More likely: swapping fans, power supplies and disks, putting a memorabilia sticker and handing out to faculty and into the labs. After all it is a piece of history and as such the fact that it is a bit out of spec will not be offensive for many people out there. If I worked there I for once would have liked to have it on my desk once it has been taken out of the machine room.
I am betting a case of beer that the BLIAR will once again suck up to Bill and open his election campaign at the MSFT UK headquaters near Reading as he did in the last elections.
I am betting ten cases of beer that this will not get anywhere. You do not expect anything but MSFT to go around in a country where the prime minister prefers to open his campaing at MSFT headquarters instead of any members of the FTSE 100 list.
Fascinating, so what do all these devices function on?
Dark magic vooooooooodooooooooooooo?
How many computers you have in the house? How many lightbulbs? How many routers do you have in the house? How many distribution boxes?
Ad fscking naseum...
The problem is that the grid in US does not have algorithms to adapt and decrease demands in times when capacity is at premium.
The control algorithms are usually extremely primitive and based on simple feedback. When you combine lack of adaptability in consumption and feedback control you ALWAYS get a cascade system failure as a main failure scenario. It can be proven mathematically.
What is needed is to cut off sections elsewhere to decrease load which consumers will immediately scream about. Let's say you have a main line cut in Ohio. In order to compensate for the increased load on other lines you cut off two residential districts in Maine. Why? Because it happens to fit the bill to decrease the consumption. Guess what will all those people vote for in the next elections. And they will not give a flying F*** that if they would have remained with power for 5 more minutes the entire East Coast would have plunged into darkness.
The methodology is something I have expected to come for a while now.
All I can say is that the right answer is the last line of the article.
It is actually the right answer to all SPAM problems period. Especially when applied to the company which is using it to promote their trade, not the spammers. The latter will die by themselves if there will be noone to buy their product.
You mean that in another century or so marketing will rule the world and people will not give a flying fuck about what are their car (software/hardware) specs are and look only at who makes it?
There ain't such thing as Toyota Camry in Japan since 1999. What is sold in ther rest of the world as Camry is actually a Daihatsu Altis. But you would not buy a pensioner brand such as Daihatsu, would you? So that is the reason Toyota group sells it as Camry. If you do not believe me go to the Daihatsu and Toyota Japanese sites and read the model lists and the specs
So, no, thanks. I like the world as it is now, where you can actually still buy software and hardware based on spec, not on Brand.
The human readable result is you need to know what you want. There is no silver bullet.
It looks like xfs wipes the floor for all but temporary (loads of create/delete) file usage. Jfs looks like the best all-rounder. Reiser looks like something that can be tuned to the specific usage, but eats CPU for breakfast, lunch and dinner and EXT3 "surprise, surprise" sucks rocks. The other "surprise, surprise" is that EXT2 is still very good for many uses.
Frankly, I do not see anything new and fascinating in the results, but they are good to throw at people who keep asking me "why not EXT3" and "Why XFS or EXT2". Here is why!
We have used it long before XP and in a completely different context - complex mathematical problems as well as data aquisition and process control systems. When the logic and the underlying math are hair-raising having another set of eyes following the development is extremely beneficial. Also, some people have a talent to drive and some people have a talent to navigate. This also means that you are usually either the pilot or navigator and there is no such thing as scheduled swap as it does not make any sense. Sometimes the navigator gets the keyboard, but this is only when the pilot does not see what the navigator is trying to explain.
Note that this is not XP at all. XP is an approach designed for cases which can be split into a multitude of extremely simple problems. In XP you need a pair in order not to stray and cut corners in development because it looks easy.
This IMO still leaves a considerable chunk of software development that is not useable for pair programming. It is not so simple that XP works. It is also not complex enough to warrant a copilot. It is the middle ground. There, writing solo, was, is and shall be the best approach. Grafting XP on it only leads to low efficiency solutions.
