This is what requirements capture is about. At the point where the guys asks for "oh, just a house, you know" then you pull out your social skills and talk to him. Perhaps offer him a sketch or two of a strawman house. Draw out his needs, make him feel that the req cap process isn't overly burdensome by being the smart person you surely are - you need his specific input, but getting there is a partnership.
This is like any other skill. You learn it with practice. Some people are naturally better at it than others, but it's possible to develop this skill if you apply yourself. And skilled, intelligent people who are interested in house building (which is where the analogy came from, right?) are perfectly capable of lerning this too. You don't just lump it when the client presents the first hurdle and do something that'll cost you and him more grief in the long run, because - hell, you're a professional. You build a professional relationship with the guy and he'll come back for repeat business, tell all his friends and associates what a good job you did, and everyone'll be happy.
Meant to add: "Content Addressable Storage" does NOT mean you can look up objects by their metadata attributes, normally. That kind of functionality is usually a part of the application that uses CAS and sits on top of it (although IIRC the Celerra which I mentioned earlier does have some kind of metadata search facility).
CAS products typically identify objects by a hash of the contents. They're usually accessed via an API: you give it the stuff to store (and, optionally, a retention policy) and you get an opaque ID in retrun. You can get the content back if you have the ID, so IDs normally are capabilities (in the authorisation sense).
These things often sit behind document or other content-management systems. The advantage is that you only store one copy of the data; the disadvantage? Well, mostly, it's just a different way to store data. You wouldn't "use" this directly as a filesystem on a desktop. It might, however, form a part of the storage layer for an HSM system. So you need to understand what it is before you realise how you might use it.
Various companies have CAS systems (it's a fairly new storage buzzword - think EMC's is called "Celerra", for instance) and you'll find whitepapers explaining what these are. Don't expect to be amazed:-)
This is a different problem area and principally aimed at unstructured or semistructured data.
JSR170 _is_ a pretty baked model, however. The major thing it lacks is the notion of a prepared statement (a la JDBC) - which is ironic, considering (as you say) the cast list on the committee.
The problem with the lack of parameterised prepared statements is that the quoting rules in the query languages are complex to explain, and easy for a developer to screw up (vide any PHP app:-) ). So we're likely to see the widespread adoption of this API in the java world (it's certainly about time) coupled with a whole slew of injection attacks against JSR170-based web apps, alas.
Don't get me wrong: I think the JSR is a very good piece of work. However, despite the pressing time constraints, I really feel that this is an issue that should be corrected before this spec hits rec.
You're trolling, right? JavaScript is a great language, and far more powerful* than you might imagine. Have a look at Dave Crockford's javascript stuff. Alternatively you might be interested in knowing that Javascript also supports (for some values of) goal-directed logic programming as a paradigm, too: js prolog
* if you're about to say anything using the words "Universal" and "Turing": yeah, right, but that's not what I'm talking about and you damn well know it.
It seems disingenuous to quote half of what the porting team say. The preceding sentence:
"The JDK 1.4.2 patchset has reached a point where it is suitable for almost all uses."
echoes my experience and that of regular reporters on the freebsd java mailing list. YMMV: what problems have you experienced recently running Java 1.4.2 on FreeBSD? (The list would like to hear, or you could simply raise a PR)
Re:Am I the only one on here who likes Netbeans?
on
Netbeans 4.1 Released
·
· Score: 1
We shifted to Eclipse in-house a while ago, when it was definitely better than Netbeans. Since then NB has gotten a lot better, and I occasionally re-evaluate it, but it would have to be _significantly_ slicker than eclipse (not "just as good as") to warrant the shift from Eclipse for our development team.
That's Eclipse plus a load of plugins, I should add: the JBossIDE, Omondo's UML stuff, and some other bits and pieces. That the plugins (especially for J2EE) need to be tracked down doesn't matter so much - since we've got them already - although the more integrated out-of-the-box nature of NB might be an advantage if you're selecting an IDE for the first time.
