Good idea, maybe, but there are some implementation problems. The advertiser has an obvious incentive to claim low sales from such ads. You don't have a good way of verifying their claims unless you can have people on their site watching every stage of their sales process. Also, it could easily lead to advertisers demanding that large numbers of (slightly) different ads be run, because if an ad doesn't work, they don't have to pay for it. You'd be doing their marketing research for them for free.
How could you keep your advertisers honest if you used such a policy?
Google's got it right. Small. Innocuous. Relevant to what you're looking for.
Yup. Like most of the geeks here, I mostly use browsers that can do things like block images from sites, so as to cut down on the more obnoxious ads. But I've also bought a fair number of things online over the years. And when I'm looking to buy something, I tend to first ask google about it. Both the matches and the accompanying ads are useful in that case.
Dunno how well it works with the general population, but google's approach is fairly good for people who are trying to find something and just get annoyed by irrelevant ads.
We oughta let them know that we appreciate their subtler approach to the whole topic.
In a few cases, commercial sites have asked me how I found them, and I've enjoyed telling them that I used google. That oughta give some of their marketing people a bit of a pause, since they probably "know" that google's approach isn't very successful at selling.
Hey, why don't you get in touch with our hero Mike Rowe, and offer him a top position. Then you can sell "MikeRoweSoft Windows". Or maybe make them with a soft vinyl frame, so they'le be "MikeRowe Soft Windows". Since the product will be real windows for buildings, your lawyers will be able to argue reasonably that they aren't competing products. But Microsoft will sue you anyway....
Instead, why don't *you* just design your webpages to W3C standards and be done with it?
Uh, which W3C standard(s) should I follow?
Have you seen how many standards they've published? Do you honestly think any merely-human brain could even start to hold all that?
Meanwhile, I'm out here doing' my best. I do feed pages to various online validators quite often. Sometimes I can even make sense of what they tell me, and I fix the problems. Sometimes I'm just baffled at what they want me to do. But in those cases, I just reason that they're obviously talking to someone smarter than I am, so I don't worry about it too much.
Even if you try hard to implement to W3C standards, any sensible web developer would still test against a list of browsers. The actual screen representation can vary widely, even for software implementiing the same standards. Lately I've been testing against a number of PDA browsers to make sure that users of Tungstens and Blackberries can use my stuff. Those devices do have "real" browsers now, but they are severely constrained by their small screen. Just saying that your HTML is standards compliant isn't enough if some of your important clients are using such small display devices.
Now if there were a way to test against all of them without setting up a million-$-per-year testing lab. That's the real problem. You can easily have a page that passes all known HTML validators' tests, but still comes up unreadable (or very confusing) on some clients' screens, even though they are using a standards-compliant browser.
Lots of us really wish the W3C would ban frames...
Heh. Actually, it's those starving artists who more and more are threatened by the "long after you're in your grave" copyrights.
It used to be that, to become a good musician, one of the important things was to copy the past masters. The first recordings enhanced this, by giving people access to actual performances of musicians that you'd never have heard before.
But now, we're seeing the start of copyright-infringement suits based on musicians copying old recordings. It sorta started with the samplers, but now its looking to effect everyone. Say that as a kid, you loved your parents' Satchmo and Dizzy recordings, and when you got your axe, you wore out the records learning to do those riffs. You'd better be careful. Someday soon, you'll be playing with a pickup band at a local club, you'll take several solos, and the next day you'll be hit with a lawsuit for unauthorized performance of a 1923 recording that's you don't even remember, but is there in your subconscious influencing your style.
A few years back, there was a sci-fi short story based on the idea that, with permanent copyrights finally enacted, the world had reached the point that all possible musical motifs had been copyrighted, and it was no longer legal to write new music. Any "performance", such as whistling a tune while walking down the street, would produce a rapid arrest and imprisonment, unless you had had the foresight to get written permission to perform that work at that time and place.
And note that almost all copyrights are now owned by corporations. In a few more years, the two remaining Beatles will die, and their music will fall into the hands of one or more corporations. It's the starving artists that will lose really big by this process, as they lose the right to produce without first getting permission from the corporations that own the bits and pieces of any new creation. The winners, as in the old feudal system of land ownership, will be the aristocracy that owns the shares of the IP corporations.
In another generation or three,, it may not be possible to produce any new art without violating a century-old copyright. But, by then we may be finally escaping our planet, and new land will be available for settling. But that probably won't be available to the poor, either.
... a small (iPAQ or BlackBerry or Tungsten sized) gadget that runs linux and can do IP over 802.11*, GSM and GPRS, I'd grab it and start developing software. Imagine what we could build if something like this were available to those of us who just want to do it...