I think it should be worth the community to ask SGI to do this and more specifically to do it versus a BSD version that is old enough to suggest a theft by SCO and not vice versa. Alternatively, SysV licensees are well known so we might as well ask every single one of them.
Re:NDAs are a necessary evil to some environments
on
The Cult of the NDA
·
· Score: 0
When competitors show up, it might be a sign that this is good stuff,
Exactly. The thing you should be most afraid of is the "no competition whatsoever" case. People are not exactly stupid and the probability that someone looked at the same thing you are doing and found a fatal flow in the business is quite high.
It is much harder to make money by doing something totally new compared to starting in an existing field with tough competition and make your way through by having an edge your competititors do not have.
it is the economics of medium and large projects. Plain and simple.
There are quite a few companies indulging in various forms of technological roccoco out there and doing grand Web projects.
In this case they have the choice of doing it in either:
1. Java
2. C# or any of the other MSFT languages using the.NET framework.
If you look at the average java developer salary in the UK in london it is hovering above 45K. If you look at the average C#.NET salary it is under 35K.
Considering that 95% of the companies out there do not understand the concept of remote management and require staff to be shackled to the desk 9x5 (and relevant shifts) the remote management of the Unix systems no longer plays a role. If people took it into account it would have been _obviously_ more efficient to use Linux, Solaris, AIX, etc for _grand_ technological roccoco. People do not use it so it comes to non-obvious reasons. And in the age when companies count pennies noone wants to try to prove to the management that saving 10K on salaries will result in 20K charge over the lifetime of the project if the 20K is not bleeding obvious. And once the remote management factor is away it is definitely no longer bleeding obvious.
Overall, we are in the era of "Savings now or die". So expect more of 2003 on the Net as it fits such PHB thinking very well.
At least in those days if I recall my 1991 MSDOS books right Microsoft was formally proclaiming OS/2 to be the future and all MS to go that way as well as recommending to program in an OS/2 compliant manner.
Not that anyone at the time gave a f**k. So I think it was noone giving a f**k which did OS2. MSFT helped it only by means of providing so slow, lame and crippled implementation of their DOS APIs that anyone in his sane mind bypassed them left right and center. As a result the programs did not work under OS2. When WARP came out it could do emulation of real mode in protected mode as per the 386 spec, but it was too bloody late.
So IMO you are mistaking common programming practice from those days with the actual MSFT party line. Just take any MSFT press book from those time (I have a few around). And MSFT just followed the flow. And accidentally won.
Now this is the only meaningfull comment for the entire article.
Actually not just IBM. Anyone with any contribution can. And frankly, I would be more afraid of Intel laywer squadrons, and let's not even mention people like Serverworks, Broadcom, 3Com, so on so forth. Compared to them IBM is very agreable and kind.
On your other comment: It is on the border of this: Western Caucasus (N ii, iv/ 1999). Possibly incorporated into it as it was a site prior to 1999. Dunno. Too many wars there and I have not been in the region for 20+ years.
There is another relict grove in Pitcunda on the Russian Black Sea coast. Due to something noone so far understands which happened over the last 600 or so years it no longer reproduces. The peninsula itself is slowly sinking into the sea after several earthquakes in the region in the 60-es.
So for now there is another grove and it is also listed as world heritage site by Unesco. Note the "for now" as you will not see any saplings from it. You are least likely to see the grove itself in a few hundred years either (it is awesome).
Not to mention that nobody has written an "auto spreader" for linux such as slammer.
Bollocks. There was one for DNS, one for apache and usually there is at least one for every major vulnerability.
Linux is the most breached OS on the net minus
Possibly true due to the fact that it is both usefull to the atacker, second most popular after MSFT and it is quite often unprotected. If Solaris was as popular with end-lusers as Linux it would have been breached twice as often. After all you cannot beat decades of insecure coding combined with the stubborned stupidity of refusing to ship tools to do ip filtering though the relevant kernel infrastructure (BPF) is in place (on SloWarez that is).
