Most people don't care. It's simpliar to explain that Verizon and AT&T use different cell technologies. It's like how Mac's and PC's used to use different formatted floppies. Sure they were both 3.5" disks, but one wouldn't work with the other.
Re:Slashdot's anti-Google schtick is out of contro
on
Google About Openness
·
· Score: 2, Insightful
Apple was the big white knight around here the first half of decade./. was cheering them on for using *iux (even it was BSDish), supporting CUPS, and then it was cheering for Webkit. Then the mood changed about 2006 - 2007 with the release of the iPhone and/. went from being pro Apple to anti-Apple and Google replaced them as the great white knight of opensource. Like Apple, that's been going on for 3 - 4 years, so now it's time for the mood to change to Google being the next evil(tm) company on/./. is no different than any other media: build something up so it's more fun to rip them apart later.
Bingo. I've always had my developers come back with the task completed and usually working ahead of time. Granted we've never set aggressive due dates unless there was a bug or really critically urgent feature that we overlooked. My only rule is, let me know if there is a problem as soon as you know it. Because we do have a couple really good programmers on contract. They are expensive, but if you need a particular problem solved, they can do it. (and charge for it) I just need to know because I have to give them usually a week or two lead time.
We're moving our infrastructure over to Perl from PHP. When we did testing we found that pages on our sites tended to load in about the same amount of time as it was the Database query and then front-end CSS/JS that took the bulk of the time to load a page, not the script execution time. (We're moving to Perl because there is no PHP connector other than ODBC for the database we'll be moving to and Perl has 2 DBI modules). Plus every time we wanted to do something/add a feature/add an external API, we found there was already a Perl Module for that...
In my experience in the past 10 years, it's usually been the database that has often been the bottle neck. Seems like any time the traffic was too much with Perl or PHP, adding more hardware with a load balancer always worked until you started to overloaded the database. (Which usually happened first anyway). AND THAT is where you started spending the $$$$.
Generally I'm there until midnight or 1AM ANYWAY, but most of my developers probably get the bulk of new work done between 10PM - 2AM. I think we have more code commits during the 1AM hour than any other. Sometimes they work at home, other times at the office, if the Application and Web Development people are working on something that involves the API. But typically our developers set their own hours. Just so long as the work gets done by the due date and are responsive to SMS if we have an "Oh shit" moment and they need to come in. It happens, but not often.
However, I HAVE to be at the office at 8AM and keep normal business hours for client meetings and if clients call with problems, they expect someone to be there and as it stands right now, the buck stops at my desk. Generally everyone is in the office by 11AM and if have meetings it is usually during lunchtime. Afternoons are usually spent fixing any issues that may have popped up and if the different development teams need to work together.
Then again, we're a small company, with 10 full-time developers plus six interns (4CS, 2ECE) at the moment. We have 4 + 3 Interns on Desktop & Mobile Java Application Development (Java Team), 4+1 intern Web Development Group, and 2 of us who are Database & Systems people with 2 interns working on a R&D project.
We've been working on native mobile apps for our systems the past 6 months. iPhone has been pretty good to work with. You build your software, if it works on one iPhone (or iPod Touch even), it will work on the next in the same exact way. Now there may be differences in OS (2, 3, 3.1, 3.2), but at least the hardware is the same and operates the same way. Same pretty much with Blackberry as well as they have the classic Blackberry style and then the Storm. There are some hardware differences between models, but basically you have to make sure it works on your normal blackberry and then the Storm series phones.
Windows Mobile is a nightmare. You can write the software, but it runs on so many different hardware platforms, each with their own difference (some have a stock UI, others a manufactures UI, others a carrier UI), that it takes a lot of time an expense to debug it. And even then we still get complaints that things don't work on XYZ model phone that we had never even heard of before. Our app maybe perfectly usable on one phone, completely unusable on the next because of screen size or one has a touch screen, one only has a keyboard interface, etc..
