What is MythTV? You claim that it is "DVR". It isn't just that.
The NAME (and its attendant database name "mythconverg") comes from the idea that the "Mythical Convergence" of entertainment media, TV, and the digital computer could be realized.
Why is the version number so low? Consider that, as of this version (0.20), MythTV is capable (at last) of dealing with DVD menus. Earlier versions needed to use Xine (an external player) to play DVDs with menus.
Consider that, as of this version, MythTV is capable (at last) of "natively" recording DVDs. Previously, we needed an export script, and mencoder/dvdrecord (external programs) to accomplish this task.
And that's just handling DVDs. So, the convergence is coming, but its not there yet. And so, not version 1 yet.
In my book, that means DONE. MythTV isn't done; not by a long shot. After all,.18 didn't even support DVD export directly (I had to use a mencoder script). It didn't do audio track backed slideshows from digital photos. It doesn't permit content to be "re-anchored" onto a different backend (or all backends). At least, not directly.
It is, however, a nice package.
But, there is no pressure to call it "Version 1". A change to "MythTV 20060910" may be in order (given that the scope of the project is so large). But that would be confusing.
If you are GOING to have a "Version 1" release, it should be (1) satisfying to the developers, (2) stable, and (3) feature complete for the users.
And after Version 1, where do you go? Version 2 must be a pretty big step. Even Linux hasn't hit Version 3 yet.
Of course you can. The "fingerprint" is being measured -- somehow. So, the characteristic that IS the fingerprint must be resolvable. If its measurable in that sense, it can be generated. How else would the testing gear itself be tested?
Now the COST of the generating gear may be prohibitive, but it certainly MUST be practical.
And, in a game, "normal" rules don't apply. Neither the "Law of Gravity", or "Thou shalt not Kill".
Take Grand Theft Auto.
Anyway, in this game, a Ponzi scheme was possible. And that was exploited. In chess, "en passant capture" can be exploited -- even though some players are not aware of its existence.
If some players feel "cheated", don't play. This will apply pressure to have the rules modified. As to "real world consequence"? Exactly HOW "politically correct" are you? Are you going to regulate my chess and backgammon play next?
The IA does exactly that -- it respects robots.txt. Further, it RETROACTIVELY applies robots.txt. Now, this may not work (which is what the complaint was about). And AFAIK the retroactive edit doesn't remove data, it simply doesn't allow visibility (which is one of the reasons it may not work -- if there are two separate paths to the data, and the data is there, it can still be retrieved).
The devils advocate argument would be that IA may be necessary to retain copyrighted works, but since everything is copyrighted, IA has no right to mirror or serve such content (not until the copyright expires, or under suitable subpoena).
The strange thing is that NO MATTER what choice you make, the software should still install. "Declining" the GPL simply removes the right to further distribute the software.
So, the correct behaviour should be: Present the GPL -> Accept/Decline. If Accept, indicate that distribution is under the terms of the GPL, and continue. If Decline, indicate that distribution is only possible under the terms of the GPL, but use is free, and continue.
In other words, why bother? I can decline the GPL, and still USE the software, legally.
For what its worth (nothing), on Fedora Core 5 (just happen to have it up) --
From the menu bar: "System/Administration/Display". Which is where I would expect it. If you are doing remote administration: "system-config-display".
Although, changing the resolutions manually is easy: change "1280x1024" (or whatever) to what you want. NOT that obscure.
All that aside, there is an expectation that the user SHOULD have some familiarity with the "system" being installed. Installation is only done once, but the configuration of a COMPLETE system is a rather involved subject.
Just let me know when I can install Windows XP or OS X on my PVR machine. (Windows XP needs special drivers for optical disk and network, and these don't fit on a floppy - catch 22: can't use the machine without the drivers, and can't load the drivers). OS X? Its not an "Apple" label machine. Apple doesn't really make a machine that I find suitable (I guess I could buy an Apple, take it out of its case, put it into a Mozart case). Linux (and BSD) "just works" on this machine.
I am one of those luddites that still uses analog cable. Basic analog cable. I won't pay the monthly premium for a digital SD box (and, my provider charges the premium EVEN IF I OWN THE BOX). I don't watch much more than the basic channels (free to air), but I figure that paying $25 per month is better than putting an antenna on my roof. Fair exchange for granting the cable company a right-of-way on my property.
I do have net access through cable; and for this service I signed up early. Early enough that the cable company offered "free converter box". Now, they want me to BUY a converter (and, in exchange, will give me 2mbps higher download rate). No thanks. Besides, they imposed a 60GB monthly cap, so I can't really use the 2mbps anyway.