PPP sends keep-alives if you do not use the link. These contain well known and often predictable plaintext.
Any protocol encrypting these with the data encryption key is offering a free ride for a known plaintext attack. Also, depending on your cypher you are offering free ride on the PPP header and IP header fields which do not change enough between frames. This is is the usual case for links without proper protocol field compression. So on so forth ad naseum until you puke and some more to boot.
Once you use SSL for a VPN protocol you will get into most of the problems which G describes anyway due to the excessive amount of well known predictable plain-text in the message. SSL is good for what it was designed for - as means of securing TCP with arbitrary contents.
You will also get a severe performance hit because every datagram will have to be ACK-ed.
CIPE is one of the best infrastructure VPN packages out there for the solely reason that it uses an un-acked protocol and offers the notion of interface to both end. As a result you can run a routing protocol on it.
CIPE has its failings, it can be improved, but if you think that IP or PPP in SSL are any better think again. Hit yourself with the aforementioned soundwave a few times if needed.
Primo: I have worked with SAIC and frankly I would question any mention of the word independent along with their name especially when it comes to open reviews.
Secundo: On the basis of the experience with code delivered by SAIC, the competence of the reviewer should be octuple checked as well.
It is a common flaw in a corporation going down to start believing their own PR. As with most things like this it is not always clear what is first the chicken or the egg.
In other words, you should have an alarm siren going off in your head the moment you see the bosses lose their healthy scepticism and paint a rosy picture of the bright future.
Slashdot Mirror
More likely: swapping fans, power supplies and disks, putting a memorabilia sticker and handing out to faculty and into the labs. After all it is a piece of history and as such the fact that it is a bit out of spec will not be offensive for many people out there. If I worked there I for once would have liked to have it on my desk once it has been taken out of the machine room.
which cost a fair bit more
Or less...
The power of this err... is nothing compared to the power of the force (marketing).
Just think of all the free publicity they have got so far.
Note this.
I am betting a case of beer that the BLIAR will once again suck up to Bill and open his election campaign at the MSFT UK headquaters near Reading as he did in the last elections.
I am betting ten cases of beer that this will not get anywhere. You do not expect anything but MSFT to go around in a country where the prime minister prefers to open his campaing at MSFT headquarters instead of any members of the FTSE 100 list.
Fascinating, so what do all these devices function on?
Dark magic vooooooooodooooooooooooo?
How many computers you have in the house? How many lightbulbs? How many routers do you have in the house? How many distribution boxes?
Ad fscking naseum...
The problem is that the grid in US does not have algorithms to adapt and decrease demands in times when capacity is at premium.
The control algorithms are usually extremely primitive and based on simple feedback. When you combine lack of adaptability in consumption and feedback control you ALWAYS get a cascade system failure as a main failure scenario. It can be proven mathematically.
What is needed is to cut off sections elsewhere to decrease load which consumers will immediately scream about. Let's say you have a main line cut in Ohio. In order to compensate for the increased load on other lines you cut off two residential districts in Maine. Why? Because it happens to fit the bill to decrease the consumption. Guess what will all those people vote for in the next elections. And they will not give a flying F*** that if they would have remained with power for 5 more minutes the entire East Coast would have plunged into darkness.
The methodology is something I have expected to come for a while now.
All I can say is that the right answer is the last line of the article.
It is actually the right answer to all SPAM problems period. Especially when applied to the company which is using it to promote their trade, not the spammers. The latter will die by themselves if there will be noone to buy their product.
You mean that in another century or so marketing will rule the world and people will not give a flying fuck about what are their car (software/hardware) specs are and look only at who makes it?
There ain't such thing as Toyota Camry in Japan since 1999. What is sold in ther rest of the world as Camry is actually a Daihatsu Altis. But you would not buy a pensioner brand such as Daihatsu, would you? So that is the reason Toyota group sells it as Camry. If you do not believe me go to the Daihatsu and Toyota Japanese sites and read the model lists and the specs
So, no, thanks. I like the world as it is now, where you can actually still buy software and hardware based on spec, not on Brand.