Unfortunately for Android, they're going down the same road as Windows Mobile. As it stands right now, we have to test against 3 different OS versions (1.5, 1.6, 2.0) AND test usability against different configuration. How does it look on AB size screen vs. CD sized screen. How well does it interface with touch screen? How well does it interface with keyboard? Does it run well on processor version X vs. Y, etc.. That adds a lot of cost to develop for in testing and QA.
We'll give Android another year and see. But if some of these problems don't look to be righted by Google, then in the future we're likely to support iPhone and Blackberry native and then develop a web-based interface for everything else to keep down costs.
The last two times I tested it for a true shared-nothing HA cluster, NDBCLUSTER failed miserably without a lot of tweaking. The optimizer was buggy to the point of being broken. And basically the response I got from MySQL AB at the time was, "If you want to use NDBCLUSTER, you'd better get the Enterprise Support Package". After pricing out what it would cost in support from MySQL AB AND the cost of having to go through and rewrite a bunch of our code to optimize it, it was cheaper to buy DB2.
Company I work for now uses PostgreSQL for main product lines. But two of their package are third party and use MySQL including their billing system. It works, but as it stands right now, neither of those systems are being taxed on a Dual-Quad Core DB server with 12GB RAM. In fact, it barely runs at 5% of resource utilization. We still use MySQL for one of our website's CMS. And it does the job well.
MySQL works well up until you need more than one box. Replication can work in some circumstances, but as a HA solution, it looses any advantages it had in terms of cost vs. extremely proven and reliable systems.
I've never been in a St. Louis Bread Co. (same company as Panera) and charged for access. They have a splash screen that you have to click "log-in" before you can get on and agree to their terms of service. But that has never required a purchase. Granted their web is useless for anything other than email and surfing to CNN and a few web sites as anything of interest seems to be block and they also seem to block ports like FTP/SSH starting a little over a year ago.
Coffee shop I'm setting at now requires you purchase a drink for 3 hours of internet time. And has for years because it is a college town and it's a popular study hang out. If you didn't you'd get people come in, sit down for 6 - 8 hours, and never buy anything. Meanwhile paying customers are leaving and going somewhere else because there aren't any seats available. (They had free wifi when they first opened. And Free was costing them too much in lost revenue. At least $150 - $200 a day.)
Fourth, if you're writing a plain web app, however fancily mobile-enhanced, how are you going to make use of the cool features of different phones? The iPhone has a camera, accelerometers, GPS, and multitouch. I admit I'm not terribly well-versed in the features of other smartphones, but a) do they all have these? b) can you access them from web apps? and c) can you access them all in the same way from web apps?
I develop and sell a program for $50. You buy the program, copy it say crack the 'copy protection', and sell it as MyProgram for $40. Can you not see the problem with this?
While I agree that Copyright and Patent system needs reform, it is still needed. If I couldn't develop and sell my program with some kind of protection, the chances are I'm not going to invest the time to do it.
Personally I think we need two IP system. One for Creative works (art/books/etc) and another one for Business related things. I agree with artists, writers, etc. retaining copyright for the duration of their life times, no longer. For businesses, they should be granted a provisional patent for 7 years. If they aren't producing a product with that patent by then, it expires. If they are, they get another 7 years of protection.
So, basically, you're mad that it won't play your pirated movies out of the box? I've got karma to burn, so time to feed the trolls.
