From his wording, it sounds as though this was simply just an out-of-tree kernel driver, which was not highly maintained, (Out of tree kernel drivers require an enormous amount of work to maintain, so they often don't get maintained properly. This is why out of tree kernel drivers are frowned upon. Drivers in the kernel tree automatically have basic maintenance (updates to code when API's change), such that only minor work is needed to maintain the driver.)
This really is not be a license issue, or anything that you could do much to fix (except perhaps pester Digium to get the driver into the kernel tree.) It may be worth reminding companies that working with the community when they use or modify Open Source software can save time and effort in some cases. Kernel drivers are a good example of that, but far from the only one.
Besides, there is no need to use hydroponics. Aeroponics generally works just fine. While I have no idea what they would be planning to grow in these places, trees are definiately not part of the plan. The two crops we use the most seem to be wheat and corn (breads and High Fructose corn syrup). Now very few people eat raw wheat, and corn is used for animal feed and HFCS much more than it is eaten directly, neither would be good candidates for such a building, since they require significant processing, which cannot occur very easily in the heart of a city.
So we are left with the small vegetables (carrots, beans, peas, etc), tomatoes, potatoes, and small fruits. Those are what would be produced in such a place.
The corn, wheat, sugar, and tree-based fruits would still be mainly real farmland. (With the corn and wheat getting processed before being brought into the heart of the city.)
At least, that is why seems most likely to me to happen.
All HTC phones phones have real standard USB connections.
Now, there is something special about drawing more than 1/2 amp (maximum allowed by USB standard), as the charging station can distinguish between the supplied 1 amp charger, and other power sources (real usb, or smaller chargers).
This much I do know: It will charge at at least 1/2 amp from dumb chargers, or will charge at whatever current it can negotiate from a usb host. It will charge at one amp with the provided cable. It might charge at 1 amp with third party chargers, or might stop at 1/2 amp. It might be willing to use more than 1/2 amp if a usb host provides it.
Any which way, half-amp is still a decent charging current, unless you are trying to place a call while charging.
The downside of HTC phones is that many of them use a very non-standard headset connector. (An extension to the mini-usb port).
The latest 3d games? Try 4 year old games that were not even grpahically top of the line at the time of release. My younger cousin received a laptop for Christmas with the Intel GMA 950 chip. It was unable to play The Sims 2 without issue, which as you can guess is not the most demanding of 3d games. I was not surprised that some of the pixel shading effects were omitted, but it visibly had trouble keeping a framerate of 25. Not to mention the game crashes when using that chip. (This is probably a driver issue, with the driver corrupting game memory when trying to load textures, or something of the sort.)
Granted that that chip is known for being crap, and Intel offers much better ones.
(On the other hand, by ATI card is known to have periods of time on some games (Half-life 2 Ep 1) with a frame rate less than 5. In s few cases the frame-rate dropped below 1 fps! I suspect it was either slightly defective, or Windows was at fault here. )
Point taken. But considering how little good content there is on cable, it is a legitimate concern for those considering switching to OTA, how much of that content actually remains.
Interesting. My understanding was that most companies would broadcast primary content from -001, and use the other possibilities for alternate content. (Show re-runs, etc) and perhaps smaller local style channels.
For local channels consider that some small towns have simple channels available that display high-school basketball games, and the like, the rest of the time broadcasting literally a PowerPoint presentation of a few textual slides in a a continuous loop, with a local radio station rebroadcast as the audio, or a Muzak subscription/equivalent.
Some PBS stations will likely become secondary channels of the larger network affiliates, as this would be substantially less expensive, allowing the station's limited funding to be used more for programming.
Is that 42 primary channels, or is that including subchannels? If those are all primary channels, than that is an awful lot. Also, surely only a few of those are major networks. What are the rest? Do they have much good content?
AIUI the android support for any HTC devices is currently limited. On the Titan (a.k.a. 6800 or Mogul) (not to be confused with the TyTN), the system runs, but is basically not usable yet due to lack of drivers for a fair amount of the hardware, but phone calls do work.