The human readable result is you need to know what you want. There is no silver bullet.
It looks like xfs wipes the floor for all but temporary (loads of create/delete) file usage. Jfs looks like the best all-rounder. Reiser looks like something that can be tuned to the specific usage, but eats CPU for breakfast, lunch and dinner and EXT3 "surprise, surprise" sucks rocks. The other "surprise, surprise" is that EXT2 is still very good for many uses.
Frankly, I do not see anything new and fascinating in the results, but they are good to throw at people who keep asking me "why not EXT3" and "Why XFS or EXT2". Here is why!
Yes, they are.
We have used it long before XP and in a completely different context - complex mathematical problems as well as data aquisition and process control systems. When the logic and the underlying math are hair-raising having another set of eyes following the development is extremely beneficial. Also, some people have a talent to drive and some people have a talent to navigate. This also means that you are usually either the pilot or navigator and there is no such thing as scheduled swap as it does not make any sense. Sometimes the navigator gets the keyboard, but this is only when the pilot does not see what the navigator is trying to explain.
Note that this is not XP at all. XP is an approach designed for cases which can be split into a multitude of extremely simple problems. In XP you need a pair in order not to stray and cut corners in development because it looks easy.
This IMO still leaves a considerable chunk of software development that is not useable for pair programming. It is not so simple that XP works. It is also not complex enough to warrant a copilot. It is the middle ground. There, writing solo, was, is and shall be the best approach. Grafting XP on it only leads to low efficiency solutions.
In other words it is delayed because the entire world has understood how halfbaked it is.
With the code out there any reviewer can do a quick search for the TODO/BUGS and reproduce the condition.
As a result if they release it without fixing them they will get slaughtered.
So IMO:
1. The leak may be deliberate as a publicity stunt which gives them a chance to delay it despite investor pressure.
2. They intended to do as some other game companies have done in the past and ship utter bogus buggy crap.
Dunno which of these is the case. But it is either.
Took the words out of my mouth.
I think it should be worth the community to ask SGI to do this and more specifically to do it versus a BSD version that is old enough to suggest a theft by SCO and not vice versa. Alternatively, SysV licensees are well known so we might as well ask every single one of them.
Exactly. The thing you should be most afraid of is the "no competition whatsoever" case. People are not exactly stupid and the probability that someone looked at the same thing you are doing and found a fatal flow in the business is quite high.
It is much harder to make money by doing something totally new compared to starting in an existing field with tough competition and make your way through by having an edge your competititors do not have.
it is the economics of medium and large projects. Plain and simple.
.NET framework.
.NET salary it is under 35K.
There are quite a few companies indulging in various forms of technological roccoco out there and doing grand Web projects.
In this case they have the choice of doing it in either:
1. Java
2. C# or any of the other MSFT languages using the
If you look at the average java developer salary in the UK in london it is hovering above 45K. If you look at the average C#
Considering that 95% of the companies out there do not understand the concept of remote management and require staff to be shackled to the desk 9x5 (and relevant shifts) the remote management of the Unix systems no longer plays a role. If people took it into account it would have been _obviously_ more efficient to use Linux, Solaris, AIX, etc for _grand_ technological roccoco. People do not use it so it comes to non-obvious reasons. And in the age when companies count pennies noone wants to try to prove to the management that saving 10K on salaries will result in 20K charge over the lifetime of the project if the 20K is not bleeding obvious. And once the remote management factor is away it is definitely no longer bleeding obvious.
Overall, we are in the era of "Savings now or die". So expect more of 2003 on the Net as it fits such PHB thinking very well.
Or have your humour gland amputated at birth.
At least in those days if I recall my 1991 MSDOS books right Microsoft was formally proclaiming OS/2 to be the future and all MS to go that way as well as recommending to program in an OS/2 compliant manner.