(Actually, all that in a laptop would be a good start. But so far, I haven't learned of a laptop with hardware for all three wireless comm systems. And there's the ongoing problem of publishing the proprietary specs so that 802.11* hardware works seamlessly on a linux laptop. But maybe we'll soon read that WiFi "almost works" on linux, just like it "almost works" on OSX.;-)
Anyway, in our house, we have a Tungsten C (PasmOS + WiFi) and a BlackBerry 7280 (java VM + IP over GSM and GPRS). We keep thinking how nice it could be to have one pocket/purse-sized toy with all three and the programmer accessability of linux. If someone were to put one together and find a way to make it legal to use all three paths to the Net, we could start a real explosion of Open-Source, standards-based portable computer/comm gadgets.
Heh. I've read any number of comments on this phenomenon. They usually include the observation that someone with true command of a language will find a way to express their ideas as simply and clearly as possible (but no more so, as Einstein pointed out).
The "style" metrics seem to invariably give a higher score for complexity of syntax and for longer words. This isn't necessarily an indication of sophisticated writing, as a few of the big-word parodies here have clearly shown.
[N.B. The above split infinitive was intentional.;-]
Part of Wodehouse's command of the English language was knowing how to express things simply and clearly. I'd guess that he would have sneered at suggestions that there was something uneducated about this.
Huh? Why would anyone bother to get a thesaurus, for themselves or for a friend? They're all online these days, and using them is faster than walking over to the bookcase and pulling down the hard copy.
Of course, if you're out of range of an Access Point, you might have a problem. But then, if you use one regularly, you've already downloaded it to your PB, right? Right?
(Actually, there's one within reach of my usual work chair, but how useful is a thesaurus when you work as a programmer?;-)
2) If I were implementing the metric, any text using "boxen" would be downgraded to "Idiocy".
And, of course, anyone reading your ratings would downgrade you to "humo(u)r impaired".
Now, granted, "boxen" is a rather old bit of wordplay that's not nearly as funny as when it was new. But it's still good for ferreting out the people who don't have anything more important to complain about. So we can expect that it will continue to appear here, until it no longer gets any comment from bored readers.
I wonder if there are any language metrics that successfully take into account things like geek wordplay humor? That's gotta be something that's difficult to measure.
Cars tend to have much higher resale values than computers and also tend to have a much longer average life.
True for resale value. But the linux crowd has profited enormously from the fact that even crappy Intel hardware often lasts a decade or more. It's the forced software upgrades that kill Windows machines while the hardware is still good. So the suckers who are forced to buy new hardware that will run the Windows upgrades are selling their "obsolete" hardware cheap, and it runs linux just fine. Carry your knoppix CD along to garage sales, and you can get yourself a nice machine for next to nothing.
In particular, there are lots of old Intel boxes out there running the firewalls and gateways for most of the Internet. There was even the story here last year about the "super-computer" that was a beowulf cluster of mostly discarded Intel boxes that had been bought for next to nothing. (Of course, the high-speed comm equipment that tied them together wasn't cheap.;-)
Mac. And no one spends $2k+ on an Intel based PC, right? I mean, certainly no one here...
Well, I have a number of friends that have recently bought $2k+ IBM Thinkpads, which seem to all be Intel based. Of course, they immediately wiped the Windows OS and replaced it with linux. And I get the impression that there are a few linux users here.
One problem with Macs, of course, is that they're designed for people who haven't learned to move their fingers independently. My wife got a Powerbook some time back, and while she likes it a lot better than her old Windows box (and gave that to me so I could have another linux box), she does sometimes wax nostalgic about the 16-button mouse on the CAD workstations that she used to work with.
Myself, I'd be happy with 3 buttons, so that cut and paste will be as fast as it is on any unix/linux machine. I think that's one of the advantages of Thinkpads. And after a few months, it could well save $2000 as measured by our billing times.
There's the ongoing question of why IBM isn't selling Thinkpads with linux installed. With all their touted linux support, how are they missing this market opportunity?
As I am sure you know we have a similar paradox in sw consulting. The better you get the less time things take and the less you make.
Yup. Case in point: I remember my first encounter with writing a device driver, years ago. A new gadget was ordered, but there was no driver. I sent a message off to the vendor's Support people, and they sent me the hardware specs. We had source for the OS, so I grabbed a driver for a similar gadget, cloned it, and modified all the magic numbers to agree with the specs for the new gadget. It took maybe an hour at most.
A few days later, the new gadget arrived. I plugged it in, ran a couple of tests, and it worked. A Customer Support guy happened to be present, and he was astounded. "We don't have a driver for that." So I said "Well, you do now." And I explained that they had done a good job in the specs, so all I had to do was what I had done, and it worked.
