There's a fundamental difference here to the oft-repeated horse and carriage comparison; for the first time in history this will be a shift removing the fundamental freedom of travel from control of the people on board. Self driving cars are essentially rolling jail cells, which can be used to convey human cargo to the destination chosen by the establishment.
As modern society finds new and creative ways to remove human rights, with basic freedoms of speech increasingly curtailed with family friendly excuses like 'offensive speech to others', 'hate speech', 'dangerous speech', etc.; removing the basic ability to travel to unapproved locations, and tracking precisely where every citizen is going, are logical next steps, and explains the government's quick embracing of these technologies instead of the much more cautious approach such a radical shift would justify.
These types of technologies are curiously advocated for by otherwise smart people who fail to see they are racing the world towards a dystopian doom. There's a line between technologies that empower humans, and technologies that enslave them; and this is it.
I was very disappointed in this video. As a stimulant to begin a conversation, it's not entirely worthless, but I expected a few actual examples of plausible scenarios.
From PBS I expected a well researched topic and a linear progression through the history and facts of the topic. It ended up being some random ADD gibberish from a forum of random thoughts on the topic. There were no coherent thoughts on what such battles would actually be like.
"traded his Linux and Mac PCs in for a Windows 7 laptop and took on a newly created job at Microsoft designed to encourage collaboration between Redmond and open source communities"
If this is Microsoft's approach to open source and product collaboration in general; to have their main inter-relations representative stop using other products before taking the job; failure is certain. The correct approach would have been to add a Windows 7 laptop to his existing systems to better coordinate collaboration and gain experience actually using them together.
The issue of public self-signed certs seems best resolved by using Perspectives (http://www.cs.cmu.edu/~perspectives/firefox.html), which solves the man-in-the-middle problem using a distributed set of auditing servers to verify you are getting the same certificate others on the internet are.
This method has advantages over paying for a certificate from a CA vendor. It is possible for a determined man-in-the-middle attack to succeed without any errors on the client using social engineering or other measures to get a validly signed copy of the certificate for a site without being the actual site owner due to the lax verification measures used by some of these vendors.
Another common issue, companies should be creating their own CA certs and deploying them to clients in situations where the client is controlled (for example intranet sites), but most instead train their users to ignore these errors. See this example, (http://www.debian-administration.org/article/Creating_and_Using_a_self_signed__SSL_Certificates_in_debian) note that these basic instructions work on any operating system, not just Debian, using openssl at the command line.
I like the Belkin Cushtop http://www.belkin.com/laptopathome/cushtop/. The empty space in the middle keeps it from getting too hot, and the angle it places the keyboard and screen work well for me. It lacks space for using a mouse however.
Another example of the problems caused by no longer having an unstable development tree of the kernel comes to the forefront. Ever since that development model ended, there has been a hesitation to try new things which this article brings to the forefront. There used to be little hesitation to toss experiments into the development kernel, as since it wasn't a stable tree it could be taken out a few releases later if it didn't work out.
Not branching a separate development tree this time was one of the worst decisions Linus has ever made.
Your example explains the reason to use inheritance quite well; it is much simpler and easier to maintain than keeping track of many cross-referencing index tables and considerably easier to code for in middleware.
You're right that for generic items where you wouldn't need unique tables for custom attributes, your method would be much better. But whenever you need those tables anyway, why not inherit and gain the benefits of those cross reference tables without the complexity?
Here's an example of table inheritance's usefulness. Let's say we have a set of inventory items that share a common header. We would like to be able to easily search the common header set, for example for rfid tag sku's, then drill down to more specific attributes.
inventoryitems
-- sku
-- description tables inherits inventoryitems
-- size
-- color
-- top material lamps inherits inventoryitems
-- numbulbs
-- bulbtype
Now we can do a select on inventoryitems to search all of the sub-tables for the scanned item at the point of sale system where we don't need the more specific information yet we want to be able to quickly search all the subitems without doing tons of selects (obviously this is useful for more item types than my example of two).
This is exactly what Debian does within a stable release; only package updates that maintain the api are allowed, and is why people who just want to use the system really need to stay with the stable release and not use testing for critical day-to-day tasks.
Especially with the huge time between releases, I do not believe it is realistic to expect upgrades from one stable release to another to work cleanly. With any set of software it is generally much safer to do a clean install, then migrate your configuration over to the new setup than to risk your data and applications trying a major version upgrade. I don't think that this is a specific package-related issue, it is going to happen any time you do an api-changing upgrade.