I have also invested in analog TV cards (two of them) for recording TV programs. If I go digital, I lose the ability to record TV programs. And, every TV I own (three) has an analog tuner -- none offer a digital tuner.
"There are many aspects of creating software that require non-technical administrative personell to handle. I don't remember ever hearing about OSA (Open Source Accountants);) These people cost money. And it's one thing to freely dedicate your time to devlopment, but it's quite another to freely donate large sums of personal money."
Probably because "Accountant" is not a creative process. There are many "Open Source Writers", "Open Source Poets", "Open Source Sculptors", and "Open Source Painters", etc.
"... programmers, who are doing some of the most mind-bending, skillful crafting of any career, get compensated for their work?"
Easy. If I write for myself, its free. If you want to tell me what to do, it costs money. Even if you want me to continue to work on something I did "for free", it'll cost you. I make a very comfortable living as a "programmer", and I am still amazed (every day) that people will pay me for my skills. I would do it for free, but then I would be working on what I wanted.
"Here's the irony in my eyes - right now hardware manufacturers pay Microsoft in order to get a little sticker that says "built for Windows XP".
This doesn't seem right. It seems totally backwards to me."
Wrong. "Hardware" isn't particularly useful (well, I guess you could use it as a boat anchor, a flower pot...
Until software is added...
Software is not "beholden" to hardware. And that's why you see that sticker.
What they were trying to determine was your ability to "think outside the box". The box is the definition of string.
Does that give you a clue?
Try something like this:
for character in string do . . . if uppercase(character) add character to uppercasestring . . . else if space(character) increment spacecount . . . else add character to otherstring print uppcasestring otherstring space x spacecount
Single pass. Sort of redefines "string".
That would be my first solution. And I can think of more... I would ask a question like that of a potential programmer (at a senior level) that is going to be doing design. Just to see how the brain juices work.
I had to think awhile on your reply. Both to what I said, and didn't say.
I didn't say anthing about my opinions on copying DVDs, or downloading movies or TV shows. All I did was comment on the economics of TV and movies.
Now, on to the oranges: are you implying that TV shows and movies "go bad"? I mentioned that some do. Others don't. But there should be no expectation of perpetual profit (or potential profit). With copyright terms set as they are now, there is such an expectation.
On to music. Where I live, we have a "peronal copy provision". In exchange for a levy, I can copy music. In a way, this is worse than a long copyright term, because the levy has the possibility of being forever. But, with the long terms, I welcome it, and want the levy extended to books and movies.
As to downloading movies -- I simply claimed that it is not practical. When copyright was first introduced, the only people who could practically copy books were other publishers; not the consumers. And this is the situation with movies now. It won't always be the case (look at what happened with CDs and MP3 compression). By the time it becomes practical to download movies, I hope that progressive legislation will allow movies under the "personal copy provision". If it doesn't -- I won't download movies.
Except for a SINGLE circumstance -- which is when I download a movie now. If I have a kids DVD (purchased), and I am unable (some DVDs I can't seem to copy) to copy to hard disk, and my kids like the movies, I break the law. I download a copy to put onto my home entertainment unit. Saves wear and tear on the DVD. And, do you know? I don't feel guilty about doing that at all.
b and g performed about the same; wired was more than twice as fast.
I started with a "b" network (SMC Barricade, DWL-810+). This works well, and I still use it, but is not fast enough to move video reliably. I added "g", and tripled the performance. I can now do video over wireless.
However, I replaced the 2dbi antennas with 7dbi units. "n" would be welcome, but I wonder if antenna replacement will be practical. I have also been considering "boosters" to bump power output (to 500mw).
The author provides no insight into this at all. Indeed, his results are hard to interpret:
2057 Kbps must be BITS per second, unless the data has been compressed. If we assume compression, the 802.11g result shows no improvement (the link is working at 11mpbs, perhaps? Due to range, interference, what?) What link quality was achieved? And WHERE ARE THE "n" NUMBERS?
Why not "/Applications/Firefox.app"? Ok. Make it that (I wouldn't, and not because I cast things in a "Linux" way, but in a "Solaris" way. FYI)./Applications resolves (for me) to an automount point./Firefox.app resolves to Solaris/9/x86/Firefox (replace 9 and x86 with 9,10, x86/sparc/sparc64) or Linux/... depending on platform and architecture. Basically,/Applications/Firefox.app doesn't exist until it is referenced, and it is indirected. So, you either have to know to look there for it to be there, or have a proxy for it. Note that your description of removing such "dead" links fails -- specifically the machine containing the program may be temporarily unavailable -- should the icon be removed? The desktop is NOT representative of any modern computing infrastructure.
Your model works for very small installations. It doesn't grow well. Maybe a possible solution is to divide the "big" into many "small" pieces that DO work you want.