Also HTC is nearly the exclusive manufacturer of Windows Mobile PPC phones, (as opposed to the touchscreen less Windows Mobile "Smartphones"). I'm aware of no more than five Windows Mobile PPC phones made by other companies.
Handango is currently severely broken, now that it requires you to specify which mobile device you are using, even if all that is relevant is what version of the OS you are using.
I should be able to specify that I want to see All applications compatible with Windows Mobile 6 PPC (as opposed to Windows Mobile Smartphone which ironically refers to phones without touchscreens (i.e. what most people would call dumb-phones)), including the applications that require the presence of a phone.
As it is right now, if I go to select my device it is listed in the system under two different names (same phone hardware, but different branding, and small case-only differences). Both of these are include only a subset of all compatible software Handango sells, and more annoyingly both names for the same phone do not result in the same list of applications.
My understanding was Texas had no partiticipation in the Ez-Pass network. That it to say, neither the Department of transportation, nor any transit authority is an EZ-pass member (and thus producing EZ-pass branded devices) or an EZ-pass partner (making it effectively part of the EZ-pass network despite the difference in transponder name.)
The wiki indicates that the TxDOT makes a TxTag transponder that TxDot and CTRMA use on their highways.
EZ Tag is used by HCTRA and FBCTRA.
TollTag is used by NTTA.
These systems are intercompatible, so they make up the Greater TxTag Network.
But there is no compatibility between the EZPass and TxTag networks.
Most of those issues are primarily issues with third party autoconf m4 modules, or with people who don't know what they are doing adding functionality to configure.ac. That is not really autoconf's fault. (The fact that it is easy to mess things up if you don't know what you are doing might be autoconf's fault). And proper use of autoconf can and does make portability easier, although I will admit little outside the GNU project uses autoconf properly, and there are even exceptions within the GNU project.
The worse issue is that many of the automake m4 scripts were buggy, running dozens of tests not useful to the package in question. (One of many examples: I often see configure scripts checking if I have a fortran compiler, despite being simple C packages.)
You are missing the point that Amazon is selling the works as both text and audio.
No they are selling the works as text. The also sell a device that can convert arbitrary text into an audio version of the text. The user is the one who makes the audio version.
This is not unlike a device that scans sheet music and attempts to produce an audio rendition of the music denoted on it (but obviously would not be nearly as good as a man made audio version).
If I have the right to sell sheet music version of some songs, then there should be no problem with me selling sheet music, as well as this device, assuming my sheet music is just as had been provided to me to sell. I would not be selling an audio version of the music. There is a big difference between me selling a cd, and selling sheet music and the device.
Now, if I embedded additional information into the sheet music, beyond what the engraver put there, such as to give hints to this device about playing the song better, then I would be doing something wrong, because the sheet music with embedded hints would definately be a derived work of the original sheet music.
Indeed, in such an event, one could reasonably argue that I am selling an audio version of the work. (If the hints were a compressed version of the the difference between what the device would normally output for this music, and what the desired output would be, then clearly this would be both sheet music and an audio storage medium.
Now if Amazon embedded extra data like pronounciation hints, temp, maarking dialog with the character's name so that the tts engine coulduse a distinct voice for each character, and or other similar metadata, then they may be selling both a text version and an audio version. Otherwise they are selling a text version and a device,
Most apt-based distributions discourage upstream from producing any binaries for the distribution in question. They should produce a source tarball, and the distribution will be responsible for packaging it, including any distribution specific patches needed. Distributions will generally disable any auto-update feature, because it will likely fail, and if it does not fail, it is likely to end up changing files in a way disruptive to the package management system.
So it is no loss that MoCo and Sun don't make Linux binaries for Firefox and OOo.
As for your concerns about old versions of packages in Ubuntu, that is what Ubuntu Backports is there for. Debian/Ubuntu Backports usually have the latest versions of most end-user oriented packages like Firefox and OOo available.
That said, I can can sympathize with you on the problem of system administrators not being willing to install backports, leaving users with old versions of important packages. At my University the Linux labs are running RHEL 4, with version 1.X of OOo! There are currently no plans to upgrade anything until the release of RHEL 6. Needless to say I avoid using those machines when possible.