Now, this isn't too amazing; it's how things should work. The interesting aspect to this is that I've gone to a number of interviews at companies that are, among other things, looking for people to write device drivers. I've never gotten such a job. The reason is always the same. The interviewer asks about driver experience, and I mention having produced a number of them on different projects over the years. But it's never good enough. They always say that they want someone with several years experience writing drivers. I can only honestly admit to a few hours experience. After all, if you have the detailed specs, it only takes an hour or so; if not, you can't do it no matter how much time you have.
When they bring up the "several years experience" requirement, I've several times asked if this means they really want someone who takes months to write a driver rather than hours. Inevitably, they respond to this as if I'm some sort of wise ass. A couple have told me to my face that, yes, this is what they want.
It doesn't surprise me that we have a lot of problems getting drivers written for new hardware, given that this appears to be the "normal" attitude toward hiring people to write the drivers.
Maybe I should be a bit less honest. After all, I could easily stretch the job out to months. But really, I'd much prefer to spend the time on difficult problems. (Something that isn't doable because I can't get the specs isn't what I'd call "difficult"; it's "impossible", and that's not the same thing at all.;-)
So, in order to survive I guess you have to make shitty sw...
Funny, perhaps, but with a core of truth. I recall a year or so back, when someone here got a high Funny rating for suggesting that consulting firms would be stupid to recommend linux, which would be just a few hours of billing time to install and configure, when they could recommend Windows, which would get them continuous billing time for support.
I showed this to several friends who worked at companies that did business consulting, and they all reacted the same way. "That's not funny; that's exactly how it works." Several said that their own in-house systems were all linux or BSD, but they'd be stupid to recommend them for clients.
I've also had occasion to discuss with employers the idea that I shouldn't produce software that "just works", because then they'll never have to hire me again. They all got worried looks on their faces as they agreed that I was right. But they didn't know what to do about it, considering their own restrictions on how company hiring was done.
If we want true Software Engineering, we need to find a way to end this conundrum. Currently we're paying people to produce software that doesn't last. We do this by the simple approach of paying them to work on software that has problems, and not paying anyone for software that doesn't have problems. Right now, I don't seem to be reading of anyone working on this. Complaining, yes. But how many people realize that when you hire people, they usually try to do things that you pay them to do?
I'll predict that we will eventually discover that we do have 200-year-old software. And it was all created by people who were writing it for their own use, not because someone was paying them to write it. This is why most Open Source software was created, after all. And it was how unix arose at Bell Labs.
I have a number of 20-year-old programs that I keep around because they're useful...
You can't worry about your software working for that long until your hardware can last that long.
Oh, nonsense. Consider the well-known "Hello, world." program in the K&R C "bible". It's been around 30 years or so, and the hardware they were working on now only exists in a few museums. But that program is still in routine use on millions off computers.
Note also that K&R included not only the code for the program, but the commands to compile and execute it. They correctly pointed out that this information covered the initial implementation details in most software, and once you had it working on a new system, you had soved the major problems in getting any software running on the new system.
And note that, after more than a quarter century, their sample code to compile and execute this valuable little program still work on millions of computers, without changing a single character.
Funny thing is, you can't say this about the equivalent programs in most other languages. I've seen attempts to do the same (code + compile if needed + invocation) for any number of other languages and OSs, but I've seen very few successes. Usually code that's only 10 years old will need tweaks. But if you have a standard-compliant (ANSI) compiler and a standard-compliant (POSIC) OS, you can type in this standard program using any editor, type in the compile command to any shell, and type in the invocation to that shell, and see the expected greeting appear on whatever "terminal" hardware you are using.
I'll bet that this little program will still be around and will still work 200 years from now. Even that bizarre "a.out" name, though I expect that using "-o hello" in the cc command will also work as it did in 1975. (And the program will still have the same subtle bug as the original did.;-) I wish I could be around then to be proved right or wrong.
But so far, the signs are that this won't be true for many programming languages or OSs. The idea that code should be reliable in the long run seems to be something that's beyond the comprehension of few language or OS implementers.
I'm not saying they're a bad paper, just that we should demand more from the US's supposed #1 paper.
Um, we can't demand anything from them. Only their advertisers and stockholders can do that. They're a for-profit corporation, after all.
If you want a news source that you personally can make demands of, you're stuck with things like NPR and VOA, where you in fact do have some influence (though it's diluted a bit by the other millions with the same influence).
Of course, there is a way to make demands of the NYT and other corporate news sources. Just give them enough money by buying ads or stock that they have to listen to you. But unless you're willing and able to do that, forget about making any demands of them.
Y'know, my first thought was "Google has to pay to index their own archive?" Then it occurred to me that you weren't referring to the google cache, but rather to the NYT archive.
So what about google's cache? I've read a number of NYT articles from there. What problems are there with using it rather than the NYT archive?
