I can't tell you how many potential Linux/UNIX users I know that have told me they're waiting for something like buttons that disappear in a puff of smoke and, until Linux has that, they'll stick with Windows.
... prohibitions against deep-linking are permitted by the DMCA, right?
Is there an analysis of their reasoning for this policy anywhere? (I'm assuming that reason played a part in its being put into place. And, yes, I know all about ``assuming''.) I'm guessing that they're finding out that brokering cheap airfares isn't the huge moneymaker that it used to be and they're ticked off that deep-linking is eating into the advertising views that they get when you come in only through their home page. My heart (sniff) bleeds for them.
``However, you can mainly hear it on Jean Michel Jarre's Oxygene and Pink Floyd's Echoes songs.''
Of course, if you want to go even farther back, you could try tracking down a copy of a Lothar and the Hand People. ("Lothar" was the Theremin.) Great stuff.
I once worked with a bunch of pilots when GPS was still in its infancy. I never heard any of them, nor have I heard of any to this day, ever say they were interested in, say, getting rid of their VOR receiver once they ever put a GPS receiver in their plane. Why make mariners navigate without a backup system? I can't imagine that they're in favor of this. Can't imagine the companies that offer insurance are crazy about eliminating the lighthouse system either.
I was thinking the same thing. Why would I want this in the kernel? (At least in the kernel proper.) It's an application. How would this benefit the desktop user? How would this benefit the embedded user? If it doesn't benefit all of these users, then it probably (IMHO) doesn't belong in the kernel. It does sound like it might make a dynamite module that someone might want to load if they have a need to schlepp around lots of data between databases (for example).
Please put your coffee down while reading the "Using your OS" section. Failure to do so may result in damage to your keyboard. This is especially true when you read the bullet points:
``Windows XP was the only operating system that couldn't recognise and open an imported Excel file...''
and
``Windows XP and Mac OS X both required drivers in order to set up our printer; the Linux distributions included the drivers.''
Remember all that talk a while back about Linux having poor driver support? (OK, that may be true to a certain extent if you're a hardcore gamer, but then you guys will still have trouble finding a lot of games that run natively on Linux anyway.)
Never had to/got to use it. A company I worked for back in the early '80s used to use a FORTRAN preprocessor called IFTRAN. It allowed you to write something like FORTRAN-77 with extra constructs; I recall that it allowed DO-WHILE, DO-UNTIL, and the like.
The output was FORTRAN-IV that was run through the compiler. If you were into squeezing more performance out of the code, you kept the intermediate code and cleaned it up. See, the odd thing about this preprocessor is that it converted all of its coding structures into standard IF-GOTOs and a ton of numbered CONTINUE statements. Which wouldn't really be so bad but the PDP FORTRAN-IV compiler turned CONTINUEs into a NOOP instructions. For programs with highly nested loops and ifs structures, you could have a whole lot of long sequences of NOOPs throughout the resulting compiled code. After a while, I got to using some scripts to find and replace a long string of CONTINUEs with a single one and finding and replacing the line numbers throughout the code for the CONTINUEs that were being removed.
Lest one think this was a waste, the effect on the execution of the program was appreciable. Back then, CPU time was getting charged back to the department. It was all funny money but the managers took it all so very seriously. Then we all got a PC on our desk and CPU time was, essentially, free from then on. And then the attitude was "Heh, we'll just let the program run all night".
Ah... but COMMON was only nasty (IMNSHO) when you depended on unnamed COMMON blocks. By using named COMMON you could restrict what subroutines had access to what variables. With a little care, you could avoid the general confusion that resulted when folks got lazy and stashed everything in an unnamed COMMON block. Debugging one of these programs could be was particularly nasty and I tended to frown a lot when people handed me a listing looking for my help and I noticed in the compiler listing that nearly all the variables were listed as being in unnamed COMMON.
And on some OSes (RSX-11M-PLUS was one of these), you passed data between separate processes via what looked to FORTRAN as a named COMMON block. Back in my FORTRAN days, I had written some large simulations that ran on a big RSX system where a couple of dozen different programs passed data among themselves via these chunks of shared memory. It was actually a piece that you wrote in MACRO-11 and loaded into memory before starting the FORTRAN tasks. I'm trying to remember what is was called now. Global section or was that the VMS equivalent? Darned cobwebs obscuring things and my RSX Executive reference is buried down in the basement.:-)
The six-character variable name makes me think that the code was written for a pretty old compiler. I recall the IBM FORTG and FORTH compilers having this limit. I remember finding the switch to DEC machines and using their compilers quite liberating. (Not to mention I didn't have to deal with XEDIT any more.)
