I hold that IT staff who are so unresponsible as to
run an intrinsically insecure OS on a mission critical system are are much a bunch of criminals
as the time-wasters who write viruses when they
could be developing Free Software instead.
[Disclaimer: I can't claim to know whether the coast guard's computers are really mission-critical, maybe they only read their fan emails on these machines anyway.]
I can only fully recommend the Mappit A4F 19-1HE, which comes in a black chassis ready for a 19" rack. I bought my first one recently, and it's doing a great job. This box is ideal for servers run in private homes, since they are low-energy consumption (low-$$ when run 24/7 like mine) and low-noise (can live in a corner of your living-room
without being noticed).
It's built by a German company, but they shipped
it to the UK quickly and hassle-free, so they might as well ship to the US.
A friend of mine, who had first recommended the Mappit series to me, runs a set of smaller ones. All solid metal cases, and Linux/BSD
installed easily (no 'Microsoft tax').
> Rackmount? For 3 pcs? Something small?... ATX!
Yes and yes -- luckily, that's not a contradiction
at all.
No, sorry, I believe you are wrong here;
both the Web and USENET are client-server based.
Web: Your Web broser (e.g. Mozilla) is a client, which sends requests to the Web server, which is a, well, server (nomen est omen).
USENET: Your newsreader, e.g. gnus is a client, the USERNET news server, is, again, what the name says: a server.
The fact that there is
data replication between servers doesn't make
them a P2P system. For a scenario to be P2P, both
parties need to be able to submit requests to
others AND answer requests with responses.
true, but google seems to be the one company that has managed to really make money with advertising on the internet.
One option is to create another Google applicance,
the GMail appliance, to again sell their service to corporate clients in the same out-of-the-box fashion as in the Google search appliance.
Sergey Brin's dad: poetry for son's 25th birthday
on
Google's Early Hardware
·
· Score: 2, Funny
Here's part of a poem mentioning written by Sergey Brin's father Michael Brin (he's a mathematician at U Maryland) on the occasion of his son's 25th birthday:
...
You are tough, you mine data,
You surf first and think later,
And your crawler fast as light
Wanders madly in the night. ...
Surf first, think later, huh? Not sure I can agree with that...;-)
The title of this post is a bit misleading--the
series doesn't really tell you how to upgrade your
system landscape from NT to Linux at all, it's
merely a Linux tutorial for Windows users.
I can see no corporate aspects discussed.
"After the Second World War the director of the Computer Laboratory, M[aurice] V. Wilkes, headed what was perhaps the most influential of Britain's postwar computer projects the building of the EDSAC. Modelled on the American storedprogram concept that Wilkes had heard outlined at the Moore School lectures in August 1946, the EDSAC (Electronic Delay Storage Automatic Calculator) ran its first calculation in May 1949. The objective of the design team Wilkes, W. Renwick, S. Barton and G. Stevens on the hardware side; and D.J. Wheeler on the programming side was to provide a useful and reliable computing service. Such a service, the first in the world using a storedprogram computer, was available from early 1950. A significant feature of the Cambridge approach was the attention paid to userconvenience and programming; hence the group's book, The Preparation of Programmes for an Electronic Digital Calculator (1951), became the first textbook on programming a storedprogram computer, and was soon regarded as a classic."
(Quote, hyperlinks added by JLL)
[Can ANYONE explain] the latest craze about social networks like Orkut or Friendster? I'm indifferent to the internet or the people on the internet so I highly doubt social networks on the net will ever be even slightly interesting for me, personally. But I really don't see what's worth the fuss about them, because they aren't exactly incorporating ground-breaking technologies
The technology behind Orkut is not in itself
revolutionary, but bear in mind that social relationships provide a source of authority and thus provide potential for job recommendations and personalized search.
For Google, Orkut is a convenient way to bind
users to its brand, widening their strategy
(search -> profiles/messaging/...).
This is essential for them in the growing
competition faced from Yahoo.
Also, think about what they can do with relationship graphs: if I stand in relationship
with person X and person X has property Y, then
chances are I might not be completely opposed to
property Y, or otherwise I'd not be friends with
Y. Using all sorts of graph properties, you can
build personality profiles and use them for better
ad targeting and personalized Web search (bear
in mind that PageRank is already graph based).
Orkut's terms of use allow them to sell or use
commercially all data you include in a profile.