Of course, we still haven't seen the end of attempts to use the copyright law to outlaw things like google's cache. It's entirely possible that, in the not-so-distant future, all news will be online. Then the main effect of copyright will be to erase all of our history, as the copyright holders make it all inaccessible or erase it.
We really do need an addendum to the copyright laws stating that anything no longer available from the copyright holder becomes public domain.
Heh. I recall a few years ago, when the place I was working at instituted RSA everywhere, and I found that I didn't need to save copies of any keys. When I needed them, I'd just google for them.
I really wish this were a joke. I repeatedly suggested that they should investigate possible leaks of the keys. I'd do this without telling them what I'd found, because I was afraid that they'd just block that path and consider the problem fixed. I did wonder what other gaping holes were around the place.
Once I even fetched a needed key this way in front of a gang of company bigwigs. They didn't bat an eye; they oviously hadn't a clue as to what I'd just done. The one other techie present looked at me with large, disbelieving eyes...
Later, on the way out, I heard him mutter "I don't believe it!"
In some technical circles, the term "postdict" has been coined for this usage. But it hasn't caught on widely. Probably because this sort of misuse of "predict" is sufficiently obvious that most people understand what you meant.
Scientists routinely "predict" what will be found when observations are made. The fact that these are usually postdictions is true, perhaps, but not sufficiently interesting or important to produce a real push for adopting the more precise term.
If you google for "postdiction" you'll see that the term is in use in some specialties.
... signs like McDonalds used to have: "over 1 billion served!
Well, I recall back when those first started appearing, my thought was "Yeah, but how many were actually eaten?"
Also, I like to mention that I have two linux boxes that were delivered with Windows. So Microsoft certainly lists me as one of their satisfied customers. Obviously, I'm so happy with Windows that I bought two of them. In fact, I eventually erased the Windows partition on both and converted them to/var partitions. (They're much more useful to me as a cage to limit the damage of an overgrown log file than as a box full of shoddy MS spyware.;-)
Just one of many reasons why industry statistics aren't to be taken seriously.
Yeah, by the US Dept of Justice, among others. And, of course, MS won that one.;-)
Anyone have a list of lawsuits in which Microsoft actually lost and paid damages?
This isn't intended as a troll. I don't think I've read of anyone actually getting money from Microsoft via the courts. It might be interesting to read of a few cases.
There have been some cases close to that in which MS settled out of court. For example, their recent settlement with Opera. But Opera didn't actually win the settlement in court.
In particular, has any mom-and-pop company ever sued Microsoft for damages and won in court? Can anyone name the cases? And did the companies actually receive the money?
(That's money as in money, not discounts on purchases of MS products.;-)
Well, considering that a number of Microsoft people had already gone public with the "It's not our problem; it's Mozilla's problem", I'd think that the obvious answer is that Microsoft's management was already very much aware of the problem. Pointing out that MS products have the same vulnerability is an obvious (if somewhat in-your-face) way to shoot down their FUD.
And, let's face it, they were using this as an opportunity to squelch the recent rash of switches from IE to Mozilla. They deserve to be hit fast and hard for such tactics.
(Not that the Mozilla people are totally innocent here. Even if you agree that it's a Windows bug, it's clear now that Mozilla could very easily catch it and pop up a warning window. That would have taken less time than was apparently spent discussing the issue and deciding to not deal with it right away.)
Software is manufacturing? I always thought I was in the service sector. I mean a lawyer doesn't manufacture contracts, does he? He provides a service.
Heh. You may not have seen the news, but the Bush administration has lately been trying to get jobs like fast-food work redefined as "manufacturing".
After all, the person at the grill is taking raw materials (ground-up cattle) and using machinery to convert them to a product (the cooked burger patty). Then someone else in the assembly line combines that with other components to build the delivered product. Sounds like manufacturing, right?
With only a small stretch, we could also reclassify marketing and sales as "manufacturing". After all, they are building public images (sometimes out of whole cloth) and assembling them in the marks' - uh, I mean valued customers' -- minds.
Of course, this is being pushed by politicians who talk of "manufacturing a concensus".
Heh, heh; if I had any mod points, I'd give you a "funny".
But, like most jokes, there's a grain of truth at the core.
I use google a fair amount, but probably a lot less than many non-geeks. I know how to bookmark things and organize my bookmarks. Periodically I move many of my bookmarks to more permanent places in several "Links.html" files. So I don't have to keep looking things up; my computers remember the good ones for me.
And, of course, I do remember a lot of URLs. I've been known to type in search URLs for google and a few other sites, eliminating a download of their main search page.
And another thing that confuses the browser statistics: It's quite common now for browsers to include a selection of ID strings to mimic other browsers. Mostly, this is used to mimic IE, so that you can see a site that only works for IE. There were a few stories recently about opera's facility for this, and the stories explained about the msn.com sites that actively looked for opera and sent damaged pages. Recently I got a BlackBerry 7280, installed its real browser, and one of the first things it did was ask me if I wanted it to pretend to be IE. Guess why they include this feature?