An above poster suggested that they licensed the Exchange ActiveSync protocol, which would allow Android phones to grow support for syncing contacts/calendar-events with your exchange server, and receiving push email. I have no doubt that a patent is involved, but the licensing also mostly likely included protocol documentation and permission to implement such a system. That makes good sense to me, so I'd put money on this being what was licensed. This has nothing to do with the PCPhone ActiveSync protocol.
True, but it is worth noting that Wine does contain some more emulator-like features than many other API reimplementation projects. These features are a necessary component of being able to run existing binaries. More specifically, they need to support existing binaries that are not targeted at a virtual machine.
Nevertheless it is definitely not an emulator. It might validly be called a simulator, as host-native implementations of embedded device APIs are often labeled simulators, to distinguish them from emulators, but that would be stretching the terminology quite a bit too.
Besides, not all virus scanners can search the contents of all installer types, and the installer is what was passed to the site. The results of the installation will probably trigger more virus scanners.
For what it is worth, the set blaster line instructs old DOS soundblaster drivers to use certain settings, which Window's NT's VDM supports. Please keep in mind that this is essential for proper operation of sound in certain 16-bit Windows applications (read Windows 3.X apps).
I know the.NET drill. I recent was trying to install a.Net SP and found it failed. I had to try to uninstall all old.NET installations (Excpet for the one that is considered an OS component). Since some of the installers were missing, I had to use the special.NET obliterating tool Microsoft makes available. I then re-install All the recent.Nets. Later a.Net 2.0 app complains with that crytpic error meesaage on startup that it could not find.NET 2.0. It turns out something had killed some random.Net related registry entries. So I try to re-install all the recent.Nets to fix this. Well some of the installers crash (uncaught exception) rather than propperly install. So I'm probably going to have to blow away all of.Net and start again.
And this was on XP! Sigh. And is it just me or is it odd that 1.1 was a replacement for 1.0, and 2.0 was a replacement for 1.1, but 3.0 was just a small collection of add-on assemblies for 2.0, and 3.5 is add-on assemblies for 3.0?
But that is fairly true of most many free or open source projects. There is the main release that people are encouraged to use, and many other development branches on which real development is done. The way I see it, nothing could be done to make the mainline releases more official, as they are as official something can get in such a project. The Linux Kernel in this case in unusual only in the sheer number of development branches in existence. (I prefer the term branches for friendly forks, especially those intended to have most or all the changes merged into the mainline (trunk) at some point. But I won't dispute that they are forks.)
The Heros was good, the Hulu commercial was mildly amusing, and the H&R block Death & Taxes was not terrible, but in general the rest were abysmal. Most of the commercials were significantly worse than the average regular season commercial which is saying something.
Consider Fords advertisement about a truck with "A full-sized tranny". I'm not sure about you, but that is really not what I want in a truck. I really don't go that way.
Ender's Game may be my favorite novel of all time.
But he has the name wrong. Ender is not a lord. If we want to be techincal his full title would be something like:
The late Adm. Andrew "Ender" Wiggen "The Xenocide" IF (Ret.) (Original) Speaker for the Dead.
(I'm guessing the Speaker for the Dead would be abbreviated like most name suffixes, in some sort of standardized way, But I'm not sure what it would be.)
The problem with LSB is that it is really meant only to make life easier for vendors of proprietary software. If your software is Free or Open Source, you just publish official upstream source tarballs, and let the distributions create the binaries with any needed customizations. (Debian is quite good at this, and Gentoo's system for packages is also extraordinary. I'm insufficiently familiar with RPM based distros to comment on them.)
The distributions are good at that. The main philosophy of most distributions is that users do not install software packages published on the web or from a purchased CD (like is common for Windows or OS X), but simply download and install any programs desired from the distribution. If you need something especially esoteric, you simply compile it from source yourself. Standards like LSB don't do that much to help this. The FHS helps, which is why most distributions follow it quite closely, and encourage upstream to do the same.
To distros the LSB is fluff that only really helps proprietary software companies, and many people involved see that as a neutral or even negative thing, so prefer to spend their time on other things.