Your api examining idea is interesting, but isn't specific to a package system or not having a package system; it could be implemented by having the package system check those apis for each dependency. Even then, though, there's not going to be a good way to see if those apis, even if they appear the same, have subtle intent differences which are not going to be easy to detect.
At any rate, I recommend clean installs and not upgrades for any major update of any operating system, unless it is a test or non-critical machine.
I think it will take time before users really get effective use out of spotlight. It takes time to get used to the concept of using the search box in the upper right in file dialogs, the file manager, etc. instead of drilling down through directories manually. Spotlight also doesn't remove the need to have things reasonably categorized in directories, but it makes things far easier to find in such well patterned layouts.
It doesn't replace directories, it just makes things easier to find in them. Once people get used to finding files in subdirectories via spotlight searches instead of drilling down, spotlight is vastly more useful.
To see where I am coming from, examine the view that porting open source applications to windows is a bad thing. An excellent posting on this viewpoint can be found here:
By encouraging people to extend a proprietary platform instead of an open one, you lock them in to that platform and discourage development and usage of open alternatives.
I don't believe that this is applicable for all software like many others do, however. I think there is a line between infrastructure and information applications that must be open and ones that serve special markets and are fair game for the proprietary software market.
As some extreme examples to exhibit the point, imagine if http had been a proprietary protocol, and the company in control of it now decided to charge a fee to use it. On the other side, look at Photoshop, a specialized graphics editing application that is useful and profitable for many, but whose existance as a proprietary application doesn't inhibit innovation in the graphics market.
I agree with APIs for open source and open standard applications. Where they start becoming problematic is when closed companies use them to encourage development on top of proprietary applications. Unfortunately google's desktop search tool is neither open source or an open standard.
I am well aware that I am heavily biased towards open source and/or open standards for basic infrastructure and information handling tools. I suppose the real question, at least to me, is where desktop search falls on the line.
My basis for the third party toolbars not being fully functional is pretty well summarized at http://googlebar.mozdev.org/pagerank.html, so you can clearly see that my basis is fact, not conspiracy theories. You can also find more factual information about their lack of platform support on their own FAQ, linked in this thread a few posts down.
As it is not evident by the facts, available on their own site, I find it hard to take hearsay about them bringing cross platform support to their applications into account. I will certainly reconsider if they post anything officially and/or actually come up with such applications, however.
And your platform argument comes once again to the reasoning behind making web pages cross-browser and cross-platform. Excluding users is not good business.
"Much the same functionality" is exactly the point. It doesn't have the same functionality.
I don't know where you've seen these announcements, it doesn't match google's own stance listed on their site at http://desktop.google.com/support/bin/answer.py?an swer=10073&topic=168 or http://desktop.google.com/support/bin/answer.py?an swer=10072&topic=168. Looks like they are just avoiding the issue to me. I do notice they have Firefox and Thunderbird support now. That rather makes sense, as you pointed out they direct users to those applications, and indeed they've hired some of the developers. Still no sign of OpenOffice support though. But they do have an API so that you can extend their closed-source application for them.
Ok, so standards support on web pages is stupid too. 90% plus uses IE, what's the point of supporting Firefox? Those web developers are just realizing market reality and ignoring the standards. I don't buy this argument. Just because it's more convenient to support one platform doesn't make it the right choice.
There are many cross platform toolkits available nowadays, wx, Java/SWT, even gtk, that they could be using to easily write apps that support all platforms with a minumum of additional work.
This is far more important for information services companies than other companies. Like online banking not supporting other platforms making those other platforms unusable for others, by making these apps windows only and continuing to do so, they lock people in to windows.
I chuckle every time I see that "don't be evil" mentioned. Let's look every non-web app they've released:
Google Toolbar: IE/windows only. They actively resist efforts by Mozilla to make a full-featured version including pagerank.
Google Desktop Search: Not only is it windows only, it only supports the major Microsoft apps on that platform. No OpenOffice, Thunderbird, or any other open application support.
Picassa: Windows only. No plans for a non-windows version announced.
Keyhole: Windows only. No plans for a non-windows version announced.
Ah, but this one is coming to google maps you say. Not even close. Look at the capabilities of the Keyhole app vs the ridiculously low res photos from google maps before you bring that up.
Any other company that releases such an array of single-platform apps would be blasted by geeks for being closed minded and evil.
In my experience this does not work; it's a radical change of hardware from the operating system's perspective, and if it even boots it still requires re-activation.
These types of ads are never as annoying as pop-up windows. You can always get rid of these floaters just by closing the tab or window the web site is in, or just hitting the back button. Pop-up windows sometimes appeared under your browser to haunt you later, opened many windows that needed to be closed separately, etc.
