New paradigm: "Laying off programmers makes Open Source software better".
I suspect there are more than 5K programmers working on Linux and related projects, which is what I'd use to compare to the MS Windows (and apps that are "part of the OS") team. How did that go? "With many eyeballs, all bugs are shallow"? Mozilla has gotten much more productive over the last few years, possibly due to the number of bored programmers (hi!) currently unemployed.
Have you got a link to that Harvard study?
10 down to 9 is only a 10% cut in staff. The sort of layoffs I've seen tend to be between 30% and 70%. These were usually accompanied by "we haven't made a profit in a long time" from management. The implication is that if they could fire enough people, they might make a profit.
Going from 10 employees to 9 isn't a layoff, it's more like "firing for cause", "terminating at-will employment", or "retirement".
The other questions I'd have about this Harvard study are: 1) How long did the productivity increase last? 2) How productive was the same team if the person wasn't "let go"? 3) If you do the same test, but cut 2 people using the same system, do you get a 60% increase in productivity? If so, you may be using full-time college students as employees. 4) How was the choice of who to let go made?
I doubt there are many MS OS Coders posting on/. much these days. Especially not the "one" who invented executable email. It seems to me that a lot of MS Things get designed by committee.
The "region-free" version sort of exists already. Just go to www.google.com instead of www.google.de. The restrictions on what is shown by google US are (at least in the stormfront case) less than google.de. No need for any "proxy" yet. Elgoog could be a possibility, as some Chinese users apparently found out after being unable to reach the real thing.
Heh, but you want the commercials playing during the break, so you know when to come back. So, as soon as you come back, it should fast forward to the end of the commercials.
I'd prefer a predicting, mind-reading TiVo that just edits out everything you don't want to watch. TV on really-fast forward. Cross cutting between scenes just before you start thinking "couldn't they say this faster?" or "they said this already" assuming you would have been correct. Hmm... The tenses in this paragraph were a bit tricky.
Of Course this is impossible now, and probably forever, but why bother thinking small?
I'd be surprised if Mobil didn't sue the first company to do that. The cost of defending against the suit, and the negative publicity, would probably negate any benefits of having the TIRIS devices already in the hands of customers. Besides, I think TIRIS (or something much like it) is available much smaller these days.
I could easily see Mobil claiming "theft of customer data", even if it is just a unique serial number. The investment in adding scanners to pumps, stations, etc. is, I think, subject to too much litigation risk for most companies. They'd probably rather use their own system, anyway.
I personally wouldn't need Mobil's permission to walk into a non-Mobil store, and provide the token number and card info to the clerk. That store probably would have to get permission.
For information on a pilot project at Stop&Shop (obviously having the permission of Mobil to use SpeedPass) http://www2.exxonmobil.com/Corporate/N ewsroom/News releases/xom_nr_120203.asp
What a picture: McDonald's, Stop&Shop, and ExxonMobil, all working together. Welcome to Standard Oil, you want Fries with That bottle of milk?
I'm sure I had a point in here somewhere, but it seems to have vanished in the ether.
I agree, cost of gas varies. YMMV, too. Ron Popeil is the one of the canonical examples of how to make money by marketing instead of engineering.
There are at least 5 gas stations in five miles around here. Mobil (with speedpass) is more expensive than most of the others, including the Exxon right across the street. There is one that's occasionally more expensive, but it seems to be one of the "gas station garage" types, where the gas seems to be more than it is elsewhere.
There's also "'n' proprietary solutions means 'n' more things to carry around all the time." I know people around here with more loyalty cards than keys on the keyring. Welcome to the future.
Speedpass is pretty much a "Mobil" only thing, though with the "ExxonMobil" acquisition, that might change. The proprietary solution problem still comes up.
Take your speedpass to BP, or Texaco, or whatever, and watch them ignore it. It would take more effort for them, AND they'd have to go into a contract with Mobil.
The other problem around here, that keeps me from using SpeedPass: All Mobil stations with speedpass charge more for gas than the non-Mobil non-speedpass ones. This may not be the case elsewhere.
Wearer of armor is walking down street. Criminal pulls gun, points it at the armor wearer. Armor activates laser, and fuses bullet to barrel of gun so that, if criminal shoots gun, it just explodes in his face.
