We all know what the law is getting at. The sensible interpretation for what code to disclose/escrow is all code that is not otherwise available in the marketplace which takes care of the subsidiary thing. This means that Diebold doesn't have to escrow Windows or Windows CE, but *does* have to disclose proprietary code written *for* them by anybody, along with all the code Diebold has written that builds/integrates the 3rd-party code into the product.
I'll second that. The second thing I did on my first use of the QuietComfort(r) headphones, right after sighing "Ahhhhh..." was turn my Aireo down. Just no need for that volume when the sound reproduction is so good and outside noise is inhibited.
For various reasons, I find Viagra ads objectionable. Can't I skip that? My Aunt Tillie may find the violence in the next Vin Diesel smash hit's tailer objectionable, but if it is an ad, she can't skip it?
I think this bill, or a similar one, will eventually pass, but founder on the shoals of the judiciary.
An excellent, concise summary, my man. Just wish I could forward to the family mailing list. Though, to be fair, they pretty much all reside in blue states, anyway, so it would be like poking 'em with a stick.
>...and read the *actual*, *effective* result in the >next day's newspaper. I can wait. Amen, bro! Until then, it is just a horserace, which is fine entertainment only if you have a beer in hand and money on a pony.
I was thinking along the same lines. Consider if a majority of MTAs out there had established keys. Simply giving priority to transferring email from "keyed/trusted" servers over non-trusted servers would *slow* spam propagation, or at least, would help non-spam not be slowed by spam.
And, since the MTAs have public keys to work with, they can securely exchange some session/pipeline-specific symmetric keys for fast and secure transfers of email, as STARTTLS does.
Public keys could be exchanged at session initiation, but that is not much different than STARTTLS today. An improvement would be to either store key fingerprints in the DNS (which scales) or use the PGP keyservers (which probably would not, initially). A private local key cache could easily be established, using signing, web-o-trust, and all that. Don't laugh -- we used to do a hell of a lot more setup for a UUCP connection back in the day.
We should all be doing that anyway. Why are we all only sending postcards on the Internet rather than letters in envelopes? Are we all on vacation all the time?
I believe that setting up "SMTP webs of trust" is the right way to go, and would at least allow email to be sorted as it travels between trusted webs of MTAs. Email originating in, and transitting through, trusted MTAs could then be treated preferentially with regards to filtering, or at least, what gets processed/queued first.
An MTA wouldn't need to be part of.mail to be trusted, but its operator would have to do whatever it takes to join particular webs of trust, which could mean that it meets policies and guidelines similar to what is suggested for.mail.
Who would run these webs? The answer isn't that hard -- we used to do that for USENET news all the time, really.
... of the Galaxy Rangers! You've reminded me that I have to find and re-read that wonderful Harry Harrison book. He knows how to take his space-opera spoofing seriously.
Lomborg is the poster child for "Lies, damn lies, and statistics."
Re:Greatest and most significant Human achievment.
on
Human Accomplishment
·
· Score: 1
I rather think it was fermentation that changed everything. Hunter/gatherers became farmers, cooperated to get larger quantities of grain and/or grapes, and so on.
Perhaps "Learning to Party" was the greatest achievment.
"People like you make me sick. If I buy music, its mine to share with whomever I choose."
No, dammit. If you buy music, it is yours to view and store however you choose *OR* yours to *transfer ownership* to whomever you choose.
Otherwise we are in today's situation, where we have lost (or are losing) what we had in the vinyl/tape age and still don't have the benefits of quality *and* choice *and* distribution we should have in the digital age.
The long view includes history as well as the future.
The function fatal() calls various *_cleanup() functions in a call-back fashion. Some of these cleanups do buffer_free() calls on Buffer structs. The buffer_free() function fills the buffer up to its now bogus alloc length with zeros and then frees it.
At least I can see that possibility. I think I will apply the patch and also fatal-out on xrealloc() returning NULL.
Slashdot Mirror
We all know what the law is getting at. The sensible interpretation for what code to disclose/escrow is all code that is not otherwise available in the marketplace which takes care of the subsidiary thing. This means that Diebold doesn't have to escrow Windows or Windows CE, but *does* have to disclose proprietary code written *for* them by anybody, along with all the code Diebold has written that builds/integrates the 3rd-party code into the product.
Without strong copyright laws, how would you grant a license for use of your own open source software, be it GPL, BSD, or whatever?
