How 'easy' it is to track down the person depends on how hard they've covered their tracks. I imagine that you could get a bank account with a PO Box and some fake ID. There's stories of 'verified' PayPal users that do their banking in the Virgin Islands which has a completely opaque banking system.
Your credit card could help, but there's lots of stories where it didn't. Check your member agreement - I have a card where Internet fraud insurance is a specific feature, but I have others where it is not.
I guess I'm just saying to use good sense and not think it that a dodgy deal will be OK because it's a PayPal or Credit. Assume the same risk as a cash deal and you will be better off.
(The basic problem is that PayPal is not a integral part of the transaction. You contract with a seller on eBay for an item. You then have a seperate contract with PayPal to send the person money. Unlike with a normal purchase, there's no legal link between the two contracts -- PayPal is acting more like Western Union than a credit card processor. That puts you in the middle when you are defrauded.)
"I backtracked a whole bunch of bad dealers in various types of merchandise"
Well, you've got the data then. Did you correlate that with a random sampling of successful eBay auctions? (I ask because I've seen a few million feedback pro sellers that use "L@@K", and I've posted mfg specs myself in an attempt to look professional.)
I don't know why people are promting this blind faith in PayPal. Anyone can get a verfied paypal account, and lots of scammers do.
There's this thing called an Escrow Service which removes the risk from a transaction. You should use it anytime you couldn't live without the money you are sending off. If you bought a $2000 computer without using one, you are either stupid or richer than I am. Yes, it costs the buyer money. No, PayPal is not an escrow service.
The only thing that PayPal seems to provide is a piece of mind that you can go and post on PayPalSucks.com when you get ripped off, instead of blaming the guy who ripped you off or yourself.
The great thing about eBay is the "garage sale" aspect, not the fact that it's an ad board for real businesses. There's lots of people who have only sold a few things and haven't got around to a PayPal account, and if you are careful you shouldn't have any problem.
Another tip -- If it doesn't say "Working" or "Not DOA" or "Guaranteed", then assume the item is broken. "Pulled from a working system" doesn't mean squat. If the thing says "not tested" or "no Guarantee" or "I don't know if this works", then assume the seller knows the item is broken.
I can't count how many obviously broken things I've seen sold on eBay to buyers without basic reading comprehension skills.
Also, assume the item has no warranty (unless it says otherwise) and bid accordingly. There's a risk that it will break in 2 weeks, and unlike retail, you don't have a recourse.
(I've never had any particular problem with the "L@@K" people or people who take only money orders. I don't consider PayPal to be any less or more risky than a money order transaction, although I hear that Postal MOs give you some USPS backup.)
Novell had this masterplan for UNIXWare to compete with the up-and-coming NT -- they were going to merge NetWare and UNIXWare into the "SuperNOS" (cue black sabbath). At the time they had about 75% of the PC server market and had an enormous amount of influence, including a legion of paper Novell certs more slavish than any MCSEs you've ever met.
For some reason, however, they got chickenshit and instead _buried_ UNIXWare and just kept cranking out versions of primitive ol' NetWare. Because NT had SMP, TCP/IP, and could serve applications (you know, stuff UNIX still is the best at), they proceeded to crush Novell like an ant. Novell didn't get core TCP/IP support until 1999!
Thank Novell for setting PC Unix back by 5 years or more. And handing "M$" a bit of that world domination on a big silver platter.
AOL will want to differentiate their distro so they can charge big bucks for it.
Actually, AOL is probably just thinking about grabbing an embedded platform that they can control for their upcoming media consumption terminals (settop boxes).
The company has no current interests in corporate server or workstation technology, and doesn't seem to be going in that direction. Hopefully they aren't insane enough to go head-to-head with MS in the (increasingly irrelevant) PC OS market. What happened to all of the "enterprise" software they picked up with Netscape? They turned it right over to Sun with iPlanet....
And that is exactly the problem. Sun will want to differentiate their distro so they can charge big bucks for it.
Ahh, it all makes sense now. No wonder Solaris x86 went away.