I'm not saying it'll ever be possible, and of course it's got problems, including requiring lots of processing power, and mind-reading.:-)
To extend the analogy a bit... I don't normally receive 40 junk snail mail messages at random times each day, with all of them appearing (or claiming) to be "important". Nor do they tend to include envelopes mentioning body parts that most people only talk about to doctors and really close friends.
The only good part of junk snail mail is that it can be used as fuel for a properly modified wood stove.
Opting out of snail mail spam has had the effect of greatly reducing the amount I receive.
Opting out of email spam has had no noticeable effect on the amount of email spam I get, other than possibly increasing it.
Larry Wall does seem to have taken seriously the "Nothing is sacred" bit. Earlier in Apocalypse 5, he wrote (about the earlier (?...) regex syntax decision) "It's not correct now, since the Perl 6 approach is to break everything that needs breaking all at once."
So, I wouldn't worry too much about it. Perl 6 will be different from 5. It appears that it should be better when you look at it from certain directions, and worse from others. Paradigm shift? Sure. I think lwall just got bored with where Perl 5 was going, and wanted to do something different.
Anyway "* signifies a glob" is one of those sacred things that was to be examined carefully. I'm guessing that most useful existing Perl 5 code doesn't use "*" that way except for filehandles. Several other 'modern' languages use "*" for type definitions, so it might make sense from the point of making it easier for a new Perl6 programmer to learn.
I've seen many spams that basically go to "every word in dictionary" at each ISP. I've seen a few that obviously tried "every 3-letter combination" as a login name (26 cubed=17576). It took spammers about a month to find a test 3-letter loginname on a medium-sized ISP, though I don't get many on it yet. There's no real need for any spammer to steal your email address from your own computer, unless your address is not found in the spammer's dictionary.
Currently, 73% of my email (by number of messages, not size) is spam. I did have my ISP turn off their "spam filter", because it tended to junk the wrong email (some relatives couldn't send me email), but still, 73% is a bit much.
From: spam@hotmail.com != went through hotmail.com
on
IETF to Look at Spam
·
· Score: 1
Hotmail.com or Yahoo.com in the headers of an email doesn't mean it came anywhere near those servers. I've gotten more than several spams that were apparently sent through open relays in China, Russia, or almost any other country, with hotmail.com or yahoo.com in the "From" field, but none of them actually touched either company's servers. Forging email is still surprisingly easy, if you can find a badly configured email server. How easy is it to do that? Just check some of the spam mail you get.
I finally went to popfile.sf.net, and though I still download the junk email, at least I can filter it out much easier. Other interesting spam filters include Mozilla(1.3beta+), and CRM114, though I haven't tried that.
None of this filtering is a real solution to the spam problem, and I think it will take something like the IETF or some standards org to get it right.
Many of the older games have timing loops, which expect to take a certain time, but newer CPUs basically optimize those loops out, so they seem to take 0 time. I know that the old Origin game Strike Commander wouldn't run on a Pentium 100 without rebooting into the BIOS, and turning off the CPU cache.
So, to go along with the "never assume the input length is less than a constant", we have "never assume that there's a minimum time to complete an operation."
Maxis has published some of the old SimWhatever games, in boxes containing 3 games. I got the one containing SimTower, City, and Isle. Interesting to see how dated SimCity 95 (real version 2.0 I think) looks compared to SimIsle and SimTower.
I've often wondered how much code Maxis can reuse in all these different "Sim" games, since the basic ideas in all of them are pretty similar.
Electronic Arts, which has done things I don't like, did at least release a whole bunch of their earlier games in simplified "manual-less" packages for about $5 (once they hit the street). BioForge
Every once in a while, I go in to Radio Shack, having forgotten how annoying they can be.
Went in a few days ago, looked around a bit, grabbed a headphone/microphone combination, and a CD cleaning kit. Total price around $30. Went to the counter.
I said "None of these need batteries" (first clue to the salesman 'I KNOW your pitch')
"What's your last name?"
"I'd rather not say." (Second clue)
"OK... Would you like to get $20 off your purchase today?"
"Err.. What do I have to do to get this?"
"Sign up for sprint long distance service at your house"
"No thanks" (third clue.. you don't stop marketting at me, you're out.)
"Do you have an ISP?"
"Yes"
"Which one is it?"
