I'm on a pretty idle cable node on a fiber loop with 512/512 service. Unfortunately, the connection appears very bursty, and actually never seems to get the rated speed. It appears that it transfers in a very quick burst for about 1/10th of a second, and then just sleeps/idles for about a second, then another burst. Problem is - overall, those delays really add up.
I've yet to see any project that is as bad as MIT kerberos as far as refusing or just plain ignoring contributions. There has been a blatant extremely inconvenient bug in the telnetd/login implementation in MIT krb5 for at least a year, that I've submittedbug fixes for numerous times. Every time it's ignored completely. Similar for other patches.
I forget who did it (there was an article in SciAm a few months ago about it), but the artcile was talking about how elegant (scientifically) a solution was to decreasing the amount of material required to build a bridge structurally, and at the same time making it aesthetically pleasing. In fact, the reduction of materials in this case actually made the bridge stronger. Someone seeing old bridge designs, and then new bridge designs might think "Wow, what a beautiful solution to the problem, and it works too!"
I would liken this to some of the evolutionary development that takes place while coding. How many times have you written something that is ugly and gets the job done, but is really inneficient, then later come across another piece of code, or someones reworking of your own code, and thought in awe "I can't believe how much better that is, look at how it steamlines all the processing, etc.". At the same time as the code is being functional (and doing the same thing that yours did) it is expressive in showing an alternative way of doing it that is a more beautful solution (at least to another programmer).
Also, there is definitely something expressive and instructional about having more than one implementation. For example, by comparing the different bridge designs, and analyzing them, someone else might be able to come up with a better design for bridges, or for that matter, for anything structural. (Think of how many developments from NASA have made it into everyday life.)
Not just the authors of EveryBuddy should be ticked off, but the authors and contributors to every GPL/LGPL library USED by EveryBuddy. i.e. libyahoo, libfaim, etc.
Re "in public domain"... Technically, once somethig is in the public domain, anyone can do ANYTHING with it, including copying it and revoking all rights from it. That doesn't stop you from doing anything with the original copy of course.
Nothing wrong with using rsync on the database files, you just have to do them at the right time.
For example, on oracle, you can do this by putting the databases in backup mode, which temporarily freezes the contents of the files, and writes all updates to redo logs. Using rsync on the db files while in backup mode would be fine. Of course, I've found that when dealing with huge (multple gigabytes) amounts of data, rsync doesn't help much.
Yeah, had forgotten about Slackware. Used to use it a while back. The tag files worked pretty well.
I'll take a look at SuSE.
Right now, I've got a server set up that we use to install HP-UX 10.20, 11.00, Solaris 2.6, Solaris 2.7, RedHat 6.2 (customized to be reiserfs based), and Irix 6.5. Adding SuSE probably would be easy enough.
Add UMR to the list of clueless schools. They have blocked Napster at the routers and have claimed/stated that 'use of technologies such as Napster and Gnutella' is against the acceptable use policies. Note - no mention at all of using it for legal/illegal activities, the univ is claiming that any use is against the rules.
Of course, they also state in their acceptable use policy that they can search any attached machine without any due process or notification being required.
This is purely an issue with the FCC and the choices/rules of individual airlines. There is NOTHING in the FARs (Federal Aviation Regulations) that restricts the use of any electronics while in flight. The only restriction is what the PIC (Pilot in Command) deems will not interfere with the safe operation of the aircraft.
There is however a FCC regulation that calls for a up to a $10,000 fine for use of a cell phones in the air due to the multiple-tower problem. It's FCC Regulation 22.925.
it's all well and good that the two groups came to a peaceful friendly solution, however, these solutions to GPL violations are setting a very bad precedent - they are saying that in a confrontation between proprietary code and GPL'd code, the GPL'd code loses, at least for a while.
Think about if the situation were reversed - and a GPL'd project appropriated some proprietary binaries from a commercial product. "Sure, we'll take it off our web site in a couple weeks." Yeah right! Can we say 'letter from the lawyers' faster than you can blink?
There needs to be some sort of penance for these companies stealing code. I'm not saying we should go out and attack - I'm saying that if they screwed up, fine, they screwed up - they should give us something in return as part of the apology.
I don't have many details, but it will be in by fall of this year if I remember correctly.
"We want open protocols" from Yahoo+MS, yeah right
on
AOL Jilts Open Source
·
· Score: 3
The way MS and Yahoo are saying that they want open protocols is such a joke, when they aren't even willing to open up their own.
I can't even get them to get me minimal information that would let me extend GTKYahoo to all of the new capabilities. I have been able to do some of the features, but others require the new login method, which I haven't been able to figure out how to generate.
It allows hooking up 4 B&W or Color coax cameras and 1 SVideo camera, completely self contained.
We use it in conjunction with a linux based server that grabs the images off the camera server every few seconds.
Biggest problem I've had is that I would like to dynamically assemble the images (can be retrieved either as gif or jpg in many sizes) into a video stream (bandwidth is not generally an object, but gif anims are way too big).
It doesn't say anything about source and libraries. Personally, I could care less about the commercial products, but I want to be able to collaborate on projects using kerberos, and that is currently a royal pain in the ass.