One good question is why Apple survived while Commodore didn't.
Mainly because Apple sold $5-10K machines with a 60% profit margin and Commodore was selling cheapo $1500 boxes to the home market.
The prices killed Apple's chances for mass market adoption, but they still have billions in the bank from the old days, which has afforded a massive number of fuckups.
I could also see hard drives and CD-ROM drives going this route as well.
A few years ago, Intel and Microsoft came up with a spec called Device Bay that would allow end users to easily slide-in USB and 1394 drives.
Never caught on with OEMs though. Probably because of the lack of integrated 1394 controllers, but also because in the white-box market every penny counts, and for things like the iPaq desktop, the vendor would prefer that you buy proprietary parts.
And what would he find? A bunch of Apple ][ parts which probably won't do what he wants.
And one (1) IBM PS/2 part that IIRC uses the only thing remotely standardized as an external floppy connector -- A giant DB connector with something like 35 pins. Let me know if you have a PC with one of those plugs around:)
(IBM also had/has a much smaller ThinkPad floppy connector, and you might be able to find a 5.25" drive for one of those. But good luck...)
"Urban Population Density" -- and you are right on the mark. Canada generally has stricter urban building controls that have lead much denser concentrations of people. Easier to wire, easier to run streetcar service, etc.
Well, I think you had it on the mark with your earlier reference to bank vaults. If you want security, you need to have small verifiable pieces with limited interfaces.
The problem is that's not what people are buying in PC space. Users want to have dodgy filesharing clients installed, complex software like mailers and browsers, and so on. An object system like COM only makes it easy - there's probably an unlimited number of local exploits in the typical hunk of PC software, and as long as you can fork processes or load libraries, you have a problem if you can't trust every bit of code on your disk, even if you are MS-free.
My feeling is that the core security mechinisms in OSes date from the timesharing era and are really just unsuitable for personal computing. The most valuable thing by far on my machine is my data, not configuration and not code.
The real answer is probably a pervasive sandboxing system on the OS level (above and beyond what scripting or Java-only sandboxing gets you). Such things exist for government-bound trusted OSes (preventing interaction between "Top Secret", "Classified", and "Unclassified" code and data), and would be an excellent feature even in a PC OS ("Untrusted P2P and Warez", "Personal", "Work")
But -- the key point is that doesn't exist yet, and when it does we probably won't like it (.NET, Passport, DRM). Until the rules change, Microsoft remains a big target for people looking for implementation faults. That's not to say that they shouldn't produce better code, saner defaults, and maybe actually look into using file permissions in the meanwhile.
It strikes me that this solution alone would have negated pretty much every major email virus out there.
SirCam contained it's own grep and SMTP code, so no, it wouldn't. As long as the OS provides the functionality, having a limited scripting environment only hurts the lamers.
There's two MS preview pane autorun viruses that I can think of -- Both used HTML mail: One used a bugged java applet that exploited the MS JVM (broken sandbox), another used a buffer overrun against a ActiveX control that came preinstalled with certain version of Windows (and therefore was 'trusted').
Other 'autorun' MS mail attacks utilized buffer overflows against the MIME type and date parsers. Can you say with certainty that your mailer is free of such bugs? I can't.
The point is that none had to do with the preview pane itself, and instead used malformed data to attack code that was already on your computer and therefore you had to trust. Maybe HTML mail is just a bad idea, but it's customer demanded feature that exists on non-Windows platforms too (recall the Netscape JVM 'feature' that turned your browser/mailer into a webserver).
So, I guess I just don't see the design flaw issue with the preview pane viruses, just bad coding. It's not on the same braindamage scale as the Word autorun macros, where delivered code is executed.
I agree that MS nearly always makes the wrong decision when it comes to security versus features, I just don't think that they are at all unusual in that respect in the consumer PC application space, with the exception that people are more actively trying to find flaws in their products.
So, the obvious question is if an affinity system is a good thing or not.