"Netplex"
"How much do they charge per month?"
clue-by-four:
At that point I stormed out the door, leaving the merchandise (unbought) on the counter.
I tried to opt out of Radio Shack, but the only way to do that is not to go into the store.
# Look for gifs with tracking info tacked on the
# back
/.*\.(gif|jpe?g)\?.*
# Block some adloggers
/.*adlog.pl\?.*
An example of a coremetrics eluminate link is at the end of this comment (taken from www.coremetrics.com). Interesting how it keeps track of everywhere I looked on the site before I looked in the page source... Yes, it really was this long... Perhaps looking for CGIs with 'http' in the middle of the link will do it.
/.*http.*
This may break some other sites, but does seem to work on coremetrics.com.
http://data.coremetrics.com/cgi-bin/eluminate.cg i?pt%3DC%26vn1%3De2.2.8%26vn2%3De2.2.8%26c i%3D22222224%26rf%3Dhttp%253A//www.coremetrics.com /home2.html%26ul%3Dhttp%253A//www.coreme trics.com/asp_model.html%26se%3D%26pn%3DThe%2520AS P%2520Model%26pi%3DThe%2520ASP%2520Model %26cn%3DThe%2520ASP%2520Model%26sc%3DSolution%26ps 1%3D%26ps2%3D%26pn1%3D%26pn2%3D%26a1%3D% 26a2%3D%26a3%3D%26pa%3D%26pc%3DYes%26ts%3Dnull%26t p%3Dnull%26rnd%3D2387440
The 105th Congress' H.R. 4176 is an anti-jamming, anti-spamming, anti-slamming thing. It never made it out of the House of Reps. See thomas.loc.gov, and remember to go to the 'previous congress' search area.
The current Congress' H.R. 4176 is the Information Technology Act of 2000.
If you're really interested in the 'legal bits', check out http://thomas.loc.gov It's got a search engine that can help track down the text and status of bills.
An interesting clause of s.1618 would have made it a 'civil' offence to send spam without putting a physical address and telephone number at the top of the spam, so the message probably wasn't in compliance in the first place. (see Title III section 301)
That said, s.1618 of the 105th congress never made it into law, so it doesn't matter anyway.
HR.4176 never even passed the house of representatives, so you can safely ignore that reference.
A reference to a real federal law should be of the form <number> USC <number>, or refer to some section of the U.S. Code (of law).
Claiming compliance to a bill that hasn't even passed congress, much less been signed by the President, is meaningless drivel, and a waste of bits.
I have checked into the details of absentee voting. It's simple. Really simple. Just call your local County Clerk or Registrar of Voters, preferably the Registrar, and say "I need an absentee ballot", and they'll help get the ball rolling. The local political machine actually sent one to me because they thought I'd voted absentee once, and might like to do so again. There is a form to fill out, and you've got to put a reason, but I'm sure "Too busy to drive to polling place" is OK.
I don't see how a truly unbreakable system (which doesn't yet exist) would encourage the movie studios to release things to the public that they wouldn't have before. I see this solely as a way for the movie studios to be able to charge more to provide essentially the same content as they provide today. Almost every new movie technology that's come out has been followed shortly by an increase in ticket price.
Haven't they learned ANYTHING from DivX? Forcing people to buy more expensive equipment that has more limitations on what can be done with the data just doesn't work very well.
I guarantee that HDCP will increase the cost of a digital display. How significantly, I don't know, the spec isn't public yet.
As someone earlier in this discussion said, there's still the Linux-based display problem. No one in the Linux world was willing to shell out $10000 for the DVD license. I don't see anyone wanting to shell out more money for more restrictions anytime soon.
I won't be buying or using anything DVD or CD related until the MPAA, DVD "open forum", etc all come to their senses (ie, when hell warms up (See Dante)). Equivalently with Intel's HDCP, I won't be buying a digital display with such encryption technology built in.
Slashdot Mirror
New paradigm: "Laying off programmers makes Open Source software better".
I suspect there are more than 5K programmers working on Linux and related projects, which is what I'd use to compare to the MS Windows (and apps that are "part of the OS") team. How did that go? "With many eyeballs, all bugs are shallow"? Mozilla has gotten much more productive over the last few years, possibly due to the number of bored programmers (hi!) currently unemployed.
Have you got a link to that Harvard study?
