Well, it seems like a lot of people would happily run financials via the web... our company is moving to that from installed Windows apps, and it wasn't my doing, it was our CFO.
I seem to remember a few years back being told by a Siebel sales guy that hosted software for sales CRM wasn't going to fly either, but I bought it anyway:-)
You've worded it gently, but this sounds like your former boss was a little corrupt:-)
When I worked for Mrs Queen, even with the most ehtical of intentions, we did have the issue that the open procurement system would result in unsuitable tenders (the process was run by procurement people who knew nothing about technology) and would cheerfully saddle you with nonsense to save a quid.
So, we would tend to do an informal market survey first, talk to vendors, make a buying decision based on normal commercial business criteria (is this vendor competent? Is the product any good?) then load the RFP in favour of the best solution.
The other issue that made things hard was that some procurements are inhrenetly single vendor - for example, what about renewing a maintenance contract for Sun servers? Sun's own Platinum maintenance is very well priced, so you'd be an idiot not to use them, but government procurement left unfettered will end up hiring Joe PC shop down the street.
They often fail to even check facts, and even when they do, the quality of their output is limited by the fact that journalsists are experts in writing, and not experts in the subject matter they write about, which they often don't really understand.
1. I'm the CTO at a leading vendor of high end online systems for non-profits. We have two CMS offerings. We may well be the "evil" vendor mentioned in this post:-)
2. I am the worlds biggest fan of open source - I'm posting here, on my work laptop which, yes, I run Linux on, because I find it to be a better fit for my personal needs (mixture of hardcore techie and traditional business user) than Windows.
We don't open source our software, we don't even ship it... we deliver everything hosted "Software as a Service"... why? Because 99% of non-profits don't have the IT resources to set up and run a complex, robust, clustered 24x7 web envrionment, and for the ones that do, we can do it better, faster, and cheaper, because we have economies of scale associated with doing it for over 1,000 clients instead of just one. The American Red Cross is like a Fortune 500, and has an IT department nearly twice the size of our company, but they use our solutions too.
Customers don't need to develop stuff, or even set up servers, to get our product running. We can deploy a new site as fast as they can work with us on the content and business level concerns. On occasions, we've deployed new sites in under a week.
We use almost all open source stuff in our production systems. The big exception is the database - we run Oracle, which costs a lot of $$$, because it *works*. We have the expertise to leverage this technology and make it work for our customers. Our customers can't afford to carry that skill set and staffing in house.
Our business exists because there is a true need for a packaged
solution
for non-profits. I'm sure the usual whiners on Slashdot will say that anyone who pays money for software is stupid or evil, but they're not - they just want to get stuff done.
It does sound like the original poster is thinking first about personal ideology of promoting open-source, and not about the business needs of his/her employer.
As many people have observed, this is nothing to do with OSS.... the real business decision here is "build vs buy", and it sounds as if the organization's ED and board have made a good call.
We too buy Dell servers bare or with a Linux preload (which we flash and reinstall anyway so we can load the RPMs we want).
The cost of the Windows 2003 Server license is indeed about $800. However, the article is about desktop/laptop Windows licenses... for a consumer machine, the only thing Microsoft provides is a licnese and a sticker, and Dell only pays $20 or so a machine.
Servers are a whole other world...
1. There is much more demand for Linux on servers 2. Generally, people like us who buy them know what they are doing 3. There isn't a need for drivers for a whole bunch of cheapo Tiwansese hardware like USB webcams and printers
Dell has been offering Linux either bundled with, or pre-installed with, on **servers** for a long time - they even have a 3 way agreement with Red Hat and Oracle to certify particular configurations and support each others products without any finger pointing.
Given the relatively limited bandwidth going in and out of Australia, and that 99% of the world's websites are at the wrong end of that, there is arguably some justification for this. Still inconvenient though. </devilsadvocate>
.... made in Europe, with Stevie Wonder driving (IIRC) a 306 convertible on a beach? Is that more or less dangerous than your toddler taking the wheel.