Slashdot's moderation and filtering system ended up having the side-effect of encouraging different "zoos" within it's readerbase, because unlike other webboard software, 'trolling' (however you define it.. posts one doesn't like) could be accommodated without removing posts. The upside to this is that's allowed the discussions to scale, attracting more readers and posters without becoming a total mess. (and a crapflooder banner hit is the same income as any other banner hit)
An affinity system would really take this to the next level. By using it, you can read exactly the slashdot that you want to read, and the crapflooders can read exactly the slashdot they want to read without some informative posts emerging above their thresholds. Combined with multiple account tracking and central trusted/untrusted user lists (for example, the editor's friends and foes), and you have a very sophisticated automated system of filtering. This could really improve discussion quality, and allow slashdot to scale even a larger user base.
But it also magnifies the big problem with the moderation system, where groupthink and being agreeable tends to score you points. Now, much like a politician smiling and shaking hands while delivering his message, your posts now will gain you "friends" (or enemies). Fine. But, an affinity system would allow "parties" to form along particular political lines, where people could be reading only what's agreeable to them - "Java Programmers", "Sysadmins", "Linux Advocates" (who would have a collective set of "MS astroturfer" foes), and so on.
Unlike your usual google search, there isn't a right answer to the question of "Is this the post I wanted to read?" (or there shouldn't be, at least). If one goes through the effort of making a post disagreeing with a point, you do so in the hope that people will read it and understand your point. That becomes pointless if there's a high likelihood that they've tuned you out due to your affiliations. (For the same reason that pure flamebait is pointless under the current system -- most people don't read below a certain score.)
I guess to sum up, there's a big difference between ranking posts by content versus who's saying it.
Microsoft had to give away Windows for several years before anyone would consider paying for it. IBM could have done the same with OS/2, but chose to sell it at a premium to enterprise customers. Much like Apple, they didn't properly anticipate the exponential growth in PC users that happened in the early 90s.
I think that's (part of) what people mean when they talk about IBM's "bad marketing" of OS/2.
"It just doesn't matter." is a terrible lesson to be learned from your experience with OS/2. For one it absolves IBM for a terribly handled product, and for another it implies that all Bill Gates had to was punch the clock every morning to become the richest man on earth. You might not like to hear it, but at the time of the IBM-Microsoft divorce, Big Blue should have been able to rubout those pipsqueaks.
We've crossed swords over OS/2 on/. a few times before, and I find your worldview depressing. You bet on the wrong horse, and your reaction has been to valorize a flawed product and develop a defeatist worldview and an eternal hatred for your product's competitors (and the tendancy to spew offtopic bile about "LookOut"). I encourage you to try to take a fresh look at things and get on with life.
I'm no more happy about a MS Monopoly than the average slashdotter, just making an effort to understand how we got here. Not to mention that if Gates would sold out to IBM in 1990, I'm pretty sure that an IBM-dominated PC landscape would be worse that what we ended up getting.
If you want a timesheet program, write a timesheet program (web frontend, VB, C+Gnome, whatever).
Using Excel+Email as a timesheet system has always been the height of Microsoft office automation braindamage in my book. Unless you are someone in accounting that likes the makework.
Besides you miss the point of Microsoft Office's popularity entirely. 90% of the spreadsheets out there may use 10% of Excel's capability, but across a larger organization, 90% of Excel's features are being used. The bloat price you pay is pre-emptive compatibility with something that you haven't needed yet.
Excel is the deadest product that Microsoft makes -- it's essentially been static since 1994. The Graphing is just about the same as what Excel 2.2c gave you on a Mac IIcx. Every couple years, MS polishes the widgets and sells a new version as $150 upgrade, but they never ever do anything to fundementally make the program any better.
(Just ask Mr Clippy "What's New In Excel", and there's so little that he tells about new features from 1995. Sad.)
Given that Excel is a static target (like Gosling Emacs or UNIX itself), it's inevitable that the taillight chasing Open Source community will eventually replicate it.