But, understand that Redhat, who is driving Gnome (along with SUN and others) is solving the problem from the upper end. In other words, the solutions that will be generated will be alien to you. Maybe a possible solution is to divide up the desktop/systems space into a "direct" GUI model (could KDE take the lead?) and a big-system solution (Gnome for this?).
OS X doesn't really play in the bigger space anyway (its "solutions" just plain don't work for large sets and heterogenous software). I guess the term "average desktop user" is partly to blame here. I don't consider the "home/casual" user to be "average". My "average" is corporate user. If a solution can be found that satisfies corporate user and administrator, can it be deployed for home? Yes, but you seem to be arguing that it is too complicated and a GUI failure to do so.
How would OS X know if a link is removed? Short of maintaining a database of file reference counts (and if its softlinks, it would be a database of file references, paralleling the directory structure). It doesn't. And the desktop icon is simply representative of the application. It isn't the application. Now, OS X has an approach to icons and launchers. Different approach for other systems. As an example, assume that the file system itself is "virtual" in a sense -- things are NOT present unless called for by name (automount - remember I mentioned that?).
That is,/opt/firefox/bin/firefox would not EXIST in your reference space unless you actually referenced it. Now, how is the icon to exist? There is a disconnect between the "desktop metaphor" and a physical file implementation.
Even at home, I have a terrabyte online. I don't want to drill down hundreds of thousands of files, nor do I want the "desktop manager" traversing this data. There needs to be a separation between these for any modern (large, possibly distributed) file system. This approach may have worked with Mac OS 9, but it doesn't play now.
Gnome handles this via XML files for menus. The desktop itself? Still follows the file system (although it is getting away from that). How do we add a desktop icon? Yes, the application can be on the desktop, or in a folder, and it can be launched. Doesn't really help when trying to generate the proper menu descriptor or launcher (with tool-tip help, and integration into Gnome itself).
Or, lets say your new OS X application can handle files of type "video/mp4v-es" (just choosing a real mime type at random). You want your web browser to create a window, and pass that window handle and the file stream to this application when the data type appears in a web page. Ok, how does THAT work? PS. No pre-launching of the application allowed here (just to make my point as to what an INSTALLER should be doing). What browser is it, anyway? You want to "launch" an existing file (which is why you just copied the application to your computer, after all) and have it come up in "edit" mode.
These issues are either (1) taken care of by assuming the existence of a single and standard stack, or (2) by assuming the existence of a "registry" or (3) by using the services of an installer, which isolates the program author from these issues. (or a combination of these things, and note that the registry can be dynamic -- take Palm OS as an example).
The problem with saving "preferences" in the application directory is that it makes the application not shareable. Big no-no there.
However, a statically linked Linux application, x86, built for a 2.4 (2.x) kernel will run on Linux, OpenBSD and Solaris. No rebuild or relink needed. It's not the issue.
And, if the application is statically linked, it can be dragged to your "thumb drive" and just run on another system. Again, why are you bringing this up. It's not the issue.
And, to boot, package management is not an issue here EITHER. Package management is only a problem if shared objects are in use. Not an issue with the class of shareable/copyable applications we are talking about here.
And, just for grins, if you add QEMU (or equivalent) into the mix, you can run statically linked binaries for ARM, SPARC, MIPS, POWERPC or other processors just as easily. This, again, is not the issue.
What is the issue? A well written x86 Linux application is about as portable as you can get. It will run on Linux, BSD, Solaris. It can even be run on PowerPC OS X (using system emulation). It will run on Intel and Sparc. It can be as simple as a "drag and drop" installation.
Anything added on beyond this is "fluff" designed to comfort some users who are aclimated to another way of doing things. The GNU Autoconf approach BROADENS this support, and permits "automatic" system level integrations.
As an example, how are your "desktop icons" and/or "program launcher" going to be done? How is this updated if the program is removed? How are links to document types created? How are extension programs for your browser installed? These are the issues that an installer should address.
The "low-level" issues have been addressed. Shouldn't be a problem -- if they ARE a problem, the software itself can be considered defective.
You presume that the Windows Installer actually works.
Not true.
I purchased a program called "123 DVD Copy" to make copies of some (legal - I own the copyright) DVDs.
The box said "Works on Windows 98... XP". Fresh install of Windows 98: it installs, but complains about missing DLLs (won't run). Windows XP? Installs, runs. Doesn't work -- but that's probably another issue.
There are approaches to installing software "universally" on Linux. Choice #1: statically linked software. This is actually a pretty good one. Choice #2: target common distributions. Again, not bad, just not as universal. Choice #3: GNU Autoconf. My preferred choice for individual workstations.