Yeah, I'm going to assume that you really have no idea how the encoder/decoder market works or video production in general. Basically, there are companies that create codecs like DivX and then require payment to encode and distribute videos using their codec because of patents. In order for Apple to ship DivX, they would have to license DivX from DivX. Why should apple pay the licensing fee when they can direct users here: http://www.divx.com/en/mac and the user can download and install for free? I mean the last time I dealt with Windows I remember having to go and fetch the DivX Codec. And the last time I set up a new Mac, QuickTime popped up when it could find the Codec, knew what kind of file it was, and provided a link to the Divx site to go download it. All of this has to do with Licensing. It's really more of a legal issue than a technical one. So how is software licensing and patents preventing them from shipping every codec known to man Apple's fault again? Because unlike a lot of F/OSS projects, Apple can't be 100% based out of Hungry or France to circumvent these licensing restrictions. The do business in the United States and other countries that do recognize and enforce these copyright treaties. Technically, downloading and using VLC and FFMPEG in the United States constitutes infringement on various Codec patents, but that's a topic for another thread and another day. Also I would check on Windows. If it shipped with DivX, chances are that was added by the PC vender in a 3rd party deal, not by Microsoft. (Actually I don't pay attention to who is licensing what these days).
MKV isn't a format, it's a container. Just like.MOV is a container as is.AVI. The quality inside a.MKV,.AVI, or.MOV is all dependent upon the compression and codecs used not the container format. And the last time I tried playing MKV files, the program took up WAY too many system resources. The only place that I've seen MKV as a popular format is with Azurus/Vuze. In fact, I've never seen it used outside of Vuze. Occasionally you'll see it in a torrent, but they are mostly DivX/Avi.
The fact that macs can't play Blu-Ray or HD-DVD has to do with the fact that Apple doesn't ship any models with said player. MAC LACKS THE HARDWARE NEEDED IN ORDER TO READ EITHER FORMAT That has nothing to do with Quicktime and Codecs. When I worked in video production, nobody was adopting either optical format. It was clear to many of us that Optical Media was going the way of the floppy disk and this was in 2003. By the time HD-DVD or Blu-Ray won the format wars, it would be too late, we'd see everything delivered via a digital stream. The only question was, is it going to be an iTunes like store where you buy the item to a set top box/hard drive or whether it was going to be via streaming like Hulu or Netflix. So far it looks like there is a market for both. But time will tell.
And I see a lot of people electing to skip the Blu-Ray player in favor of a media PC/Mac hooked up to their TV and streaming Netflix or Hulu to their TV. I've been using a Mac MIni for this purpose since 2005. I know I have. To me, the extra quality for twice the price of a DVD plus the cost of the player isn't worth it to me. I'll put up with watching my TV shows from Hulu when I have extra time at my connivence, even if the quality isn't the same as on HD. It' good enough for me and plenty of other people too.
Now, I have to say that I'm generally happy with Quicktime X. Quicktime had been languishing for years and this was a much needed up date. It runs faster and smoother than the previous versions.
It doesn't kill OSS, it kills the use of GPL'ed code. With our proprietary products, we were extremely careful that if we did use other code/libraries that they were either BSD or MIT style licenses.
Why do people forget that OSS is more than just the GPL?
Well, firefox needs to get its act together and remember what their original purpose was because I've noticed a lot of average users complaining about the last couple Mozilla releases being buggy and slow across all platforms. On the windows side, quite a few have already flocked to Chrome an a few to Opera. OSX, a lot of folks have gone back to Safari.
Seems like I remember Apple being this great white knight circa 2004/2005 and then after a couple years they are "evil" and "propitiatory", yet Apple is one of the more OSS friendly companies out there by buying and supporting CUPS (Those who like the fact their printer works with Linux has Apple to thank for that), and then Webkit, and a number of other things.
Basically they got the same treatment that Google has the past couple years, but now it looks like the tide is starting to turn as I predict that 2010 will the year that Google becomes widely regarded as "evil" around/.
Depends on the type of shares they purchased. If they were preferred shares, you get first dibs on dividends, but no votes. If they were Common Shares, then eBay would have a say in the vote of board of directors, etc..
At our company employees do get their own cell phone, company chips in $30 or $60 a month depending on whether they need a data plan (which we cover the cost of) into the first paycheck of the month.
Works extremely well.
I don't really see the point of this in the real world. I could see where this could be useful where we would have 1 phone and test in Windows Mobile and Android on one hardware platform. Outside of that, I see no real value.
