to getting software on and off a system. Standardizing this should be able to include compiling from source as well as dealing with binaries. Heck, if WinZip can track where a group of files go why can't Linux?
2) I happen to like KDE for some things, but I end up mostly using Gnome because of how it handles virtual desktops. Bouncing between the two of them there's a couple of things I just don't get. Why can't these two managers actually see eachothers "Start" (for lack of a better term) menu? Why are the file associations mutually exclusive between the two? There's quite a bit that could be done here to standardize things without impacting the unique features each bring to the game.
3) The article did get into specifying some of the problems that lack of certain standards has had. Stuff like having to figure out which version of X, glibc, Gnome, KDE, kernel, etc. you happen to be running. Unless you did the building Linux From Scratch route, you're most likely going to have to scramble about your system to figure this stuff out. For a newbie, their going to have to spend a couple of hours researching how to go about scrambling through their system. I don't know if simply stating what components are to be standard is the solution, but there is definitely a problem here that needs addressed.
4) I remember the issue of distros coming together or fading away from eachother talked about over a year ago. On one hand, if they start coming together too closely they can no longer differentiate themselves from the rest of the pack. Going the other way, we end up down the road that split the Unix world in the past. It's a delicate balancing act that all the distros are playing.
5) I fully agree, "let's be like windows" is a poor way to go about things. Linux should be looking to grab the good stuff that is in Windows and move beyond. I think we're still in the grabbing stage for the most part (when talking about UI) and now is definitely the time to encourage setting the stage for moving beyond.
From an NT user here who actually knows how to get around in NT from the command prompt. Actually pretty good at getting around the entire NT GUI without a mouse as well, but that's another story.
.ini files from a DOS prompt. Definite plus for Linux, but now I just need to figure out where they're all at!
:) Yeah, I know somebody out there has probably already done it, I've just been too lazy to go look. I think I may have 5 too many text editors on here now as it is.
My Impressions of the Linux CLI
1) Love the tab complete!!
2) Nice roll back history. Between sessions even!
3) Why can't the Delete key NOT act like a backspace? ACK!
4) Still haven't figured out how to deal with spaces in a file name or directory.
5) Rather like the hi-light with left, paste with middle mouse button thing. Not 100% sure I prefer it to NT's clipboard.
6) I hate case sensitivity. I have yet to hear a single reasonable argument as to why this is a "feature". pizza != Pizza != PiZZa. Has a logic train like this ever done anything but cause errors?
7) Don't much care for launching GUI apps from the CLI. Locks out that CLI from being used for anything else.
Impressions of NT's CLI
1) Nice roll back history, but it doesn't remember between CLI sessions.
2) "Insert", "Delete", "Home", and "End" do exactly what they are supposed to do here.
3) Copying is very easy (hi-light, press Enter), but pasting is a pain.
4) Spaces in file names and directories not a problem. Supports either "PROGRA~1" or "cd c:\Program Files\".
5) No case sensitivity issues!!
6) Launching GUI apps gives the user control of the CLI after the app has launched.
Windows 98 CLI (aka Win95SR-3, DOS 7.0 Shell, etc.)
1) Run screaming.
2) Don't turn around.
One major strike against Windows in my mind is that f'ing registry. For some apparently morbid reason I keep finding myself missing some of the stuff in Win 3.1, like being able to edit the
CLI programs for NT are all but dead when it comes to development. About the only new stuff I see there is some cheesy driver config. It is nice to see that Linux is still actively working on CLI apps. Now if I could just get a port of "edit.com" on over to Linux!
LOL! How long ya gonna be waiting for Microsoft Flight Simulator to get a Linux version done? Oh wow, why does laughter like this cause tears?
Hmmm, pizza arrived!! WooHoo!
What is pif you ask? I haven't got the foggiest idea.
Sounds like you just asked. I forget what the letters stand for, but this is a carry over from Win 3.1 that is still in use. Probably in Win2k, but not certain. What this file does is store the attributes of a shortcut. Stuff like where the file is stored, what kind of memory to allocate to it, how it will run, and other stuff like that. Under 3.1 you used to have to go into a seperate application to edit this, or tweak it directly with a text editor. With every version after 95 you edit those files by going into the properties of the shortcut. Roughly the equivalent of setting up symlinks in Unix, but with a bunch of extra stuff tossed in so you can actually get them old DOS apps to run under Windows.
Wow, have to go back 6-7 years to appreciate why those files exist. Now imagine how tough it is for folks like myself just now getting accustomed to Unix having to go back better than 20 years to comprehend why certain things are done in the way they are.
Along these lines, I can appreciate where the original poster is coming from. Being brought up from the world of DOS I'm pretty much used to defining whatever darn directory structure I want. I'm just now beginning to get a feel for where key items are in Linux, but I have no idea WHY they need to be there. I am slowly learning it, but I can fully appreciate exactly how intimidating that directory structure looks to a newbie.
The user shouldn't care how the filesystem is laid out so long as he/she knows how to find that document he wrote.
On this point we are in disagreement. I truly believe that the user should be able to understand what the heck is going on with their directory structure. The often mentioned "Program Files" is actually a great idea, though there are devils in them details to be sure. It sure makes a lot more sense than "/usr/bin" and all it's related cousins. Both Windows and Unix have a lot of legacy they need to bring along with them, so I don't expect either platform to get this truly friendly and/or organized any time soon.