Graphic installer? Sure, why not: but please give me the choice of WHERE the software goes. Unix/Linux is not a "one size fits all". I may want it in/opt, or/usr/local/bin -- or somewhere else. I may want extra fonts isolated from the X server, or even shared with XFS.
Fixing Windows applications is such a pain that no-one really does it (there are ways, but it is a "black art"). Generally, the solution is "let it install where it wants to be". With Unix/Linux I (as administrator) want the control.
Some examples: SUN java installation is not bad -- it doesn't install stuff in the tree outside of its designated spot. nVidias driver? Sticks stuff (the control panel) where IT wants. Why? To make it "simpler" I guess. Makes it annoying, though. Thank god the driver works... And there are different rules for the distribution itself. Redhat Linux 9 (and RHEL) is contained, Fedora Core 5 isn't (and is arguably worse from an administration standpoint).
There are some basic rules for Unix (Linux):/usr and/opt should not be needed to boot the system (Redhat 9 breaks this rule with the "kudzu" stuff -- this can be disabled). //bin/sbin/etc should not be written (certainly not by an application). Logging should go through standard facilities. Binary applications should be statically linked (especially if written in C++), or the libraries used should be distributed WITH the application -- only the standard C library can be relied on. The application should accomodate automount and GUI remoting (automount means that a directory need not exist until it is requested, which means keyboard entry must be allowed in file open dialogs). Assume that LD_PRELOAD overriding of API entry points is being done. Maybe some more considerations...
If followed, the binary will be "portable" across multiple distributions (at least on the same processor and major OS version). The "installer" can the be trivial -- and can simply be a self-run shell script.
Basically, "Linux Installers" are trying to solve problems that don't really exist.
You are wrong about TV The producer/actors/etc involved in a TV production are paid on the first airing. By the commercial inserts. If the program HAPPENS to be successful and go into syndication, there will be additional payouts, but that is NOT guaranteed. And, in some cases, is rather ludicrous (examples off the top of my head: reruns of Jeopardy, American Idol, or Survivor?).
In other words, some shows MUST have made all the money on the first airing; we can presume that most others do so as well.
TV programs and movies cannot be, as a result, directly compared. "Piracy" cannot really hurt TV, unless the program is pirated BEFORE it is released with its commercials. Indeed, PVRs with commercial skip are a greater "threat" to the TV content producers. Which is why product placement becomes such a big deal. I watch "American Idol" (a guilty pleasure). I rarely (never) watch it "live", but from my PVR, with commercials deleted. However, I know Coke and Ford sponsor the program - product placement. (and, yes, I *am* influenced by the ads.
YMMV Ratboy
"Piracy" can hurt movies; but not to a big extent. Specifically, it is still very costly to download a full-resolution (DVD) quality movie. I compute the typical cost in Toronto to be $5. Given that this is directly comparable to DVD rental, and considerably less convenient (days to download), AND is only DVD (SD) resolution, I don't think the theater experience is really threatened. And that is where the movies should be paid for -- the theatrical release.
The original Apple ][ did everything in software to save a few bucks.
The Macs? Same thing. Unshielded cables, cases... caused me a LOT of grief. Networking over serial? An attempt to save money by not putting ethernet into Apple kit.
Apple stands for interesting (cool) concepts implemented cheaply.
Reasonable software, though. But, its always been a bandaid over inferior hardware.
Slashdot Mirror
What is MythTV? You claim that it is "DVR". It isn't just that.
The NAME (and its attendant database name "mythconverg") comes from the idea that the "Mythical Convergence" of entertainment media, TV, and the digital computer could be realized.
Why is the version number so low? Consider that, as of this version (0.20), MythTV is capable (at last) of dealing with DVD menus. Earlier versions needed to use Xine (an external player) to play DVDs with menus.
Consider that, as of this version, MythTV is capable (at last) of "natively" recording DVDs. Previously, we needed an export script, and mencoder/dvdrecord (external programs) to accomplish this task.
And that's just handling DVDs. So, the convergence is coming, but its not there yet. And so, not version 1 yet.
YMMV
Ratboy.
What is version 1?
.18 didn't even support DVD export directly (I had to use a mencoder script). It didn't do audio track backed slideshows from digital photos. It doesn't permit content to be "re-anchored" onto a different backend (or all backends). At least, not directly.
In my book, that means DONE. MythTV isn't done; not by a long shot. After all,
It is, however, a nice package.
But, there is no pressure to call it "Version 1". A change to "MythTV 20060910" may be in order (given that the scope of the project is so large). But that would be confusing.
If you are GOING to have a "Version 1" release, it should be (1) satisfying to the developers, (2) stable, and (3) feature complete for the users.