I still prefer IDEA, it has to be said, but use Eclipse to maintain IDE compatibility with the devs I work with. This is more than just the ability to sit down at someone's desk and help them out - project layout, build strategies, etc, vary between IDEs. After selecting and getting used to an IDE one of the major stumbling blocks with shifting is the subtle difference between it and the thing you're thinking about moving to. So we keep an eye open, but will probably be using Eclipse for some time. But we're not blinkered about it.
The generation of a headless Ant file sounds really handy. It _is_ possible to invoke Eclipse in a headless mode to drive a build; but I must confess that I don't do things that way here*. We tend to have our projects in an eclipse-friendly layout but maintain a build script to automate deployment more-or-less by hand. (Eclipse doesn't particularly get in the way of this.)
* I tried it, I liked it, but the project we were thinking about using it for had well-established Ant build scripts already and migrating would've been a pain.
It's also curious what this "anywhere" really means. Last I checked, FreeBSD still did not have a stable version of Java 1.4, let alone Java 5.0
Check again. I've been running the native 1.4 JSDK on FreeBSD for ages, and have found it to be rock-solid for all that time. (EJB/J2EE development and deployment, using Eclipse (as it happens) to remain compatible with the other developers on the team - although I have to say I prefer IDEA)
The Java 5 port's currently in alpha on that platform.
"Only deal with food distributors that embrace healthy eating."
It's nice in principle, but unfortunately in the UK the cat's already out of the bag due to something called the PFI (Private Finance Initiative). This is basically a scheme where fatcat companies (like Jarvis) get ludicrous contracts to supply "public" services. Many schools, post the Jamie Oliver wake-up call, discovered that they were stuck with the cheap and nasty greasy lardburger-serving PFI suppliers foisted upon them by local councils.
Unfortunately, these contracts have usually been negotiated with whopping penalty clauses should the council decide to pull out of the scheme. This means that councils and schools cannot _afford_ to do the right thing.
Great if you're on the board of Jarvis; not so great if you're anyone who actually cares aobut the state of public services: the privatisation of these services was intended to cost-cut and promote competition. Exactly the opposite occurred: every privatisation thus far has been botched. I'm not claiming that public services are paragons of efficiency - however, I've never heard a coherent explanation of why a private company could be able to provide the same level of service for less money AND make a return for its shareholders. Something doesn't compute.
As someone with a colleague who works with a RH satellite server, I'd urge you to not just believe their press. The box turned up misconfigured; the support service was useless and my cow-orker had to figure it out himself. Not a good start for a site shelling out 25k (GBP) for apparently top-line support.
Then recently we had another support incident over an apparently "known issue". RH support give a list of instructions on the phone. "How do I back this out?" Well, you don't need to, done this a thousand times, it's a doddle. Hey presto: one f*cked satellite server. And the backout process they recommend? "Restore from backups."
We're less than impressed. If you're like us (25k users, about 10k desktops) I'd recommend rolling your own with debian. Since you wind up rolling your own with RH anyway:-/
That's what I was wondering about. You can minimise the modification of shared library pages, I suppose, by (a) using jump table pages for relocation; (b) by having "preferred" locations that libraries can hint they like to be loaded at. Or (c), which is stick all your functionality into one executable and set of libraries, relocate once, and fork-and-call rather than fork-and-exec for child processes.
A CA can _always_ impersonate you, whether or not they have a copy of your certificate. How? Well, they just create a new one that claims to be yours and sign it.
Incidentally something along these lines hit the news recently: more conflict of interest with Verisign, who as well as holding root certificates AND managing a large swathe of the DNS have just gotten into the interception business. Since they have all the pieces to grab "secure" traffic to sites in.com, this is obviously cause for concern.
Google can find you an explanation (I believe Bruce Schneier's newsletter contained pointers to this recently).
This is off-topic, but it's not intended as flaimbait. I'm curious where you stand on the issue of "the woman's right to choose". Because your argument rings a bell.
I would broadly (and probably inaccurately, please correct me) characterise your defence of the GPL as a defence of the "pro-life" stance: the right to choose is better at preserving a woman's freedom to take away the freedoms of her (as yet merely potential) descendants, not to mention the rights of the father and other family. In other words, the needs of the many, etc.