Don't get me wrong. I think there are areas where more work is really needed. For example, take fonts. There pretty much no reason to have fonts install on most system in any format other than the opentype font format. (Freetype fonts are in Opentype format already). The OpenType format has support for fixed sized fonts in the rare event those are actually needed. It supports both cubic Bezier fonts (Postscript fonts), and quadratic Bezier fonts. Overall it can basically support any fonts losslessly (including the kerning/hinting information). (Except perhaps for MetaFont fonts).
Yet when I install Ghostscript my distro installs many Type-1 fonts. When I install X11 my distro installs old bitmapped fonts. Why not use the ISO standardized OpenType?
I'm pretty sure the packaged releases found at http://kernel.org/pub/linux/kernel/v2.6/* are considered the official sources, by pretty much everybody involved. Now, realistically most people deeply familiar with the kernel will pull from git rather than just download the published tarballs, but the are still the most official releases. (I'm really not sure if the are just periodic packaging of Linus's tree or some special staging tree. I don't follow kernel development closely enough to know that.)
Of course recent 2.6 kernels work far better with initramfs, since by default all modern kernels have a built-in ramfs image, and external initram fs images (optional, but can be supplied via the old initrd mechanism) just get merged with the internal one on boot.
This is really nice for embedded systems since the kernel can thus contain all of the root filesystem internally (single file), and also obliterates the need for a external file for LVM use.
I recommend reading linux-2.6/Documentation/filesystems/ramfs-rootfs-initramfs.txt if you are not familiar with this, as it is quite nice.
I did only limited testing. I'm guessing there are bugs in Google's filter system or something really screwy going on with filter interactions. Such a filter should work, as you will surely agree, so the fact that it does not surprises and disappoints me.
I apologize for wasting your time with this. I certainly was only trying to help.
For what it is worth, one can avoid having to use the Webmail interface by using GMail's Mutant somewhat IMAP compatible "IMAP" interface, but that certainly has its own share of issues. This is all rather unfortunate.
Slashdot Mirror
From his wording, it sounds as though this was simply just an out-of-tree kernel driver, which was not highly maintained, (Out of tree kernel drivers require an enormous amount of work to maintain, so they often don't get maintained properly. This is why out of tree kernel drivers are frowned upon. Drivers in the kernel tree automatically have basic maintenance (updates to code when API's change), such that only minor work is needed to maintain the driver.)
This really is not be a license issue, or anything that you could do much to fix (except perhaps pester Digium to get the driver into the kernel tree.) It may be worth reminding companies that working with the community when they use or modify Open Source software can save time and effort in some cases. Kernel drivers are a good example of that, but far from the only one.
Besides, there is no need to use hydroponics. Aeroponics generally works just fine. While I have no idea what they would be planning to grow in these places, trees are definiately not part of the plan. The two crops we use the most seem to be wheat and corn (breads and High Fructose corn syrup). Now very few people eat raw wheat, and corn is used for animal feed and HFCS much more than it is eaten directly, neither would be good candidates for such a building, since they require significant processing, which cannot occur very easily in the heart of a city.
So we are left with the small vegetables (carrots, beans, peas, etc), tomatoes, potatoes, and small fruits. Those are what would be produced in such a place.
The corn, wheat, sugar, and tree-based fruits would still be mainly real farmland. (With the corn and wheat getting processed before being brought into the heart of the city.)
At least, that is why seems most likely to me to happen.
All HTC phones phones have real standard USB connections.
Now, there is something special about drawing more than 1/2 amp (maximum allowed by USB standard), as the charging station can distinguish between the supplied 1 amp charger, and other power sources (real usb, or smaller chargers).
This much I do know:
It will charge at at least 1/2 amp from dumb chargers, or will charge at whatever current it can negotiate from a usb host. It will charge at one amp with the provided cable. It might charge at 1 amp with third party chargers, or might stop at 1/2 amp. It might be willing to use more than 1/2 amp if a usb host provides it.
Any which way, half-amp is still a decent charging current, unless you are trying to place a call while charging.