Users were at first unsure of a hierarchical file system. They all seemed to adopt the room and mentality for file access. They understood the concept of a "file" as a name of a box where the computer will store some data which is placed in their personal "room" but directories proved a difficult concept for them to handle.
I can only confirm this: in a course I once taught,
participants found the hierarchical file system
(commands: pwd, cd, special directories:..,.,/, ~/, and its concepts of relative/absolute paths)
more difficult to grasp than extended
regular expressions, much to my surprise.
For those hackers who are wondering what the radiation displays mean:
The unit micro-Rontgen is the ISO-unit of radiation.
1 uR ('u': read mu, the Greek letter) is the quantity of radiation means 10E-06 times the associated corpuscular emission per 0.001293 g of dry air (of volume 1 cm^3 at 0 degrees Celsius temperature and 760 mm Hg presure) produces in air ions carrying 1 electrostatic unit of electrical charge.
PS: sorry that I can't use proper symbols here, as./ seems not to render neither UNICODE SGML entities properly nor accept ISO 8859-1 input.:/
... but I really wish SOMEONE would do some "duplication and evolution;" maybe THAT would light a fire under some a[....]s at [G]oogle.
Not duplication, revolution is the notion you
want: Google was successful because its founders believed in a
completely new paradigm, that graph-based methods
(PageRank, HITS) could outperform dusty (but
effective) vector-space retrieval.
Many people have a shady intuition of what
information retrieval really is ("Um.. yeah, you look the pages up in which the keywords occur"), trivializing the area. Go to any top-500 company and try their site search if you want to have a good laugh.
What we need is once more something completely
different. It still holds that there is more than one way to do it!
One way is to go ahead and build a distributed
indexing scheme (see my earlier posting on this theme), borrowing conepts from SETI@home or Freenet, because an index that cannot be located anywhere cannot be controlled. It might also be a better
test-bed for large-scale experiments, but where
only few developers want to try out new algorithms ("at home"), using the distributed indices built on distributed, donated diskspace around the world.
It's often not clear whether a problem is hardware or software (graphics card versus its driver), but that's not my point.
If I am the customer, I demand to be served. If I
am willing to pay, I want my problems fixed.
It would be fine for DELL to say that they can't
do it for free, but my point is they didn't offer
any help nor did they seem to have any expertise.
The German "Manager Magazin" had a report a couple
of years ago about a new cinema that used a similar mechanism with cartridges beneath each
seat.
There are three main problems: 1. It's expensive because you have to refill the cartridges. 2. The set of smells is too restricted to make it interesting. 3. Humans are very sensitive to smells. It's easy to get it wrong and put off
the customer!
"PSCP" is not new at all. Some Commodore C= 64
games back in 1982 were delivered in an encrypted
form. Only a short window around the program
counter's (PC) current position would be
decrypted on demand by an interrupt-driven
procedure.
Empirical study shows that such protection mechanisms are very weak.
The new Perl regex notation might be slightly
less horrible than the old one, but I think
regular expressions are to be rejected altogether,
as they make software un-maintainable.
The nice thing about it is that it's fully declarative
and bi-directional, i.e. a:b can be applied
substituting a by b or vice versa (if run
backwards), whereas the traditional
if (/pattern/) { action; }
pattern-action paradigm that Perl has inherited
from awk is only partially declarative (only
the pattern).
Another interesting property is that xfst, one
of Xerox' compilers, allows naming and re-use
of subexpressions:
(all without any ugly '$'s:-)
Abstraction by naming is a powerful feature, as
Abelson and Sussman (SICP) and other good
elementary textbooks point out.
Maybe anybody wants to volunteed to build something like this into Perl6 instead?
It's due to what some consider the extreme American "work ethic." Here, you are expected to work hard, all the time, preferably six or seven days a week, until you "retire" (more and more people now work during "retirement").
Well spotted. It's worth remembering that humans work to live, not live to work.
Except geeks, for which of course the distinction between work and play does not exist.:)
Brief Historic Note
on
How C# Was Made
·
· Score: 2, Informative
"After 13 years with Borland, Hejlsberg joined Microsoft in 1996 [...]"
If you remember what happened a while ago, the
story goes: Microsoft was facing competition from
Borland, so they hired away all their lead
developers to break the company's neck.