1. Although it's technically O(n) compared to QuickSort's O(n log n), consider the fact that
a. the number of passes required by radix sort is proportional to the key size, and that b. if the keys are primarily unique, then the key length will be O(log n).... thus radix sort is actually O(n log n) when applied to practical cases, hence why everyone uses quicksort or mergesort instead.
2. One cute feature of radix sort for very large data sets is that you only need linear access to the data, so it can be done with non-random-access external storage. The one practical implementation I've ever heard of used four 9-track tape drives.
However, considering use in main memory, both radix sort and quicksort make efficient use of a D-cache that is 2-way set-associative or better, and both hit the whole data set once per pass -> if it swaps, it'll swap a lot:-)
There is something very wrong with your fuel consumption measurements...
Although there is a certain overhead due to the inefficiency of a petrol engine at low RPM, at those kinds of speeds, fuel consumption is about linear with power, and air resistance is the dominant use of power.
The power required to keep a car moving at a given speed against air resistance goes up as the cube of the speed, while the ground covered increases only linearly with speed. Thus, distance covered per unit of fuel (mpg) is reduced quadratically with speed.
Surely the readers of this august forum have been responsible for much of those (currently) 22k+ votes.
I've never bought desktops from Dell pre-loaded with Linux, it was always easier just to reinstall than fight them on it, and you generally want the Windows license for VMware use.
We do get some servers preloaded with RHAS but we tend to config our own installs of that too - Dell's config people don't always "get" non-desktop IT (example: order dual channel hardware RAID controller, and they'll create a RAID-1 LUN with both drives on the same SCSI bus), but to be 100% fair I prefer us to reinstall any server system just to (a) get it exactly the way I want it, and (b) to be sure we have the knowledge to do so.
A very common form of DDoS attack is a SYN flood where the source IP in the packets is NOT the IP of the bot being used.
Last time someone had a go at our servers, the forged IPs traced back to well known locations which obviously weren't the real source (mostly US government labs like LLNL and Sandia).
I see a risk here where DDoS is used specifically to frame the real owner of an IP.
In any event, a moderately competent hacker will use a botnet which is managed using wardriving sessions, or from a server in a non-authority-friendly country, ensuring they are untraceable.
.... but I'm not sure it makes a huge difference in real software development. I do however like it as the safe test for an exact match without worrying about null....
By "practical", I was responding to the prior poster's comment that electric vehicles could only go a short distance. Perhaps "viable" is a better way to put it;-)
The Tesla is certainly not ready for the mass market, but they are going about things the right way by bootstrapping their business and producing a sports car first, something for which people are prepared to pay the kind of money that it currently costs to build an EV. The use of off-the-shelf battery tech (where the artcile brought us in) is also critical to business success.
If you do, and as is typical the patent office drops the ball and issues the patent, then that prior art is lost forever to you as an anti-patent defence, and cannot be used in a court case.
This is why companies rarely challenge inappropriate patent filings via the USPTO, and save prior art until they need it in a court of law to challenege enforcement of a bad patent, so they can have it argued by their own experts.
The system is broken in many ways, this is just one more.
Actually, cinemas do both "network" and local advertising, just like television. Our local multiplex (Regal, big US chain) runs ads for local firms.
I don't think that the movie studio that released the main feature gets much input, beyond the fact they do special cross-promotions like running trailers.... a phenomenon here in the US when there is a big anticipated release like one of the Lord of the Rings or Star Wars series is to to release the trailer for showing only before some low grade flick. Die hard fans will pay their $7.00 for a ticket to a movie they don't care about, see the trailer and leave:-)
For added convenience compared to passwords, but similar levels of security, the fingerprint reader built into current Thinkpad laptops works very nicely.
For a bit of added security without too much grief over drivers and special hardware, RSA SecurID is the gold standard... it's not true public key crypto, and it is quite pricey at c. $130 a user, but it works with a normal keyboard, defends against replay and can be integrated into anything.