As for making something better? Just ask Lotus and Corel how well that's doing. Excel pretty much sucks, but I can't see the market for something that's actually worse.
Notes was the original scriptable mailclient, so don't laugh too hard.
I wrote my own little version of ILUVYOU in LotusScript and it propagated just fine using the default Domino/Notes configuration. It could be stopped with tweaked Domain ECLs.
Within an Organization, that is -- it couldn't migrate outside. But still, there's plenty of 10K+ Notes installs and a few 100K+ shops, which is a bit of havock.
I don't want to get into a OS/2 flamewar, but I did _specifically_ mention v 2.1 which dates from 1992 or so.
+ It needed 16MB of memory. Windows 3 could get away with 4MB. That added $500 or more to the cost of a machine.
+ The default UI was 'featureful', but the execution was terrible. Ugly icons scattered hither and fro in 4-level deep nested folders scattered randomly around your desktop, gigantic ugly oddly-colored dialogs, terrible terrible filemanager, key features like "Shutdown" hidden in obscure places, etc.
If IBM would have hired somebody with a little artistic and usability training, this could have been significantly improved in a short amount of time. However, they didn't, and OS/2 (true to it's name) had a half-done GUI until Warp 4 shipped years later, after the product had been defeated.
Windows 3? Simple. Didn't do much. Obvious options. Therefore, cheap to support for OEMs and a more 'rational' product for economic reasons.
Yeah, but consider the point of these statistics. It's not to determine the marketshare of "all users", but to determine the marketshare of "users who count".
Meaning, if I'm trying to decide to target my internet-enabled, multimedia desktop application to Windows, Mac, or Linux, I don't give a shit about 3rd World DOS users, people with Mac LCs, or the guy who is surfing the web using Lynx on his 486 Linux box.
Those people are irrelevant from a market perspective -- they've made their software choices already and aren't in the market.
You don't seem to get that the OEM's technical support costs vastly outweigh what they are paying on the Microsoft Tax.
Meaning as soon as a user accidentally boots into the wrong OS and has to call support, the OEM has probably lost money on the box. That does not make economic sense.
I understand that you feel burnt over OS/2, but if you are going to talk economics, the 'rational' configuration is the one that is the most uniform and the cheapest to support (has the fewest options). That circumstance created the Microsoft monopoly far more than any dirty OEM dealings did.
It's also rather insulting to think that users are being buffooned into running Windows because the OEMs ship. Users run Applications and are mostly uninterested in OSes. They don't buy "Windows machines", they buy "Excel machines", and if you can't provide Excel (or a facimile), they won't be interested.
It's too bad that IBM missed the mark by inches. If OS/2 2.1 had been a little lighter weight and gotten a little UI touchup, it probably would have been the "supportable" choice over Windows. Linux ain't even close.
Here's why not:
Because it would cost millions of dollars to develop, if not hundreds of millions.
Because nobody is going to do UI heavy lifting for free.
Because once you built such a product, the existing Linux users would reject it.
Furthermore, it wouldn't be [useful|compatible] enough to get many to switch from Windows, and the thing would go down like another IBM Warp or BeOS.
Everybody knows it's technically possible to start with Linux and end up with something that's competititve with modern consumer desktops. It's just rather low on most people's priority lists (except the Advocates moaning outside).
Character mode Word and Excel did ship for OS/2. They just weren't very popular applications there either.
It's an old story - MS bet on the next level of hardware sophistication (GUI, extended memory, i386), and WordPerfect and Lotus were sitting around figuring out how to cram more features into a 640K XT and missed the boat.
The above has nothing to do with OS/2 versus Windows. Therefore I don't see how MS suckered them.
(For example, Lotus shipped a GUI integrated Office Suite years before MS called Lotus Jazz for the Mac. They dropped it after a year, and found themselves without a product to compete with MS Office in the early 90s.)
Slashdot Mirror
How 'easy' it is to track down the person depends on how hard they've covered their tracks. I imagine that you could get a bank account with a PO Box and some fake ID. There's stories of 'verified' PayPal users that do their banking in the Virgin Islands which has a completely opaque banking system.