I'll second that. The second thing I did on my first use of the QuietComfort(r) headphones, right after sighing "Ahhhhh..." was turn my Aireo down. Just no need for that volume when the sound reproduction is so good and outside noise is inhibited.
I've waited for the school bus in sub -40 weather (and remember, 40 below is 40 below no matter which scale you use).
"But it's a dry cold." -- any average Minnesotan explaing why we live here to any average Californian.
'Nuff said!
For various reasons, I find Viagra ads objectionable. Can't I skip that? My Aunt Tillie may find the violence in the next Vin Diesel smash hit's tailer objectionable, but if it is an ad, she can't skip it?
I think this bill, or a similar one, will eventually pass, but founder on the shoals of the judiciary.
An excellent, concise summary, my man. Just wish I could forward to the family mailing list. Though, to be fair, they pretty much all reside in blue states, anyway, so it would be like poking 'em with a stick.
And, hasn't the massive amount of deforestation over the past century already had the opposite effect?
>...and read the *actual*, *effective* result in the >next day's newspaper. I can wait.
Amen, bro! Until then, it is just a horserace, which is fine entertainment only if you have a beer in hand and money on a pony.
I was thinking along the same lines. Consider if a majority of MTAs out there had established keys. Simply giving priority to transferring email from "keyed/trusted" servers over non-trusted servers would *slow* spam propagation, or at least, would help non-spam not be slowed by spam.
And, since the MTAs have public keys to work with, they can securely exchange some session/pipeline-specific symmetric keys for fast and secure transfers of email, as STARTTLS does.
Public keys could be exchanged at session initiation, but that is not much different than STARTTLS today. An improvement would be to either store key fingerprints in the DNS (which scales) or use the PGP keyservers (which probably would not, initially). A private local key cache could easily be established, using signing, web-o-trust, and all that. Don't laugh -- we used to do a hell of a lot more setup for a UUCP connection back in the day.
Indeed! I've had the same email address for a dozen years - how much spam will I get daily after 120?
We should all be doing that anyway. Why are we all only sending postcards on the Internet rather than letters in envelopes? Are we all on vacation all the time?
*resume OT* The current governor pisses-off plenty of suburbanites, you can be sure. *end OT* :-)
I believe that setting up "SMTP webs of trust" is the right way to go, and would at least allow email to be sorted as it travels between trusted webs of MTAs. Email originating in, and transitting through, trusted MTAs could then be treated preferentially with regards to filtering, or at least, what gets processed/queued first.
.mail to be trusted, but its operator would have to do whatever it takes to join particular webs of trust, which could mean that it meets policies and guidelines similar to what is suggested for .mail.
An MTA wouldn't need to be part of
Who would run these webs? The answer isn't that hard -- we used to do that for USENET news all the time, really.
And here I was just thinking that "... bypassing browser restrictions," etc. was a security flaw just waiting to be exploited.
... of the Galaxy Rangers! You've reminded me that I have to find and re-read that wonderful Harry Harrison book. He knows how to take his space-opera spoofing seriously.
Seriously, you all need to read more Golden Age Sci-Fi....
"Mr. Harrison on line 1, sir...."
Ah, I see you've read "Fahrenheit 451".
Lomborg is the poster child for "Lies, damn lies, and statistics."
I rather think it was fermentation that changed everything. Hunter/gatherers became farmers, cooperated to get larger quantities of grain and/or grapes, and so on.
Perhaps "Learning to Party" was the greatest achievment.
As root: update-menus -v
Puts it all back.
I agree it shouldn't have gone away with updates, but it ain't fatal.
"People like you make me sick. If I buy music, its mine to share with whomever I choose."
No, dammit. If you buy music, it is yours to view and store however you choose *OR* yours to *transfer ownership* to whomever you choose.
Otherwise we are in today's situation, where we have lost (or are losing) what we had in the vinyl/tape age and still don't have the benefits of quality *and* choice *and* distribution we should have in the digital age.
The long view includes history as well as the future.
You are correct! Thanks for following up.
The function fatal() calls various *_cleanup() functions in a
call-back fashion. Some of these cleanups do buffer_free() calls on
Buffer structs. The buffer_free() function fills the buffer up to its
now bogus alloc length with zeros and then frees it.
At least I can see that possibility. I think I will apply the patch and also fatal-out on xrealloc() returning NULL.
Yeah, and the return value of xrealloc isn't
checked against NULL, neither.