I ran an earlier version Communigate Pro at a previous job. Simply put, it is the only closed-source software package I actively recommend. Its just that good.
The web mail is slick. IMAP works beautifully. The API for customer-added functionality is extensive. The system is rock solid reliable, and FAST FAST FAST.
If you have too many accounts, they support clustering on multiple servers. Here's a quote from their manual:
When your site serves more than 150,000-200,000 accounts, or when you expect really heavy IMAP/WebMail/MAPI traffic, you should consider using a Cluster configuration.
Huh. So if you have less than 150,000 accounts you can do it with just one server. I'd like to see an open source mail package that can live up to that particular boast.
Yeah, I hate that about Word. I have a perl script that strips out the worst of junk that MS Word seems to add. It does the job for me. Your milage may vary.
The present invention teaches a method and apparatus for creating and managing custom Web sites. Specifically, one embodiment of the present invention claims a computer-implemented method for managing a dynamic Web page generation request to a Web server, the computer-implemented method comprising the steps of routing the request from the Web server to a page server, the page server receiving the request and releasing the Web server to process other requests, processing the request, the processing being performed by the page server concurrently with the Web server, as the Web server processes the other requests, and dynamically generating a Web page in response to the request, the Web page including data dynamically retrieved from one or more data sources.
A reverse proxy, in other words, that contacts more than one back end server for the same server name. That's a pretty easy one to bust. Reverse proxies were already in the open source Squid and Harvest caches prior to this patent's filing in 1996.
Nice idea, but it doesn't work. I'd have to prove misfeasance -- that they knew or should have known that their case was little more than a pack of lies. Even if I prevail (and its a tough case to prove) I'm unlikely to get more than my legal expenses back. Who pays me for the loss of my time?
I got sued once. Copyright infringement and breach of contract. The case was nonsuited about six months later. Dropped, in other words, prior to discovery or any court appearances. It seems my contract was with a company they bought their assets from, twice removed, it had a dispute mediation clause they weren't honoring and the software (which I wrote) was no longer in my posession anyway.
The moral of the story is: anybody can sue anybody for anything. It means nothing. If they obviously have no case, just tell 'em to get bent. More precisely, pay a lawyer for two hours of time to very formally tell them to get bent, and then get on with your life.
Unfortunately for Linux, mi2g also confirmed that the Linux operating system has become somewhat of a hacker's paradise. In a study conducted only seven months ago they found that overall, the most vulnerable operating system for manual hacker attacks was Linux, accounting for 65.64% of all hacker breaches reported.
Search for "mi2g" on Google. The second result is a Register article titled, "Why is mi2g so unpopular?" According to the article, "The chief charge against mi2g is its regular predictions of withering cyber-assaults which, critics say, rarely seem to materialise." It goes on to say, "most of its staff appear to be without significant operational IT security experience".
Most of the rest of the google links are news storys about experts debunking the a mi2g "study" from about 9 months ago which reports Darl's numbers. Here's a choice quote from an article at http://nwc.serverpipeline.com/52500233 :
Mi2g appeared to anticipate criticism of its study. "We would urge caution when reading negative commentary against mi2g, which may have been clandestinely funded, aided or abetted by a vendor or a special interest group," it said in a press release publicizing the study.
Its not a question of vendors ignoring vulnerabilities. Vendors rarely ignore security issues. Its a question of:
* Inadequate security planning during the development process.
* Vulnerability reports that get stuck in tier 1 tech support instead of reaching someone who can fix them.
* Venders who allow marketing and other non-technical matters to improperly influence security oriented decisions.
If you've ever done commercial software development then you know exactly what I'm talking about.
The security researchers' solution is to instigate a marketing/public relations pressure on the vendors which compels technically reasonable handling of security matters. Its a counterweight to the other improper pressures, and a healthy one.