So we should expect that the statistics count lots of other browsers as IE. It takes some rather sophisticated programming to avoid this, and in some cases, you can't determine the true nature of a web client at all. Many of the counters are commercial operations doing the counting as part of a marketing operation, and they aren't well-motivated to try to uncover all the browser ID-string deceptions.
Of course, this use of fake ID strings was started years ago, when the first IE versions told servers that they were Mozilla. They still do this, actually, though they now include "IE" in a later part of the ID string. Anyone trying to extract such data from server logs is familiar with this.
In any case, anyone who takes the browser statistics seriously is either seriously ignorant of the problems, or knows the problems but wants to push some PR agenda.
Slashdot Mirror
Don't charge per click but per sale generated.
Good idea, maybe, but there are some implementation problems. The advertiser has an obvious incentive to claim low sales from such ads. You don't have a good way of verifying their claims unless you can have people on their site watching every stage of their sales process. Also, it could easily lead to advertisers demanding that large numbers of (slightly) different ads be run, because if an ad doesn't work, they don't have to pay for it. You'd be doing their marketing research for them for free.
How could you keep your advertisers honest if you used such a policy?
Google's got it right. Small. Innocuous. Relevant to what you're looking for.
Yup. Like most of the geeks here, I mostly use browsers that can do things like block images from sites, so as to cut down on the more obnoxious ads. But I've also bought a fair number of things online over the years. And when I'm looking to buy something, I tend to first ask google about it. Both the matches and the accompanying ads are useful in that case.
Dunno how well it works with the general population, but google's approach is fairly good for people who are trying to find something and just get annoyed by irrelevant ads.
We oughta let them know that we appreciate their subtler approach to the whole topic.
In a few cases, commercial sites have asked me how I found them, and I've enjoyed telling them that I used google. That oughta give some of their marketing people a bit of a pause, since they probably "know" that google's approach isn't very successful at selling.
Hey, why don't you get in touch with our hero Mike Rowe, and offer him a top position. Then you can sell "MikeRoweSoft Windows". Or maybe make them with a soft vinyl frame, so they'le be "MikeRowe Soft Windows". Since the product will be real windows for buildings, your lawyers will be able to argue reasonably that they aren't competing products. But Microsoft will sue you anyway. ...
3) Profit!
Instead, why don't *you* just design your webpages to W3C standards and be done with it?
...
Uh, which W3C standard(s) should I follow?
Have you seen how many standards they've published? Do you honestly think any merely-human brain could even start to hold all that?
Meanwhile, I'm out here doing' my best. I do feed pages to various online validators quite often. Sometimes I can even make sense of what they tell me, and I fix the problems. Sometimes I'm just baffled at what they want me to do. But in those cases, I just reason that they're obviously talking to someone smarter than I am, so I don't worry about it too much.
Even if you try hard to implement to W3C standards, any sensible web developer would still test against a list of browsers. The actual screen representation can vary widely, even for software implementiing the same standards. Lately I've been testing against a number of PDA browsers to make sure that users of Tungstens and Blackberries can use my stuff. Those devices do have "real" browsers now, but they are severely constrained by their small screen. Just saying that your HTML is standards compliant isn't enough if some of your important clients are using such small display devices.
Now if there were a way to test against all of them without setting up a million-$-per-year testing lab. That's the real problem. You can easily have a page that passes all known HTML validators' tests, but still comes up unreadable (or very confusing) on some clients' screens, even though they are using a standards-compliant browser.
Lots of us really wish the W3C would ban frames
Heh. Actually, it's those starving artists who more and more are threatened by the "long after you're in your grave" copyrights.
It used to be that, to become a good musician, one of the important things was to copy the past masters. The first recordings enhanced this, by giving people access to actual performances of musicians that you'd never have heard before.
But now, we're seeing the start of copyright-infringement suits based on musicians copying old recordings. It sorta started with the samplers, but now its looking to effect everyone. Say that as a kid, you loved your parents' Satchmo and Dizzy recordings, and when you got your axe, you wore out the records learning to do those riffs. You'd better be careful. Someday soon, you'll be playing with a pickup band at a local club, you'll take several solos, and the next day you'll be hit with a lawsuit for unauthorized performance of a 1923 recording that's you don't even remember, but is there in your subconscious influencing your style.
A few years back, there was a sci-fi short story based on the idea that, with permanent copyrights finally enacted, the world had reached the point that all possible musical motifs had been copyrighted, and it was no longer legal to write new music. Any "performance", such as whistling a tune while walking down the street, would produce a rapid arrest and imprisonment, unless you had had the foresight to get written permission to perform that work at that time and place.