The DC is 24V, which is not high voltage in anyone's world. The ring signal is a bit more, IIRC 120V AC which I have gotten zapped by once:-) There were older phone switches (1970's) which did NOT provide the 24V DC power, and where a customer premises power supply was needed for outgoing calls.
An analogue phone line works entirely off the two wires, as many posters have noted the second pair is normally used for a second line, or is left open.
The red / black / green / yellow colour scheme is only used for endpoint cables (those silver ones) and, unusually compared ot other wiring standards, they actually flip over and pins 2-3-4-5 connect to 5-4-3-2 respectively.
Once you get into the real wiring, typically RJ-11 jacks in a building will be wired with the USOC code... like most telco wiring codes, this is based on pairs; the insulation on each pair has a complementary colour scheme, where the first wire is mostly white and has a coloured dash, and its partner is mostly coloured and has a white dash.
For an RJ-11 phone line under the USOC scheme, pins 4&3 get "Wb/Bw" (white-blue, blue-white) and pins 2&5 (note the order) get the line 2 orange pair "Ow/Wo". Pins 1&6 get the green (if the cable provides a 3rd pair) or re left unconnected.
For ethernet (Cat 5, 5e, 6) schemes like 508-B they use four pairs, the 4th pair is brown. Only GigE and up uses more than 2 pairs.
In a trunk cable, there will typically be 40 pairs, carrying 40 customer lines... once they use up all the main colours with white, IIUC the next batch of pairs are based on either grey or red.
Google's stock went up by more than $1.65b in market cap the very day they bought YouTube, which was done as a pure stock deal. This means that while Youtuve's founders, investors and staff still made a pile of money, it was effectively free for Google. Since that time, their market cap has gone up another $15bn or so, and a good proportion of that is due to the buzz from the YouTube deal.
Google didn't buy YouTube to kill it, they bought it because it was a bunch of free eyeballs. Yes, they will have to fund its operating costs (I'd guess at about $3m-$4m a month including that rather large internet bill) but that's pocket change. At minimum, it's a very reasonable experiment for Google to enter into, and it's a way to corner the audience for a whole new media phenomenon. It would have been stupid NOT to buy it.
Copyrighted content is an issue, but a minor one - the nay-sayers like Mark Cuban are quite frankly jealous. As long as YouTube makes a reasonable effort to take that stuff down, they'll never get taken into court - the copright holders know it's helping their sales, not hurting, they are just making sure they are seen to be protecting their IP so that they can go after real pirates.
As other posters have observed, piracy and porn isn't the main thrust of YouTube, it's about the network effect of being THE portal for all of those wannabe filmakers out there and then getting the benefit of all the advertising, marketing and distribution around it.
Self-Deprecating Humour is the Key
on
An Ode To Al
·
· Score: 5, Insightful
As a foreigner living here, I find that American culture often needs a little help to laugh at itself. Parker and Stone may have made doing so mainstream, but Al was ahead of the curve when he started it.
Aside from that, the simple fact is that using different words matching the same rhythm and rhyme of well known music is simply intrinsically funny, and he does a very good job of it - he also has a decent amount of musical talent, and he has shown superb diversity in keeping up with the times and parodying whatever is currently hip.
In short, Al is a key contributor to American culture, and deserves to be lauded for it.
In the past CAGW was without question shilling for corporate interests... the letters from dead people advocating against Microsoft's anti-trust prosecution were certainly that:-)
However, the organization has been cleaned up, and has returned to its original mission, and if you look at some of their more recent work and argument, they are doing what their name implies, and as a hardcore liberal / libertarian European who finds even the US **Democratic** party too right wing for my taste, I find myself agreeing with a lot of what CAGW has to say nowadays.
There is no way for AOL to tell the difference between a legitimate non-filtering forwarding service, an open relay, and a spammer that just fakes headers to make themselves look like a forwarder. All three look identical from a receiving ISP's perspective.
If there was a secure way for the forwarder to tag the forwarded email as something you deliberately asked them to do (spam and all) then ISPs would be able to tell the difference, but there isn't so they have to make do with the tools they have, with the result that a provider is held accountable for the stuff that eminates from their IP addresses.