10 down to 9 is only a 10% cut in staff. The sort of layoffs I've seen tend to be between 30% and 70%. These were usually accompanied by "we haven't made a profit in a long time" from management. The implication is that if they could fire enough people, they might make a profit.
Going from 10 employees to 9 isn't a layoff, it's more like "firing for cause", "terminating at-will employment", or "retirement".
The other questions I'd have about this Harvard study are:
1) How long did the productivity increase last?
2) How productive was the same team if the person wasn't "let go"?
3) If you do the same test, but cut 2 people using the same system, do you get a 60% increase in productivity? If so, you may be using full-time college students as employees.
4) How was the choice of who to let go made?
I doubt there are many MS OS Coders posting on /. much these days. Especially not the "one" who invented executable email. It seems to me that a lot of MS Things get designed by committee.
The "region-free" version sort of exists already. Just go to www.google.com instead of www.google.de. The restrictions on what is shown by google US are (at least in the stormfront case) less than google.de. No need for any "proxy" yet. Elgoog could be a possibility, as some Chinese users apparently found out after being unable to reach the real thing.
For those missing the joke, Bush today announced that S.H. and sons had 48 hours to leave Iraq, or the US would start the attack.
Heh, but you want the commercials playing during the break, so you know when to come back. So, as soon as you come back, it should fast forward to the end of the commercials.
I'd prefer a predicting, mind-reading TiVo that just edits out everything you don't want to watch. TV on really-fast forward. Cross cutting between scenes just before you start thinking "couldn't they say this faster?" or "they said this already" assuming you would have been correct. Hmm... The tenses in this paragraph were a bit tricky.
Of Course this is impossible now, and probably forever, but why bother thinking small?
I'd be surprised if Mobil didn't sue the first company to do that. The cost of defending against the suit, and the negative publicity, would probably negate any benefits of having the TIRIS devices already in the hands of customers. Besides, I think TIRIS (or something much like it) is available much smaller these days.
N ewsroom/News releases/xom_nr_120203.asp
I could easily see Mobil claiming "theft of customer data", even if it is just a unique serial number. The investment in adding scanners to pumps, stations, etc. is, I think, subject to too much litigation risk for most companies. They'd probably rather use their own system, anyway.
I personally wouldn't need Mobil's permission to walk into a non-Mobil store, and provide the token number and card info to the clerk. That store probably would have to get permission.
For information on a pilot project at Stop&Shop (obviously having the permission of Mobil to use SpeedPass)
http://www2.exxonmobil.com/Corporate/
What a picture: McDonald's, Stop&Shop, and ExxonMobil, all working together.
Welcome to Standard Oil, you want Fries with That bottle of milk?
I'm sure I had a point in here somewhere, but it seems to have vanished in the ether.
I agree, cost of gas varies. YMMV, too. Ron Popeil is the one of the canonical examples of how to make money by marketing instead of engineering.
There are at least 5 gas stations in five miles
around here. Mobil (with speedpass) is more expensive than most of the others, including the Exxon right across the street. There is one that's occasionally more expensive, but it seems to be one of the "gas station garage" types, where the gas seems to be more than it is elsewhere.
There's also "'n' proprietary solutions means 'n' more things to carry around all the time." I know people around here with more loyalty cards than keys on the keyring. Welcome to the future.
Speedpass is pretty much a "Mobil" only thing, though with the "ExxonMobil" acquisition, that might change. The proprietary solution problem still comes up.
Take your speedpass to BP, or Texaco, or whatever, and watch them ignore it. It would take more effort for them, AND they'd have to go into a contract with Mobil.
The other problem around here, that keeps me from using SpeedPass: All Mobil stations with speedpass charge more for gas than the non-Mobil non-speedpass ones. This may not be the case elsewhere.
--
This message no sig.
Smart Active Armor.
:-)
Wearer of armor is walking down street. Criminal pulls gun, points it at the armor wearer. Armor activates laser, and fuses bullet to barrel of gun so that, if criminal shoots gun, it just explodes in his face.
I'm not saying it'll ever be possible, and of course it's got problems, including requiring lots of processing power, and mind-reading.
To extend the analogy a bit... I don't normally receive 40 junk snail mail messages at random times each day, with all of them appearing (or claiming) to be "important". Nor do they tend to include envelopes mentioning body parts that most people only talk about to doctors and really close friends.