I've heard IP over power lines for local bandwidth delivery described as "Internet Fools Gold." Its an apt description -- so far everyone who has put money into it has lost their investment. Further, anyone with a basic understanding of radio should understand that a long unshielded wire is also known as an antenna. IP over power lines is fated to deliver unlawful "harmful interference" everywhere its attempted.
OKay, I'll cop to an ignorance of physics. Why did the aluminum cylinder implode? Doesn't increasing the energy in an object normally cause it to expand?
I did one better than that: I took a course in the subject.
You're not allowed to transmit with a radio system that has not been put through the FCC's approval process and received a certification and FCC ID. That's radio _system_, not the individual antenna or card. The system even has to include the lengths of cable that'll be attached, although you can specify a range.
This means that its also unlawful to use a store-bought antenna with a transmitter that it hasn't been certified with. Since cantennas and the wifi antennas we're discussing are effectively never used in a receive-only application, this rule applies.
As certification requires some rather expensive testing and documentation by an approved lab, nobody has bothered certifying a cantenna with any
Residential Verizon FiOS is 15mbps for $50/mo. That's more than 4 terabytes per month. If you have a flat rate pipe at your business, put a disk-based network backup server at an employee's home.
Your "truly object oriented" OSes didn't catch on for a large number of excellent reasons. The chief two are: It doesn't enhance the user experience and it doesn't make the machine run faster.
Persistant virtual memory has hardly been passed by. It exists to greater or lesser degrees in most modern OSes in the form of "suspend mode." Lots of room for it to evolve, especially with respect to hardware support, but no remarkable concepts in the offing.
"Presentation types" is meaninglessly vague. If there's a clever but abandoned presentation type out there, be specific.
Batting 0 for 3 boss. Try again?
Nice little dig about being lazy by the way. Doesn't make much sense as my claim is that the information isn't out there, thus no research will turn it up. But hey what's a little ad hominem between friends?
And this is paranoia on crack... it assumes that people will ALWAYS do the wrong thing and will ALWAYS try and screw you about, and that customer profiling NEVER results in a better service.
If SOME of the people do it SOME of the time, its still a problem.
Also bear in mind that the purpose of customer profiling is improved sales conversion, that is getting you to part with more of your money. Customer service is an incidental goal, used only to the extent that it encourages you to come back and spend more.
Do you know your bank?
While I don't "know" the folks at the various banks I use, there are several things I DO know:
1. The credit card folks can't yet correlate WHAT I purchased; they can only correlate WHO I bought it from and use that to create abstract categories.
2. Since my deposit accounts, mortgage and credit cards are all held by different banks, they can't correlate information between them.
3. When I don't want anyone to know at all, I can and do use cash. Perhaps this foils my credit card's attempt to suggest other purchases I might be interested in. Good!
The trouble begins when you want to add things to it...(due to) the brittle nature of the platform, when you do that, other things break.
In other news, Michael Taylor retracted his earlier claim saying, "I misread the note from the engineers. They were in fact referring to Microsoft Windows, not Linux."
There is plenty of great stuff left to invent. Amazon One-Click should not have been one of them.
I'm not suggesting (and I won't suggest) that there will never be a major improvement in OSes. What I am suggesting is that OS improvement has plateaued. The theory that underlies operating systems currently has no major unsolved problems, at least not that anyone has identified. Until it does, there is nothing to instigate a revolutionary change.
Einstein didn't come up with Relativity out of the blue. The Michaelson-Morley experiements proved the then current physics theories wrong. Relativity was the solution to that very vexing problem.
Where are the new operating systems likely to come from?
Where are the new wheels likely to come from? I mean the wheel was invented by the cave man and tires have been around since the 19th century. So who will make a better wheel?
Maybe Unix is still around because back in 1968 those engineers got it right. Maybe there aren't any revolutionary OSes around the corner -- just evolutionary changes to the ones we have.
In 1986 I wrote a Commodore 64 terminal program that allowed BBS' to download and run bits of assembly code onto the user's machine in order to enhance the user's experience. It took about 48 hours before someon posted a message that executed a jump to address 64738 -- system reset.