As for having to deal with someone's really old software... I know what you're going through. An office mate left behind some code that we were to use in a later project. While he was working on it, he'd asked my advice on how to make it easier to manage as he'd just been plugging away and the thing grew to be a spaghetti code monster filled with GOTOs. He slaved away on it and was pretty satisfied with himself. When he left and I got the listings to begin on the followup project. I found that all he'd really accomplished was to pull out about a page and a half that became the main program. The remainder of the 27 pages or so became one giant subroutine. My job was to pull the guts of this hideous thing out and run it on a PDP-11. Finally got it to work but it took just a little doing. Ever since I've really loathed having to work on someone else's code. In the end, though, it does feel sort of good getting that old rat's nest of code cleaned up and the needed changes incorporated.
A recent "stupid code comment" story: I recently inherited a shell script from a departing consultant that contained the following:
# Add one to count
count=`expr $count + 1`
The worst of it, though, was that a few lines later he wrote:
if [ count -le 5 ]; then
Now I get to tell his former employers that, maybe we should go over all the stuff he wrote.
Speaking of silly "increment counter" comments... Does anyone remember when SuperSoft's FORTRAN compiler for PC-DOS came out? You know, the one with the INCREMENT statement? (Because, you know, having to code "I = I + 1" is just so difficult.)
I jumped for joy when I moved out of the IBM supermini arena and moved to VAXes. Being able to use FORTRAN variables longer than 6 characters, allowed underscores in variables, actual tabs for indenting, actual CHARACTER arrays (no more using EQUIVALANCEs to LOGICAL*1 arrays), etc. It was like going to heaven (well, at the time it seemed so). Even though the older IBM code would still compile and run, I found myself "porting" much of it to VAX FORTRAN as I was making mods to it just to make is easier to read.
The article's comments on whitespace were funny to. Contrary to the author, though, I do prefer that there be some sort of style manual for programmers to follow. Years ago, a coworker of mine wrote a ton of FORTRAN using statements like the following:
DO10I=1,100
just because the compiler would accept it (and he claimed he was saving disk space, as well). And he refused to use CONTINUE statements to terminate his DO loops. Let me tell ya, it was no picnic reading through his code and find out where loops started and ended.
It wasn't uncommon to find that he had three or four DO loops all ending on the same statement.
``Actually I don't know that line numbers were ever required, just often put in so you could re-assemble your cards if you dropped them.''
If anyone used them to aid in reordering their dropped card deck, they were nuts.
I remember the IBM editors (this back in the early '80s before I shifted over to DEC systems) would put the sequence numbers into columns 73-80 automatically. And that was past the days when anyone actually used cards. The IBM editors could use those numbers for commands to insert/delete lines and some of the system tools would let you modify files (sort of like "patch") by inserting into or deleting from a source file based on those sequence numbers.
Once we found out that they weren't required, though, we found ways to prevent the editor from inserting them or ran CMS EXECs on the source files to strip them off just so that compressing our source files would save even more disk space. (An important thing to know back when a typical disk allocation on an IBM supermini was something like a couple of megabytes.)
... as ``The greatest boon to transportation since the wheel''. So will read the joint press release from the Concrete Industry and the Rebar Manufacturers Association.
I'd surely love to see how they plan on getting this monstrosity to pass muster with just about every environmental regulation on the books. What am I thinking? Why they have a former governor who will all-too happily waive those regulations!
True-to-form Microsoft. Years ago, Microsoft (IIRC, it was in a Bill Gates speech/interview) actually claimed that Windows was an Open System. The reason? Because you could run it on a PC from any IBM-compatible manufacturer, of course. Sure made sense to me. (Gag!)
... is that most of the people who are even aware of this class action lawsuit are thinking ``Oh, great! Just what I need: a check for $50 or voucher from Microsoft for some more of their buggy software''. At that point they decide that the suit isn't worth the effort to track down the receipt for the PC/software and, from there on, their brain's ``waste of my time'' filter kicks in whenever it's mentioned.