Your credit card could help, but there's lots of stories where it didn't. Check your member agreement - I have a card where Internet fraud insurance is a specific feature, but I have others where it is not.
I guess I'm just saying to use good sense and not think it that a dodgy deal will be OK because it's a PayPal or Credit. Assume the same risk as a cash deal and you will be better off.
(The basic problem is that PayPal is not a integral part of the transaction. You contract with a seller on eBay for an item. You then have a seperate contract with PayPal to send the person money. Unlike with a normal purchase, there's no legal link between the two contracts -- PayPal is acting more like Western Union than a credit card processor. That puts you in the middle when you are defrauded.)
"I backtracked a whole bunch of bad dealers in various types of merchandise"
Well, you've got the data then. Did you correlate that with a random sampling of successful eBay auctions? (I ask because I've seen a few million feedback pro sellers that use "L@@K", and I've posted mfg specs myself in an attempt to look professional.)
I don't know why people are promting this blind faith in PayPal. Anyone can get a verfied paypal account, and lots of scammers do.
There's this thing called an Escrow Service which removes the risk from a transaction. You should use it anytime you couldn't live without the money you are sending off. If you bought a $2000 computer without using one, you are either stupid or richer than I am. Yes, it costs the buyer money. No, PayPal is not an escrow service.
The only thing that PayPal seems to provide is a piece of mind that you can go and post on PayPalSucks.com when you get ripped off, instead of blaming the guy who ripped you off or yourself.
The great thing about eBay is the "garage sale" aspect, not the fact that it's an ad board for real businesses. There's lots of people who have only sold a few things and haven't got around to a PayPal account, and if you are careful you shouldn't have any problem.
Another tip -- If it doesn't say "Working" or "Not DOA" or "Guaranteed", then assume the item is broken. "Pulled from a working system" doesn't mean squat. If the thing says "not tested" or "no Guarantee" or "I don't know if this works", then assume the seller knows the item is broken.
I can't count how many obviously broken things I've seen sold on eBay to buyers without basic reading comprehension skills.
Also, assume the item has no warranty (unless it says otherwise) and bid accordingly. There's a risk that it will break in 2 weeks, and unlike retail, you don't have a recourse.
(I've never had any particular problem with the "L@@K" people or people who take only money orders. I don't consider PayPal to be any less or more risky than a money order transaction, although I hear that Postal MOs give you some USPS backup.)
Novell had this masterplan for UNIXWare to compete with the up-and-coming NT -- they were going to merge NetWare and UNIXWare into the "SuperNOS" (cue black sabbath). At the time they had about 75% of the PC server market and had an enormous amount of influence, including a legion of paper Novell certs more slavish than any MCSEs you've ever met.
For some reason, however, they got chickenshit and instead _buried_ UNIXWare and just kept cranking out versions of primitive ol' NetWare. Because NT had SMP, TCP/IP, and could serve applications (you know, stuff UNIX still is the best at), they proceeded to crush Novell like an ant. Novell didn't get core TCP/IP support until 1999!
Thank Novell for setting PC Unix back by 5 years or more. And handing "M$" a bit of that world domination on a big silver platter.
AOL will want to differentiate their distro so they can charge big bucks for it.
Actually, AOL is probably just thinking about grabbing an embedded platform that they can control for their upcoming media consumption terminals (settop boxes).
The company has no current interests in corporate server or workstation technology, and doesn't seem to be going in that direction. Hopefully they aren't insane enough to go head-to-head with MS in the (increasingly irrelevant) PC OS market. What happened to all of the "enterprise" software they picked up with Netscape? They turned it right over to Sun with iPlanet....
And that is exactly the problem. Sun will want to differentiate their distro so they can charge big bucks for it.
Ahh, it all makes sense now. No wonder Solaris x86 went away.
One good question is why Apple survived while Commodore didn't.
Mainly because Apple sold $5-10K machines with a 60% profit margin and Commodore was selling cheapo $1500 boxes to the home market.