Bad idea then. Worse idea now, no matter how much supposed security you surround it with.
they still laugh about how "Emacs Makes A Computer Slow".
Ah, yes, My first experience with Emacs was how some dipwit grad student using it on the Sparc 690 would make the machine drag for the other 20 users. I've hated it ever since.:-)
I live in the outskirts of Washington DC and about 80% of the local programming jobs call either for Java or for a mix of languages including Java. Some want C or Visual Basic and even Perl but almost nobody servicing the Federal Government calls for C++.
IMHO, C++ is among the worst languages. The OO stuff encourages the programmer to move away from C's efficiency while pointers continue to saddle you with C's most troublesome trait. But don't take my word for it, judge for yourself. Windows is written in C++. Unix and Linux are written in C. Which runs faster with fewer problems?
If you want the benefits of OO, use a language like Java that also has good error catching and few buffer overflow problems. I'm not a big fan of Java -- its super clunky on text processing jobs -- but I'll give it this: Java does Object Oriented right.
If you need the raw speed, stick with C like the real gurus do.
Over the weekend I got about 20 messages fed back from AOL members who tagged a message from a mailing list I manage as spam. This is an opt-in US government mailing list with subscribe confirmation and a clear unsubscribe link. The message was US government content. In other words, a list that does everything right.
Vigilantism relies on the vigilante's ability to accurately identify the evildoers. Such an ability is woefully lacking, even among smart people.
Slashdot Mirror
Looks interesting. I'll be sure to check it out when the authors reach 1.0 and declare it production-ready.
I ran an earlier version Communigate Pro at a previous job. Simply put, it is the only closed-source software package I actively recommend. Its just that good.
The web mail is slick. IMAP works beautifully. The API for customer-added functionality is extensive. The system is rock solid reliable, and FAST FAST FAST.
If you have too many accounts, they support clustering on multiple servers. Here's a quote from their manual:
When your site serves more than 150,000-200,000 accounts, or when you expect really heavy IMAP/WebMail/MAPI traffic, you should consider using a Cluster configuration.
Huh. So if you have less than 150,000 accounts you can do it with just one server. I'd like to see an open source mail package that can live up to that particular boast.
Yeah, I hate that about Word. I have a perl script that strips out the worst of junk that MS Word seems to add. It does the job for me. Your milage may vary.
http://bill.herrin.us/freebies/striphtml.pl
The present invention teaches a method and apparatus for creating and managing custom Web sites. Specifically, one embodiment of the present invention claims a computer-implemented method for managing a dynamic Web page generation request to a Web server, the computer-implemented method comprising the steps of routing the request from the Web server to a page server, the page server receiving the request and releasing the Web server to process other requests, processing the request, the processing being performed by the page server concurrently with the Web server, as the Web server processes the other requests, and dynamically generating a Web page in response to the request, the Web page including data dynamically retrieved from one or more data sources.
A reverse proxy, in other words, that contacts more than one back end server for the same server name. That's a pretty easy one to bust. Reverse proxies were already in the open source Squid and Harvest caches prior to this patent's filing in 1996.
Nice idea, but it doesn't work. I'd have to prove misfeasance -- that they knew or should have known that their case was little more than a pack of lies. Even if I prevail (and its a tough case to prove) I'm unlikely to get more than my legal expenses back. Who pays me for the loss of my time?
I got sued once. Copyright infringement and breach of contract. The case was nonsuited about six months later. Dropped, in other words, prior to discovery or any court appearances. It seems my contract was with a company they bought their assets from, twice removed, it had a dispute mediation clause they weren't honoring and the software (which I wrote) was no longer in my posession anyway.
The moral of the story is: anybody can sue anybody for anything. It means nothing. If they obviously have no case, just tell 'em to get bent. More precisely, pay a lawyer for two hours of time to very formally tell them to get bent, and then get on with your life.