Now if it turned out that you'd get a full refund of the retail price of your MS software plus, oh, $100-$200 for every time your PC had to be rebuilt from scratch due to the swiss cheese nature of their OS/application security, then it starts becoming worthwhile to join in on the suit.
Ah, but what would you give just to have even LAVC-like clustering available for Linux? Is there anything better than the silly failover-to-a-hot-spare style clustering available for Linux?
``It's peer to peer disk sharing with all the lockin problems sorted.''
Heh, heh. Anyone ever tell you that you have a gift for understatement?
``Dec's AdvFS is not really any better than Veritas, except that it's so nicely integrated with Truclusters. I don't know how well Veritas' clustered filesystem works, but it runs on solaris and linux. Thus you can run both linux and hp-ux on vpars within the same hp server, and share data. Though I really liked trucluster/cfs, it would only be really helpful if they ported to both linux and hp-ux.
Appart from making the Tru64 -> HP-UX transition harder, I don't see that they lose any features by picking veritas over CFS.''
Imagine how many more people get to enjoy the finger pointing that'll take place when their database vendor, OS vendor, and clustering vendor decide it's the other guys' fault that the database is toast because of some oddball bug. It's annoying enough having to deal with just two of them. I can't wait until I have to have a four-way conference call to get complex systems problems solved. Oops! Forgot! Toss in a hardware vendor for good measure.
Heck it was bad enough when we had these sort of troubles 20 years ago when the DEC HW and SW folks pointed fingers at each other but at least then you could get a hold of a manager to make the two sit down and solve the problem. What option do you have when all these players work for different companies? Who do you call? Legal? (That'll have you measuring downtime with a calendar.)
I haven't forgotten about that, young feller. It's the first thing that came to mind, in fact. I think the Datamation web site (whew! does that even exist any more) had it posted for a while.
You guys got crossed off my list of potential replacement printers as a result of your using the DMCA to try to lock out toner cartridge competition. Now this? Has anyone at Lexmark even heard of public relations let alone what it means?
Invest in a UPS for this system. Any uncontrolled shutdown (i.e. power outage) will take quite a long time to fully recover from. It will take a lo-o-o-ong time for the resync to finish on an 8x250GB raidset. Then you really ought to force an fsck as well on the filesystems you've created on that monster.
``Does SCO even PRODUCE anything nowadays, other than FUD and lawsuits?''
Well, they have at least a few technical people left. Otherwise, who'd be bundling the new version of Samba and other OSS packages with their crummy UNIX?
``Claiming you can't find source code is one of the most rediculous things I've ever heard.''
Not that it's much of an excuse, but didn't that argument work nicely in at least one trial that Microsoft was involved in? (Well at least I recall them trying that one. Dont't remember how well it worked.)
``Does he think that cheaper hardware will make copying software harder to do?''
Nah. My theory is that Ballmer sees their problem as people are spending money on hardware that Ballmer thinks they ought to be spending on software. Every dollar that Intel/Dell/HP/etc makes is a dollar that Microsoft didn't make. No one should be expecting Microsoft to be lowering their prices in lockstep with any price reductions that the hardware vendors put into place.
What Microsoft is hoping for is a situation where cheaper computers are in the hands of the masses resulting there will be an even bigger demand for software. What they don't seem to realize is that, for poorer countries, that could easily mean an even bigger demand for counterfeited software. Or the folks in those countries will just load a free OS on those computers. Microsoft would like very much to prevent that from happening. (Good luck with that.) The trouble is that someone from a poor country has a small set of choices. Say they've got $500 to spend. They'd like a computer (perhaps so they can become computer literate and work for one of those outsourcing companies):
Option 1: They look at a computer with Windows+Office. It costs $500 which includes a $100 computer and $400 worth of Microsoft software.
Option 2: They can get a $100 computer with $40 worth of a commercial Linux distribution and have an extra $360 left over to feed their family.
Option 3: Get the $100 computer and borrow Linux from the guy down the street. Now he's got $400 more for food.
Option 4: Get the $100 computer, use a ripped off copy of Windows. He's still got the extra $400 for groceries. And if the buyer is a struggling small business, a chance to win a visit by the local police and a representative of the BSA. Oddly enough, this can happen if they choose Option 1. (What an incentive to go with Microsoft. Where do I sign up?:-) )
Ballmer seems to think that people will choose Option 1. I would contend that Option 1 is the last choice people will make. (Me personally? I'd go with Option 2 since it allows me to support the OSS "industry".)