And after Version 1, where do you go? Version 2 must be a pretty big step. Even Linux hasn't hit Version 3 yet.
YMMV
Ratboy
I do that too!
And, as I type my password, I hum along:
m i c . . . k e y . m o u s e
YMMV
Ratboy
Of course you can. The "fingerprint" is being measured -- somehow. So, the characteristic that IS the fingerprint must be resolvable. If its measurable in that sense, it can be generated. How else would the testing gear itself be tested?
Now the COST of the generating gear may be prohibitive, but it certainly MUST be practical.
Ratboy
It's a GAME.
And, in a game, "normal" rules don't apply. Neither the "Law of Gravity", or "Thou shalt not Kill".
Take Grand Theft Auto.
Anyway, in this game, a Ponzi scheme was possible. And that was exploited. In chess, "en passant capture" can be exploited -- even though some players are not aware of its existence.
If some players feel "cheated", don't play. This will apply pressure to have the rules modified. As to "real world consequence"? Exactly HOW "politically correct" are you? Are you going to regulate my chess and backgammon play next?
YMMV
Ratboy
Datajack: To whom ae you playing devils advocate?
The IA does exactly that -- it respects robots.txt. Further, it RETROACTIVELY applies robots.txt. Now, this may not work (which is what the complaint was about). And AFAIK the retroactive edit doesn't remove data, it simply doesn't allow visibility (which is one of the reasons it may not work -- if there are two separate paths to the data, and the data is there, it can still be retrieved).
The devils advocate argument would be that IA may be necessary to retain copyrighted works, but since everything is copyrighted, IA has no right to mirror or serve such content (not until the copyright expires, or under suitable subpoena).
Ratboy
YMMV
The strange thing is that NO MATTER what choice you make, the software should still install. "Declining" the GPL simply removes the right to further distribute the software.
So, the correct behaviour should be: Present the GPL -> Accept/Decline. If Accept, indicate that distribution is under the terms of the GPL, and continue. If Decline, indicate that distribution is only possible under the terms of the GPL, but use is free, and continue.
In other words, why bother? I can decline the GPL, and still USE the software, legally.
YMMV
Ratboy
So, complain to SuSe. Its the enterprise edition.
Don't paint "Linux" with a negative over this.
For what its worth (nothing), on Fedora Core 5 (just happen to have it up) --
From the menu bar: "System/Administration/Display". Which is where I would expect it. If you are doing remote administration: "system-config-display".
Although, changing the resolutions manually is easy: change "1280x1024" (or whatever) to what you want. NOT that obscure.
All that aside, there is an expectation that the user SHOULD have some familiarity with the "system" being installed. Installation is only done once, but the configuration of a COMPLETE system is a rather involved subject.
Just let me know when I can install Windows XP or OS X on my PVR machine. (Windows XP needs special drivers for optical disk and network, and these don't fit on a floppy - catch 22: can't use the machine without the drivers, and can't load the drivers). OS X? Its not an "Apple" label machine. Apple doesn't really make a machine that I find suitable (I guess I could buy an Apple, take it out of its case, put it into a Mozart case). Linux (and BSD) "just works" on this machine.
YMMV
Ratboy
Huh.
I am one of those luddites that still uses analog cable. Basic analog cable. I won't pay the monthly premium for a digital SD box (and, my provider charges the premium EVEN IF I OWN THE BOX). I don't watch much more than the basic channels (free to air), but I figure that paying $25 per month is better than putting an antenna on my roof. Fair exchange for granting the cable company a right-of-way on my property.
I do have net access through cable; and for this service I signed up early. Early enough that the cable company offered "free converter box". Now, they want me to BUY a converter (and, in exchange, will give me 2mbps higher download rate). No thanks. Besides, they imposed a 60GB monthly cap, so I can't really use the 2mbps anyway.
I have also invested in analog TV cards (two of them) for recording TV programs. If I go digital, I lose the ability to record TV programs. And, every TV I own (three) has an analog tuner -- none offer a digital tuner.
YMMV
Ratboy
"There are many aspects of creating software that require non-technical administrative personell to handle. I don't remember ever hearing about OSA (Open Source Accountants) ;) These people cost money. And it's one thing to freely dedicate your time to devlopment, but it's quite another to freely donate large sums of personal money."
Probably because "Accountant" is not a creative process. There are many "Open Source Writers", "Open Source Poets", "Open Source Sculptors", and "Open Source Painters", etc.
"... programmers, who are doing some of the most mind-bending, skillful crafting of any career, get compensated for their work?"