..? What you "know" is wrong. Ethanol doesn't have a left- and right- isomer. Ethanol is CH3 CH2 OH. There's only one way to make it. There's another compound with the formula C2H6O, and that's dimethyl ether: CH3 O CH3.
However, you're right in general that some biological processes only produce one particular stereoisomer.
Professional development is one of the leading reasons for open source contribution cited by a large number of developers. You learn by working with hackers who are better than you. In other words, there already _is_ an "apprentice system".
You don't "just" have to get another hard disk:-) You also need to be able to store and manipulate arbitrarily large values just to figure out which disk you're addressing. See the infinite monkeys RFC for an example.
The statement's derived (I'd expect) from a utilitarian viewpoint: moral value can only be derived by moral creatures. The moral value (or utility) of anything else is derived solely from its relationship or perceived worth as viewed by moral creatures.
Arguments along these lines are popular with modern ecologists (eg, Monbiot, who isn't particularly eloquent when it comes to expressing them). For instance: that biodiversity and the preservation of species should be considered important not just because of the (spurious) arguments like "but god knows how many cancer cures are being lost with every acre of Amazonian rainforest" - but instead, because of the aesthetic appeal to humans. It makes thinking, feeling people sad when species are lost. That's sufficient reason to do something about it.
Or so the argument goes. I personally find it persuasive; YMMV.
That last claim's an interesting one. D'you have a pointer to the relevant papers?
Although this: "At that point finding two of these strings with the same (other hash) is no harder than finding any two strings with the same (other hash)" isn't much of an improvement; in fact, I'd claim it's actually wrong. Because there may well be (intelligent) attacks to find two strings that MD4 the same; whereas the approach you outline relies on effectively discovering such a hit using a brute force attack.
Slashdot Mirror
This is what requirements capture is about. At the point where the guys asks for "oh, just a house, you know" then you pull out your social skills and talk to him. Perhaps offer him a sketch or two of a strawman house. Draw out his needs, make him feel that the req cap process isn't overly burdensome by being the smart person you surely are - you need his specific input, but getting there is a partnership.
This is like any other skill. You learn it with practice. Some people are naturally better at it than others, but it's possible to develop this skill if you apply yourself. And skilled, intelligent people who are interested in house building (which is where the analogy came from, right?) are perfectly capable of lerning this too. You don't just lump it when the client presents the first hurdle and do something that'll cost you and him more grief in the long run, because - hell, you're a professional. You build a professional relationship with the guy and he'll come back for repeat business, tell all his friends and associates what a good job you did, and everyone'll be happy.
Right?
Meant to add: "Content Addressable Storage" does NOT mean you can look up objects by their metadata attributes, normally. That kind of functionality is usually a part of the application that uses CAS and sits on top of it (although IIRC the Celerra which I mentioned earlier does have some kind of metadata search facility).
CAS products typically identify objects by a hash of the contents. They're usually accessed via an API: you give it the stuff to store (and, optionally, a retention policy) and you get an opaque ID in retrun. You can get the content back if you have the ID, so IDs normally are capabilities (in the authorisation sense).
:-)
These things often sit behind document or other content-management systems. The advantage is that you only store one copy of the data; the disadvantage? Well, mostly, it's just a different way to store data. You wouldn't "use" this directly as a filesystem on a desktop. It might, however, form a part of the storage layer for an HSM system. So you need to understand what it is before you realise how you might use it.
Various companies have CAS systems (it's a fairly new storage buzzword - think EMC's is called "Celerra", for instance) and you'll find whitepapers explaining what these are. Don't expect to be amazed
This is a different problem area and principally aimed at unstructured or semistructured data.
:-) ). So we're likely to see the widespread adoption of this API in the java world (it's certainly about time) coupled with a whole slew of injection attacks against JSR170-based web apps, alas.
JSR170 _is_ a pretty baked model, however. The major thing it lacks is the notion of a prepared statement (a la JDBC) - which is ironic, considering (as you say) the cast list on the committee.
