That delay would deny hard-lobbying^Wworking companies fruits of the law they already paid for.
It's a matter of forcing people to ditch a solution that has been working for over 50 years, something that is dated but does its job, and is a lot cheaper. Old, cost-efficient things are what the industry hates. I run a server off a Pentium 120Mhz box -- do I need anything more for a minor WWW server that doubles as a border router for a small company LAN and an ISDN dial-in box for several employees? It works just perfectly. I get more from this ancient machine than you get from your P4 6Ghz if you waste your CPU cycles for running a spiffy GUI that blue-screens once a week.
The poor who watch TV can't afford HDTV. What they need, is low-cost entertainment, not high-end displays. I'm sorry if it cuts your company's bottom line -- but using legislation to force people to throw out what's working well just so they have to pay the upgrade costs is just wrong.
Someone may be uneducated or poor but this does not necessarily mean that their genetic stock is any worse. Of course, there are many smart poor people, and there are many who are dumb but had the luck of being born in a higher social class. What I mean is, there is a significant correlation between intelligence+education and the social stratum. And, the amount of offspring is inversely proportional to the social level.
The truly mentally retarded are often in medicalised or institutionalised situations and have few offspring. Only in developed countries and only in better city districts. There are villages in central Africa where the average IQ is around 40. It's not only the lack of education, but actual hereditary retardation of whole populations. I can also spot similar cases on a lesser scale even in my home town. In this case, it's just my feeling rather than a scientific study, but I doubt those people can be degenerated to such a level purely because of the lack of education.
And these are the kind of people who tend to breed families of 10+ kids.
I'm using a GPLed font (Antykwa Poltawskiego) thoroughly in a proprietary program I'm doing at work. I assumed this is ok, as the font is not linked in in any way -- the lack of it in the operating system's fonts directory will just make the program's appearance different.
The program in question is developed for a single customer, and doesn't even come with an installer at all -- so, I don't have any issues at the moment. I install everything by hand, so the font isn't even explicitely distributed.
However, it's an interesting question how closely will I be allowed to bundle this font when (or if) our company will sell the program to other customers than the one whom it was developed for. Obviously, putting both the font and the program on a single CD is ok. Otherwise, no commercial Linux distributions would be able to exist. But, can I have it as a part of a self-contained installer?
Well, all my evidence here is extremely anecdotal. My whole statistic is based on 1 (one) case where it took just filling out a form, paying the fee and waving the passport. And the number of 10 years, while pulled not from my ass but from my personal passport, doesn't say a lot about policies. I also come from Poland, a country that is said to have one of biggest percentages of denied visa apps.
Perhaps, they deny some visas "just because" but accept some for the very same reason. Or, it may depend on the phase of moon and its relation to Neptun's 3rd ring. Who knows.
If I email you a PDF document, and now it's on your PC, and you open it, and flash is inbedded in it, and you're connected to the net via your ever-present dsl line, flash can actually go to the web and pull content.
Yeah, this is a very strong argument against Flash integration. You see, I'm not going to let Flash-enabled e-mail clients within ten miles of my machine, for pretty obvious reasons. HTML mail is a monstrosity bad enough, and even most business types around me learned that it's better to stick to plain text.
Also, the usefullness-to-annoyance factor of Flash has a lot of zeros in front of it. I've seen Peasant's Quest some time ago, and the rest is, well, basically just pure adverts/spyware.
Even when US visas are concerned, getting a B (tourist/business) one is trivial. You just submit your application, pay a fee and that's it. No interviews, no need to provide a stack of paperwork, invitations or anything. And, you get it for 10 years outright.
On the other hand, getting any other kind of US visa is a very hard, long process. Getting my F1 took many months and several trips to the embassy (a >5 hours long trip for me, then a day spent in a queue, followed by 5h for returning home). I had to submit an invitation, a number of papers from US offices, papers from my home university, and so on, so on. And once on American soil, I had additional paperwork to do -- like the entry card you have to keep with you and return on your way out. All for a short stay for a research project I missed the most of because of visa-related delays.
Now, compare it to the wave-your-passport kind of security required for B visas...
Having your own DNS server on your border router isn't a bad idea, even if your LAN is relatively poorly connected. My ISP's DNS server keeps going apeshit -- my own one at home works when and only when my connection is up. And uhm, that means it's working exactly when the response is meaningful anyway.