Charles Stiles** is a good guy, who wants to deliver the mail people want, and not the stuff that they don't.
Your ISP needs to get out of the spam forwarding business, either by not forwarding at all, or by filtering the spam, and then sign up with AOL's whitelist.
.... with the real world and the needs of the open source community at large can be seen in the fact that the redline between the second and third drafts of GPLv3 is enormous; more than half the text had to be rewritten, and changes are not something that a purist like Stallman makes readily.
Don't get me wrong, I'm an open source zealot myself, and I have huge respect for what RMS has done to put open software on the map, and having someone on the extreme acts as a foil for moderates, but the GPL is not a soapbox. Embrace pragmatism!
Wearing my business hat as CTO of a USA-based internet startup, where naturally we use a ton of open source stuff, my concern is that RMS's extremism is handing a gift to the anti-open-source community, especially to Microsoft in their battle with Linux.
I deal with venture capitalists and other financial types: their lawyers don't understand open source licensing, and lawyers always fear what they don't understand. Even now, their standard going in position in deal negotiations is "tell us about all the open source your company uses, what the licenses are, and what it would cost to replace it with closed source". The normal GPL gives a business less commercial protection than a proprietary license which, for example, typically inclues indemnification against patent lawsuits.
We do use GPL'ed stuff, but as a business person I prefer almost any other major open source license (Apache, BSD, Artistic, etc.)
If RMS has his way, there is going to be massive confusion between GPLv2 (just liveable with, if you're careful) and GPLv3 (totally unsuitable for business use) and the FUD is going to be used against every open source project out there. He's going to force me off of Linux and onto a closed OS. How does that help open source?
RMS needs to step up to his natural leadership position as elder statesman of the open source movement: this is a political role, which means embracing compromise.
He also needs to get over the Linux / HURD thing and move on... 9 years later and half the front page at fsf.org is still an anti-Linux rant. Not helpful.
The restrictiveness of the TiVo situation is caused by the closed hardware, not the GPL software. The software is doing its job under the GPL.
The intent of the GPL (even RMS would have to admin) was to ensure freedom to tinker
on general purpose computers
, not to patch an embedded system.
The "right to tinker" may be central to FSF philosophy, but it's the least important of the GPL rights to the general public. Strange as it may seem, 99.9% of the world doesn't write code. And further, that right is still available.... those who wish to tinker with TiVo code are free to port it onto an open hardware platform (x86 PC) and have at it.
The TiVo software is fulfilling the primary GPL right of making dervative improvements freely available - anyone who wants to can take a cool new codec that TiVo wrote and port it into mplayer. TiVo has made a permanent contribution to the pool of freely available source code.
Slashdot Mirror
Well, it seems like a lot of people would happily run financials via the web ... our company is moving to that from installed Windows apps, and it wasn't my doing, it was our CFO.
:-)
I seem to remember a few years back being told by a Siebel sales guy that hosted software for sales CRM wasn't going to fly either, but I bought it anyway
You've worded it gently, but this sounds like your former boss was a little corrupt :-)
When I worked for Mrs Queen, even with the most ehtical of intentions, we did have the issue that the open procurement system would result in unsuitable tenders (the process was run by procurement people who knew nothing about technology) and would cheerfully saddle you with nonsense to save a quid.
So, we would tend to do an informal market survey first, talk to vendors, make a buying decision based on normal commercial business criteria (is this vendor competent? Is the product any good?) then load the RFP in favour of the best solution.
The other issue that made things hard was that some procurements are inhrenetly single vendor - for example, what about renewing a maintenance contract for Sun servers? Sun's own Platinum maintenance is very well priced, so you'd be an idiot not to use them, but government procurement left unfettered will end up hiring Joe PC shop down the street.
They often fail to even check facts, and even when they do, the quality of their output is limited by the fact that journalsists are experts in writing, and not experts in the subject matter they write about, which they often don't really understand.