The only good part of junk snail mail is that it can be used as fuel for a properly modified wood stove.
Opting out of snail mail spam has had the effect of greatly reducing the amount I receive.
Opting out of email spam has had no noticeable effect on the amount of email spam I get, other than possibly increasing it.
See the difference yet?
George Orwell Bush
Pathologically Eclectic Rubbish Lister... Second definition, from the man-page (man perl) itself.
But it sure is a useful piece of code. Any programming language will have bits in it that aren't perfect. Perl has some, Tcl has others.
All languages suck. All OSes suck. Everything sucks (except MS Vacuum), and that's not gravity you're feeling.
(Maybe I should have gotten some rest before typing that)
Larry Wall does seem to have taken seriously the "Nothing is sacred" bit. Earlier in Apocalypse 5, he wrote (about the earlier (?...) regex syntax decision) "It's not correct now, since the Perl 6 approach is to break everything that needs breaking all at once."
So, I wouldn't worry too much about it. Perl 6 will be different from 5. It appears that it should be better when you look at it from certain directions, and worse from others. Paradigm shift? Sure. I think lwall just got bored with where Perl 5 was going, and wanted to do something different.
Anyway "* signifies a glob" is one of those sacred things that was to be examined carefully. I'm guessing that most useful existing Perl 5 code doesn't use "*" that way except for filehandles. Several other 'modern' languages use "*" for type definitions, so it might make sense from the point of making it easier for a new Perl6 programmer to learn.
I've seen many spams that basically go to "every word in dictionary" at each ISP. I've seen a few that obviously tried "every 3-letter combination" as a login name (26 cubed=17576). It took spammers about a month to find a test 3-letter loginname on a medium-sized ISP, though I don't get many on it yet. There's no real need for any spammer to steal your email address from your own computer, unless your address is not found in the spammer's dictionary.
Currently, 73% of my email (by number of messages, not size) is spam. I did have my ISP turn off their "spam filter", because it tended to junk the wrong email (some relatives couldn't send me email), but still, 73% is a bit much.
Hotmail.com or Yahoo.com in the headers of an email doesn't mean it came anywhere near those servers. I've gotten more than several spams that were apparently sent through open relays in China, Russia, or almost any other country, with hotmail.com or yahoo.com in the "From" field, but none of them actually touched either company's servers. Forging email is still surprisingly easy, if you can find a badly configured email server. How easy is it to do that? Just check some of the spam mail you get.
I finally went to popfile.sf.net, and though I still download the junk email, at least I can filter it out much easier. Other interesting spam filters include Mozilla(1.3beta+), and CRM114, though I haven't tried that.
None of this filtering is a real solution to the spam problem, and I think it will take something like the IETF or some standards org to get it right.
Many of the older games have timing loops, which expect to take a certain time, but newer CPUs basically optimize those loops out, so they seem to take 0 time. I know that the old Origin game Strike Commander wouldn't run on a Pentium 100 without rebooting into the BIOS, and turning off the CPU cache.
So, to go along with the "never assume the input length is less than a constant", we have "never assume that there's a minimum time to complete an operation."
Maxis has published some of the old SimWhatever games, in boxes containing 3 games. I got the one containing SimTower, City, and Isle. Interesting to see how dated SimCity 95 (real version 2.0 I think) looks compared to SimIsle and SimTower.
I've often wondered how much code Maxis can reuse in all these different "Sim" games, since the basic ideas in all of them are pretty similar.
Electronic Arts, which has done things I don't like, did at least release a whole bunch of their earlier games in simplified "manual-less" packages for about $5 (once they hit the street). BioForge
/.*http.* /.*http.*
Breaks much of www.fool.com...
so, add
~fool.com
~bigcharts.com
just after
so that any other blocks further down still
take effect.
Check out the graphic nav. on their home page, near the top, right side: "Important information about coremetrics"
According to T'r'us. April 20 through August 2 was how long they used the service, and they don't anymore.
Every once in a while, I go in to Radio Shack, having forgotten how annoying they can be.
Went in a few days ago, looked around a bit, grabbed a headphone/microphone combination, and a CD cleaning kit. Total price around $30. Went to the counter.
I said "None of these need batteries" (first clue to the salesman 'I KNOW your pitch')
"What's your last name?"