The prices killed Apple's chances for mass market adoption, but they still have billions in the bank from the old days, which has afforded a massive number of fuckups.
I could also see hard drives and CD-ROM drives going this route as well.
A few years ago, Intel and Microsoft came up with a spec called Device Bay that would allow end users to easily slide-in USB and 1394 drives.
Never caught on with OEMs though. Probably because of the lack of integrated 1394 controllers, but also because in the white-box market every penny counts, and for things like the iPaq desktop, the vendor would prefer that you buy proprietary parts.
And what would he find? A bunch of Apple ][ parts which probably won't do what he wants.
:)
And one (1) IBM PS/2 part that IIRC uses the only thing remotely standardized as an external floppy connector -- A giant DB connector with something like 35 pins. Let me know if you have a PC with one of those plugs around
(IBM also had/has a much smaller ThinkPad floppy connector, and you might be able to find a 5.25" drive for one of those. But good luck...)
"Urban Population Density" -- and you are right on the mark. Canada generally has stricter urban building controls that have lead much denser concentrations of people. Easier to wire, easier to run streetcar service, etc.
Well, I think you had it on the mark with your earlier reference to bank vaults. If you want security, you need to have small verifiable pieces with limited interfaces.
The problem is that's not what people are buying in PC space. Users want to have dodgy filesharing clients installed, complex software like mailers and browsers, and so on. An object system like COM only makes it easy - there's probably an unlimited number of local exploits in the typical hunk of PC software, and as long as you can fork processes or load libraries, you have a problem if you can't trust every bit of code on your disk, even if you are MS-free.
My feeling is that the core security mechinisms in OSes date from the timesharing era and are really just unsuitable for personal computing. The most valuable thing by far on my machine is my data, not configuration and not code.
The real answer is probably a pervasive sandboxing system on the OS level (above and beyond what scripting or Java-only sandboxing gets you). Such things exist for government-bound trusted OSes (preventing interaction between "Top Secret", "Classified", and "Unclassified" code and data), and would be an excellent feature even in a PC OS ("Untrusted P2P and Warez", "Personal", "Work")
But -- the key point is that doesn't exist yet, and when it does we probably won't like it (.NET, Passport, DRM). Until the rules change, Microsoft remains a big target for people looking for implementation faults. That's not to say that they shouldn't produce better code, saner defaults, and maybe actually look into using file permissions in the meanwhile.
It strikes me that this solution alone would have negated pretty much every major email virus out there.
SirCam contained it's own grep and SMTP code, so no, it wouldn't. As long as the OS provides the functionality, having a limited scripting environment only hurts the lamers.
There's two MS preview pane autorun viruses that I can think of -- Both used HTML mail: One used a bugged java applet that exploited the MS JVM (broken sandbox), another used a buffer overrun against a ActiveX control that came preinstalled with certain version of Windows (and therefore was 'trusted').
Other 'autorun' MS mail attacks utilized buffer overflows against the MIME type and date parsers. Can you say with certainty that your mailer is free of such bugs? I can't.
The point is that none had to do with the preview pane itself, and instead used malformed data to attack code that was already on your computer and therefore you had to trust. Maybe HTML mail is just a bad idea, but it's customer demanded feature that exists on non-Windows platforms too (recall the Netscape JVM 'feature' that turned your browser/mailer into a webserver).
So, I guess I just don't see the design flaw issue with the preview pane viruses, just bad coding. It's not on the same braindamage scale as the Word autorun macros, where delivered code is executed.
I agree that MS nearly always makes the wrong decision when it comes to security versus features, I just don't think that they are at all unusual in that respect in the consumer PC application space, with the exception that people are more actively trying to find flaws in their products.
Does anyone know if MS's C# compiler is written in C#?
So, the obvious question is if an affinity system is a good thing or not.