The problem with the lack of parameterised prepared statements is that the quoting rules in the query languages are complex to explain, and easy for a developer to screw up (vide any PHP app
Don't get me wrong: I think the JSR is a very good piece of work. However, despite the pressing time constraints, I really feel that this is an issue that should be corrected before this spec hits rec.
You're trolling, right? JavaScript is a great language, and far more powerful* than you might imagine. Have a look at Dave Crockford's javascript stuff. Alternatively you might be interested in knowing that Javascript also supports (for some values of) goal-directed logic programming as a paradigm, too: js prolog
* if you're about to say anything using the words "Universal" and "Turing": yeah, right, but that's not what I'm talking about and you damn well know it.
It seems disingenuous to quote half of what the porting team say. The preceding sentence:
"The JDK 1.4.2 patchset has reached a point where it is suitable for almost all uses."
echoes my experience and that of regular reporters on the freebsd java mailing list. YMMV: what problems have you experienced recently running Java 1.4.2 on FreeBSD? (The list would like to hear, or you could simply raise a PR)
We shifted to Eclipse in-house a while ago, when it was definitely better than Netbeans. Since then NB has gotten a lot better, and I occasionally re-evaluate it, but it would have to be _significantly_ slicker than eclipse (not "just as good as") to warrant the shift from Eclipse for our development team.
That's Eclipse plus a load of plugins, I should add: the JBossIDE, Omondo's UML stuff, and some other bits and pieces. That the plugins (especially for J2EE) need to be tracked down doesn't matter so much - since we've got them already - although the more integrated out-of-the-box nature of NB might be an advantage if you're selecting an IDE for the first time.
I still prefer IDEA, it has to be said, but use Eclipse to maintain IDE compatibility with the devs I work with. This is more than just the ability to sit down at someone's desk and help them out - project layout, build strategies, etc, vary between IDEs. After selecting and getting used to an IDE one of the major stumbling blocks with shifting is the subtle difference between it and the thing you're thinking about moving to. So we keep an eye open, but will probably be using Eclipse for some time. But we're not blinkered about it.
The generation of a headless Ant file sounds really handy. It _is_ possible to invoke Eclipse in a headless mode to drive a build; but I must confess that I don't do things that way here*. We tend to have our projects in an eclipse-friendly layout but maintain a build script to automate deployment more-or-less by hand. (Eclipse doesn't particularly get in the way of this.)
* I tried it, I liked it, but the project we were thinking about using it for had well-established Ant build scripts already and migrating would've been a pain.
Check again. I've been running the native 1.4 JSDK on FreeBSD for ages, and have found it to be rock-solid for all that time. (EJB/J2EE development and deployment, using Eclipse (as it happens) to remain compatible with the other developers on the team - although I have to say I prefer IDEA)
The Java 5 port's currently in alpha on that platform.
"Only deal with food distributors that embrace healthy eating."
It's nice in principle, but unfortunately in the UK the cat's already out of the bag due to something called the PFI (Private Finance Initiative). This is basically a scheme where fatcat companies (like Jarvis) get ludicrous contracts to supply "public" services. Many schools, post the Jamie Oliver wake-up call, discovered that they were stuck with the cheap and nasty greasy lardburger-serving PFI suppliers foisted upon them by local councils.
Unfortunately, these contracts have usually been negotiated with whopping penalty clauses should the council decide to pull out of the scheme. This means that councils and schools cannot _afford_ to do the right thing.
Great if you're on the board of Jarvis; not so great if you're anyone who actually cares aobut the state of public services: the privatisation of these services was intended to cost-cut and promote competition. Exactly the opposite occurred: every privatisation thus far has been botched. I'm not claiming that public services are paragons of efficiency - however, I've never heard a coherent explanation of why a private company could be able to provide the same level of service for less money AND make a return for its shareholders. Something doesn't compute.
As someone with a colleague who works with a RH satellite server, I'd urge you to not just believe their press. The box turned up misconfigured; the support service was useless and my cow-orker had to figure it out himself. Not a good start for a site shelling out 25k (GBP) for apparently top-line support.