Unfortunately for Linux, mi2g also confirmed that the Linux operating system has become somewhat of a hacker's paradise. In a study conducted only seven months ago they found that overall, the most vulnerable operating system for manual hacker attacks was Linux, accounting for 65.64% of all hacker breaches reported.
2 g_so_unpopular/
Search for "mi2g" on Google. The second result is a Register article titled, "Why is mi2g so unpopular?" According to the article, "The chief charge against mi2g is its regular predictions of withering cyber-assaults which, critics say, rarely seem to materialise." It goes on to say, "most of its staff appear to be without significant operational IT security experience".
http://www.theregister.co.uk/2002/11/21/why_is_mi
Most of the rest of the google links are news storys about experts debunking the a mi2g "study" from about 9 months ago which reports Darl's numbers. Here's a choice quote from an article at http://nwc.serverpipeline.com/52500233 :
Mi2g appeared to anticipate criticism of its study. "We would urge caution when reading negative commentary against mi2g, which may have been clandestinely funded, aided or abetted by a vendor or a special interest group," it said in a press release publicizing the study.
Wow. Darl's been cloned.
Its not a question of vendors ignoring vulnerabilities. Vendors rarely ignore security issues. Its a question of:
* Inadequate security planning during the development process.
* Vulnerability reports that get stuck in tier 1 tech support instead of reaching someone who can fix them.
* Venders who allow marketing and other non-technical matters to improperly influence security oriented decisions.
If you've ever done commercial software development then you know exactly what I'm talking about.
The security researchers' solution is to instigate a marketing/public relations pressure on the vendors which compels technically reasonable handling of security matters. Its a counterweight to the other improper pressures, and a healthy one.
hopefully IP over power lines
I've heard IP over power lines for local bandwidth delivery described as "Internet Fools Gold." Its an apt description -- so far everyone who has put money into it has lost their investment. Further, anyone with a basic understanding of radio should understand that a long unshielded wire is also known as an antenna. IP over power lines is fated to deliver unlawful "harmful interference" everywhere its attempted.
OKay, I'll cop to an ignorance of physics. Why did the aluminum cylinder implode? Doesn't increasing the energy in an object normally cause it to expand?
I did one better than that: I took a course in the subject.
You're not allowed to transmit with a radio system that has not been put through the FCC's approval process and received a certification and FCC ID. That's radio _system_, not the individual antenna or card. The system even has to include the lengths of cable that'll be attached, although you can specify a range.
This means that its also unlawful to use a store-bought antenna with a transmitter that it hasn't been certified with. Since cantennas and the wifi antennas we're discussing are effectively never used in a receive-only application, this rule applies.
As certification requires some rather expensive testing and documentation by an approved lab, nobody has bothered certifying a cantenna with any
"They're (Pringles cans fashioned into antennas) unsophisticated but reliable, and it's illegal to possess them,"
That's not strictly accurate, but it contains a grain of truth.
It is an unlawful violation of the FCC regs to USE a cantenna, as it has not been certified for use with any radio broadcast systems.
Violators may be forced to immediately and permanantly cease use of their uncertified system. That is the extent of the possibile penalties.
play my radio audibly
That was you? Damn it, turn it down!
Residential Verizon FiOS is 15mbps for $50/mo. That's more than 4 terabytes per month. If you have a flat rate pipe at your business, put a disk-based network backup server at an employee's home.
Your "truly object oriented" OSes didn't catch on for a large number of excellent reasons. The chief two are: It doesn't enhance the user experience and it doesn't make the machine run faster.
Persistant virtual memory has hardly been passed by. It exists to greater or lesser degrees in most modern OSes in the form of "suspend mode." Lots of room for it to evolve, especially with respect to hardware support, but no remarkable concepts in the offing.
"Presentation types" is meaninglessly vague. If there's a clever but abandoned presentation type out there, be specific.
Batting 0 for 3 boss. Try again?
