Actually, if the people who wanted to bash out the routine would do so in a library with a proof-of-concept app to go with it, then others could pick that up and run with documenting and interfacing it.
PS, (harps again), I'm tired of poor NULL handling in glibc.
strcpy(NULL, "bah") segfaults instead of doing something handlable...
For your _fact_ list, why don't you tell me how many Fortune 500 companies use off-the-shelf software to run their businesses? I mean, the database applications, not the back-end software.
I don't care if they use Oracle or DB2, I care who wrote the app their secretaries and automation people use every day. I'm talking about the stuff that drives the robotics equipment in the shop or the order tracking software for their online systems. You're trying to tell me that they use pre-packaged pretty printed shrink-wrapped software to do that?
One of the companies I work with is a smaller database design group where three partners design and code stock and custom database solutions for a fairly large user base. Standard policy at the office is that the programmers handle any difficult question; who else would understand it as thoroughly? What I find interesting is how often they recode the program in question online and ask the customer to try it again to their satisfaction...
Funny, reminds me of the "IBM PC"... which we all still run according to some pundits...
... but IBM sure isn't getting all the revenue;-)
Building clones to a product _can_ be successful, but there has to be a good reason like cost coupled with a lack of difference in functionality. Lots of people still bought real IBM PCs while the IBM PC Clone was taking over, but the slow information push that there was little difference between the two has meant that IBM no longer gets its name in front of the PC anymore.
Getting your head out of the sand is in order too.
In an industrial environment, training is always a necessity; the average user uses customised database and control software they've never seen before anyway. Designing such software to be user friendly in a Unix / Windows environment is wholely seperate from the platform it is running on.
Hiring experienced Unix admins can be costly, but with only one long-running Unix server, Windows admins seem to have the misguided impression the school / company would need a team of admins. In fact, one or two solid unix administrators would do fine for most of the Unix maintenance and an external company might even be sufficient for 90% of their needs.
As for "full-featured" bloatware like Office vs. stripped-down but has every feature I've needed software like StarOffice, well, that was my comment.
An actual application compatibility layer _would do_ that, but since this isn't a real compatibility layer but simply falisfying the information returned from a single system call, I don't expect much.
Considering a 'route' to a destination is pronounced with a soft 'ou' as in "oooh", a 'router' would be properly pronounced 'rooter' since that's exactly its job.
However, 'rowter' has stuck in N-A, so I guess its here to stay.
Common usage has a lot to do with spelling and pronunciation in English, which is why its different in Canada from the U.K. and Jersey from Tennessee from Texas.
They seem, more importantly, to think that Linux people in general (that's 80% of Linux people, for example) _care_ if it becomes a desktop OS for the masses.
I agree that Gilmore is wrong on the issues, but I agree with the EFF on freedom of speech vs. spam. I don't think that spam is protected speech, but I do think that MAPS blocks E-mail and E-communications too arbitrarily to not be abusing peoples' free speech.
I was responding to your post about whitelists as I remember it now and my comments on hashes were w.r.t. that.
I think the EFF is unfortunately right. If we allow independant groups who have no external accountability to create and administer lists like MAPS which, if used widely, can arbitrarily cut off a person's ability to communicate with the outside, we're setting ourselves up for very large freedom of speech problems.
If a person is too loud and someone wanted them silenced, getting them onto all the MAPS-like lists by way of pursuasion (in the future) might be a good method. Being able to silence that person on the Internet shouldn't be possible without public review -- that's why we have court systems for crimes.
MAPS is too vigilante for the EFF is what it comes down to for me, and unfortunately MAPS-like services are very useful but I think closing down spammers themselves is a much better long-term solution.
You contradict your first statement with your second, so why didn't you just delete it?
And what's the "twice"? The whitelist sends you one confirmation, once. It then adds you to their list, assuming they didn't give you a sender-specific address in the first place.
I'm going to ignore most of your message because the first paragraph shows a lack of understanding so deep that your message can't be useful to me.
Read how those filtering systems work. Using the time-based or sender-based hashes means the user's E-mail doesn't have to be verified before getting through. Leaving your E-mail lying around for people to pick up and having a 3 month hash on it will prevent the spam that comes a year from now from long-term collection bots... read how the tools work.