.. posts one doesn't like) could be accommodated without removing posts. The upside to this is that's allowed the discussions to scale, attracting more readers and posters without becoming a total mess. (and a crapflooder banner hit is the same income as any other banner hit)
Slashdot's moderation and filtering system ended up having the side-effect of encouraging different "zoos" within it's readerbase, because unlike other webboard software, 'trolling' (however you define it
An affinity system would really take this to the next level. By using it, you can read exactly the slashdot that you want to read, and the crapflooders can read exactly the slashdot they want to read without some informative posts emerging above their thresholds. Combined with multiple account tracking and central trusted/untrusted user lists (for example, the editor's friends and foes), and you have a very sophisticated automated system of filtering. This could really improve discussion quality, and allow slashdot to scale even a larger user base.
But it also magnifies the big problem with the moderation system, where groupthink and being agreeable tends to score you points. Now, much like a politician smiling and shaking hands while delivering his message, your posts now will gain you "friends" (or enemies). Fine. But, an affinity system would allow "parties" to form along particular political lines, where people could be reading only what's agreeable to them - "Java Programmers", "Sysadmins", "Linux Advocates" (who would have a collective set of "MS astroturfer" foes), and so on.
Unlike your usual google search, there isn't a right answer to the question of "Is this the post I wanted to read?" (or there shouldn't be, at least). If one goes through the effort of making a post disagreeing with a point, you do so in the hope that people will read it and understand your point. That becomes pointless if there's a high likelihood that they've tuned you out due to your affiliations. (For the same reason that pure flamebait is pointless under the current system -- most people don't read below a certain score.)
I guess to sum up, there's a big difference between ranking posts by content versus who's saying it.
Microsoft had to give away Windows for several years before anyone would consider paying for it. IBM could have done the same with OS/2, but chose to sell it at a premium to enterprise customers. Much like Apple, they didn't properly anticipate the exponential growth in PC users that happened in the early 90s.
/. a few times before, and I find your worldview depressing. You bet on the wrong horse, and your reaction has been to valorize a flawed product and develop a defeatist worldview and an eternal hatred for your product's competitors (and the tendancy to spew offtopic bile about "LookOut"). I encourage you to try to take a fresh look at things and get on with life.
I think that's (part of) what people mean when they talk about IBM's "bad marketing" of OS/2.
"It just doesn't matter." is a terrible lesson to be learned from your experience with OS/2. For one it absolves IBM for a terribly handled product, and for another it implies that all Bill Gates had to was punch the clock every morning to become the richest man on earth. You might not like to hear it, but at the time of the IBM-Microsoft divorce, Big Blue should have been able to rubout those pipsqueaks.
We've crossed swords over OS/2 on
I'm no more happy about a MS Monopoly than the average slashdotter, just making an effort to understand how we got here. Not to mention that if Gates would sold out to IBM in 1990, I'm pretty sure that an IBM-dominated PC landscape would be worse that what we ended up getting.
If you want a timesheet program, write a timesheet program (web frontend, VB, C+Gnome, whatever).
Using Excel+Email as a timesheet system has always been the height of Microsoft office automation braindamage in my book. Unless you are someone in accounting that likes the makework.
Besides you miss the point of Microsoft Office's popularity entirely. 90% of the spreadsheets out there may use 10% of Excel's capability, but across a larger organization, 90% of Excel's features are being used. The bloat price you pay is pre-emptive compatibility with something that you haven't needed yet.
Excel is the deadest product that Microsoft makes -- it's essentially been static since 1994. The Graphing is just about the same as what Excel 2.2c gave you on a Mac IIcx. Every couple years, MS polishes the widgets and sells a new version as $150 upgrade, but they never ever do anything to fundementally make the program any better.
(Just ask Mr Clippy "What's New In Excel", and there's so little that he tells about new features from 1995. Sad.)
Given that Excel is a static target (like Gosling Emacs or UNIX itself), it's inevitable that the taillight chasing Open Source community will eventually replicate it.
As for making something better? Just ask Lotus and Corel how well that's doing. Excel pretty much sucks, but I can't see the market for something that's actually worse.
Oh, and that's just scratching the surface ...