Some day they'll understand that the world is not always looking for ways to send their hard-earned money to Microsoft.
IMHO, it was when the Internet was opened up for commercial purposes when all the crud started flowing. Maybe, when businesses see it as unsuitable for their mass marketing needs and go away, the 'net will start to function properly again. Perhaps raising the cost of admission to the ".com" domains might work to keep out some of the garbage that's clogging up the works. Who knows? But something's got to happen. Thirty-six emails in the Inbox this morning at work and only one of them was something I asked to receive or from someone that I know. I'm getting a bit sick of having to go through all the junk. It's way too much like the daily ritual of sifting through all the junk snail mail that I get at home.
Slashdot Mirror
I can't tell you how many potential Linux/UNIX users I know that have told me they're waiting for something like buttons that disappear in a puff of smoke and, until Linux has that, they'll stick with Windows.
Darn it! I forgot the "sarcasm" tags.
... prohibitions against deep-linking are permitted by the DMCA, right?
Is there an analysis of their reasoning for this policy anywhere? (I'm assuming that reason played a part in its being put into place. And, yes, I know all about ``assuming''.) I'm guessing that they're finding out that brokering cheap airfares isn't the huge moneymaker that it used to be and they're ticked off that deep-linking is eating into the advertising views that they get when you come in only through their home page. My heart (sniff) bleeds for them.
Which is important because, as everyone knows, security is a product. And now Microsoft offers more security than other vendors.
(Now excuse me while I go and gouge out my eyes after having read this article.)
Of course, if you want to go even farther back, you could try tracking down a copy of a Lothar and the Hand People. ("Lothar" was the Theremin.) Great stuff.
I once worked with a bunch of pilots when GPS was still in its infancy. I never heard any of them, nor have I heard of any to this day, ever say they were interested in, say, getting rid of their VOR receiver once they ever put a GPS receiver in their plane. Why make mariners navigate without a backup system? I can't imagine that they're in favor of this. Can't imagine the companies that offer insurance are crazy about eliminating the lighthouse system either.
I was thinking the same thing. Why would I want this in the kernel? (At least in the kernel proper.) It's an application. How would this benefit the desktop user? How would this benefit the embedded user? If it doesn't benefit all of these users, then it probably (IMHO) doesn't belong in the kernel. It does sound like it might make a dynamite module that someone might want to load if they have a need to schlepp around lots of data between databases (for example).
There's an episode of Futurama that points out what a huge mistake that would be.
Please put your coffee down while reading the "Using your OS" section. Failure to do so may result in damage to your keyboard. This is especially true when you read the bullet points:
and
Remember all that talk a while back about Linux having poor driver support? (OK, that may be true to a certain extent if you're a hardcore gamer, but then you guys will still have trouble finding a lot of games that run natively on Linux anyway.)
Never had to/got to use it. A company I worked for back in the early '80s used to use a FORTRAN preprocessor called IFTRAN. It allowed you to write something like FORTRAN-77 with extra constructs; I recall that it allowed DO-WHILE, DO-UNTIL, and the like.
The output was FORTRAN-IV that was run through the compiler. If you were into squeezing more performance out of the code, you kept the intermediate code and cleaned it up. See, the odd thing about this preprocessor is that it converted all of its coding structures into standard IF-GOTOs and a ton of numbered CONTINUE statements. Which wouldn't really be so bad but the PDP FORTRAN-IV compiler turned CONTINUEs into a NOOP instructions. For programs with highly nested loops and ifs structures, you could have a whole lot of long sequences of NOOPs throughout the resulting compiled code. After a while, I got to using some scripts to find and replace a long string of CONTINUEs with a single one and finding and replacing the line numbers throughout the code for the CONTINUEs that were being removed.
Lest one think this was a waste, the effect on the execution of the program was appreciable. Back then, CPU time was getting charged back to the department. It was all funny money but the managers took it all so very seriously. Then we all got a PC on our desk and CPU time was, essentially, free from then on. And then the attitude was "Heh, we'll just let the program run all night".