That depends. 60% of orders from our website still comes from customers using MSIE 6. Until that changes, I will continue to curse, kick, and scream that MSIE is broken.
Honestly, I could heat my old 1 bedroom apartment with my Quad-Core G5 powermac tower. I set thermostat to 60 and the only time the heat ran that winter was when it was less than 25 degrees outside.
I worked for a company about 10 years ago that had a set of video editing tools for NT and IRIX. We knew SGI was dying and looked into porting our tools to Linux and we re ran into the same problem. So we officially supported Red Hat 4/5 at the time. It also happened to work with SuSE as well. Sales to Linux based systems was less than 10% of all sales and was accounting for more than 35% of all Technical support requests. And a lot of the support requests were, "Why don't you support my custom leet slackware userland with debian kernel?" If you billed that out in man hour cost in tech support vs. sales, we lost like $25k on the Linux version. Probably more if you included the 8 of us that worked on that version.
We supported Linux until OSX 10.1 was released and ported the tools over to Mac. For a year we supported Windows, Linux and Mac, but it was pretty clear that our IRIX customers were largely going to Apple or Windows as their replacement systems. When I left the company, sales of the linux version was down to 5% and 40% of sales/upgrades were for OSX and the rest windows. I know I looked a year or so later and they had dropped the linux version all together.
Of course, it would help if they asked what the users actually do.
Bingo.
We had the advantage of a small business owner wanting our software developed because he thought "It should work like this". So we made it work like that and a lot of other small business owners found it to make sense and relatively easy to use. There were a couple quirks, but that's not good enough. Not for me.
And this is where so many others fails. After the phase 1 deployment of our product (about 100 installs), I drove/flew around to our customers 6 months later, stopped by in person and asked as the first question: "What doesn't work?" followed by "How can it work better?"
Slashdot Mirror
Most people don't care. It's simpliar to explain that Verizon and AT&T use different cell technologies. It's like how Mac's and PC's used to use different formatted floppies. Sure they were both 3.5" disks, but one wouldn't work with the other.
Apple was the big white knight around here the first half of decade. /. was cheering them on for using *iux (even it was BSDish), supporting CUPS, and then it was cheering for Webkit. Then the mood changed about 2006 - 2007 with the release of the iPhone and /. went from being pro Apple to anti-Apple and Google replaced them as the great white knight of opensource. Like Apple, that's been going on for 3 - 4 years, so now it's time for the mood to change to Google being the next evil(tm) company on /. /. is no different than any other media: build something up so it's more fun to rip them apart later.
Apple Computer paid Apple Records some $500M or so to buy the trademark rights from the record company in 2007.
Bingo. I've always had my developers come back with the task completed and usually working ahead of time. Granted we've never set aggressive due dates unless there was a bug or really critically urgent feature that we overlooked. My only rule is, let me know if there is a problem as soon as you know it. Because we do have a couple really good programmers on contract. They are expensive, but if you need a particular problem solved, they can do it. (and charge for it) I just need to know because I have to give them usually a week or two lead time.
We're moving our infrastructure over to Perl from PHP. When we did testing we found that pages on our sites tended to load in about the same amount of time as it was the Database query and then front-end CSS/JS that took the bulk of the time to load a page, not the script execution time. (We're moving to Perl because there is no PHP connector other than ODBC for the database we'll be moving to and Perl has 2 DBI modules). Plus every time we wanted to do something/add a feature/add an external API, we found there was already a Perl Module for that...
In my experience in the past 10 years, it's usually been the database that has often been the bottle neck. Seems like any time the traffic was too much with Perl or PHP, adding more hardware with a load balancer always worked until you started to overloaded the database. (Which usually happened first anyway). AND THAT is where you started spending the $$$$.
Generally I'm there until midnight or 1AM ANYWAY, but most of my developers probably get the bulk of new work done between 10PM - 2AM. I think we have more code commits during the 1AM hour than any other. Sometimes they work at home, other times at the office, if the Application and Web Development people are working on something that involves the API. But typically our developers set their own hours. Just so long as the work gets done by the due date and are responsive to SMS if we have an "Oh shit" moment and they need to come in. It happens, but not often.