:-/
Then recently we had another support incident over an apparently "known issue". RH support give a list of instructions on the phone. "How do I back this out?" Well, you don't need to, done this a thousand times, it's a doddle. Hey presto: one f*cked satellite server. And the backout process they recommend? "Restore from backups."
We're less than impressed. If you're like us (25k users, about 10k desktops) I'd recommend rolling your own with debian. Since you wind up rolling your own with RH anyway
Perhaps. I'm a scot who's proud of being British, however. Are you familiar with the period of history surrounding the Act of Union?
You can imagine it: "Aleph Zero was a great hit! Let's make the prequel!"
That's what I was wondering about. You can minimise the modification of shared library pages, I suppose, by (a) using jump table pages for relocation; (b) by having "preferred" locations that libraries can hint they like to be loaded at. Or (c), which is stick all your functionality into one executable and set of libraries, relocate once, and fork-and-call rather than fork-and-exec for child processes.
That sounds great. Genuinely curious: how does the interaction with shared libraries affect this picture?
A CA can _always_ impersonate you, whether or not they have a copy of your certificate. How? Well, they just create a new one that claims to be yours and sign it.
.com, this is obviously cause for concern.
Incidentally something along these lines hit the news recently: more conflict of interest with Verisign, who as well as holding root certificates AND managing a large swathe of the DNS have just gotten into the interception business. Since they have all the pieces to grab "secure" traffic to sites in
Google can find you an explanation (I believe Bruce Schneier's newsletter contained pointers to this recently).
Well, it offers a simple rebuttal to Microsoft's whinging about software patents in Europe.
"Protecting development, blah blah, investment, blah blah, we'll take our ball away, whinge..."
"IsNot, Bill."
That's not streamable. However, this:
b=MD5(a+s)
is the basis of HMAC - shared secret key used to salt the hash.
This is off-topic, but it's not intended as flaimbait. I'm curious where you stand on the issue of "the woman's right to choose". Because your argument rings a bell.
I would broadly (and probably inaccurately, please correct me) characterise your defence of the GPL as a defence of the "pro-life" stance: the right to choose is better at preserving a woman's freedom to take away the freedoms of her (as yet merely potential) descendants, not to mention the rights of the father and other family. In other words, the needs of the many, etc.
..? What you "know" is wrong. Ethanol doesn't have a left- and right- isomer. Ethanol is CH3 CH2 OH. There's only one way to make it. There's another compound with the formula C2H6O, and that's dimethyl ether: CH3 O CH3.
However, you're right in general that some biological processes only produce one particular stereoisomer.
But no, methane is just CH4. One version only.
Which is what you're doing if you continue to drive a car. Good point, well made.
Professional development is one of the leading reasons for open source contribution cited by a large number of developers. You learn by working with hackers who are better than you. In other words, there already _is_ an "apprentice system".
You don't "just" have to get another hard disk :-) You also need to be able to store and manipulate arbitrarily large values just to figure out which disk you're addressing. See the infinite monkeys RFC for an example.
The statement's derived (I'd expect) from a utilitarian viewpoint: moral value can only be derived by moral creatures. The moral value (or utility) of anything else is derived solely from its relationship or perceived worth as viewed by moral creatures.
Arguments along these lines are popular with modern ecologists (eg, Monbiot, who isn't particularly eloquent when it comes to expressing them). For instance: that biodiversity and the preservation of species should be considered important not just because of the (spurious) arguments like "but god knows how many cancer cures are being lost with every acre of Amazonian rainforest" - but instead, because of the aesthetic appeal to humans. It makes thinking, feeling people sad when species are lost. That's sufficient reason to do something about it.
Or so the argument goes. I personally find it persuasive; YMMV.
That last claim's an interesting one. D'you have a pointer to the relevant papers?
Although this: "At that point finding two of these strings with the same (other hash) is no harder than finding any two strings with the same (other hash)" isn't much of an improvement; in fact, I'd claim it's actually wrong. Because there may well be (intelligent) attacks to find two strings that MD4 the same; whereas the approach you outline relies on effectively discovering such a hit using a brute force attack.