Ah... but COMMON was only nasty (IMNSHO) when you depended on unnamed COMMON blocks. By using named COMMON you could restrict what subroutines had access to what variables. With a little care, you could avoid the general confusion that resulted when folks got lazy and stashed everything in an unnamed COMMON block. Debugging one of these programs could be was particularly nasty and I tended to frown a lot when people handed me a listing looking for my help and I noticed in the compiler listing that nearly all the variables were listed as being in unnamed COMMON.
And on some OSes (RSX-11M-PLUS was one of these), you passed data between separate processes via what looked to FORTRAN as a named COMMON block. Back in my FORTRAN days, I had written some large simulations that ran on a big RSX system where a couple of dozen different programs passed data among themselves via these chunks of shared memory. It was actually a piece that you wrote in MACRO-11 and loaded into memory before starting the FORTRAN tasks. I'm trying to remember what is was called now. Global section or was that the VMS equivalent? Darned cobwebs obscuring things and my RSX Executive reference is buried down in the basement. :-)
The six-character variable name makes me think that the code was written for a pretty old compiler. I recall the IBM FORTG and FORTH compilers having this limit. I remember finding the switch to DEC machines and using their compilers quite liberating. (Not to mention I didn't have to deal with XEDIT any more.)
As for having to deal with someone's really old software... I know what you're going through. An office mate left behind some code that we were to use in a later project. While he was working on it, he'd asked my advice on how to make it easier to manage as he'd just been plugging away and the thing grew to be a spaghetti code monster filled with GOTOs. He slaved away on it and was pretty satisfied with himself. When he left and I got the listings to begin on the followup project. I found that all he'd really accomplished was to pull out about a page and a half that became the main program. The remainder of the 27 pages or so became one giant subroutine. My job was to pull the guts of this hideous thing out and run it on a PDP-11. Finally got it to work but it took just a little doing. Ever since I've really loathed having to work on someone else's code. In the end, though, it does feel sort of good getting that old rat's nest of code cleaned up and the needed changes incorporated.
A recent "stupid code comment" story: I recently inherited a shell script from a departing consultant that contained the following:
The worst of it, though, was that a few lines later he wrote:
Now I get to tell his former employers that, maybe we should go over all the stuff he wrote.
Speaking of silly "increment counter" comments... Does anyone remember when SuperSoft's FORTRAN compiler for PC-DOS came out? You know, the one with the INCREMENT statement? (Because, you know, having to code "I = I + 1" is just so difficult.)
I jumped for joy when I moved out of the IBM supermini arena and moved to VAXes. Being able to use FORTRAN variables longer than 6 characters, allowed underscores in variables, actual tabs for indenting, actual CHARACTER arrays (no more using EQUIVALANCEs to LOGICAL*1 arrays), etc. It was like going to heaven (well, at the time it seemed so). Even though the older IBM code would still compile and run, I found myself "porting" much of it to VAX FORTRAN as I was making mods to it just to make is easier to read.
The article's comments on whitespace were funny to. Contrary to the author, though, I do prefer that there be some sort of style manual for programmers to follow. Years ago, a coworker of mine wrote a ton of FORTRAN using statements like the following:
just because the compiler would accept it (and he claimed he was saving disk space, as well). And he refused to use CONTINUE statements to terminate his DO loops. Let me tell ya, it was no picnic reading through his code and find out where loops started and ended. It wasn't uncommon to find that he had three or four DO loops all ending on the same statement.
If anyone used them to aid in reordering their dropped card deck, they were nuts.
I remember the IBM editors (this back in the early '80s before I shifted over to DEC systems) would put the sequence numbers into columns 73-80 automatically. And that was past the days when anyone actually used cards. The IBM editors could use those numbers for commands to insert/delete lines and some of the system tools would let you modify files (sort of like "patch") by inserting into or deleting from a source file based on those sequence numbers.
Once we found out that they weren't required, though, we found ways to prevent the editor from inserting them or ran CMS EXECs on the source files to strip them off just so that compressing our source files would save even more disk space. (An important thing to know back when a typical disk allocation on an IBM supermini was something like a couple of megabytes.)
Look at the size of those shoes!
... as ``The greatest boon to transportation since the wheel''. So will read the joint press release from the Concrete Industry and the Rebar Manufacturers Association.
I'd surely love to see how they plan on getting this monstrosity to pass muster with just about every environmental regulation on the books. What am I thinking? Why they have a former governor who will all-too happily waive those regulations!