1. I'm the CTO at a leading vendor of high end online systems for non-profits. We have two CMS offerings. We may well be the "evil" vendor mentioned in this post
2. I am the worlds biggest fan of open source - I'm posting here, on my work laptop which, yes, I run Linux on, because I find it to be a better fit for my personal needs (mixture of hardcore techie and traditional business user) than Windows.
We don't open source our software, we don't even ship it
Customers don't need to develop stuff, or even set up servers, to get our product running. We can deploy a new site as fast as they can work with us on the content and business level concerns. On occasions, we've deployed new sites in under a week.
We use almost all open source stuff in our production systems. The big exception is the database - we run Oracle, which costs a lot of $$$, because it *works*. We have the expertise to leverage this technology and make it work for our customers. Our customers can't afford to carry that skill set and staffing in house.
Our business exists because there is a true need for a packaged
- solution
for non-profits. I'm sure the usual whiners on Slashdot will say that anyone who pays money for software is stupid or evil, but they're not - they just want to get stuff done.It does sound like the original poster is thinking first about personal ideology of promoting open-source, and not about the business needs of his/her employer.
As many people have observed, this is nothing to do with OSS
We too buy Dell servers bare or with a Linux preload (which we flash and reinstall anyway so we can load the RPMs we want).
... for a consumer machine, the only thing Microsoft provides is a licnese and a sticker, and Dell only pays $20 or so a machine.
The cost of the Windows 2003 Server license is indeed about $800. However, the article is about desktop/laptop Windows licenses
Servers are a whole other world...
1. There is much more demand for Linux on servers
2. Generally, people like us who buy them know what they are doing
3. There isn't a need for drivers for a whole bunch of cheapo Tiwansese hardware like USB webcams and printers
Dell has been offering Linux either bundled with, or pre-installed with, on **servers** for a long time - they even have a 3 way agreement with Red Hat and Oracle to certify particular configurations and support each others products without any finger pointing.
Given the relatively limited bandwidth going in and out of Australia, and that 99% of the world's websites are at the wrong end of that, there is arguably some justification for this. Still inconvenient though.
</devilsadvocate>
1. Although it's technically O(n) compared to QuickSort's O(n log n), consider the fact that
a. the number of passes required by radix sort is proportional to the key size, and that
b. if the keys are primarily unique, then the key length will be O(log n)
2. One cute feature of radix sort for very large data sets is that you only need linear access to the data, so it can be done with non-random-access external storage. The one practical implementation I've ever heard of used four 9-track tape drives.
However, considering use in main memory, both radix sort and quicksort make efficient use of a D-cache that is 2-way set-associative or better, and both hit the whole data set once per pass -> if it swaps, it'll swap a lot
He is in fact the long lost twin brother of Jeremy Guscott, former England rugy international.
There is something very wrong with your fuel consumption measurements ...
Although there is a certain overhead due to the inefficiency of a petrol engine at low RPM, at those kinds of speeds, fuel consumption is about linear with power, and air resistance is the dominant use of power.
The power required to keep a car moving at a given speed against air resistance goes up as the cube of the speed, while the ground covered increases only linearly with speed. Thus, distance covered per unit of fuel (mpg) is reduced quadratically with speed.
At 90 mph you get a lot less mpg than at 60 mph.
Surely the readers of this august forum have been responsible for much of those (currently) 22k+ votes.
I've never bought desktops from Dell pre-loaded with Linux, it was always easier just to reinstall than fight them on it, and you generally want the Windows license for VMware use.
We do get some servers preloaded with RHAS but we tend to config our own installs of that too - Dell's config people don't always "get" non-desktop IT (example: order dual channel hardware RAID controller, and they'll create a RAID-1 LUN with both drives on the same SCSI bus), but to be 100% fair I prefer us to reinstall any server system just to (a) get it exactly the way I want it, and (b) to be sure we have the knowledge to do so.
A very common form of DDoS attack is a SYN flood where the source IP in the packets is NOT the IP of the bot being used.