Not that anyone at the time gave a f**k. So I think it was noone giving a f**k which did OS2. MSFT helped it only by means of providing so slow, lame and crippled implementation of their DOS APIs that anyone in his sane mind bypassed them left right and center. As a result the programs did not work under OS2. When WARP came out it could do emulation of real mode in protected mode as per the 386 spec, but it was too bloody late.
So IMO you are mistaking common programming practice from those days with the actual MSFT party line. Just take any MSFT press book from those time (I have a few around). And MSFT just followed the flow. And accidentally won.
Now this is the only meaningfull comment for the entire article.
...
Actually not just IBM. Anyone with any contribution can. And frankly, I would be more afraid of Intel laywer squadrons, and let's not even mention people like Serverworks, Broadcom, 3Com, so on so forth. Compared to them IBM is very agreable and kind.
Fun fun fun
Or they will die (in business sense) on purpose so that the two fscking arseholes footing the bill in first place can finance 10 new SCOs.
Assuming IBM is in OSS for the long run they would not like anyone getting any silly ideas to try the SCO route anytime in the future.
They could have bought SCO ten times so far and can still buy it, but this will create 10 new SCOs. And this not something even IBM can afford.
On your other comment: It is on the border of this: Western Caucasus (N ii, iv/ 1999). Possibly incorporated into it as it was a site prior to 1999. Dunno. Too many wars there and I have not been in the region for 20+ years.
Corect, not the same species. But relict as well.
There is another relict grove in Pitcunda on the Russian Black Sea coast. Due to something noone so far understands which happened over the last 600 or so years it no longer reproduces. The peninsula itself is slowly sinking into the sea after several earthquakes in the region in the 60-es.
So for now there is another grove and it is also listed as world heritage site by Unesco. Note the "for now" as you will not see any saplings from it. You are least likely to see the grove itself in a few hundred years either (it is awesome).
Bollocks. There was one for DNS, one for apache and usually there is at least one for every major vulnerability. Linux is the most breached OS on the net minus
Possibly true due to the fact that it is both usefull to the atacker, second most popular after MSFT and it is quite often unprotected. If Solaris was as popular with end-lusers as Linux it would have been breached twice as often. After all you cannot beat decades of insecure coding combined with the stubborned stupidity of refusing to ship tools to do ip filtering though the relevant kernel infrastructure (BPF) is in place (on SloWarez that is).
PPP sends keep-alives if you do not use the link. These contain well known and often predictable plaintext.
Any protocol encrypting these with the data encryption key is offering a free ride for a known plaintext attack. Also, depending on your cypher you are offering free ride on the PPP header and IP header fields which do not change enough between frames. This is is the usual case for links without proper protocol field compression. So on so forth ad naseum until you puke and some more to boot.
Nope it is not.
Once you use SSL for a VPN protocol you will get into most of the problems which G describes anyway due to the excessive amount of well known predictable plain-text in the message. SSL is good for what it was designed for - as means of securing TCP with arbitrary contents.
You will also get a severe performance hit because every datagram will have to be ACK-ed.
CIPE is one of the best infrastructure VPN packages out there for the solely reason that it uses an un-acked protocol and offers the notion of interface to both end. As a result you can run a routing protocol on it.
CIPE has its failings, it can be improved, but if you think that IP or PPP in SSL are any better think again. Hit yourself with the aforementioned soundwave a few times if needed.
Primo: I have worked with SAIC and frankly I would question any mention of the word independent along with their name especially when it comes to open reviews.
Secundo: On the basis of the experience with code delivered by SAIC, the competence of the reviewer should be octuple checked as well.
It is a common flaw in a corporation going down to start believing their own PR. As with most things like this it is not always clear what is first the chicken or the egg.
In other words, you should have an alarm siren going off in your head the moment you see the bosses lose their healthy scepticism and paint a rosy picture of the bright future.