It almost worked.
PS: Don't use anything that has a '#' in it and isn't music!
Good news. But at the same time, and as usual, this is a grass roots movement, while officials can't be bothered too much.
I asked DELL support about a Linux-related problem
and they told me they don't support Linux.
They said my laptop was shipped with Windows 2000,
so they can't be bothered. While this is certainly
true, it's not MY fault that they didn't ship
Linux in 2001 when I bought it.
I've also contacted them about RAID systems,
and they corporate sales folks I had contact with
didn't really know what they were talking about,
so I had to get in touch with a RedHat developer
to answer my question (which he instantly did).
Maybe here's a good way to make money for distro
companies: try get a service contract with
hardware vendors like DELL, who haven't got enough
inhouse expertise (at support level).
I do hope this engineering effort is part of
a wider wave in the company.
Actually hell, the U.K. may already have that. I forgot you don't have a written Constitution that prevents such invasions of privacy and self-incrimination.
True, but on the other hand the UK don't have ID
cards either; in fact, a policeman is not even
allowed to stop you and ask you for your name.
Besides the degree of monitoring, there is also
the question of balance: there's no point being
over-strict in one realm and sloppy in another;
there needs to be a consistent policy.
... did Virginia not realize that 1100 full-size G5's might cause problems?...
I've criticized the whole idea before,
especially the hype around it, when I can't
see any big new ideas happening.
Statements like
The price of the upgrade has not yet settled on, but Varadarajan said it would be minimal compared to the cost of building a new supercomputer from scratch
are just ridiculous, since why would you need to build a new multi-million supercomputer from scratch if you have never even used your brand-new "old" one? That's like saying
"tearing down our just-completed new villa and
building the new house another way will save us
so much more compared to building the same one
again the same way!"
As Pike (of UNIX fame) and -- more recently --
Jobbs have noted, there's not a lot of innovation going on anymore.
Having said this the Xserves are very nice designs, so if anywhere, the cup for cool ideas
goes to industry (Apple's engineers), not academia (Virginia Tech) -- also in the second round.
Here's my proposal: Why don't the guy at VTech
not build a new user interface that goes beyond
the useful, but aged desktop metaphor that the
Mac introduced to the masses twenty years ago?
Or how about some serious study of automatic load balancing on the "old" supercomputer? They might "save even more" money by taking some time to learn from mistakes in the first
round before diving blindly into the next generation of their Uber-Mac project.
(Sorry for the rant, but it seems such a waste of resources when not too far away
people don't even have their jobs anymore.)
It would be interesting to hear about plans
for integrating "STANDARD COMPLIANCE" sections
into the Linux version of the POSIX man pages
(or other Linux-specific material).
How are the two document bases (GNU man documents vs. POSIX man documents) going to be integrated?
In terms of copyright protection, in order to bring suit your copyright must be registered with the US Copyright Office.
Note that such issues are very country-specific.
According to German law, for instance, no formal
'registration' is necessary. ANY evidence that
somebody is the author of a work is valid in
court.
I disagree with the initial poster about the
informal email. I recommend a written letter
be sent (and don't forget to send it asking
for a receipt that it has arrived).
If you are serious about it, informal
emails can be a waste of time (in the sense that you mostly will have to write a letter afterwards anyway).
However, they can be a way to create a more
productive atmosphere,
preparing the ground for resolving the conflict - while a more formal letter is already on its way.
Slashdot Mirror
[Disclaimer: I can't claim to know whether the coast guard's computers are really mission-critical, maybe they only read their fan emails on these machines anyway.]
I fed it with the query "dog" and expected to get back all hits for "god", but instead it redirected me to another engine :(
The first URL should have read: http://www.mappit.de/content/computer/bilder/a4f19 /a4f19_cd.jpg (image) / http://www.mappit.de/ (homepage)
I can only fully recommend the Mappit A4F 19-1HE , which comes in a black chassis ready for a 19" rack. I bought my first one recently, and it's doing a great job. This box is ideal for servers run in private homes, since they are low-energy consumption (low-$$ when run 24/7 like mine) and low-noise (can live in a corner of your living-room without being noticed).
It's built by a German company, but they shipped it to the UK quickly and hassle-free, so they might as well ship to the US.
A friend of mine, who had first recommended the Mappit series to me, runs a set of smaller ones. All solid metal cases, and Linux/BSD installed easily (no 'Microsoft tax').