Last time someone had a go at our servers, the forged IPs traced back to well known locations which obviously weren't the real source (mostly US government labs like LLNL and Sandia).
I see a risk here where DDoS is used specifically to frame the real owner of an IP.
In any event, a moderately competent hacker will use a botnet which is managed using wardriving sessions, or from a server in a non-authority-friendly country, ensuring they are untraceable.
Well, you can also do it in Java .....
String a = "Hello".toLowerCase().toUpperCase();
.... but I'm not sure it makes a huge difference in real software development. I do however like it as the safe test for an exact match without worrying about null ....
if ("FOO".equals(a)) {
By "practical", I was responding to the prior poster's comment that electric vehicles could only go a short distance. Perhaps "viable" is a better way to put it
The Tesla is certainly not ready for the mass market, but they are going about things the right way by bootstrapping their business and producing a sports car first, something for which people are prepared to pay the kind of money that it currently costs to build an EV. The use of off-the-shelf battery tech (where the artcile brought us in) is also critical to business success.
People are already making very practical electric cars using existing off the shelf battery technology from laptop computers.
If you do, and as is typical the patent office drops the ball and issues the patent, then that prior art is lost forever to you as an anti-patent defence, and cannot be used in a court case.
This is why companies rarely challenge inappropriate patent filings via the USPTO, and save prior art until they need it in a court of law to challenege enforcement of a bad patent, so they can have it argued by their own experts.
The system is broken in many ways, this is just one more.
Actually, cinemas do both "network" and local advertising, just like television. Our local multiplex (Regal, big US chain) runs ads for local firms.
I don't think that the movie studio that released the main feature gets much input, beyond the fact they do special cross-promotions like running trailers
For added convenience compared to passwords, but similar levels of security, the fingerprint reader built into current Thinkpad laptops works very nicely.
... it's not true public key crypto, and it is quite pricey at c. $130 a user, but it works with a normal keyboard, defends against replay and can be integrated into anything.
For a bit of added security without too much grief over drivers and special hardware, RSA SecurID is the gold standard
The DC is 24V, which is not high voltage in anyone's world. The ring signal is a bit more, IIRC 120V AC which I have gotten zapped by once
An analogue phone line works entirely off the two wires, as many posters have noted the second pair is normally used for a second line, or is left open.
The red / black / green / yellow colour scheme is only used for endpoint cables (those silver ones) and, unusually compared ot other wiring standards, they actually flip over and pins 2-3-4-5 connect to 5-4-3-2 respectively.
Once you get into the real wiring, typically RJ-11 jacks in a building will be wired with the USOC code
For an RJ-11 phone line under the USOC scheme, pins 4&3 get "Wb/Bw" (white-blue, blue-white) and pins 2&5 (note the order) get the line 2 orange pair "Ow/Wo". Pins 1&6 get the green (if the cable provides a 3rd pair) or re left unconnected.
For ethernet (Cat 5, 5e, 6) schemes like 508-B they use four pairs, the 4th pair is brown. Only GigE and up uses more than 2 pairs.
In a trunk cable, there will typically be 40 pairs, carrying 40 customer lines
Google's stock went up by more than $1.65b in market cap the very day they bought YouTube, which was done as a pure stock deal. This means that while Youtuve's founders, investors and staff still made a pile of money, it was effectively free for Google. Since that time, their market cap has gone up another $15bn or so, and a good proportion of that is due to the buzz from the YouTube deal.
Google didn't buy YouTube to kill it, they bought it because it was a bunch of free eyeballs. Yes, they will have to fund its operating costs (I'd guess at about $3m-$4m a month including that rather large internet bill) but that's pocket change. At minimum, it's a very reasonable experiment for Google to enter into, and it's a way to corner the audience for a whole new media phenomenon. It would have been stupid NOT to buy it.
Copyrighted content is an issue, but a minor one - the nay-sayers like Mark Cuban are quite frankly jealous. As long as YouTube makes a reasonable effort to take that stuff down, they'll never get taken into court - the copright holders know it's helping their sales, not hurting, they are just making sure they are seen to be protecting their IP so that they can go after real pirates.