However, I HAVE to be at the office at 8AM and keep normal business hours for client meetings and if clients call with problems, they expect someone to be there and as it stands right now, the buck stops at my desk. Generally everyone is in the office by 11AM and if have meetings it is usually during lunchtime. Afternoons are usually spent fixing any issues that may have popped up and if the different development teams need to work together.
Then again, we're a small company, with 10 full-time developers plus six interns (4CS, 2ECE) at the moment. We have 4 + 3 Interns on Desktop & Mobile Java Application Development (Java Team), 4+1 intern Web Development Group, and 2 of us who are Database & Systems people with 2 interns working on a R&D project.
We've been working on native mobile apps for our systems the past 6 months. iPhone has been pretty good to work with. You build your software, if it works on one iPhone (or iPod Touch even), it will work on the next in the same exact way. Now there may be differences in OS (2, 3, 3.1, 3.2), but at least the hardware is the same and operates the same way. Same pretty much with Blackberry as well as they have the classic Blackberry style and then the Storm. There are some hardware differences between models, but basically you have to make sure it works on your normal blackberry and then the Storm series phones.
Windows Mobile is a nightmare. You can write the software, but it runs on so many different hardware platforms, each with their own difference (some have a stock UI, others a manufactures UI, others a carrier UI), that it takes a lot of time an expense to debug it. And even then we still get complaints that things don't work on XYZ model phone that we had never even heard of before. Our app maybe perfectly usable on one phone, completely unusable on the next because of screen size or one has a touch screen, one only has a keyboard interface, etc..
Unfortunately for Android, they're going down the same road as Windows Mobile. As it stands right now, we have to test against 3 different OS versions (1.5, 1.6, 2.0) AND test usability against different configuration. How does it look on AB size screen vs. CD sized screen. How well does it interface with touch screen? How well does it interface with keyboard? Does it run well on processor version X vs. Y, etc.. That adds a lot of cost to develop for in testing and QA.
We'll give Android another year and see. But if some of these problems don't look to be righted by Google, then in the future we're likely to support iPhone and Blackberry native and then develop a web-based interface for everything else to keep down costs.
At least in the states, the DMCA only opens one up to civil action. The EUCD opens the possibility of criminal charges in addition to civil ones.
The last two times I tested it for a true shared-nothing HA cluster, NDBCLUSTER failed miserably without a lot of tweaking. The optimizer was buggy to the point of being broken. And basically the response I got from MySQL AB at the time was, "If you want to use NDBCLUSTER, you'd better get the Enterprise Support Package". After pricing out what it would cost in support from MySQL AB AND the cost of having to go through and rewrite a bunch of our code to optimize it, it was cheaper to buy DB2.
Company I work for now uses PostgreSQL for main product lines. But two of their package are third party and use MySQL including their billing system. It works, but as it stands right now, neither of those systems are being taxed on a Dual-Quad Core DB server with 12GB RAM. In fact, it barely runs at 5% of resource utilization. We still use MySQL for one of our website's CMS. And it does the job well.
MySQL works well up until you need more than one box. Replication can work in some circumstances, but as a HA solution, it looses any advantages it had in terms of cost vs. extremely proven and reliable systems.
I've never been in a St. Louis Bread Co. (same company as Panera) and charged for access. They have a splash screen that you have to click "log-in" before you can get on and agree to their terms of service. But that has never required a purchase. Granted their web is useless for anything other than email and surfing to CNN and a few web sites as anything of interest seems to be block and they also seem to block ports like FTP/SSH starting a little over a year ago.