> Rackmount? For 3 pcs? Something small? ... ATX!
Yes and yes -- luckily, that's not a contradiction at all.
Happy Easter, all!
Web: Your Web broser (e.g. Mozilla) is a client, which sends requests to the Web server, which is a, well, server (nomen est omen).
USENET: Your newsreader, e.g. gnus is a client, the USERNET news server, is, again, what the name says: a server.
The fact that there is data replication between servers doesn't make them a P2P system. For a scenario to be P2P, both parties need to be able to submit requests to others AND answer requests with responses.
One option is to create another Google applicance, the GMail appliance, to again sell their service to corporate clients in the same out-of-the-box fashion as in the Google search appliance.
You are tough, you mine data,
You surf first and think later,
And your crawler fast as light
Wanders madly in the night.
Surf first, think later, huh? Not sure I can agree with that... ;-)
The title of this post is a bit misleading--the series doesn't really tell you how to upgrade your system landscape from NT to Linux at all, it's merely a Linux tutorial for Windows users. I can see no corporate aspects discussed.
"After the Second World War the director of the Computer Laboratory, M[aurice] V. Wilkes, headed what was perhaps the most influential of Britain's postwar computer projects the building of the EDSAC. Modelled on the American storedprogram concept that Wilkes had heard outlined at the Moore School lectures in August 1946, the EDSAC (Electronic Delay Storage Automatic Calculator) ran its first calculation in May 1949. The objective of the design team Wilkes, W. Renwick, S. Barton and G. Stevens on the hardware side; and D.J. Wheeler on the programming side was to provide a useful and reliable computing service. Such a service, the first in the world using a storedprogram computer, was available from early 1950. A significant feature of the Cambridge approach was the attention paid to userconvenience and programming; hence the group's book, The Preparation of Programmes for an Electronic Digital Calculator (1951), became the first textbook on programming a storedprogram computer, and was soon regarded as a classic."
(Quote, hyperlinks added by JLL)
The technology behind Orkut is not in itself revolutionary, but bear in mind that social relationships provide a source of authority and thus provide potential for job recommendations and personalized search.
For Google, Orkut is a convenient way to bind users to its brand, widening their strategy (search -> profiles/messaging/...). This is essential for them in the growing competition faced from Yahoo.
Also, think about what they can do with relationship graphs: if I stand in relationship with person X and person X has property Y, then chances are I might not be completely opposed to property Y, or otherwise I'd not be friends with Y. Using all sorts of graph properties, you can build personality profiles and use them for better ad targeting and personalized Web search (bear in mind that PageRank is already graph based). Orkut's terms of use allow them to sell or use commercially all data you include in a profile.
I can only confirm this: in a course I once taught, participants found the hierarchical file system (commands: pwd, cd, special directories: .., ., /, ~/, and its concepts of relative/absolute paths)
more difficult to grasp than extended
regular expressions, much to my surprise.
Brave girl.
./ seems not to render neither UNICODE SGML entities properly nor accept ISO 8859-1 input. :/
For those hackers who are wondering what the radiation displays mean:
The unit micro-Rontgen is the ISO-unit of radiation.
1 uR ('u': read mu, the Greek letter) is the quantity of radiation means 10E-06 times the associated corpuscular emission per 0.001293 g of dry air (of volume 1 cm^3 at 0 degrees Celsius temperature and 760 mm Hg presure) produces in air ions carrying 1 electrostatic unit of electrical charge.
PS: sorry that I can't use proper symbols here, as
Not duplication, revolution is the notion you want: Google was successful because its founders believed in a completely new paradigm, that graph-based methods (PageRank, HITS) could outperform dusty (but effective) vector-space retrieval.
Many people have a shady intuition of what information retrieval really is ("Um.. yeah, you look the pages up in which the keywords occur"), trivializing the area. Go to any top-500 company and try their site search if you want to have a good laugh.
What we need is once more something completely different. It still holds that there is more than one way to do it!
One way is to go ahead and build a distributed indexing scheme (see my earlier posting on this theme), borrowing conepts from SETI@home or Freenet, because an index that cannot be located anywhere cannot be controlled. It might also be a better test-bed for large-scale experiments, but where only few developers want to try out new algorithms ("at home"), using the distributed indices built on distributed, donated diskspace around the world.