The downside of HTC phones is that many of them use a very non-standard headset connector. (An extension to the mini-usb port).
The latest 3d games? Try 4 year old games that were not even grpahically top of the line at the time of release. My younger cousin received a laptop for Christmas with the Intel GMA 950 chip. It was unable to play The Sims 2 without issue, which as you can guess is not the most demanding of 3d games. I was not surprised that some of the pixel shading effects were omitted, but it visibly had trouble keeping a framerate of 25. Not to mention the game crashes when using that chip. (This is probably a driver issue, with the driver corrupting game memory when trying to load textures, or something of the sort.)
Granted that that chip is known for being crap, and Intel offers much better ones.
(On the other hand, by ATI card is known to have periods of time on some games (Half-life 2 Ep 1) with a frame rate less than 5. In s few cases the frame-rate dropped below 1 fps! I suspect it was either slightly defective, or Windows was at fault here. )
I have Cable TV. Does it have much good content?
Point taken. But considering how little good content there is on cable, it is a legitimate concern for those considering switching to OTA, how much of that content actually remains.
Interesting. My understanding was that most companies would broadcast primary content from -001, and use the other possibilities for alternate content. (Show re-runs, etc) and perhaps smaller local style channels.
For local channels consider that some small towns have simple channels available that display high-school basketball games, and the like, the rest of the time broadcasting literally a PowerPoint presentation of a few textual slides in a a continuous loop, with a local radio station rebroadcast as the audio, or a Muzak subscription/equivalent.
Some PBS stations will likely become secondary channels of the larger network affiliates, as this would be substantially less expensive, allowing the station's limited funding to be used more for programming.
Is that 42 primary channels, or is that including subchannels? If those are all primary channels, than that is an awful lot. Also, surely only a few of those are major networks. What are the rest? Do they have much good content?
AIUI the android support for any HTC devices is currently limited. On the Titan (a.k.a. 6800 or Mogul) (not to be confused with the TyTN), the system runs, but is basically not usable yet due to lack of drivers for a fair amount of the hardware, but phone calls do work.
Also HTC is nearly the exclusive manufacturer of Windows Mobile PPC phones, (as opposed to the touchscreen less Windows Mobile "Smartphones"). I'm aware of no more than five Windows Mobile PPC phones made by other companies.
Handango is currently severely broken, now that it requires you to specify which mobile device you are using, even if all that is relevant is what version of the OS you are using.
I should be able to specify that I want to see All applications compatible with Windows Mobile 6 PPC (as opposed to Windows Mobile Smartphone which ironically refers to phones without touchscreens (i.e. what most people would call dumb-phones)), including the applications that require the presence of a phone.
As it is right now, if I go to select my device it is listed in the system under two different names (same phone hardware, but different branding, and small case-only differences). Both of these are include only a subset of all compatible software Handango sells, and more annoyingly both names for the same phone do not result in the same list of applications.
My understanding was Texas had no partiticipation in the Ez-Pass network. That it to say, neither the Department of transportation, nor any transit authority is an EZ-pass member (and thus producing EZ-pass branded devices) or an EZ-pass partner (making it effectively part of the EZ-pass network despite the difference in transponder name.)
The wiki indicates that the TxDOT makes a TxTag transponder that TxDot and CTRMA use on their highways.
EZ Tag is used by HCTRA and FBCTRA.
TollTag is used by NTTA.
These systems are intercompatible, so they make up the Greater TxTag Network.
But there is no compatibility between the EZPass and TxTag networks.
Most of those issues are primarily issues with third party autoconf m4 modules, or with people who don't know what they are doing adding functionality to configure.ac. That is not really autoconf's fault. (The fact that it is easy to mess things up if you don't know what you are doing might be autoconf's fault). And proper use of autoconf can and does make portability easier, although I will admit little outside the GNU project uses autoconf properly, and there are even exceptions within the GNU project.
The worse issue is that many of the automake m4 scripts were buggy, running dozens of tests not useful to the package in question. (One of many examples: I often see configure scripts checking if I have a fortran compiler, despite being simple C packages.)
You are missing the point that Amazon is selling the works as both text and audio.