There's no reason that 'held for confirmation' mail couldn't be left in a different folder (Maildir / IMAP / mbox) and still be visible to the user if they were bored and wanted to check.
http://madhaus.utcs.utoronto.ca/qmail/spam-filte r is one example that simply creates outgoing addresses that are either only good for a small amount of time or for a specific sender.
http://www.xns.org/xns/whitepapers/filtering/ describes XNS for white-list instead of black-list filtering.
There's always this one: http://software.libertine.org/tmda/ the Tagged Mail delivery agent, my personal favorite.
High-speed ISPs do a horrible job of routing wherever I've looked.
What path do your packets take to get from your house to your neighbour's? How many collisions or how much backplane bandwidth does that cause that are unnecessary?
Connecting neighbours to each other for the sake of their interconnectivity is different than connecting them to each other to save costs on sharing a cable modem.
I have had customers purchase volume licenses of PGP. I stopped recommending it when they stopped releasing source code.
Without the support of the tech community that built up PGP and the web of trust in the first place, they will fail. Maybe Phil's new venture can purchase his software back from NAI?
One thing I've discovered is that beyond civil disobedience, politicians love help more than money. Volunteer and get in their face. Find a local politician with some influence and tell him or her that you'll do the research for them on some totally unrelated issue they're working on. Become a volunteer staffer somewhere with clout. Get involved so that your face is seen and you willingness to help out is known. Then make sure your views are expressed every step of the way -- you'll get heard inherently.
Slashdot Mirror
Their digital video cameras (here in Canada at least) already have firewire transfer on them (works great too).
Actually, if the people who wanted to bash out the routine would do so in a library with a proof-of-concept app to go with it, then others could pick that up and run with documenting and interfacing it.
...
PS, (harps again), I'm tired of poor NULL handling in glibc.
strcpy(NULL, "bah") segfaults instead of doing something handlable
For your _fact_ list, why don't you tell me how many Fortune 500 companies use off-the-shelf software to run their businesses? I mean, the database applications, not the back-end software.
I don't care if they use Oracle or DB2, I care who wrote the app their secretaries and automation people use every day. I'm talking about the stuff that drives the robotics equipment in the shop or the order tracking software for their online systems. You're trying to tell me that they use pre-packaged pretty printed shrink-wrapped software to do that?
FUD.
One of the companies I work with is a smaller database design group where three partners design and code stock and custom database solutions for a fairly large user base. Standard policy at the office is that the programmers handle any difficult question; who else would understand it as thoroughly? What I find interesting is how often they recode the program in question online and ask the customer to try it again to their satisfaction ...
Funny, reminds me of the "IBM PC" ... which we all still run according to some pundits ...
;-)
... but IBM sure isn't getting all the revenue
Building clones to a product _can_ be successful, but there has to be a good reason like cost coupled with a lack of difference in functionality. Lots of people still bought real IBM PCs while the IBM PC Clone was taking over, but the slow information push that there was little difference between the two has meant that IBM no longer gets its name in front of the PC anymore.
Why do you think that software developers on Windows for good customised database software are easier to find than on Unix?
...
Oh, you might have missed it; I said 'good'. Much database back-end code is system-independant. FUD spreader
Getting your head out of the sand is in order too.
In an industrial environment, training is always a necessity; the average user uses customised database and control software they've never seen before anyway. Designing such software to be user friendly in a Unix / Windows environment is wholely seperate from the platform it is running on.
Hiring experienced Unix admins can be costly, but with only one long-running Unix server, Windows admins seem to have the misguided impression the school / company would need a team of admins. In fact, one or two solid unix administrators would do fine for most of the Unix maintenance and an external company might even be sufficient for 90% of their needs.
As for "full-featured" bloatware like Office vs. stripped-down but has every feature I've needed software like StarOffice, well, that was my comment.
NCD, the company that makes the thin clients in question, also makes Windows terminal server clients as well.
The comparison in costs would be almost identical because of software licensing costs anyway though.
An actual application compatibility layer _would do_ that, but since this isn't a real compatibility layer but simply falisfying the information returned from a single system call, I don't expect much.
The easy-to-find official comment is that Microsoft doesn't think its code for Windows XP was compromised.