And note that almost all copyrights are now owned by corporations. In a few more years, the two remaining Beatles will die, and their music will fall into the hands of one or more corporations. It's the starving artists that will lose really big by this process, as they lose the right to produce without first getting permission from the corporations that own the bits and pieces of any new creation. The winners, as in the old feudal system of land ownership, will be the aristocracy that owns the shares of the IP corporations.
In another generation or three,, it may not be possible to produce any new art without violating a century-old copyright. But, by then we may be finally escaping our planet, and new land will be available for settling. But that probably won't be available to the poor, either.
Hey, any reason we couldn't build, say, 1024 of these things, and make a beowulf cluster of them?
... a small (iPAQ or BlackBerry or Tungsten sized) gadget that runs linux and can do IP over 802.11*, GSM and GPRS, I'd grab it and start developing software. Imagine what we could build if something like this were available to those of us who just want to do it ...
;-)
(Actually, all that in a laptop would be a good start. But so far, I haven't learned of a laptop with hardware for all three wireless comm systems. And there's the ongoing problem of publishing the proprietary specs so that 802.11* hardware works seamlessly on a linux laptop. But maybe we'll soon read that WiFi "almost works" on linux, just like it "almost works" on OSX.
Anyway, in our house, we have a Tungsten C (PasmOS + WiFi) and a BlackBerry 7280 (java VM + IP over GSM and GPRS). We keep thinking how nice it could be to have one pocket/purse-sized toy with all three and the programmer accessability of linux. If someone were to put one together and find a way to make it legal to use all three paths to the Net, we could start a real explosion of Open-Source, standards-based portable computer/comm gadgets.
Heh. I've read any number of comments on this phenomenon. They usually include the observation that someone with true command of a language will find a way to express their ideas as simply and clearly as possible (but no more so, as Einstein pointed out).
;-]
The "style" metrics seem to invariably give a higher score for complexity of syntax and for longer words. This isn't necessarily an indication of sophisticated writing, as a few of the big-word parodies here have clearly shown.
[N.B. The above split infinitive was intentional.
Part of Wodehouse's command of the English language was knowing how to express things simply and clearly. I'd guess that he would have sneered at suggestions that there was something uneducated about this.
Somebody got a thesaurus for Christmas!
;-)
Huh? Why would anyone bother to get a thesaurus, for themselves or for a friend? They're all online these days, and using them is faster than walking over to the bookcase and pulling down the hard copy.
Of course, if you're out of range of an Access Point, you might have a problem. But then, if you use one regularly, you've already downloaded it to your PB, right? Right?
(Actually, there's one within reach of my usual work chair, but how useful is a thesaurus when you work as a programmer?
2) If I were implementing the metric, any text using "boxen" would be downgraded to "Idiocy".
And, of course, anyone reading your ratings would downgrade you to "humo(u)r impaired".
Now, granted, "boxen" is a rather old bit of wordplay that's not nearly as funny as when it was new. But it's still good for ferreting out the people who don't have anything more important to complain about. So we can expect that it will continue to appear here, until it no longer gets any comment from bored readers.
I wonder if there are any language metrics that successfully take into account things like geek wordplay humor? That's gotta be something that's difficult to measure.
And you can now run vi on your Mac.
(OK, you can run emacs on your Mac, too. Maybe it should be eMac? eMax?)
Cars tend to have much higher resale values than computers and also tend to have a much longer average life.
;-)
True for resale value. But the linux crowd has profited enormously from the fact that even crappy Intel hardware often lasts a decade or more. It's the forced software upgrades that kill Windows machines while the hardware is still good. So the suckers who are forced to buy new hardware that will run the Windows upgrades are selling their "obsolete" hardware cheap, and it runs linux just fine. Carry your knoppix CD along to garage sales, and you can get yourself a nice machine for next to nothing.
In particular, there are lots of old Intel boxes out there running the firewalls and gateways for most of the Internet. There was even the story here last year about the "super-computer" that was a beowulf cluster of mostly discarded Intel boxes that had been bought for next to nothing. (Of course, the high-speed comm equipment that tied them together wasn't cheap.
Mac. And no one spends $2k+ on an Intel based PC, right? I mean, certainly no one here...
Well, I have a number of friends that have recently bought $2k+ IBM Thinkpads, which seem to all be Intel based. Of course, they immediately wiped the Windows OS and replaced it with linux. And I get the impression that there are a few linux users here.
One problem with Macs, of course, is that they're designed for people who haven't learned to move their fingers independently. My wife got a Powerbook some time back, and while she likes it a lot better than her old Windows box (and gave that to me so I could have another linux box), she does sometimes wax nostalgic about the 16-button mouse on the CAD workstations that she used to work with.