Coffee shop I'm setting at now requires you purchase a drink for 3 hours of internet time. And has for years because it is a college town and it's a popular study hang out. If you didn't you'd get people come in, sit down for 6 - 8 hours, and never buy anything. Meanwhile paying customers are leaving and going somewhere else because there aren't any seats available. (They had free wifi when they first opened. And Free was costing them too much in lost revenue. At least $150 - $200 a day.)
Fourth, if you're writing a plain web app, however fancily mobile-enhanced, how are you going to make use of the cool features of different phones? The iPhone has a camera, accelerometers, GPS, and multitouch. I admit I'm not terribly well-versed in the features of other smartphones, but a) do they all have these? b) can you access them from web apps? and c) can you access them all in the same way from web apps?
PhoneGap?
I develop and sell a program for $50. You buy the program, copy it say crack the 'copy protection', and sell it as MyProgram for $40. Can you not see the problem with this?
While I agree that Copyright and Patent system needs reform, it is still needed. If I couldn't develop and sell my program with some kind of protection, the chances are I'm not going to invest the time to do it.
Personally I think we need two IP system. One for Creative works (art/books/etc) and another one for Business related things. I agree with artists, writers, etc. retaining copyright for the duration of their life times, no longer. For businesses, they should be granted a provisional patent for 7 years. If they aren't producing a product with that patent by then, it expires. If they are, they get another 7 years of protection.
So, basically, you're mad that it won't play your pirated movies out of the box? I've got karma to burn, so time to feed the trolls.
Yeah, I'm going to assume that you really have no idea how the encoder/decoder market works or video production in general. Basically, there are companies that create codecs like DivX and then require payment to encode and distribute videos using their codec because of patents. In order for Apple to ship DivX, they would have to license DivX from DivX. Why should apple pay the licensing fee when they can direct users here: http://www.divx.com/en/mac and the user can download and install for free? I mean the last time I dealt with Windows I remember having to go and fetch the DivX Codec. And the last time I set up a new Mac, QuickTime popped up when it could find the Codec, knew what kind of file it was, and provided a link to the Divx site to go download it. All of this has to do with Licensing. It's really more of a legal issue than a technical one. So how is software licensing and patents preventing them from shipping every codec known to man Apple's fault again? Because unlike a lot of F/OSS projects, Apple can't be 100% based out of Hungry or France to circumvent these licensing restrictions. The do business in the United States and other countries that do recognize and enforce these copyright treaties. Technically, downloading and using VLC and FFMPEG in the United States constitutes infringement on various Codec patents, but that's a topic for another thread and another day. Also I would check on Windows. If it shipped with DivX, chances are that was added by the PC vender in a 3rd party deal, not by Microsoft. (Actually I don't pay attention to who is licensing what these days).
MKV isn't a format, it's a container. Just like .MOV is a container as is .AVI. The quality inside a .MKV, .AVI, or .MOV is all dependent upon the compression and codecs used not the container format. And the last time I tried playing MKV files, the program took up WAY too many system resources. The only place that I've seen MKV as a popular format is with Azurus/Vuze. In fact, I've never seen it used outside of Vuze. Occasionally you'll see it in a torrent, but they are mostly DivX/Avi.
The fact that macs can't play Blu-Ray or HD-DVD has to do with the fact that Apple doesn't ship any models with said player. MAC LACKS THE HARDWARE NEEDED IN ORDER TO READ EITHER FORMAT That has nothing to do with Quicktime and Codecs. When I worked in video production, nobody was adopting either optical format. It was clear to many of us that Optical Media was going the way of the floppy disk and this was in 2003. By the time HD-DVD or Blu-Ray won the format wars, it would be too late, we'd see everything delivered via a digital stream. The only question was, is it going to be an iTunes like store where you buy the item to a set top box/hard drive or whether it was going to be via streaming like Hulu or Netflix. So far it looks like there is a market for both. But time will tell.