No they are selling the works as text. The also sell a device that can convert arbitrary text into an audio version of the text. The user is the one who makes the audio version.
This is not unlike a device that scans sheet music and attempts to produce an audio rendition of the music denoted on it (but obviously would not be nearly as good as a man made audio version).
If I have the right to sell sheet music version of some songs, then there should be no problem with me selling sheet music, as well as this device, assuming my sheet music is just as had been provided to me to sell. I would not be selling an audio version of the music. There is a big difference between me selling a cd, and selling sheet music and the device.
Now, if I embedded additional information into the sheet music, beyond what the engraver put there, such as to give hints to this device about playing the song better, then I would be doing something wrong, because the sheet music with embedded hints would definately be a derived work of the original sheet music.
Indeed, in such an event, one could reasonably argue that I am selling an audio version of the work. (If the hints were a compressed version of the the difference between what the device would normally output for this music, and what the desired output would be, then clearly this would be both sheet music and an audio storage medium.
Now if Amazon embedded extra data like pronounciation hints, temp, maarking dialog with the character's name so that the tts engine coulduse a distinct voice for each character, and or other similar metadata, then they may be selling both a text version and an audio version. Otherwise they are selling a text version and a device,
Most apt-based distributions discourage upstream from producing any binaries for the distribution in question. They should produce a source tarball, and the distribution will be responsible for packaging it, including any distribution specific patches needed. Distributions will generally disable any auto-update feature, because it will likely fail, and if it does not fail, it is likely to end up changing files in a way disruptive to the package management system.
So it is no loss that MoCo and Sun don't make Linux binaries for Firefox and OOo.
As for your concerns about old versions of packages in Ubuntu, that is what Ubuntu Backports is there for. Debian/Ubuntu Backports usually have the latest versions of most end-user oriented packages like Firefox and OOo available.
That said, I can can sympathize with you on the problem of system administrators not being willing to install backports, leaving users with old versions of important packages. At my University the Linux labs are running RHEL 4, with version 1.X of OOo! There are currently no plans to upgrade anything until the release of RHEL 6. Needless to say I avoid using those machines when possible.
An above poster suggested that they licensed the Exchange ActiveSync protocol, which would allow Android phones to grow support for syncing contacts/calendar-events with your exchange server, and receiving push email. I have no doubt that a patent is involved, but the licensing also mostly likely included protocol documentation and permission to implement such a system. That makes good sense to me, so I'd put money on this being what was licensed. This has nothing to do with the PCPhone ActiveSync protocol.
True, but it is worth noting that Wine does contain some more emulator-like features than many other API reimplementation projects. These features are a necessary component of being able to run existing binaries. More specifically, they need to support existing binaries that are not targeted at a virtual machine.
Nevertheless it is definitely not an emulator. It might validly be called a simulator, as host-native implementations of embedded device APIs are often labeled simulators, to distinguish them from emulators, but that would be stretching the terminology quite a bit too.
Besides, not all virus scanners can search the contents of all installer types, and the installer is what was passed to the site. The results of the installation will probably trigger more virus scanners.
For what it is worth, the set blaster line instructs old DOS soundblaster drivers to use certain settings, which Window's NT's VDM supports. Please keep in mind that this is essential for proper operation of sound in certain 16-bit Windows applications (read Windows 3.X apps).
I know the .NET drill. I recent was trying to install a .Net SP and found it failed. I had to try to uninstall all old .NET installations (Excpet for the one that is considered an OS component). Since some of the installers were missing, I had to use the special .NET obliterating tool Microsoft makes available. I then re-install All the recent .Nets. Later a .Net 2.0 app complains with that crytpic error meesaage on startup that it could not find .NET 2.0. It turns out something had killed some random .Net related registry entries. So I try to re-install all the recent .Nets to fix this. Well some of the installers crash (uncaught exception) rather than propperly install. So I'm probably going to have to blow away all of .Net and start again.
And this was on XP! Sigh. And is it just me or is it odd that 1.1 was a replacement for 1.0, and 2.0 was a replacement for 1.1, but 3.0 was just a small collection of add-on assemblies for 2.0, and 3.5 is add-on assemblies for 3.0?