Establish some standards that programmers can use to know where to place files.
Believe it or not, Bill Gates is in agreement with you. The problem here goes back to legacy support. Windows to this day has DOS apps to consider, not to mention how they set up Win 3.1 originally. I believe that if MS had the chance to start over from scratch there would be certain folders set aside that would not allow 3rd party files to be added. As I understand it (never played with it myself) BeOS does something very similar to this. BeOS presently enjoys not having to support a huge legacy of apps like MS and Unix presently have to deal with, so they can define these things without breaking anything.
These are not simple changes, but they are necessary to gaming development on the Linux platform.
For all the good points in this post, this one misses the mark. These kinds of things are going to be needed for ANY use of Linux on the desktop. It's these kinds of innovations that Linux should be pointing at as why it is so much better than Windows. Thing is, at this point it isn't.
1) I'm running RH 6.1 here myself, and at first I really loved the updating agent it had. Oh boy, then did that love ever go sour. Their FTP server they provide without cost is never available, and I have yet to get any of the supposed mirrors to work. To make matters worse, as I got more comfortable where to find the actual updates I discover that the stuff from the RH updater is usually more than a couple of versions back.
<rant val="on">
Just a stupid newbie question here, but couldn't an agent be written that utilizes FreshMeat as a backend? FreshMeat has all the latest stuff, summaries of the changes, usually functional links even. Instead of having a system that relies upon the distributor to keep things up to date, it seems that a 3rd party like FreshMeat would be perfect. Wonder what Andover would think of it?
The biggest problem I see with this is that Linux has just too many package managers. I like variety as much as the next fella, but not in this case. I fully realize the religous aspects to this debate, but this is a huge obstacle on the road to Linux ease of use. Why can't the Linux community bring in the best of RedHat's, Debian's, and even BSD's package management and label it a standard?
In the wild world of MS there's 3 or 4 major installer apps that most folks use. On occasion I do see a few that appear to be custom written, but they are the exception. Thing is, they all pretty much function the same. Boring as that may be, it provides a level of consistancy that is presently lacking on Linux. Oh god, then we could go on a while about the fun of uninstalling apps. Where the heck did that "make install" put everything? ACK! This is the kind of stuff that an OS should be keeping track of for me. Of course, it won't ever be able to until there is some level of standardization to application installation.
<rant val="off">
2) From what little I understand about the architecture X it may need to be replaced to support everything you suggest. There are a couple of projects in their infancy now that are attempting to address this. Thing is, they are VERY much in their infancy, and there is also a huge legacy that comes along with X.
3) To my original point, OpenGL at the OS level would have implications that go beyond gaming. I could envision new GUI technologies being developed around this, not to mention advanced support of graphics editing packages. We would certainly see the impact on the games before anything else though.
I've been running a dual monitor set up now under NT for a number of years. Been doing it for a long time at work, and more recently got set up at home to do the same. I've got to say, it is something I definitely miss when trying to work on someone else's PC.
Thing of it is, when I upgraded my NT to this Matrox G400, I moved my old Voodoo3000 into my Linux box. I'm finding that I really love having all these virtual desktops under Linux, even with just one screen.
The dual screen comes in really handy for apps like Photoshop, Imageready, Dreamweaver, Homesite, or darn near anything else that has a lot of floating tools. On the other hand, I've gotten quite used to dragging lots of opened browser windows to different desktops on Linux. Gnome and Sawfish handle this really nicely.
I don't know if there's much of a point to this post, other than I believe that each kind of environment has it's pros and cons. I haven't seen a dual monitor set up for Linux yet. I'd have to imagine that dual head with virtual desktops would get pretty confusing real quick. Sounds like it'd be fun to play with if such a thing were possible though.
Okay, here's a point of view that's been rarely expressed here, but here goes: font antialiasing is a crufty outdated process which isn't really very useful on today's monitors.
Not even close to truth. Whatever performance hit you might take from anti-aliasing fonts, it is absolutely worth it. I'm running an NT Workstation next to this Linux box I'm typing this on now, and there is a marked difference in the font quality that is quite visible. Both are on 17" monitors running 1152x864. And yes, I can quite easily tell when AA is turned off on the NT side, even with true type fonts being used.
In trying to see if Linux could act as my primary desktop I've tried using Abiword and StarOffice to simply write up a document. Every time I try I get about a paragraph into it, then realizing I can't read what I'm writing! I'd rather use a CLI based word processor than one in X, because at least there the text is readable.
Even with TrueType fonts there is a marked difference in visual quality as the size of the font grows. Might explain why the Mac only supports font smoothing for large type fonts. Even still, NT actually presents far clearer fonts than Mac as well, primarily for support for font smoothing down to small fonts.
If Microsoft ever figures out how to move that new font rendering technology of theirs to CRT support, it will raise the bar on visual quality quite a bit. If you haven't seen what they've done with this, you might want to go check it out. Smooth, crisp, clean fonts with no anti-aliasing involved. Only working on LCD based screens at this point though.
Bottom line, some day down the road Linux may work out to be my primary desktop. Aside from missing a few key apps, one of the biggest show stoppers for me is the readability of the fonts. I hope that either a future version of X corrects this, or X gets replaced by something that can. Heck, I'd love to see Linux one up Windows by working in things like kerning and tracking of fonts at the OS level. That'd be an awsome display, and a compelling reason for folks doing print work to seriously consider Linux.