...Their support largely consists of other users in support forums, with the majority of the cost absorbed by the client organisation.
In most cases I'd agree with you: vendor support is a dumbed down user forum much inferior to Stackoverflow.
However, it has been a couple of years since I last managed larger teams using this software, but the support from SAS was simply outstanding. Nothing like it anywhere else. You call, they answer and you speak immediately to someone really knowledgeable in the tool, in stats (the main use case for SAS), and, more often than not, in your industry. Never had to do more than one transfer to get anything resolved.
You pay for support but this is the only vendor where I ever felt you got your money's worth.
These days even the client realises that it is more data science and knowledge discovery, and less about stats which more or less zeros the value of the SAS support. I wish them well; by all accounts they were a great company to work for back in the day when they still mattered.
was most surprised when I found that Redhat... appears to allow [root to ssh into your machine] as the default.
Useful when you are setting up a brand new (remote) server. You are supposed to use the access only once to (a) create a new account and (b) disable root ssh access (by setting 'PermitRootLogin no' in/etc/sshd/sshd_config).
In both the UK and Netherlands at least, if you browse the web using a Vodafone 3G data card or an O2 GPRS connection, the images will be changed to lower-resolution copies (at the original URL!), the full-resolution images will be in a different place, and your web page will have inserted a whole load of JavaScript code that in principle allows you to get from the corrupted -- sorry: optimized -- version to the original. Assuming JavaScript works on your browser and with your settings, and assuming that the hot-keys they are using have not been used for something else on the web page you are browsing.
Of course the JavaScript code could also sniff out all your credit card numbers and send them to the company, but I had a look and this version does not appear to do that....
And what's the point of listening to lossless audio formats on earbud headphones...
That's why the Etymotic ER-4P Earphones were invented. Granted, you need to re-mortgage your house to afford them, but you can't beat their 35-40 dB noise reduction that allows you to hear you music at normal volumes while you are out and about. And it sounds great: the response of these earphones is great.
But otherwise I agree. What is the point of anything other than low-quality mp3 if you are not going to be able to listen to it anyhow.
The average/. reader is an idiot. Half of/. readers are below average. Are you scared yet?
That's the median you are thinking off, not the average. But your general point is probably valid: If intelligence in the general population is normal distributed and/. has an average that is below that of the general population, you would expect more than half of the readers to be below the/. average (because there is more range available above the average so those people have a higher weight in the average).
Thought you wanted to know....
This announcement brought to you by another average/. reader
Writing a spec is the easy part (and this one seems particularly trivial). Implementing it is a lot harder.
The main missing information seems to be a DNS equivalent function. It is one thing to introduce yet another central registrar to insure "against hostile usurpation or inappropriate usage of registered service marks" (groan!) but how are we going to access the information? Section 4.2 says
The "info" Registry will be publicly accessible and will support discovery (by both humans and machines) of [...lots of stuff...]
and it is clear from section 6 (Normalization and Comparison of "info" URIs) that any sensible implementation would need access, but how?
The information they want to return is much more complex than what DNS returns and DNS is not a trivial infrastructure.
The one problem I've seen with Amazon's affiliate program is that while it covers customization and can display ads based on keywords, those ads aren't very attractive. The best way to target your customers (I've seen) is through very subtle, individual item links. It's not too hard to manage but it's another thing on the list of stuff to check every week (out of stock, discontinued, etc.).
Ehm, but that's exactly the point! That's why the Amazon Web Services (which is what we are discussing here) is so useful: by giving you an API to the Amazon database and shopping functionality, it allows you to make the ads in the way that works best on your site, and to maintain them automatically.
You did read the ar.... nevermind. Slashdot. I forgot:-)
The core of Verisign's defence seems to be that domain names are not property rights. From the BBC article on the same ruling:
Forcing Verisign to accept blame for transferring the domain name in the first place could prove equally difficult.
Verisign maintains that domain names are not legal property and as such it cannot be held accountable for giving it away.
If it loses, as legal experts expect, Verisign would face a huge legal bill and fines of up to £100m.
Two questions: what excatly am I buying when I buy a domain name from Verisign and why do "legal experts" think they'll loose that battle -- presumably they have an extensive user agreement that clears them of responsibility for all and any wrongdoings?
Where is this energy actually coming from? In our normal ecosystem, it comes from the sun via photosynthesis.