As other posters have observed, piracy and porn isn't the main thrust of YouTube, it's about the network effect of being THE portal for all of those wannabe filmakers out there and then getting the benefit of all the advertising, marketing and distribution around it.
As a foreigner living here, I find that American culture often needs a little help to laugh at itself. Parker and Stone may have made doing so mainstream, but Al was ahead of the curve when he started it.
Aside from that, the simple fact is that using different words matching the same rhythm and rhyme of well known music is simply intrinsically funny, and he does a very good job of it - he also has a decent amount of musical talent, and he has shown superb diversity in keeping up with the times and parodying whatever is currently hip.
In short, Al is a key contributor to American culture, and deserves to be lauded for it.
In the past CAGW was without question shilling for corporate interests ... the letters from dead people advocating against Microsoft's anti-trust prosecution were certainly that :-)
... on Linux.
However, the organization has been cleaned up, and has returned to its original mission, and if you look at some of their more recent work and argument, they are doing what their name implies, and as a hardcore liberal / libertarian European who finds even the US **Democratic** party too right wing for my taste, I find myself agreeing with a lot of what CAGW has to say nowadays.
Oh, and our company hosts their website
There is no way for AOL to tell the difference between a legitimate non-filtering forwarding service, an open relay, and a spammer that just fakes headers to make themselves look like a forwarder. All three look identical from a receiving ISP's perspective.
If there was a secure way for the forwarder to tag the forwarded email as something you deliberately asked them to do (spam and all) then ISPs would be able to tell the difference, but there isn't so they have to make do with the tools they have, with the result that a provider is held accountable for the stuff that eminates from their IP addresses.
Charles Stiles** is a good guy, who wants to deliver the mail people want, and not the stuff that they don't.
Your ISP needs to get out of the spam forwarding business, either by not forwarding at all, or by filtering the spam, and then sign up with AOL's whitelist.
** AOL chief Postmaster
Don't get me wrong, I'm an open source zealot myself, and I have huge respect for what RMS has done to put open software on the map, and having someone on the extreme acts as a foil for moderates, but the GPL is not a soapbox. Embrace pragmatism!
Wearing my business hat as CTO of a USA-based internet startup, where naturally we use a ton of open source stuff, my concern is that RMS's extremism is handing a gift to the anti-open-source community, especially to Microsoft in their battle with Linux.
I deal with venture capitalists and other financial types: their lawyers don't understand open source licensing, and lawyers always fear what they don't understand. Even now, their standard going in position in deal negotiations is "tell us about all the open source your company uses, what the licenses are, and what it would cost to replace it with closed source". The normal GPL gives a business less commercial protection than a proprietary license which, for example, typically inclues indemnification against patent lawsuits.
We do use GPL'ed stuff, but as a business person I prefer almost any other major open source license (Apache, BSD, Artistic, etc.)
If RMS has his way, there is going to be massive confusion between GPLv2 (just liveable with, if you're careful) and GPLv3 (totally unsuitable for business use) and the FUD is going to be used against every open source project out there. He's going to force me off of Linux and onto a closed OS. How does that help open source?
RMS needs to step up to his natural leadership position as elder statesman of the open source movement: this is a political role, which means embracing compromise.
He also needs to get over the Linux / HURD thing and move on
The restrictiveness of the TiVo situation is caused by the closed hardware, not the GPL software. The software is doing its job under the GPL.
The intent of the GPL (even RMS would have to admin) was to ensure freedom to tinker
- on general purpose computers
, not to patch an embedded system.The "right to tinker" may be central to FSF philosophy, but it's the least important of the GPL rights to the general public. Strange as it may seem, 99.9% of the world doesn't write code. And further, that right is still available
The TiVo software is fulfilling the primary GPL right of making dervative improvements freely available - anyone who wants to can take a cool new codec that TiVo wrote and port it into mplayer. TiVo has made a permanent contribution to the pool of freely available source code.