Slashdot Mirror
I'm on a pretty idle cable node on a fiber loop with 512/512 service. Unfortunately, the connection appears very bursty, and actually never seems to get the rated speed. It appears that it transfers in a very quick burst for about 1/10th of a second, and then just sleeps/idles for about a second, then another burst. Problem is - overall, those delays really add up.
I've yet to see any project that is as bad as MIT kerberos as far as refusing or just plain ignoring contributions. There has been a blatant extremely inconvenient bug in the telnetd/login implementation in MIT krb5 for at least a year, that I've submittedbug fixes for numerous times. Every time it's ignored completely. Similar for other patches.
I forget who did it (there was an article in SciAm a few months ago about it), but the artcile was talking about how elegant (scientifically) a solution was to decreasing the amount of material required to build a bridge structurally, and at the same time making it aesthetically pleasing. In fact, the reduction of materials in this case actually made the bridge stronger. Someone seeing old bridge designs, and then new bridge designs might think "Wow, what a beautiful solution to the problem, and it works too!"
I would liken this to some of the evolutionary development that takes place while coding. How many times have you written something that is ugly and gets the job done, but is really inneficient, then later come across another piece of code, or someones reworking of your own code, and thought in awe "I can't believe how much better that is, look at how it steamlines all the processing, etc.". At the same time as the code is being functional (and doing the same thing that yours did) it is expressive in showing an alternative way of doing it that is a more beautful solution (at least to another programmer).
Also, there is definitely something expressive and instructional about having more than one implementation. For example, by comparing the different bridge designs, and analyzing them, someone else might be able to come up with a better design for bridges, or for that matter, for anything structural. (Think of how many developments from NASA have made it into everyday life.)
I'd like to know how "getting them to sit down at a computer running a sniffer" counts as "hacking into their computers".
Not just the authors of EveryBuddy should be ticked off, but the authors and contributors to every GPL/LGPL library USED by EveryBuddy. i.e. libyahoo, libfaim, etc.
Re "in public domain"... Technically, once somethig is in the public domain, anyone can do ANYTHING with it, including copying it and revoking all rights from it. That doesn't stop you from doing anything with the original copy of course.
Nothing wrong with using rsync on the database files, you just have to do them at the right time.
For example, on oracle, you can do this by putting the databases in backup mode, which temporarily freezes the contents of the files, and writes all updates to redo logs. Using rsync on the db files while in backup mode would be fine. Of course, I've found that when dealing with huge (multple gigabytes) amounts of data, rsync doesn't help much.
Yeah, had forgotten about Slackware. Used to use it a while back. The tag files worked pretty well.
I'll take a look at SuSE.
Right now, I've got a server set up that we use to install HP-UX 10.20, 11.00, Solaris 2.6, Solaris 2.7, RedHat 6.2 (customized to be reiserfs based), and Irix 6.5. Adding SuSE probably would be easy enough.
It built and runs just fine for me on several linux kernels, and on both solaris 2.6 and solaris 2.7.
It was released several days ago... Oct 30th I believe.
Of course, they also state in their acceptable use policy that they can search any attached machine without any due process or notification being required.
This is purely an issue with the FCC and the choices/rules of individual airlines. There is NOTHING in the FARs (Federal Aviation Regulations) that restricts the use of any electronics while in flight. The only restriction is what the PIC (Pilot in Command) deems will not interfere with the safe operation of the aircraft.
There is however a FCC regulation that calls for a up to a $10,000 fine for use of a cell phones in the air due to the multiple-tower problem. It's FCC Regulation 22.925.
-- Nathan (PPSEL)
it's all well and good that the two groups came to a peaceful friendly solution, however, these solutions to GPL violations are setting a very bad precedent - they are saying that in a confrontation between proprietary code and GPL'd code, the GPL'd code loses, at least for a while.
Think about if the situation were reversed - and a GPL'd project appropriated some proprietary binaries from a commercial product. "Sure, we'll take it off our web site in a couple weeks." Yeah right! Can we say 'letter from the lawyers' faster than you can blink?
There needs to be some sort of penance for these companies stealing code. I'm not saying we should go out and attack - I'm saying that if they screwed up, fine, they screwed up - they should give us something in return as part of the apology.
A more apt analogy would be - if you own the CD, and she owns the CD, and she makes a backup copy - can she give you her backup copy?
I don't have many details, but it will be in by fall of this year if I remember correctly.
The way MS and Yahoo are saying that they want open protocols is such a joke, when they aren't even willing to open up their own.
I can't even get them to get me minimal information that would let me extend GTKYahoo to all of the new capabilities. I have been able to do some of the features, but others require the new login method, which I haven't been able to figure out how to generate.
It allows hooking up 4 B&W or Color coax cameras and 1 SVideo camera, completely self contained.
We use it in conjunction with a linux based server that grabs the images off the camera server every few seconds.
Biggest problem I've had is that I would like to dynamically assemble the images (can be retrieved either as gif or jpg in many sizes) into a video stream (bandwidth is not generally an object, but gif anims are way too big).
It doesn't say anything about source and libraries. Personally, I could care less about the commercial products, but I want to be able to collaborate on projects using kerberos, and that is currently a royal pain in the ass.