By forcing advertisers to use this type of in-window advertising, I believe pop-up blockers have accomplished their mission and put the control over the browsing experience back in the user's hands. You can now just hit back and forget about visiting the site if you decide the content isn't worth dealing with the forest of floaters.
I believe that web site designers have a legitimate right to control the look and feel of their page, as long as they stay within the expected bounds given to them by the user. The use of floaters keeps them within this expected window. If a site decides to use these floaters and it annoys the users enough not to look at their content, then it is up to the users to go to alternate sources and the designers to realize their site is horrible when their visitors never bother wading through the floaters to see the content. This is an open internet after all, if you don't want to go through the floaters it is now easy to hit back and get your data from another source by selecting another search result, an alternate link, etc.
As an avid VMWare user, it looks like this is going to bite me. I blow away the OEM windows install, install Linux, then install the copy of Windows that came with the PC in a VMWare session.
So far this has been easy, but it sounds like in the future I'm going to have to call to ask permission to do this.
Even Windows-using developers installing VMWare for application testing are going to end up being bitten by this one.
It's only free for non-commercial use; from the License Agreement:
Non-Commercial Use Only
Picasa Software is made available to you for your non-commercial use only. This means that you may use it at work or at home. But you first need to obtain Google's permission if you want to sell the Picasa Software or any information, services, or software associated with or derived from it, or if you want to modify, copy (except as provided below), license, or create derivative works from the Picasa Software.
The way this is worded, it looks like any pictures you edit or store in it (which would be "information" "associated or derived" using the software) can't be sold or used for any for-profit purpose. This certainly doesn't match my definition of "100% free".
I find it highly amusing how Microsoft whines about the GPL being viral, yet they release things like this; where you can develop code and applications, yet they aren't legal to distribute. This is far worse than the GPL, as apps touched by this trapware isn't safe to use for freedom OR commercial purposes. I wonder if this just a trap to try to suck in free software developers, as once the app has been touched by this IDE, it is illegal to distribute it.
Slashdot Mirror
There's a fundamental difference here to the oft-repeated horse and carriage comparison; for the first time in history this will be a shift removing the fundamental freedom of travel from control of the people on board. Self driving cars are essentially rolling jail cells, which can be used to convey human cargo to the destination chosen by the establishment.
As modern society finds new and creative ways to remove human rights, with basic freedoms of speech increasingly curtailed with family friendly excuses like 'offensive speech to others', 'hate speech', 'dangerous speech', etc.; removing the basic ability to travel to unapproved locations, and tracking precisely where every citizen is going, are logical next steps, and explains the government's quick embracing of these technologies instead of the much more cautious approach such a radical shift would justify.
These types of technologies are curiously advocated for by otherwise smart people who fail to see they are racing the world towards a dystopian doom. There's a line between technologies that empower humans, and technologies that enslave them; and this is it.
I was very disappointed in this video. As a stimulant to begin a conversation, it's not entirely worthless, but I expected a few actual examples of plausible scenarios.
From PBS I expected a well researched topic and a linear progression through the history and facts of the topic. It ended up being some random ADD gibberish from a forum of random thoughts on the topic. There were no coherent thoughts on what such battles would actually be like.
"traded his Linux and Mac PCs in for a Windows 7 laptop and took on a newly created job at Microsoft designed to encourage collaboration between Redmond and open source communities"
If this is Microsoft's approach to open source and product collaboration in general; to have their main inter-relations representative stop using other products before taking the job; failure is certain. The correct approach would have been to add a Windows 7 laptop to his existing systems to better coordinate collaboration and gain experience actually using them together.
What a prime ribbing.
The issue of public self-signed certs seems best resolved by using Perspectives (http://www.cs.cmu.edu/~perspectives/firefox.html), which solves the man-in-the-middle problem using a distributed set of auditing servers to verify you are getting the same certificate others on the internet are.
This method has advantages over paying for a certificate from a CA vendor. It is possible for a determined man-in-the-middle attack to succeed without any errors on the client using social engineering or other measures to get a validly signed copy of the certificate for a site without being the actual site owner due to the lax verification measures used by some of these vendors.
Another common issue, companies should be creating their own CA certs and deploying them to clients in situations where the client is controlled (for example intranet sites), but most instead train their users to ignore these errors. See this example, (http://www.debian-administration.org/article/Creating_and_Using_a_self_signed__SSL_Certificates_in_debian) note that these basic instructions work on any operating system, not just Debian, using openssl at the command line.