(Offtopic, but this is a common fallacy and this is news for nerds...)
Actually, the Earth radiates slightly more energy than it receives from the Sun. If it didn't radiate all the energy, it would get very hot, very quickly. (The excess ratiation os from nuclear decays in the Earth's crust.)
The difference is that the Sun is hot and the Earth is (comparatively) cold. The difference is in the quality of the energy - what physicists call entropy.
Think of it this way: With the energy from a hot plate I can boil water while with the energy from a plate that is only warm I cannot, no matter how long I leave it turned on (i.e. no matter how much energy I spend).
We live off the entropy from the Sun, not the energy.
It doesn't appear to be an IE parsing error, it looks like it was blocked from IE at mozilla.org. As in if I use proxomitron to fake my user-agent it displays.
It appears to be valid XHTML and valid use of CSS, so your comments seem strange, unless they serve different documents for different user agents. Have anybody tested this?
You can't boot from a USB device, can you? Can the BIOS keep up with all these new devices?
About the only time I use the floppy is to boot a reduced Linux system to try an recover a (usually Windows) machine that has goot itself in a twist.
Very handy. I guess I could make a bootable CD. And a bootable USB device. And hope that the BIOS allows me to boot from either.
I guess it had to go, but I'd like to see a universally available replacement first, supported by every (new, I guess) BIOS.
Maybe it is all a plot to make sure we only boot pre-installed operating systems?:-)
for instance if you take a picture of Big Ben at 2:13 on a given day and there's one tiny person at the bottom
Probably a bad example: Unlike most other European countries, the United Kingdom does not have and provisions in law that gives you a right to privacy. If you are in a public place, then you are in public, and I can take your photo and publish it to my heart's content.
There is a code that the newspapers tend to follw which says that you shouldn't publish pictures of people taken with "very" long telephoto lenses without their consent, but that is just a code of practice, not law.
All of this is likely to change at the European convention on human rights -- which does have a provision guaranteeing some privacy -- is incorporated into British law.
Always remember that (1) not all the world is like the US and (2) if you take any advice given on/., in particular legal advice, serious, then you deserve everything you get...
...since XML is just text, and text is not compressed (usually).. how can a XML based db even be plausible...
Nevermind that it is text, the important point is that is is a tree structure.
Standard databases are relational, and are great at storing simple attributes for an object. They are absolutely horrible at storing relationships between these objects and, more importantly, in managing those relationships.
So, for example, if you have a grommet that can consist of multiple other grommets, each of couse consisting of grommets etc., then in XML you are laughing:
In a standard relational database you end up with a grommet table and, perhaps, an attribute that is the parent grommet. To get the list I just suggested above, you need to do a self-join on the grommet table an unknown number of times, something SQL just can't do.
Object-oriented databases are good at this (and much more), and it is funny that the old style of databases that preceeded the relational databases, were often hierachical, i.e. tree structures!
I love Poland but is it really essential to fix the Polish language bugs for a 1.0 release? Aren't there more important priorities? Isn't 1.0 about a stable API (and product!) and such, and if so, couldn't fixing spelling mistakes in the Polish language pack wait until 1.0.1 or something?
The document outlines some really good principles for managing software, but this entry confuses it for me. Any Polish people here to explain why it is critical?:-)
It isn't the scripting per se. It's the fact that the scripts are actually stored in the document files. In other words, they mix data and code.
On Unix, lots of applications have extremely powerfull scripting languages. Just think about the stuff you can do with Emacs (elisp)...
Actually, Emacs mixes data and code in the same way. Check the File Variables section in the info system, and in particular the enable-local-eval variable. Basically, you can set buffer local variables by embedding the commands for this at the end of the file. One of these variables is 'eval':-). Thus spake RMS:
The `eval' "variable," and certain actual variables, create a
special risk; when you visit someone else's file, local variable
specifications for these could affect your Emacs in arbitrary ways.
Therefore, the option `enable-local-eval' controls whether Emacs
processes `eval' variables, as well variables with names that end in
`-hook', `-hooks', `-function' or `-functions', and certain other
variables. The three possibilities for the option's value are `t',
`nil', and anything else, just as for `enable-local-variables'. The
default is `maybe', which is neither `t' nor `nil', so normally Emacs
does ask for confirmation about file settings for these variables.