However, you can bypass only queries done by _you_. If you want to have a name for your dynamic IP box (using Dyndns or the equivalent), you'll name will be sabotages by providers guily of the practice mentioned in the article.
An example: my home LAN gets its IP forcibly changed every >24h. Whenever this happens, I update my zone file on my domain's primary NS and let bind do its magic to notify the secondary. With a TTL of 10 minutes.
Now, let's assume that 10 minutes get changed to 7 days. It means that I can ssh home... a week after the IP has been invalidated. Schweet.
Is that BonziWhatever, perhaps? You see, it has a lot of value-added stuff bundled with it, and it needs root access to install them and then re-install them after every boot if something bad happened to them...
I guess you'll want to take a look at Antykwa Poltawskiego. It's a font that looks better than Bitstream Vera Serif on the FreeType renderer.
And the license is... GPL. Hmm, do we have any articles about GPLed fonts nearby?
My edits and repackaging can be found here -- don't use the original version unless you're using it on Windows. The original version looks good only on paper -- it sucks on Windows and on FreeType without "forced autohinter", my one looks good on FreeType but really, really blows on Win32. (typing this on a Windows machine:p -- thus, I've attached no screenshots on the webpage)
deb ftp://ftp.nerim.net/debian-marillat/ stable main
to/etc/apt/sources.list before installing mplayer. The version in the main archive is crippled down because of legal concerns.
Usually, out-of-Debian apt archives are of poor quality, but, as Christian Marillat is a Debian Developer himself, he knows how to make it work right. His archive is widely known as debian-non-legal.
You see, I haven't ever bought something because of an ad in my entire life. I also don't read any ads in newspapers, mute the TV during commercials (oh well, haven't watched the TV in more than a year...), get mightily annoyed at buildings being defaced with ads and so on.
I also live in a country with a banking system worse than Albania or Nigeria. I can't use my VISA card for Internet transactions for "security" reasons, and when I applied for a card without that limitation, I was denied even though I have an account in that bank for 7 years now, I'm employeed at the same place for 1.5y and I have 6 months income on the account there. Companies like PayPal stay as far away from Poland as possible. It leads to ridiculous situations like when to buy a DNS name the registrar mailed me the bill and I responded with a paper bank transfer. Thus, with that many hoops, guess if I'm going to purchase anything from US companies. Yeah, sure.
So... they're wasting my time, and cannot earn a broken penny off me. Unless they are paying per ad displayed (and not clickthoughs or actual purchases), it's a pure waste of bandwidth.
From your tone, I guess that the ad revenue on your personal site has declined. I'm sorry, but you can still count on those people who actually buy advertised junk.
Of course. The thing is, the users would be unhappy if they can't use some buggy GUI browser that can also get their spyware for them.
As someone said, "Those who don't understand Lynx are condemned to reinvent it, poorly".
Too bad, most websites these days are designed as if GUI browsers were the only thing. I found myself going from Links 90%, Firefox 10% to Links 10%, Firefox 90%, with Links being used mostly for querybts -w.
It's feasible (as Morse Code is a valid layer 1 network protocol), but would reduce the bandwidth. You can design something a lot more efficient for a 3-state communication channel...
It's a good idea to heavily use adblock as well. As an average page refers to several advertising services, knocking them out will usually reduce the number of DNS queries by 2/3 and bandwidth use by like half. The key is to not limit it to just images, but gratuiously give wildcard bans to entire domains that have something with "counter" or "ad" in name.
And as a side effect... yeah, you'll have no ads as well.
It's not open source as it can't be used commercially in any way -- so it's useless for projects like Linux distributions and so on. Or even, you can't air them in a shop (at least according to Polish law).
But, it's open for private listening, open for modifications and open for sharing. These are important freedoms, and they're not something to snooze at. Plus, making a blow against RIAA just can't be bad:p
When it comes to Creative Commons licenses, they're way too easy to mistake one for another. The CC family of licenses provides both an open one, some "half-free" (that is, free for looking and copying) and some entirely proprietary licenses. Thus, it's better to use GPL or any other free license instead of CC ones.
I am a "Unix head". I'm an Unix head, too. In fact, the server at the biggest customer who bought the program in question is a Debian machine. We used MySQL, so the only issue is the client (the entire program, really) running on Windows.
I presume that MS has followed standard solutions here Er, what? Now this suggests you're a 100% Unix head who was lucky enough to never see the worse sides of life:p
broadcast It was TCP -- and Windows doesn't respond to broadcast ICMP.