I like the Belkin Cushtop http://www.belkin.com/laptopathome/cushtop/. The empty space in the middle keeps it from getting too hot, and the angle it places the keyboard and screen work well for me. It lacks space for using a mouse however.
Another example of the problems caused by no longer having an unstable development tree of the kernel comes to the forefront. Ever since that development model ended, there has been a hesitation to try new things which this article brings to the forefront. There used to be little hesitation to toss experiments into the development kernel, as since it wasn't a stable tree it could be taken out a few releases later if it didn't work out.
Not branching a separate development tree this time was one of the worst decisions Linus has ever made.
Your example explains the reason to use inheritance quite well; it is much simpler and easier to maintain than keeping track of many cross-referencing index tables and considerably easier to code for in middleware.
You're right that for generic items where you wouldn't need unique tables for custom attributes, your method would be much better. But whenever you need those tables anyway, why not inherit and gain the benefits of those cross reference tables without the complexity?
Here's an example of table inheritance's usefulness. Let's say we have a set of inventory items that share a common header. We would like to be able to easily search the common header set, for example for rfid tag sku's, then drill down to more specific attributes.
inventoryitems
-- sku
-- description
tables inherits inventoryitems
-- size
-- color
-- top material
lamps inherits inventoryitems
-- numbulbs
-- bulbtype
Now we can do a select on inventoryitems to search all of the sub-tables for the scanned item at the point of sale system where we don't need the more specific information yet we want to be able to quickly search all the subitems without doing tons of selects (obviously this is useful for more item types than my example of two).
This is exactly what Debian does within a stable release; only package updates that maintain the api are allowed, and is why people who just want to use the system really need to stay with the stable release and not use testing for critical day-to-day tasks.
Especially with the huge time between releases, I do not believe it is realistic to expect upgrades from one stable release to another to work cleanly. With any set of software it is generally much safer to do a clean install, then migrate your configuration over to the new setup than to risk your data and applications trying a major version upgrade. I don't think that this is a specific package-related issue, it is going to happen any time you do an api-changing upgrade.
Your api examining idea is interesting, but isn't specific to a package system or not having a package system; it could be implemented by having the package system check those apis for each dependency. Even then, though, there's not going to be a good way to see if those apis, even if they appear the same, have subtle intent differences which are not going to be easy to detect.
At any rate, I recommend clean installs and not upgrades for any major update of any operating system, unless it is a test or non-critical machine.
I think it will take time before users really get effective use out of spotlight. It takes time to get used to the concept of using the search box in the upper right in file dialogs, the file manager, etc. instead of drilling down through directories manually. Spotlight also doesn't remove the need to have things reasonably categorized in directories, but it makes things far easier to find in such well patterned layouts.
It doesn't replace directories, it just makes things easier to find in them. Once people get used to finding files in subdirectories via spotlight searches instead of drilling down, spotlight is vastly more useful.
I hope IBM provided the source code as 80GB of unsearchable PDF's in a cursive font to throw off OCR software.
SCO would be bankrupt long before they were able to afford the manpower to go through that much unsearchable data manually.
Targus has a model that fits the Zaurus excellently, is available in most stores, and is quite inexpensive.
C HUS01
http://www.targus.com/us/product_details.asp?sku=
It even has a little sleeve inside for your wifi or other accesory card.
To see where I am coming from, examine the view that porting open source applications to windows is a bad thing. An excellent posting on this viewpoint can be found here:
e n-source-on-desktop.html
http://aseigo.blogspot.com/2004/12/how-to-kill-op
By encouraging people to extend a proprietary platform instead of an open one, you lock them in to that platform and discourage development and usage of open alternatives.
I don't believe that this is applicable for all software like many others do, however. I think there is a line between infrastructure and information applications that must be open and ones that serve special markets and are fair game for the proprietary software market.
As some extreme examples to exhibit the point, imagine if http had been a proprietary protocol, and the company in control of it now decided to charge a fee to use it. On the other side, look at Photoshop, a specialized graphics editing application that is useful and profitable for many, but whose existance as a proprietary application doesn't inhibit innovation in the graphics market.
I agree with APIs for open source and open standard applications. Where they start becoming problematic is when closed companies use them to encourage development on top of proprietary applications.
Unfortunately google's desktop search tool is neither open source or an open standard.
I am well aware that I am heavily biased towards open source and/or open standards for basic infrastructure and information handling tools. I suppose the real question, at least to me, is where desktop search falls on the line.