And I see a lot of people electing to skip the Blu-Ray player in favor of a media PC/Mac hooked up to their TV and streaming Netflix or Hulu to their TV. I've been using a Mac MIni for this purpose since 2005. I know I have. To me, the extra quality for twice the price of a DVD plus the cost of the player isn't worth it to me. I'll put up with watching my TV shows from Hulu when I have extra time at my connivence, even if the quality isn't the same as on HD. It' good enough for me and plenty of other people too.
Now, I have to say that I'm generally happy with Quicktime X. Quicktime had been languishing for years and this was a much needed up date. It runs faster and smoother than the previous versions.
It doesn't kill OSS, it kills the use of GPL'ed code. With our proprietary products, we were extremely careful that if we did use other code/libraries that they were either BSD or MIT style licenses.
Why do people forget that OSS is more than just the GPL?
Well, firefox needs to get its act together and remember what their original purpose was because I've noticed a lot of average users complaining about the last couple Mozilla releases being buggy and slow across all platforms. On the windows side, quite a few have already flocked to Chrome an a few to Opera. OSX, a lot of folks have gone back to Safari.
I think it's past my bed time.
But now I have the sudden urge to work on a new DARPA grant proposal....
Seems like I remember Apple being this great white knight circa 2004/2005 and then after a couple years they are "evil" and "propitiatory", yet Apple is one of the more OSS friendly companies out there by buying and supporting CUPS (Those who like the fact their printer works with Linux has Apple to thank for that), and then Webkit, and a number of other things.
Basically they got the same treatment that Google has the past couple years, but now it looks like the tide is starting to turn as I predict that 2010 will the year that Google becomes widely regarded as "evil" around /.
I'm waiting for full Cylon Resurrection. They kill you and you download into a new body....
Depends on the type of shares they purchased. If they were preferred shares, you get first dibs on dividends, but no votes. If they were Common Shares, then eBay would have a say in the vote of board of directors, etc..
At our company employees do get their own cell phone, company chips in $30 or $60 a month depending on whether they need a data plan (which we cover the cost of) into the first paycheck of the month.
Works extremely well.
I don't really see the point of this in the real world. I could see where this could be useful where we would have 1 phone and test in Windows Mobile and Android on one hardware platform. Outside of that, I see no real value.
That's why we're very careful that any libraries we use in distributed software are licensed under BSD or MIT style licenses.
That depends. 60% of orders from our website still comes from customers using MSIE 6. Until that changes, I will continue to curse, kick, and scream that MSIE is broken.
Honestly, I could heat my old 1 bedroom apartment with my Quad-Core G5 powermac tower. I set thermostat to 60 and the only time the heat ran that winter was when it was less than 25 degrees outside.
I worked for a company about 10 years ago that had a set of video editing tools for NT and IRIX. We knew SGI was dying and looked into porting our tools to Linux and we re ran into the same problem. So we officially supported Red Hat 4/5 at the time. It also happened to work with SuSE as well. Sales to Linux based systems was less than 10% of all sales and was accounting for more than 35% of all Technical support requests. And a lot of the support requests were, "Why don't you support my custom leet slackware userland with debian kernel?" If you billed that out in man hour cost in tech support vs. sales, we lost like $25k on the Linux version. Probably more if you included the 8 of us that worked on that version.
We supported Linux until OSX 10.1 was released and ported the tools over to Mac. For a year we supported Windows, Linux and Mac, but it was pretty clear that our IRIX customers were largely going to Apple or Windows as their replacement systems. When I left the company, sales of the linux version was down to 5% and 40% of sales/upgrades were for OSX and the rest windows. I know I looked a year or so later and they had dropped the linux version all together.
Of course, it would help if they asked what the users actually do.
Bingo.
We had the advantage of a small business owner wanting our software developed because he thought "It should work like this". So we made it work like that and a lot of other small business owners found it to make sense and relatively easy to use. There were a couple quirks, but that's not good enough. Not for me.
And this is where so many others fails. After the phase 1 deployment of our product (about 100 installs), I drove/flew around to our customers 6 months later, stopped by in person and asked as the first question: "What doesn't work?" followed by "How can it work better?"