In this sense Emacs is just as guilty as Microsoft Office. Just because it's Free doesn't mean it is without security free. (But the fact that the average person using Emacs is more clued in than you Power Point suit, does help...)
Slashdot Mirror
...Their support largely consists of other users in support forums, with the majority of the cost absorbed by the client organisation.
In most cases I'd agree with you: vendor support is a dumbed down user forum much inferior to Stackoverflow.
However, it has been a couple of years since I last managed larger teams using this software, but the support from SAS was simply outstanding. Nothing like it anywhere else. You call, they answer and you speak immediately to someone really knowledgeable in the tool, in stats (the main use case for SAS), and, more often than not, in your industry. Never had to do more than one transfer to get anything resolved.
You pay for support but this is the only vendor where I ever felt you got your money's worth.
These days even the client realises that it is more data science and knowledge discovery, and less about stats which more or less zeros the value of the SAS support. I wish them well; by all accounts they were a great company to work for back in the day when they still mattered.
Useful when you are setting up a brand new (remote) server. You are supposed to use the access only once to (a) create a new account and (b) disable root ssh access (by setting 'PermitRootLogin no' in /etc/sshd/sshd_config).
In both the UK and Netherlands at least, if you browse the web using a Vodafone 3G data card or an O2 GPRS connection, the images will be changed to lower-resolution copies (at the original URL!), the full-resolution images will be in a different place, and your web page will have inserted a whole load of JavaScript code that in principle allows you to get from the corrupted -- sorry: optimized -- version to the original. Assuming JavaScript works on your browser and with your settings, and assuming that the hot-keys they are using have not been used for something else on the web page you are browsing.
Of course the JavaScript code could also sniff out all your credit card numbers and send them to the company, but I had a look and this version does not appear to do that....
Have a look at Magnatune - http://www.magnatune.com/ Doesn't do everything you want (yet) but They Are Not Evil.
That's why the Etymotic ER-4P Earphones were invented. Granted, you need to re-mortgage your house to afford them, but you can't beat their 35-40 dB noise reduction that allows you to hear you music at normal volumes while you are out and about. And it sounds great: the response of these earphones is great.
But otherwise I agree. What is the point of anything other than low-quality mp3 if you are not going to be able to listen to it anyhow.
Email takes way too much valuable time. See this post from Marc Eisenstadt who collected 8 years of personal email data.
1 /eight_years_of_email_stats_pass_1.php
http://www.corante.com/getreal/archives/2005/02/1
He doesn't get much mail, but it still adds up to 2.5 hours per day, assuming you are very disciplined about it. And we all know that we are not.
That's the median you are thinking off, not the average. But your general point is probably valid: If intelligence in the general population is normal distributed and /. has an average that is below that of the general population, you would expect more than half of the readers to be below the /. average (because there is more range available above the average so those people have a higher weight in the average).
Thought you wanted to know....
This announcement brought to you by another average /. reader
> So who do I pay to get a namespace registered? ICANN? Verisign?
:
National Information Standards Organization (NISO)
http://info-uri.niso.org/info-uri-policy
(Section 4 of the document)
Writing a spec is the easy part (and this one seems particularly trivial). Implementing it is a lot harder.
The main missing information seems to be a DNS equivalent function. It is one thing to introduce yet another central registrar to insure "against hostile usurpation or inappropriate usage of registered service marks" (groan!) but how are we going to access the information? Section 4.2 says
and it is clear from section 6 (Normalization and Comparison of "info" URIs) that any sensible implementation would need access, but how?
The information they want to return is much more complex than what DNS returns and DNS is not a trivial infrastructure.
Suggestions? Volunteers? :-)
Amazon.co.uk has it on pre-order for 338.23 which currently is US $561.66 (excluding exchange costs).
Obligatory affiliate link to item here.
Ehm, but that's exactly the point! That's why the Amazon Web Services (which is what we are discussing here) is so useful: by giving you an API to the Amazon database and shopping functionality, it allows you to make the ads in the way that works best on your site, and to maintain them automatically.
You did read the ar.... nevermind. Slashdot. I forgot :-)
Oh, they manage. They financial statements show that they spend just over $1.1bn per quarter on R&D.
Because people will buy it regardless? The effort is not primarily to make their "stuff" better, but to develop new stuff - think X-box, DRM, etc.