My basis for the third party toolbars not being fully functional is pretty well summarized at http://googlebar.mozdev.org/pagerank.html, so you can clearly see that my basis is fact, not conspiracy theories. You can also find more factual information about their lack of platform support on their own FAQ, linked in this thread a few posts down.
As it is not evident by the facts, available on their own site, I find it hard to take hearsay about them bringing cross platform support to their applications into account. I will certainly reconsider if they post anything officially and/or actually come up with such applications, however.
And your platform argument comes once again to the reasoning behind making web pages cross-browser and cross-platform. Excluding users is not good business.
"Much the same functionality" is exactly the point. It doesn't have the same functionality.
n swer=10073&topic=168 or http://desktop.google.com/support/bin/answer.py?an swer=10072&topic=168. Looks like they are just avoiding the issue to me. I do notice they have Firefox and Thunderbird support now. That rather makes sense, as you pointed out they direct users to those applications, and indeed they've hired some of the developers. Still no sign of OpenOffice support though. But they do have an API so that you can extend their closed-source application for them.
I don't know where you've seen these announcements, it doesn't match google's own stance listed on their site at http://desktop.google.com/support/bin/answer.py?a
Ok, so standards support on web pages is stupid too. 90% plus uses IE, what's the point of supporting Firefox? Those web developers are just realizing market reality and ignoring the standards. I don't buy this argument. Just because it's more convenient to support one platform doesn't make it the right choice.
There are many cross platform toolkits available nowadays, wx, Java/SWT, even gtk, that they could be using to easily write apps that support all platforms with a minumum of additional work.
This is far more important for information services companies than other companies. Like online banking not supporting other platforms making those other platforms unusable for others, by making these apps windows only and continuing to do so, they lock people in to windows.
I chuckle every time I see that "don't be evil" mentioned. Let's look every non-web app they've released:
Google Toolbar: IE/windows only. They actively resist efforts by Mozilla to make a full-featured version including pagerank.
Google Desktop Search: Not only is it windows only, it only supports the major Microsoft apps on that platform. No OpenOffice, Thunderbird, or any other open application support.
Picassa: Windows only. No plans for a non-windows version announced.
Keyhole: Windows only. No plans for a non-windows version announced.
Ah, but this one is coming to google maps you say. Not even close. Look at the capabilities of the Keyhole app vs the ridiculously low res photos from google maps before you bring that up.
Any other company that releases such an array of single-platform apps would be blasted by geeks for being closed minded and evil.
Thinkgeek has Soft Polo
I think it's a trick to tempt the lady-folk in your life to buy polo shirts for you on a subconscious level.
In my experience this does not work; it's a radical change of hardware from the operating system's perspective, and if it even boots it still requires re-activation.
These types of ads are never as annoying as pop-up windows. You can always get rid of these floaters just by closing the tab or window the web site is in, or just hitting the back button. Pop-up windows sometimes appeared under your browser to haunt you later, opened many windows that needed to be closed separately, etc.
By forcing advertisers to use this type of in-window advertising, I believe pop-up blockers have accomplished their mission and put the control over the browsing experience back in the user's hands. You can now just hit back and forget about visiting the site if you decide the content isn't worth dealing with the forest of floaters.
I believe that web site designers have a legitimate right to control the look and feel of their page, as long as they stay within the expected bounds given to them by the user. The use of floaters keeps them within this expected window. If a site decides to use these floaters and it annoys the users enough not to look at their content, then it is up to the users to go to alternate sources and the designers to realize their site is horrible when their visitors never bother wading through the floaters to see the content. This is an open internet after all, if you don't want to go through the floaters it is now easy to hit back and get your data from another source by selecting another search result, an alternate link, etc.
As an avid VMWare user, it looks like this is going to bite me. I blow away the OEM windows install, install Linux, then install the copy of Windows that came with the PC in a VMWare session.
So far this has been easy, but it sounds like in the future I'm going to have to call to ask permission to do this.
Even Windows-using developers installing VMWare for application testing are going to end up being bitten by this one.
The way this is worded, it looks like any pictures you edit or store in it (which would be "information" "associated or derived" using the software) can't be sold or used for any for-profit purpose. This certainly doesn't match my definition of "100% free".
I find it highly amusing how Microsoft whines about the GPL being viral, yet they release things like this; where you can develop code and applications, yet they aren't legal to distribute. This is far worse than the GPL, as apps touched by this trapware isn't safe to use for freedom OR commercial purposes. I wonder if this just a trap to try to suck in free software developers, as once the app has been touched by this IDE, it is illegal to distribute it.