Easy. If I write for myself, its free. If you want to tell me what to do, it costs money. Even if you want me to continue to work on something I did "for free", it'll cost you. I make a very comfortable living as a "programmer", and I am still amazed (every day) that people will pay me for my skills. I would do it for free, but then I would be working on what I wanted.
"Here's the irony in my eyes - right now hardware manufacturers pay Microsoft in order to get a little sticker that says "built for Windows XP".
This doesn't seem right. It seems totally backwards to me."
Wrong. "Hardware" isn't particularly useful (well, I guess you could use it as a boat anchor, a flower pot...
Until software is added...
Software is not "beholden" to hardware. And that's why you see that sticker.
YMMV
Ratboy
Rearrange in one pass:
What they were trying to determine was your ability to "think outside the box". The box is the definition of string.
Does that give you a clue?
Try something like this:
for character in string do
. . . if uppercase(character) add character to uppercasestring
. . . else if space(character) increment spacecount
. . . else add character to otherstring
print uppcasestring otherstring space x spacecount
Single pass. Sort of redefines "string".
That would be my first solution. And I can think of more...
I would ask a question like that of a potential programmer (at a senior level) that is going to be doing design. Just to see how the brain juices work.
YMMV
Ratboy
Really.
Take the NDA (form document). Strike out, and initial, the clauses that are offensive to you. Sign and return.
If they don't "notice", it's good. And, a lot of the time, they won't.
I struck the "prior IP" and "stock trading blackout" clauses from a contract. It went through.
If it doesn't, the updated contract becomes a point of discussion.
Ratboy.
donutello
I had to think awhile on your reply. Both to what I said, and didn't say.
I didn't say anthing about my opinions on copying DVDs, or downloading movies or TV shows. All I did was comment on the economics of TV and movies.
Now, on to the oranges: are you implying that TV shows and movies "go bad"? I mentioned that some do. Others don't. But there should be no expectation of perpetual profit (or potential profit). With copyright terms set as they are now, there is such an expectation.
On to music. Where I live, we have a "peronal copy provision". In exchange for a levy, I can copy music. In a way, this is worse than a long copyright term, because the levy has the possibility of being forever. But, with the long terms, I welcome it, and want the levy extended to books and movies.
As to downloading movies -- I simply claimed that it is not practical. When copyright was first introduced, the only people who could practically copy books were other publishers; not the consumers. And this is the situation with movies now. It won't always be the case (look at what happened with CDs and MP3 compression). By the time it becomes practical to download movies, I hope that progressive legislation will allow movies under the "personal copy provision". If it doesn't -- I won't download movies.
Except for a SINGLE circumstance -- which is when I download a movie now. If I have a kids DVD (purchased), and I am unable (some DVDs I can't seem to copy) to copy to hard disk, and my kids like the movies, I break the law. I download a copy to put onto my home entertainment unit. Saves wear and tear on the DVD. And, do you know? I don't feel guilty about doing that at all.
Just sayin'
YMMV
Ratboy
The author tried b, g, and wired.
b and g performed about the same; wired was more than twice as fast.
I started with a "b" network (SMC Barricade, DWL-810+). This works well, and I still use it, but is not fast enough to move video reliably. I added "g", and tripled the performance. I can now do video over wireless.
However, I replaced the 2dbi antennas with 7dbi units. "n" would be welcome, but I wonder if antenna replacement will be practical. I have also been considering "boosters" to bump power output (to 500mw).
The author provides no insight into this at all. Indeed, his results are hard to interpret:
802.11b 802.11g 100 Base-T
2057 Kbps 2369 Kbps 5042 Kbps
2057 Kbps must be BITS per second, unless the data has been compressed. If we assume compression, the 802.11g result shows no improvement (the link is working at 11mpbs, perhaps? Due to range, interference, what?) What link quality was achieved? And WHERE ARE THE "n" NUMBERS?
All in all, a very poor article.
YMMV
Ratboy
Why not "/Applications/Firefox.app"? Ok. Make it that (I wouldn't, and not because I cast things in a "Linux" way, but in a "Solaris" way. FYI). /Applications resolves (for me) to an automount point. /Firefox.app resolves to Solaris/9/x86/Firefox (replace 9 and x86 with 9,10, x86/sparc/sparc64) or Linux/... depending on platform and architecture. Basically, /Applications/Firefox.app doesn't exist until it is referenced, and it is indirected. So, you either have to know to look there for it to be there, or have a proxy for it. Note that your description of removing such "dead" links fails -- specifically the machine containing the program may be temporarily unavailable -- should the icon be removed? The desktop is NOT representative of any modern computing infrastructure.
Your model works for very small installations. It doesn't grow well. Maybe a possible solution is to divide the "big" into many "small" pieces that DO work you want.