(IIRC, the SMTP full headers are stuck in a hidden field which you can see either with the props browser or a custom form..)
Notes was the original scriptable mailclient, so don't laugh too hard.
I wrote my own little version of ILUVYOU in LotusScript and it propagated just fine using the default Domino/Notes configuration. It could be stopped with tweaked Domain ECLs.
Within an Organization, that is -- it couldn't migrate outside. But still, there's plenty of 10K+ Notes installs and a few 100K+ shops, which is a bit of havock.
Oh - Point #2 -- I threw that "facsimile" bit in as a bone to the non-Windows folks. I agree with your point.
I don't want to get into a OS/2 flamewar, but I did _specifically_ mention v 2.1 which dates from 1992 or so.
+ It needed 16MB of memory. Windows 3 could get away with 4MB. That added $500 or more to the cost of a machine.
+ The default UI was 'featureful', but the execution was terrible. Ugly icons scattered hither and fro in 4-level deep nested folders scattered randomly around your desktop, gigantic ugly oddly-colored dialogs, terrible terrible filemanager, key features like "Shutdown" hidden in obscure places, etc.
If IBM would have hired somebody with a little artistic and usability training, this could have been significantly improved in a short amount of time. However, they didn't, and OS/2 (true to it's name) had a half-done GUI until Warp 4 shipped years later, after the product had been defeated.
Windows 3? Simple. Didn't do much. Obvious options. Therefore, cheap to support for OEMs and a more 'rational' product for economic reasons.
Yeah, but consider the point of these statistics. It's not to determine the marketshare of "all users", but to determine the marketshare of "users who count".
Meaning, if I'm trying to decide to target my internet-enabled, multimedia desktop application to Windows, Mac, or Linux, I don't give a shit about 3rd World DOS users, people with Mac LCs, or the guy who is surfing the web using Lynx on his 486 Linux box.
Those people are irrelevant from a market perspective -- they've made their software choices already and aren't in the market.
You don't seem to get that the OEM's technical support costs vastly outweigh what they are paying on the Microsoft Tax.
Meaning as soon as a user accidentally boots into the wrong OS and has to call support, the OEM has probably lost money on the box. That does not make economic sense.
I understand that you feel burnt over OS/2, but if you are going to talk economics, the 'rational' configuration is the one that is the most uniform and the cheapest to support (has the fewest options). That circumstance created the Microsoft monopoly far more than any dirty OEM dealings did.
It's also rather insulting to think that users are being buffooned into running Windows because the OEMs ship. Users run Applications and are mostly uninterested in OSes. They don't buy "Windows machines", they buy "Excel machines", and if you can't provide Excel (or a facimile), they won't be interested.
It's too bad that IBM missed the mark by inches. If OS/2 2.1 had been a little lighter weight and gotten a little UI touchup, it probably would have been the "supportable" choice over Windows. Linux ain't even close.
"Why not just give them Linux XP"
Here's why not:
Because it would cost millions of dollars to develop, if not hundreds of millions.
Because nobody is going to do UI heavy lifting for free.
Because once you built such a product, the existing Linux users would reject it.
Furthermore, it wouldn't be [useful|compatible] enough to get many to switch from Windows, and the thing would go down like another IBM Warp or BeOS.
Everybody knows it's technically possible to start with Linux and end up with something that's competititve with modern consumer desktops. It's just rather low on most people's priority lists (except the Advocates moaning outside).
Character mode Word and Excel did ship for OS/2. They just weren't very popular applications there either.
It's an old story - MS bet on the next level of hardware sophistication (GUI, extended memory, i386), and WordPerfect and Lotus were sitting around figuring out how to cram more features into a 640K XT and missed the boat.
The above has nothing to do with OS/2 versus Windows. Therefore I don't see how MS suckered them.
(For example, Lotus shipped a GUI integrated Office Suite years before MS called Lotus Jazz for the Mac. They dropped it after a year, and found themselves without a product to compete with MS Office in the early 90s.)