Myself, I'd be happy with 3 buttons, so that cut and paste will be as fast as it is on any unix/linux machine. I think that's one of the advantages of Thinkpads. And after a few months, it could well save $2000 as measured by our billing times.
There's the ongoing question of why IBM isn't selling Thinkpads with linux installed. With all their touted linux support, how are they missing this market opportunity?
As I am sure you know we have a similar paradox in sw consulting. The better you get the less time things take and the less you make.
;-)
Yup. Case in point: I remember my first encounter with writing a device driver, years ago. A new gadget was ordered, but there was no driver. I sent a message off to the vendor's Support people, and they sent me the hardware specs. We had source for the OS, so I grabbed a driver for a similar gadget, cloned it, and modified all the magic numbers to agree with the specs for the new gadget. It took maybe an hour at most.
A few days later, the new gadget arrived. I plugged it in, ran a couple of tests, and it worked. A Customer Support guy happened to be present, and he was astounded. "We don't have a driver for that." So I said "Well, you do now." And I explained that they had done a good job in the specs, so all I had to do was what I had done, and it worked.
Now, this isn't too amazing; it's how things should work. The interesting aspect to this is that I've gone to a number of interviews at companies that are, among other things, looking for people to write device drivers. I've never gotten such a job. The reason is always the same. The interviewer asks about driver experience, and I mention having produced a number of them on different projects over the years. But it's never good enough. They always say that they want someone with several years experience writing drivers. I can only honestly admit to a few hours experience. After all, if you have the detailed specs, it only takes an hour or so; if not, you can't do it no matter how much time you have.
When they bring up the "several years experience" requirement, I've several times asked if this means they really want someone who takes months to write a driver rather than hours. Inevitably, they respond to this as if I'm some sort of wise ass. A couple have told me to my face that, yes, this is what they want.
It doesn't surprise me that we have a lot of problems getting drivers written for new hardware, given that this appears to be the "normal" attitude toward hiring people to write the drivers.
Maybe I should be a bit less honest. After all, I could easily stretch the job out to months. But really, I'd much prefer to spend the time on difficult problems. (Something that isn't doable because I can't get the specs isn't what I'd call "difficult"; it's "impossible", and that's not the same thing at all.
So, in order to survive I guess you have to make shitty sw ...
...
Funny, perhaps, but with a core of truth. I recall a year or so back, when someone here got a high Funny rating for suggesting that consulting firms would be stupid to recommend linux, which would be just a few hours of billing time to install and configure, when they could recommend Windows, which would get them continuous billing time for support.
I showed this to several friends who worked at companies that did business consulting, and they all reacted the same way. "That's not funny; that's exactly how it works." Several said that their own in-house systems were all linux or BSD, but they'd be stupid to recommend them for clients.
I've also had occasion to discuss with employers the idea that I shouldn't produce software that "just works", because then they'll never have to hire me again. They all got worried looks on their faces as they agreed that I was right. But they didn't know what to do about it, considering their own restrictions on how company hiring was done.
If we want true Software Engineering, we need to find a way to end this conundrum. Currently we're paying people to produce software that doesn't last. We do this by the simple approach of paying them to work on software that has problems, and not paying anyone for software that doesn't have problems. Right now, I don't seem to be reading of anyone working on this. Complaining, yes. But how many people realize that when you hire people, they usually try to do things that you pay them to do?
I'll predict that we will eventually discover that we do have 200-year-old software. And it was all created by people who were writing it for their own use, not because someone was paying them to write it. This is why most Open Source software was created, after all. And it was how unix arose at Bell Labs.
I have a number of 20-year-old programs that I keep around because they're useful
You can't worry about your software working for that long until your hardware can last that long.
;-) I wish I could be around then to be proved right or wrong.
Oh, nonsense. Consider the well-known "Hello, world." program in the K&R C "bible". It's been around 30 years or so, and the hardware they were working on now only exists in a few museums. But that program is still in routine use on millions off computers.
Note also that K&R included not only the code for the program, but the commands to compile and execute it. They correctly pointed out that this information covered the initial implementation details in most software, and once you had it working on a new system, you had soved the major problems in getting any software running on the new system.
And note that, after more than a quarter century, their sample code to compile and execute this valuable little program still work on millions of computers, without changing a single character.
Funny thing is, you can't say this about the equivalent programs in most other languages. I've seen attempts to do the same (code + compile if needed + invocation) for any number of other languages and OSs, but I've seen very few successes. Usually code that's only 10 years old will need tweaks. But if you have a standard-compliant (ANSI) compiler and a standard-compliant (POSIC) OS, you can type in this standard program using any editor, type in the compile command to any shell, and type in the invocation to that shell, and see the expected greeting appear on whatever "terminal" hardware you are using.
I'll bet that this little program will still be around and will still work 200 years from now. Even that bizarre "a.out" name, though I expect that using "-o hello" in the cc command will also work as it did in 1975. (And the program will still have the same subtle bug as the original did.