But, understand that Redhat, who is driving Gnome (along with SUN and others) is solving the problem from the upper end. In other words, the solutions that will be generated will be alien to you. Maybe a possible solution is to divide up the desktop/systems space into a "direct" GUI model (could KDE take the lead?) and a big-system solution (Gnome for this?).
OS X doesn't really play in the bigger space anyway (its "solutions" just plain don't work for large sets and heterogenous software). I guess the term "average desktop user" is partly to blame here. I don't consider the "home/casual" user to be "average". My "average" is corporate user. If a solution can be found that satisfies corporate user and administrator, can it be deployed for home? Yes, but you seem to be arguing that it is too complicated and a GUI failure to do so.
Divide the space then.
YMMV
Ratboy
The application is the desktop icon? Or a link?
/opt/firefox/bin/firefox would not EXIST in your reference space unless you actually referenced it. Now, how is the icon to exist? There is a disconnect between the "desktop metaphor" and a physical file implementation.
How would OS X know if a link is removed? Short of maintaining a database of file reference counts (and if its softlinks, it would be a database of file references, paralleling the directory structure). It doesn't. And the desktop icon is simply representative of the application. It isn't the application. Now, OS X has an approach to icons and launchers. Different approach for other systems. As an example, assume that the file system itself is "virtual" in a sense -- things are NOT present unless called for by name (automount - remember I mentioned that?).
That is,
Even at home, I have a terrabyte online. I don't want to drill down hundreds of thousands of files, nor do I want the "desktop manager" traversing this data. There needs to be a separation between these for any modern (large, possibly distributed) file system. This approach may have worked with Mac OS 9, but it doesn't play now.
Gnome handles this via XML files for menus. The desktop itself? Still follows the file system (although it is getting away from that). How do we add a desktop icon? Yes, the application can be on the desktop, or in a folder, and it can be launched. Doesn't really help when trying to generate the proper menu descriptor or launcher (with tool-tip help, and integration into Gnome itself).
Or, lets say your new OS X application can handle files of type "video/mp4v-es" (just choosing a real mime type at random). You want your web browser to create a window, and pass that window handle and the file stream to this application when the data type appears in a web page. Ok, how does THAT work? PS. No pre-launching of the application allowed here (just to make my point as to what an INSTALLER should be doing). What browser is it, anyway? You want to "launch" an existing file (which is why you just copied the application to your computer, after all) and have it come up in "edit" mode.
These issues are either (1) taken care of by assuming the existence of a single and standard stack, or (2) by assuming the existence of a "registry" or (3) by using the services of an installer, which isolates the program author from these issues.
(or a combination of these things, and note that the registry can be dynamic -- take Palm OS as an example).
How do you want it?
YMMV
Ratboy
The problem with saving "preferences" in the application directory is that it makes the application not shareable. Big no-no there.
However, a statically linked Linux application, x86, built for a 2.4 (2.x) kernel will run on Linux, OpenBSD and Solaris. No rebuild or relink needed. It's not the issue.
And, if the application is statically linked, it can be dragged to your "thumb drive" and just run on another system. Again, why are you bringing this up. It's not the issue.
And, to boot, package management is not an issue here EITHER. Package management is only a problem if shared objects are in use. Not an issue with the class of shareable/copyable applications we are talking about here.
And, just for grins, if you add QEMU (or equivalent) into the mix, you can run statically linked binaries for ARM, SPARC, MIPS, POWERPC or other processors just as easily. This, again, is not the issue.
What is the issue? A well written x86 Linux application is about as portable as you can get. It will run on Linux, BSD, Solaris. It can even be run on PowerPC OS X (using system emulation). It will run on Intel and Sparc. It can be as simple as a "drag and drop" installation.
Anything added on beyond this is "fluff" designed to comfort some users who are aclimated to another way of doing things. The GNU Autoconf approach BROADENS this support, and permits "automatic" system level integrations.
As an example, how are your "desktop icons" and/or "program launcher" going to be done? How is this updated if the program is removed? How are links to document types created? How are extension programs for your browser installed? These are the issues that an installer should address.
The "low-level" issues have been addressed. Shouldn't be a problem -- if they ARE a problem, the software itself can be considered defective.
YMMV
Ratboy
You presume that the Windows Installer actually works.
... XP". Fresh install of Windows 98: it installs, but complains about missing DLLs (won't run). Windows XP? Installs, runs. Doesn't work -- but that's probably another issue.
/opt, or /usr/local/bin -- or somewhere else. I may want extra fonts isolated from the X server, or even shared with XFS.