NIS maps, &etc Perhaps it's possible to tie it to SMB in some ways, but I'm not educated in MS things.
if you *really* want to do it your way -- look for the machine that worked the last time first It's an _installer_. Once the server was found, the IP gets hard-coded in the registry. To change it, you need to run the installer again and choose "change settings".
The sockets in my installer were non-blocking. It was a single-threaded application that tried to open a nonblocking TCP connection on port 3306, to every IP in A.B.C.* (for every network interface) and 127.0.0.1.
I also added a nasty hack of adding an user account to the server with the login name and the password being hashes of the application's serial number, embedded in the binary. That account was tightly stripped down and had no rights except connecting and running SELECT on a table saying the database is installed. It's an ugly kludge, but worth it as it can tell people whether they're connecting to Mars vs connecting to a wrong server vs using a wrong username -- things that BDE can't otherwise differentiate between.
And no, spawning additional threads doesn't do anything for XP SP2. The limit of 10 pending connections is system-wide.
I've once written a piece of code that probes all addresses on the local class C subnet, looking for the MySQL server. Yeah, this is a lame-ish solution, but it's much better than trying to explain what an "IP address" or even "server name" is to your average accounting drone.
On 98: the limit of available TCP sockets is pretty low, but Windows will tell your program that the call failed. Ok. On XP SP1: the limit of available sockets is a lot higher. Everything works fine. On XP SP2: Windows will start _10_ or so connections, and then lie to your process that the extra connections are pending... They won't actually start until after the first 10 completed and/or timed out.
The above means, on an XP SP2 box, you can't do a legitimate scan faster than 10 IPs/120 seconds, and this pretty much broken down my installer. The alternative, having to ask someone competent what the server is and typing the address in was pretty unviable due to the customer company lacking enough competent IT people.
Slashdot Mirror
As an earlier poster mentioned, 85% of TV viewers have their signal fed by cable. This means, it's just a matter of interpretation.
That delay would deny hard-lobbying^Wworking companies fruits of the law they already paid for.
It's a matter of forcing people to ditch a solution that has been working for over 50 years, something that is dated but does its job, and is a lot cheaper. Old, cost-efficient things are what the industry hates. I run a server off a Pentium 120Mhz box -- do I need anything more for a minor WWW server that doubles as a border router for a small company LAN and an ISDN dial-in box for several employees? It works just perfectly. I get more from this ancient machine than you get from your P4 6Ghz if you waste your CPU cycles for running a spiffy GUI that blue-screens once a week.
The poor who watch TV can't afford HDTV. What they need, is low-cost entertainment, not high-end displays. I'm sorry if it cuts your company's bottom line -- but using legislation to force people to throw out what's working well just so they have to pay the upgrade costs is just wrong.
No. The whole concept has a merit, and it works. You're just missing what it is.
It's not a viable programming technique, but it's a viable method for siphoning money into their account. And as thus, a valid business method.
Someone may be uneducated or poor but this does not necessarily mean that their genetic stock is any worse.
Of course, there are many smart poor people, and there are many who are dumb but had the luck of being born in a higher social class. What I mean is, there is a significant correlation between intelligence+education and the social stratum.
And, the amount of offspring is inversely proportional to the social level.
The truly mentally retarded are often in medicalised or institutionalised situations and have few offspring.
Only in developed countries and only in better city districts.
There are villages in central Africa where the average IQ is around 40. It's not only the lack of education, but actual hereditary retardation of whole populations.
I can also spot similar cases on a lesser scale even in my home town. In this case, it's just my feeling rather than a scientific study, but I doubt those people can be degenerated to such a level purely because of the lack of education.
And these are the kind of people who tend to breed families of 10+ kids.
At this point, our civilization has a very strongly negative evolutionary direction.
The smart people tend to have one child per couple if any at all.
The poor, uneducated, often retarded, tend to produce ten or more per family.
End result...?
Being deaf to the colon trumpet is not enough.
The gases used here sound louder.
I'm using a GPLed font (Antykwa Poltawskiego) thoroughly in a proprietary program I'm doing at work. I assumed this is ok, as the font is not linked in in any way -- the lack of it in the operating system's fonts directory will just make the program's appearance different.
The program in question is developed for a single customer, and doesn't even come with an installer at all -- so, I don't have any issues at the moment. I install everything by hand, so the font isn't even explicitely distributed.