Nice little dig about being lazy by the way. Doesn't make much sense as my claim is that the information isn't out there, thus no research will turn it up. But hey what's a little ad hominem between friends?
there have been plenty of innovations in the past that have been by-passed by the industry and forgotten.
Name three.
And this is paranoia on crack... it assumes that people will ALWAYS do the wrong thing and will ALWAYS try and screw you about, and that customer profiling NEVER results in a better service.
If SOME of the people do it SOME of the time, its still a problem.
Also bear in mind that the purpose of customer profiling is improved sales conversion, that is getting you to part with more of your money. Customer service is an incidental goal, used only to the extent that it encourages you to come back and spend more.
Do you know your bank?
While I don't "know" the folks at the various banks I use, there are several things I DO know:
1. The credit card folks can't yet correlate WHAT I purchased; they can only correlate WHO I bought it from and use that to create abstract categories.
2. Since my deposit accounts, mortgage and credit cards are all held by different banks, they can't correlate information between them.
3. When I don't want anyone to know at all, I can and do use cash. Perhaps this foils my credit card's attempt to suggest other purchases I might be interested in. Good!
The trouble begins when you want to add things to it...(due to) the brittle nature of the platform, when you do that, other things break.
In other news, Michael Taylor retracted his earlier claim saying, "I misread the note from the engineers. They were in fact referring to Microsoft Windows, not Linux."
There is plenty of great stuff left to invent. Amazon One-Click should not have been one of them.
I'm not suggesting (and I won't suggest) that there will never be a major improvement in OSes. What I am suggesting is that OS improvement has plateaued. The theory that underlies operating systems currently has no major unsolved problems, at least not that anyone has identified. Until it does, there is nothing to instigate a revolutionary change.
Einstein didn't come up with Relativity out of the blue. The Michaelson-Morley experiements proved the then current physics theories wrong. Relativity was the solution to that very vexing problem.
Where are the new operating systems likely to come from?
Where are the new wheels likely to come from? I mean the wheel was invented by the cave man and tires have been around since the 19th century. So who will make a better wheel?
Maybe Unix is still around because back in 1968 those engineers got it right. Maybe there aren't any revolutionary OSes around the corner -- just evolutionary changes to the ones we have.
I thought, incorrectly, that 20 years later folks wouldn't be making the same stupid mistake.
In 1986 I wrote a Commodore 64 terminal program that allowed BBS' to download and run bits of assembly code onto the user's machine in order to enhance the user's experience. It took about 48 hours before someon posted a message that executed a jump to address 64738 -- system reset.
Bad idea then. Worse idea now, no matter how much supposed security you surround it with.
they still laugh about how "Emacs Makes A Computer Slow".
:-)
Ah, yes, My first experience with Emacs was how some dipwit grad student using it on the Sparc 690 would make the machine drag for the other 20 users. I've hated it ever since.
I live in the outskirts of Washington DC and about 80% of the local programming jobs call either for Java or for a mix of languages including Java. Some want C or Visual Basic and even Perl but almost nobody servicing the Federal Government calls for C++.
IMHO, C++ is among the worst languages. The OO stuff encourages the programmer to move away from C's efficiency while pointers continue to saddle you with C's most troublesome trait. But don't take my word for it, judge for yourself. Windows is written in C++. Unix and Linux are written in C. Which runs faster with fewer problems?
If you want the benefits of OO, use a language like Java that also has good error catching and few buffer overflow problems. I'm not a big fan of Java -- its super clunky on text processing jobs -- but I'll give it this: Java does Object Oriented right.
If you need the raw speed, stick with C like the real gurus do.
Over the weekend I got about 20 messages fed back from AOL members who tagged a message from a mailing list I manage as spam. This is an opt-in US government mailing list with subscribe confirmation and a clear unsubscribe link. The message was US government content. In other words, a list that does everything right.
Vigilantism relies on the vigilante's ability to accurately identify the evildoers. Such an ability is woefully lacking, even among smart people.