/usr and /opt should not be needed to boot the system (Redhat 9 breaks this rule with the "kudzu" stuff -- this can be disabled). / /bin /sbin /etc should not be written (certainly not by an application). Logging should go through standard facilities. Binary applications should be statically linked (especially if written in C++), or the libraries used should be distributed WITH the application -- only the standard C library can be relied on. The application should accomodate automount and GUI remoting (automount means that a directory need not exist until it is requested, which means keyboard entry must be allowed in file open dialogs). Assume that LD_PRELOAD overriding of API entry points is being done. Maybe some more considerations...
Not true.
I purchased a program called "123 DVD Copy" to make copies of some (legal - I own the copyright) DVDs.
The box said "Works on Windows 98
There are approaches to installing software "universally" on Linux. Choice #1: statically linked software. This is actually a pretty good one. Choice #2: target common distributions. Again, not bad, just not as universal. Choice #3: GNU Autoconf. My preferred choice for individual workstations.
Graphic installer? Sure, why not: but please give me the choice of WHERE the software goes. Unix/Linux is not a "one size fits all". I may want it in
Fixing Windows applications is such a pain that no-one really does it (there are ways, but it is a "black art"). Generally, the solution is "let it install where it wants to be". With Unix/Linux I (as administrator) want the control.
Some examples: SUN java installation is not bad -- it doesn't install stuff in the tree outside of its designated spot. nVidias driver? Sticks stuff (the control panel) where IT wants. Why? To make it "simpler" I guess. Makes it annoying, though. Thank god the driver works... And there are different rules for the distribution itself. Redhat Linux 9 (and RHEL) is contained, Fedora Core 5 isn't (and is arguably worse from an administration standpoint).
There are some basic rules for Unix (Linux):
If followed, the binary will be "portable" across multiple distributions (at least on the same processor and major OS version). The "installer" can the be trivial -- and can simply be a self-run shell script.
Basically, "Linux Installers" are trying to solve problems that don't really exist.
YMMV
Ratboy
What is the speed of an OS? How fast is a VM that is just an OS? How do you think running ClearCase under a VM to your VPN WAN will improve anything?
Just Curious
Ratboy
You are wrong about TV The producer/actors/etc involved in a TV production are paid on the first airing. By the commercial inserts. If the program HAPPENS to be successful and go into syndication, there will be additional payouts, but that is NOT guaranteed. And, in some cases, is rather ludicrous (examples off the top of my head: reruns of Jeopardy, American Idol, or Survivor?).
In other words, some shows MUST have made all the money on the first airing; we can presume that most others do so as well.
TV programs and movies cannot be, as a result, directly compared. "Piracy" cannot really hurt TV, unless the program is pirated BEFORE it is released with its commercials. Indeed, PVRs with commercial skip are a greater "threat" to the TV content producers. Which is why product placement becomes such a big deal. I watch "American Idol" (a guilty pleasure). I rarely (never) watch it "live", but from my PVR, with commercials deleted. However, I know Coke and Ford sponsor the program - product placement. (and, yes, I *am* influenced by the ads.
YMMV
Ratboy
"Piracy" can hurt movies; but not to a big extent. Specifically, it is still very costly to download a full-resolution (DVD) quality movie. I compute the typical cost in Toronto to be $5. Given that this is directly comparable to DVD rental, and considerably less convenient (days to download), AND is only DVD (SD) resolution, I don't think the theater experience is really threatened. And that is where the movies should be paid for -- the theatrical release.
Not only that; please mention that we supplied the software in SOURCE format. With a theory of operation manual.
And, it was portable -- to Z80 (Radio Shack) and Commodore PET (better graphics).
The same company (Personal Software) marketed (didn't write, that was DanB), VisiCalc (but it did NAME VisiCalc).
And, produced "VisiOn", the environment that inspired Microsoft to produce Windows.
Back in the day...
Ratboy
done stuff like this.
The original Apple ][ did everything in software to save a few bucks.
The Macs? Same thing. Unshielded cables, cases... caused me a LOT of grief. Networking over serial? An attempt to save money by not putting ethernet into Apple kit.
Apple stands for interesting (cool) concepts implemented cheaply.
Reasonable software, though. But, its always been a bandaid over inferior hardware.
Ratboy
Sure... I'll take your challenge...
Mac OS X - doesn't install at all (no Apple hardware).
XP - installs, no sound, network, or CD access.
FreeBSD - actually installs and does something.
Don't need the other 19 people, I guess. And, why IS installing taken to be a Rosetta Stone, anyway?
Ratboy
But most business use doesn't need "Flash", "codecs" or the ability to play DVDs. Really.
Grey area - a business with a suitable (>10,000 seat) install base could negotiate royalties themselves. (if they REALLY have a need for MP3 or DVD).
Just sayin
Ratboy.
No
But I question your premise.
Ratboy