...
It specifically mentioned one product
Considering a 'route' to a destination is pronounced with a soft 'ou' as in "oooh", a 'router' would be properly pronounced 'rooter' since that's exactly its job.
However, 'rowter' has stuck in N-A, so I guess its here to stay.
Common usage has a lot to do with spelling and pronunciation in English, which is why its different in Canada from the U.K. and Jersey from Tennessee from Texas.
They seem, more importantly, to think that Linux people in general (that's 80% of Linux people, for example) _care_ if it becomes a desktop OS for the masses.
Tack on $20 CDN for near-unlimited long distance evening + weekened calling in Canada ... ;-)
... even if they don't have customer service.
If you use a lot of in-Canada long-distance after-hours, $90 a month gets you even better services.
Yes, Bell Canada is still a 'big player' in the world telecom industry. We have it good my friend
I should have kept all the ones I received; I miss blacknet ;-)
I agree that Gilmore is wrong on the issues, but I agree with the EFF on freedom of speech vs. spam. I don't think that spam is protected speech, but I do think that MAPS blocks E-mail and E-communications too arbitrarily to not be abusing peoples' free speech.
I was responding to your post about whitelists as I remember it now and my comments on hashes were w.r.t. that.
I think the EFF is unfortunately right. If we allow independant groups who have no external accountability to create and administer lists like MAPS which, if used widely, can arbitrarily cut off a person's ability to communicate with the outside, we're setting ourselves up for very large freedom of speech problems.
If a person is too loud and someone wanted them silenced, getting them onto all the MAPS-like lists by way of pursuasion (in the future) might be a good method. Being able to silence that person on the Internet shouldn't be possible without public review -- that's why we have court systems for crimes.
MAPS is too vigilante for the EFF is what it comes down to for me, and unfortunately MAPS-like services are very useful but I think closing down spammers themselves is a much better long-term solution.
You contradict your first statement with your second, so why didn't you just delete it?
And what's the "twice"? The whitelist sends you one confirmation, once. It then adds you to their list, assuming they didn't give you a sender-specific address in the first place.
I'm going to ignore most of your message because the first paragraph shows a lack of understanding so deep that your message can't be useful to me.
... read how the tools work.
Read how those filtering systems work. Using the time-based or sender-based hashes means the user's E-mail doesn't have to be verified before getting through. Leaving your E-mail lying around for people to pick up and having a 3 month hash on it will prevent the spam that comes a year from now from long-term collection bots
There's no reason that 'held for confirmation' mail couldn't be left in a different folder (Maildir / IMAP / mbox) and still be visible to the user if they were bored and wanted to check.
This has already been done, but better.
e r is one example that simply creates outgoing addresses that are either only good for a small amount of time or for a specific sender.
http://madhaus.utcs.utoronto.ca/qmail/spam-filt
http://www.xns.org/xns/whitepapers/filtering/ describes XNS for white-list instead of black-list filtering.
There's always this one: http://software.libertine.org/tmda/ the Tagged Mail delivery agent, my personal favorite.
Toss me an E-mail with a ".vbs" file attached for our default bounce message ;-).
...
It can be empty
High-speed ISPs do a horrible job of routing wherever I've looked.
What path do your packets take to get from your house to your neighbour's? How many collisions or how much backplane bandwidth does that cause that are unnecessary?
Connecting neighbours to each other for the sake of their interconnectivity is different than connecting them to each other to save costs on sharing a cable modem.
I missed the part in this article that said everyone should have one in their home.
High-speed CPUs are very useful to our clients who run large database implementations with voice-recognition data-entry systems, FYI.
I have had customers purchase volume licenses of PGP. I stopped recommending it when they stopped releasing source code.
Without the support of the tech community that built up PGP and the web of trust in the first place, they will fail. Maybe Phil's new venture can purchase his software back from NAI?
One thing I've discovered is that beyond civil disobedience, politicians love help more than money. Volunteer and get in their face. Find a local politician with some influence and tell him or her that you'll do the research for them on some totally unrelated issue they're working on. Become a volunteer staffer somewhere with clout. Get involved so that your face is seen and you willingness to help out is known. Then make sure your views are expressed every step of the way -- you'll get heard inherently.