If I am the customer, I demand to be served. If I am willing to pay, I want my problems fixed. It would be fine for DELL to say that they can't do it for free, but my point is they didn't offer any help nor did they seem to have any expertise.
There are three main problems: 1. It's expensive because you have to refill the cartridges. 2. The set of smells is too restricted to make it interesting. 3. Humans are very sensitive to smells. It's easy to get it wrong and put off the customer!
Empirical study shows that such protection mechanisms are very weak.
They should be replaced by something better, such as Xerox' Extended Regular Relation Calculus: Here are some examples.
The nice thing about it is that it's fully declarative and bi-directional, i.e. a:b can be applied substituting a by b or vice versa (if run backwards), whereas the traditional
pattern-action paradigm that Perl has inherited from awk is only partially declarative (only the pattern).Another interesting property is that xfst, one of Xerox' compilers, allows naming and re-use of subexpressions:
(all without any ugly '$'sAbstraction by naming is a powerful feature, as Abelson and Sussman (SICP) and other good elementary textbooks point out.
Maybe anybody wants to volunteed to build something like this into Perl6 instead?
Well spotted. It's worth remembering that humans work to live, not live to work.
Except geeks, for which of course the distinction between work and play does not exist. :)
Here's what the National Security Agency says about Security-Enhanced Linux.
If you remember what happened a while ago, the story goes: Microsoft was facing competition from Borland, so they hired away all their lead developers to break the company's neck.
It almost worked.
PS: Don't use anything that has a '#' in it and isn't music!
Good news. But at the same time, and as usual, this is a grass roots movement, while officials can't be bothered too much.
I asked DELL support about a Linux-related problem and they told me they don't support Linux. They said my laptop was shipped with Windows 2000, so they can't be bothered. While this is certainly true, it's not MY fault that they didn't ship Linux in 2001 when I bought it.
I've also contacted them about RAID systems, and they corporate sales folks I had contact with didn't really know what they were talking about, so I had to get in touch with a RedHat developer to answer my question (which he instantly did).
Maybe here's a good way to make money for distro companies: try get a service contract with hardware vendors like DELL, who haven't got enough inhouse expertise (at support level).
I do hope this engineering effort is part of a wider wave in the company.
True, but on the other hand the UK don't have ID cards either; in fact, a policeman is not even allowed to stop you and ask you for your name. Besides the degree of monitoring, there is also the question of balance: there's no point being over-strict in one realm and sloppy in another; there needs to be a consistent policy.
I've criticized the whole idea before, especially the hype around it, when I can't see any big new ideas happening.
Statements like The price of the upgrade has not yet settled on, but Varadarajan said it would be minimal compared to the cost of building a new supercomputer from scratch are just ridiculous, since why would you need to build a new multi-million supercomputer from scratch if you have never even used your brand-new "old" one? That's like saying "tearing down our just-completed new villa and building the new house another way will save us so much more compared to building the same one again the same way!"
As Pike (of UNIX fame) and -- more recently -- Jobbs have noted, there's not a lot of innovation going on anymore.
Having said this the Xserves are very nice designs, so if anywhere, the cup for cool ideas goes to industry (Apple's engineers), not academia (Virginia Tech) -- also in the second round.
Here's my proposal: Why don't the guy at VTech not build a new user interface that goes beyond the useful, but aged desktop metaphor that the Mac introduced to the masses twenty years ago? Or how about some serious study of automatic load balancing on the "old" supercomputer? They might "save even more" money by taking some time to learn from mistakes in the first round before diving blindly into the next generation of their Uber-Mac project.
(Sorry for the rant, but it seems such a waste of resources when not too far away people don't even have their jobs anymore.)
How are the two document bases (GNU man documents vs. POSIX man documents) going to be integrated?
Note that such issues are very country-specific.
According to German law, for instance, no formal 'registration' is necessary. ANY evidence that somebody is the author of a work is valid in court.
I disagree with the initial poster about the informal email. I recommend a written letter be sent (and don't forget to send it asking for a receipt that it has arrived). If you are serious about it, informal emails can be a waste of time (in the sense that you mostly will have to write a letter afterwards anyway).
However, they can be a way to create a more productive atmosphere, preparing the ground for resolving the conflict - while a more formal letter is already on its way.