However, it's an interesting question how closely will I be allowed to bundle this font when (or if) our company will sell the program to other customers than the one whom it was developed for.
Obviously, putting both the font and the program on a single CD is ok. Otherwise, no commercial Linux distributions would be able to exist. But, can I have it as a part of a self-contained installer?
Well, all my evidence here is extremely anecdotal. My whole statistic is based on 1 (one) case where it took just filling out a form, paying the fee and waving the passport. And the number of 10 years, while pulled not from my ass but from my personal passport, doesn't say a lot about policies.
I also come from Poland, a country that is said to have one of biggest percentages of denied visa apps.
Perhaps, they deny some visas "just because" but accept some for the very same reason. Or, it may depend on the phase of moon and its relation to Neptun's 3rd ring. Who knows.
If I email you a PDF document, and now it's on your PC, and you open it, and flash is inbedded in it, and you're connected to the net via your ever-present dsl line, flash can actually go to the web and pull content.
Yeah, this is a very strong argument against Flash integration.
You see, I'm not going to let Flash-enabled e-mail clients within ten miles of my machine, for pretty obvious reasons. HTML mail is a monstrosity bad enough, and even most business types around me learned that it's better to stick to plain text.
Also, the usefullness-to-annoyance factor of Flash has a lot of zeros in front of it. I've seen Peasant's Quest some time ago, and the rest is, well, basically just pure adverts/spyware.
And more important, when the US will repent?
Even when US visas are concerned, getting a B (tourist/business) one is trivial. You just submit your application, pay a fee and that's it. No interviews, no need to provide a stack of paperwork, invitations or anything. And, you get it for 10 years outright.
On the other hand, getting any other kind of US visa is a very hard, long process. Getting my F1 took many months and several trips to the embassy (a >5 hours long trip for me, then a day spent in a queue, followed by 5h for returning home). I had to submit an invitation, a number of papers from US offices, papers from my home university, and so on, so on. And once on American soil, I had additional paperwork to do -- like the entry card you have to keep with you and return on your way out. All for a short stay for a research project I missed the most of because of visa-related delays.
Now, compare it to the wave-your-passport kind of security required for B visas...
Having your own DNS server on your border router isn't a bad idea, even if your LAN is relatively poorly connected.
My ISP's DNS server keeps going apeshit -- my own one at home works when and only when my connection is up. And uhm, that means it's working exactly when the response is meaningful anyway.
However, you can bypass only queries done by _you_. If you want to have a name for your dynamic IP box (using Dyndns or the equivalent), you'll name will be sabotages by providers guily of the practice mentioned in the article.
An example: my home LAN gets its IP forcibly changed every >24h. Whenever this happens, I update my zone file on my domain's primary NS and let bind do its magic to notify the secondary. With a TTL of 10 minutes.
Now, let's assume that 10 minutes get changed to 7 days. It means that I can ssh home... a week after the IP has been invalidated. Schweet.
weather display application
Is that BonziWhatever, perhaps? You see, it has a lot of value-added stuff bundled with it, and it needs root access to install them and then re-install them after every boot if something bad happened to them...
FreeOrion? Where???????
Oh, www.freeorion.org. I see it's still in very early alpha stages.
You see, I still consider MOO2 to be the very best strategy game ever (and MOO3 to stink so badly to be next to unplayable).
I guess you'll want to take a look at Antykwa Poltawskiego. It's a font that looks better than Bitstream Vera Serif on the FreeType renderer.
:p -- thus, I've attached no screenshots on the webpage)
And the license is... GPL. Hmm, do we have any articles about GPLed fonts nearby?
My edits and repackaging can be found here -- don't use the original version unless you're using it on Windows. The original version looks good only on paper -- it sucks on Windows and on FreeType without "forced autohinter", my one looks good on FreeType but really, really blows on Win32.
(typing this on a Windows machine
to
Usually, out-of-Debian apt archives are of poor quality, but, as Christian Marillat is a Debian Developer himself, he knows how to make it work right. His archive is widely known as debian-non-legal.
You see, I haven't ever bought something because of an ad in my entire life. I also don't read any ads in newspapers, mute the TV during commercials (oh well, haven't watched the TV in more than a year...), get mightily annoyed at buildings being defaced with ads and so on.