"I'd rather not say." (Second clue)
"OK... Would you like to get $20 off your purchase today?"
"Err.. What do I have to do to get this?"
"Sign up for sprint long distance service at your house"
"No thanks" (third clue.. you don't stop marketting at me, you're out.)
"Do you have an ISP?"
"Yes"
"Which one is it?"
"Netplex"
"How much do they charge per month?"
clue-by-four:
At that point I stormed out the door, leaving the merchandise (unbought) on the counter.
I tried to opt out of Radio Shack, but the only way to do that is not to go into the store.
# Look for gifs with tracking info tacked on the
g i?pt%3DC%26vn1%3De2.2.8%26vn2%3De2.2.8%26c i%3D22222224%26rf%3Dhttp%253A//www.coremetrics.com /home2.html%26ul%3Dhttp%253A//www.coreme trics.com/asp_model.html%26se%3D%26pn%3DThe%2520AS P%2520Model%26pi%3DThe%2520ASP%2520Model %26cn%3DThe%2520ASP%2520Model%26sc%3DSolution%26ps 1%3D%26ps2%3D%26pn1%3D%26pn2%3D%26a1%3D% 26a2%3D%26a3%3D%26pa%3D%26pc%3DYes%26ts%3Dnull%26t p%3Dnull%26rnd%3D2387440
# back
/.*\.(gif|jpe?g)\?.*
# Block some adloggers
/.*adlog.pl\?.*
An example of a coremetrics eluminate link is at the end of this comment (taken from www.coremetrics.com). Interesting how it keeps track of everywhere I looked on the site before I looked in the page source... Yes, it really was this long... Perhaps looking for CGIs with 'http' in the middle of the link will do it.
/.*http.*
This may break some other sites, but does seem to work on coremetrics.com.
http://data.coremetrics.com/cgi-bin/eluminate.c
The 105th Congress' H.R. 4176 is an anti-jamming, anti-spamming, anti-slamming thing. It never made it out of the House of Reps. See thomas.loc.gov, and remember to go to the 'previous congress' search area.
The current Congress' H.R. 4176 is the Information Technology Act of 2000.
Strange, but apparently true.
If you're really interested in the 'legal bits',
check out http://thomas.loc.gov
It's got a search engine that can help track down the text and status of bills.
An interesting clause of s.1618 would have made it a 'civil' offence to send spam without putting a physical address and telephone number at the top of the spam, so the message probably wasn't in compliance in the first place. (see Title III section 301)
That said, s.1618 of the 105th congress never made it into law, so it doesn't matter anyway.
HR.4176 never even passed the house of representatives, so you can safely ignore that reference.
A reference to a real federal law should be of the form <number> USC <number>, or refer to some section of the U.S. Code (of law).
Claiming compliance to a bill that hasn't even passed congress, much less been signed by the President, is meaningless drivel, and a waste of bits.
Rule number 1: Spammers lie.
I have checked into the details of absentee voting. It's simple. Really simple. Just call your local County Clerk or Registrar of Voters, preferably the Registrar, and say "I need an absentee ballot", and they'll help get the ball rolling. The local political machine actually sent one to me because they thought I'd voted absentee once, and might like to do so again. There is a form to fill out, and you've got to put a reason, but I'm sure "Too busy to drive to polling place" is OK.
You may have to supply a stamp for the envelope.
Err.. Not exactly.
I don't see how a truly unbreakable system (which doesn't yet exist) would encourage the movie studios to release things to the public that they wouldn't have before. I see this solely as a way for the movie studios to be able to charge more to provide essentially the same content as they provide today. Almost every new movie technology that's come out has been followed shortly by an increase in ticket price.
Haven't they learned ANYTHING from DivX? Forcing people to buy more expensive equipment that has more limitations on what can be done with the data just doesn't work very well.
I guarantee that HDCP will increase the cost of a digital display. How significantly, I don't know, the spec isn't public yet.
As someone earlier in this discussion said, there's still the Linux-based display problem. No one in the Linux world was willing to shell out $10000 for the DVD license. I don't see anyone wanting to shell out more money for more restrictions anytime soon.
I won't be buying or using anything DVD or CD related until the MPAA, DVD "open forum", etc all come to their senses (ie, when hell warms up (See Dante)). Equivalently with Intel's HDCP, I won't be buying a digital display with such encryption technology built in.