But that is fairly true of most many free or open source projects. There is the main release that people are encouraged to use, and many other development branches on which real development is done. The way I see it, nothing could be done to make the mainline releases more official, as they are as official something can get in such a project. The Linux Kernel in this case in unusual only in the sheer number of development branches in existence. (I prefer the term branches for friendly forks, especially those intended to have most or all the changes merged into the mainline (trunk) at some point. But I won't dispute that they are forks.)
The Heros was good, the Hulu commercial was mildly amusing, and the H&R block Death & Taxes was not terrible, but in general the rest were abysmal. Most of the commercials were significantly worse than the average regular season commercial which is saying something.
Consider Fords advertisement about a truck with "A full-sized tranny". I'm not sure about you, but that is really not what I want in a truck. I really don't go that way.
Ender's Game may be my favorite novel of all time.
But he has the name wrong. Ender is not a lord.
If we want to be techincal his full title would be something like:
The late Adm. Andrew "Ender" Wiggen "The Xenocide" IF (Ret.) (Original) Speaker for the Dead.
(I'm guessing the Speaker for the Dead would be abbreviated like most name suffixes, in some sort of standardized way, But I'm not sure what it would be.)
The problem with LSB is that it is really meant only to make life easier for vendors of proprietary software. If your software is Free or Open Source, you just publish official upstream source tarballs, and let the distributions create the binaries with any needed customizations. (Debian is quite good at this, and Gentoo's system for packages is also extraordinary. I'm insufficiently familiar with RPM based distros to comment on them.)
The distributions are good at that. The main philosophy of most distributions is that users do not install software packages published on the web or from a purchased CD (like is common for Windows or OS X), but simply download and install any programs desired from the distribution. If you need something especially esoteric, you simply compile it from source yourself. Standards like LSB don't do that much to help this. The FHS helps, which is why most distributions follow it quite closely, and encourage upstream to do the same.
To distros the LSB is fluff that only really helps proprietary software companies, and many people involved see that as a neutral or even negative thing, so prefer to spend their time on other things.
Don't get me wrong. I think there are areas where more work is really needed. For example, take fonts. There pretty much no reason to have fonts install on most system in any format other than the opentype font format. (Freetype fonts are in Opentype format already). The OpenType format has support for fixed sized fonts in the rare event those are actually needed. It supports both cubic Bezier fonts (Postscript fonts), and quadratic Bezier fonts. Overall it can basically support any fonts losslessly (including the kerning/hinting information). (Except perhaps for MetaFont fonts).
Yet when I install Ghostscript my distro installs many Type-1 fonts. When I install X11 my distro installs old bitmapped fonts. Why not use the ISO standardized OpenType?
I'm pretty sure the packaged releases found at http://kernel.org/pub/linux/kernel/v2.6/* are considered the official sources, by pretty much everybody involved. Now, realistically most people deeply familiar with the kernel will pull from git rather than just download the published tarballs, but the are still the most official releases. (I'm really not sure if the are just periodic packaging of Linus's tree or some special staging tree. I don't follow kernel development closely enough to know that.)
Of course recent 2.6 kernels work far better with initramfs, since by default all modern kernels have a built-in ramfs image, and external initram fs images (optional, but can be supplied via the old initrd mechanism) just get merged with the internal one on boot.
This is really nice for embedded systems since the kernel can thus contain all of the root filesystem internally (single file), and also obliterates the need for a external file for LVM use.
I recommend reading linux-2.6/Documentation/filesystems/ramfs-rootfs-initramfs.txt if you are not familiar with this, as it is quite nice.
I did only limited testing. I'm guessing there are bugs in Google's filter system or something really screwy going on with filter interactions. Such a filter should work, as you will surely agree, so the fact that it does not surprises and disappoints me.
I apologize for wasting your time with this. I certainly was only trying to help.
For what it is worth, one can avoid having to use the Webmail interface by using GMail's Mutant somewhat IMAP compatible "IMAP" interface, but that certainly has its own share of issues. This is all rather unfortunate.