I also live in a country with a banking system worse than Albania or Nigeria. I can't use my VISA card for Internet transactions for "security" reasons, and when I applied for a card without that limitation, I was denied even though I have an account in that bank for 7 years now, I'm employeed at the same place for 1.5y and I have 6 months income on the account there. Companies like PayPal stay as far away from Poland as possible. It leads to ridiculous situations like when to buy a DNS name the registrar mailed me the bill and I responded with a paper bank transfer.
Thus, with that many hoops, guess if I'm going to purchase anything from US companies. Yeah, sure.
So... they're wasting my time, and cannot earn a broken penny off me. Unless they are paying per ad displayed (and not clickthoughs or actual purchases), it's a pure waste of bandwidth.
From your tone, I guess that the ad revenue on your personal site has declined. I'm sorry, but you can still count on those people who actually buy advertised junk.
http://ad.*/
http://*.ad.*/
It's not really going to hit anything legitimate...
Of course. The thing is, the users would be unhappy if they can't use some buggy GUI browser that can also get their spyware for them.
As someone said, "Those who don't understand Lynx are condemned to reinvent it, poorly".
Too bad, most websites these days are designed as if GUI browsers were the only thing. I found myself going from Links 90%, Firefox 10% to Links 10%, Firefox 90%, with Links being used mostly for querybts -w.
It's feasible (as Morse Code is a valid layer 1 network protocol), but would reduce the bandwidth. You can design something a lot more efficient for a 3-state communication channel...
It's a good idea to heavily use adblock as well. As an average page refers to several advertising services, knocking them out will usually reduce the number of DNS queries by 2/3 and bandwidth use by like half. The key is to not limit it to just images, but gratuiously give wildcard bans to entire domains that have something with "counter" or "ad" in name.
And as a side effect... yeah, you'll have no ads as well.
It's not open source as it can't be used commercially in any way -- so it's useless for projects like Linux distributions and so on. Or even, you can't air them in a shop (at least according to Polish law).
:p
But, it's open for private listening, open for modifications and open for sharing. These are important freedoms, and they're not something to snooze at. Plus, making a blow against RIAA just can't be bad
When it comes to Creative Commons licenses, they're way too easy to mistake one for another. The CC family of licenses provides both an open one, some "half-free" (that is, free for looking and copying) and some entirely proprietary licenses. Thus, it's better to use GPL or any other free license instead of CC ones.
I am a "Unix head".
:p
I'm an Unix head, too. In fact, the server at the biggest customer who bought the program in question is a Debian machine. We used MySQL, so the only issue is the client (the entire program, really) running on Windows.
I presume that MS has followed standard solutions here
Er, what? Now this suggests you're a 100% Unix head who was lucky enough to never see the worse sides of life
broadcast
It was TCP -- and Windows doesn't respond to broadcast ICMP.
NIS maps, &etc
Perhaps it's possible to tie it to SMB in some ways, but I'm not educated in MS things.
if you *really* want to do it your way -- look for the machine that worked the last time first
It's an _installer_. Once the server was found, the IP gets hard-coded in the registry. To change it, you need to run the installer again and choose "change settings".
The sockets in my installer were non-blocking. It was a single-threaded application that tried to open a nonblocking TCP connection on port 3306, to every IP in A.B.C.* (for every network interface) and 127.0.0.1.
I also added a nasty hack of adding an user account to the server with the login name and the password being hashes of the application's serial number, embedded in the binary. That account was tightly stripped down and had no rights except connecting and running SELECT on a table saying the database is installed. It's an ugly kludge, but worth it as it can tell people whether they're connecting to Mars vs connecting to a wrong server vs using a wrong username -- things that BDE can't otherwise differentiate between.
And no, spawning additional threads doesn't do anything for XP SP2. The limit of 10 pending connections is system-wide.
I've once written a piece of code that probes all addresses on the local class C subnet, looking for the MySQL server. Yeah, this is a lame-ish solution, but it's much better than trying to explain what an "IP address" or even "server name" is to your average accounting drone.
On 98: the limit of available TCP sockets is pretty low, but Windows will tell your program that the call failed. Ok.
On XP SP1: the limit of available sockets is a lot higher. Everything works fine.
On XP SP2: Windows will start _10_ or so connections, and then lie to your process that the extra connections are pending... They won't actually start until after the first 10 completed and/or timed out.
The above means, on an XP SP2 box, you can't do a legitimate scan faster than 10 IPs/120 seconds, and this pretty much broken down my installer. The alternative, having to ask someone competent what the server is and typing the address in was pretty unviable due to the customer company lacking enough competent IT people.