The core of Verisign's defence seems to be that domain names are not property rights. From the BBC article on the same ruling:
Two questions: what excatly am I buying when I buy a domain name from Verisign and why do "legal experts" think they'll loose that battle -- presumably they have an extensive user agreement that clears them of responsibility for all and any wrongdoings?
Confused.
(Offtopic, but this is a common fallacy and this is news for nerds...)
Actually, the Earth radiates slightly more energy than it receives from the Sun. If it didn't radiate all the energy, it would get very hot, very quickly. (The excess ratiation os from nuclear decays in the Earth's crust.)
The difference is that the Sun is hot and the Earth is (comparatively) cold. The difference is in the quality of the energy - what physicists call entropy.
Think of it this way: With the energy from a hot plate I can boil water while with the energy from a plate that is only warm I cannot, no matter how long I leave it turned on (i.e. no matter how much energy I spend).
We live off the entropy from the Sun, not the energy.
It appears to be valid XHTML and valid use of CSS, so your comments seem strange, unless they serve different documents for different user agents. Have anybody tested this?
Not to mention that you have to pay if you are using a multiple-CPU system:
I guess that rules me out, then... :-P
"you may not make copies of Software, other than a single copy of Software for archival purposes."
From Binary Code License Agreement.
You can't boot from a USB device, can you? Can the BIOS keep up with all these new devices? About the only time I use the floppy is to boot a reduced Linux system to try an recover a (usually Windows) machine that has goot itself in a twist. Very handy. I guess I could make a bootable CD. And a bootable USB device. And hope that the BIOS allows me to boot from either. I guess it had to go, but I'd like to see a universally available replacement first, supported by every (new, I guess) BIOS. Maybe it is all a plot to make sure we only boot pre-installed operating systems? :-)
Probably a bad example: Unlike most other European countries, the United Kingdom does not have and provisions in law that gives you a right to privacy. If you are in a public place, then you are in public, and I can take your photo and publish it to my heart's content.
There is a code that the newspapers tend to follw which says that you shouldn't publish pictures of people taken with "very" long telephoto lenses without their consent, but that is just a code of practice, not law.
All of this is likely to change at the European convention on human rights -- which does have a provision guaranteeing some privacy -- is incorporated into British law.
Always remember that (1) not all the world is like the US and (2) if you take any advice given on /., in particular legal advice, serious, then you deserve everything you get...
Nevermind that it is text, the important point is that is is a tree structure.
Standard databases are relational, and are great at storing simple attributes for an object. They are absolutely horrible at storing relationships between these objects and, more importantly, in managing those relationships.
So, for example, if you have a grommet that can consist of multiple other grommets, each of couse consisting of grommets etc., then in XML you are laughing:
In a standard relational database you end up with a grommet table and, perhaps, an attribute that is the parent grommet. To get the list I just suggested above, you need to do a self-join on the grommet table an unknown number of times, something SQL just can't do.
Object-oriented databases are good at this (and much more), and it is funny that the old style of databases that preceeded the relational databases, were often hierachical, i.e. tree structures!
So the scoop is this: the trees are back.
Sheesh:
Easy :-)
For general reference, the HTML4 LINK tags are defined here
You can add your own, but if you do, you should use a profile statement. See the Dublin Core for the usual example.
I love Poland but is it really essential to fix the Polish language bugs for a 1.0 release? Aren't there more important priorities? Isn't 1.0 about a stable API (and product!) and such, and if so, couldn't fixing spelling mistakes in the Polish language pack wait until 1.0.1 or something?
The document outlines some really good principles for managing software, but this entry confuses it for me. Any Polish people here to explain why it is critical? :-)
Given the size of the dependency tree for the 1.0 milestone target it looks like 1.0 could be a little way off??
Does anybody want to take a stab at a date? Does anyboy even want to count the number of bugs on that page? ;-)
Actually, Emacs mixes data and code in the same way. Check the File Variables section in the info system, and in particular the enable-local-eval variable. Basically, you can set buffer local variables by embedding the commands for this at the end of the file. One of these variables is 'eval' :-). Thus spake RMS:
In this sense Emacs is just as guilty as Microsoft Office. Just because it's Free doesn't mean it is without security free. (But the fact that the average person using Emacs is more clued in than you Power Point suit, does help...)