True-to-form Microsoft. Years ago, Microsoft (IIRC, it was in a Bill Gates speech/interview) actually claimed that Windows was an Open System. The reason? Because you could run it on a PC from any IBM-compatible manufacturer, of course. Sure made sense to me. (Gag!)
... is that most of the people who are even aware of this class action lawsuit are thinking ``Oh, great! Just what I need: a check for $50 or voucher from Microsoft for some more of their buggy software''. At that point they decide that the suit isn't worth the effort to track down the receipt for the PC/software and, from there on, their brain's ``waste of my time'' filter kicks in whenever it's mentioned.
Now if it turned out that you'd get a full refund of the retail price of your MS software plus, oh, $100-$200 for every time your PC had to be rebuilt from scratch due to the swiss cheese nature of their OS/application security, then it starts becoming worthwhile to join in on the suit.
Ah, but what would you give just to have even LAVC-like clustering available for Linux? Is there anything better than the silly failover-to-a-hot-spare style clustering available for Linux?
Uh, I hope you meant "locking". :-)
Heh, heh. Anyone ever tell you that you have a gift for understatement?
Imagine how many more people get to enjoy the finger pointing that'll take place when their database vendor, OS vendor, and clustering vendor decide it's the other guys' fault that the database is toast because of some oddball bug. It's annoying enough having to deal with just two of them. I can't wait until I have to have a four-way conference call to get complex systems problems solved. Oops! Forgot! Toss in a hardware vendor for good measure.
Heck it was bad enough when we had these sort of troubles 20 years ago when the DEC HW and SW folks pointed fingers at each other but at least then you could get a hold of a manager to make the two sit down and solve the problem. What option do you have when all these players work for different companies? Who do you call? Legal? (That'll have you measuring downtime with a calendar.)
I haven't forgotten about that, young feller. It's the first thing that came to mind, in fact. I think the Datamation web site (whew! does that even exist any more) had it posted for a while.
You guys got crossed off my list of potential replacement printers as a result of your using the DMCA to try to lock out toner cartridge competition. Now this? Has anyone at Lexmark even heard of public relations let alone what it means?
Invest in a UPS for this system. Any uncontrolled shutdown (i.e. power outage) will take quite a long time to fully recover from. It will take a lo-o-o-ong time for the resync to finish on an 8x250GB raidset. Then you really ought to force an fsck as well on the filesystems you've created on that monster.
Well, they have at least a few technical people left. Otherwise, who'd be bundling the new version of Samba and other OSS packages with their crummy UNIX?
Not that it's much of an excuse, but didn't that argument work nicely in at least one trial that Microsoft was involved in? (Well at least I recall them trying that one. Dont't remember how well it worked.)
... and Microsoft wants a bigger piece.
Nah. My theory is that Ballmer sees their problem as people are spending money on hardware that Ballmer thinks they ought to be spending on software. Every dollar that Intel/Dell/HP/etc makes is a dollar that Microsoft didn't make. No one should be expecting Microsoft to be lowering their prices in lockstep with any price reductions that the hardware vendors put into place.
What Microsoft is hoping for is a situation where cheaper computers are in the hands of the masses resulting there will be an even bigger demand for software. What they don't seem to realize is that, for poorer countries, that could easily mean an even bigger demand for counterfeited software. Or the folks in those countries will just load a free OS on those computers. Microsoft would like very much to prevent that from happening. (Good luck with that.) The trouble is that someone from a poor country has a small set of choices. Say they've got $500 to spend. They'd like a computer (perhaps so they can become computer literate and work for one of those outsourcing companies):
Ballmer seems to think that people will choose Option 1. I would contend that Option 1 is the last choice people will make. (Me personally? I'd go with Option 2 since it allows me to support the OSS "industry".)
Some day they'll understand that the world is not always looking for ways to send their hard-earned money to Microsoft.
IMHO, it was when the Internet was opened up for commercial purposes when all the crud started flowing. Maybe, when businesses see it as unsuitable for their mass marketing needs and go away, the 'net will start to function properly again. Perhaps raising the cost of admission to the ".com" domains might work to keep out some of the garbage that's clogging up the works. Who knows? But something's got to happen. Thirty-six emails in the Inbox this morning at work and only one of them was something I asked to receive or from someone that I know. I'm getting a bit sick of having to go through all the junk. It's way too much like the daily ritual of sifting through all the junk snail mail that I get at home.