But so far, the signs are that this won't be true for many programming languages or OSs. The idea that code should be reliable in the long run seems to be something that's beyond the comprehension of few language or OS implementers.
I'm not saying they're a bad paper, just that we should demand more from the US's supposed #1 paper.
Um, we can't demand anything from them. Only their advertisers and stockholders can do that. They're a for-profit corporation, after all.
If you want a news source that you personally can make demands of, you're stuck with things like NPR and VOA, where you in fact do have some influence (though it's diluted a bit by the other millions with the same influence).
Of course, there is a way to make demands of the NYT and other corporate news sources. Just give them enough money by buying ads or stock that they have to listen to you. But unless you're willing and able to do that, forget about making any demands of them.
Y'know, my first thought was "Google has to pay to index their own archive?" Then it occurred to me that you weren't referring to the google cache, but rather to the NYT archive.
So what about google's cache? I've read a number of NYT articles from there. What problems are there with using it rather than the NYT archive?
Of course, we still haven't seen the end of attempts to use the copyright law to outlaw things like google's cache. It's entirely possible that, in the not-so-distant future, all news will be online. Then the main effect of copyright will be to erase all of our history, as the copyright holders make it all inaccessible or erase it.
We really do need an addendum to the copyright laws stating that anything no longer available from the copyright holder becomes public domain.
Heh. I recall a few years ago, when the place I was working at instituted RSA everywhere, and I found that I didn't need to save copies of any keys. When I needed them, I'd just google for them.
...
I really wish this were a joke. I repeatedly suggested that they should investigate possible leaks of the keys. I'd do this without telling them what I'd found, because I was afraid that they'd just block that path and consider the problem fixed. I did wonder what other gaping holes were around the place.
Once I even fetched a needed key this way in front of a gang of company bigwigs. They didn't bat an eye; they oviously hadn't a clue as to what I'd just done. The one other techie present looked at me with large, disbelieving eyes
Later, on the way out, I heard him mutter "I don't believe it!"
I wonder how many other cases like this exist?
In some technical circles, the term "postdict" has been coined for this usage. But it hasn't caught on widely. Probably because this sort of misuse of "predict" is sufficiently obvious that most people understand what you meant.
Scientists routinely "predict" what will be found when observations are made. The fact that these are usually postdictions is true, perhaps, but not sufficiently interesting or important to produce a real push for adopting the more precise term.
If you google for "postdiction" you'll see that the term is in use in some specialties.
... signs like McDonalds used to have: "over 1 billion served!
;-)
Well, I recall back when those first started appearing, my thought was "Yeah, but how many were actually eaten?"
Also, I like to mention that I have two linux boxes that were delivered with Windows. So Microsoft certainly lists me as one of their satisfied customers. Obviously, I'm so happy with Windows that I bought two of them. In fact, I eventually erased the Windows partition on both and converted them to/var partitions. (They're much more useful to me as a cage to limit the damage of an overgrown log file than as a box full of shoddy MS spyware.
Just one of many reasons why industry statistics aren't to be taken seriously.
MS can and has been sued.
;-)
;-)
Yeah, by the US Dept of Justice, among others. And, of course, MS won that one.
Anyone have a list of lawsuits in which Microsoft actually lost and paid damages?
This isn't intended as a troll. I don't think I've read of anyone actually getting money from Microsoft via the courts. It might be interesting to read of a few cases.
There have been some cases close to that in which MS settled out of court. For example, their recent settlement with Opera. But Opera didn't actually win the settlement in court.
In particular, has any mom-and-pop company ever sued Microsoft for damages and won in court? Can anyone name the cases? And did the companies actually receive the money?
(That's money as in money, not discounts on purchases of MS products.
Well, considering that a number of Microsoft people had already gone public with the "It's not our problem; it's Mozilla's problem", I'd think that the obvious answer is that Microsoft's management was already very much aware of the problem. Pointing out that MS products have the same vulnerability is an obvious (if somewhat in-your-face) way to shoot down their FUD.
And, let's face it, they were using this as an opportunity to squelch the recent rash of switches from IE to Mozilla. They deserve to be hit fast and hard for such tactics.
(Not that the Mozilla people are totally innocent here. Even if you agree that it's a Windows bug, it's clear now that Mozilla could very easily catch it and pop up a warning window. That would have taken less time than was apparently spent discussing the issue and deciding to not deal with it right away.)
Software is manufacturing? I always thought I was in the service sector.
I mean a lawyer doesn't manufacture contracts, does he? He provides a service.
Heh. You may not have seen the news, but the Bush administration has lately been trying to get jobs like fast-food work redefined as "manufacturing".
After all, the person at the grill is taking raw materials (ground-up cattle) and using machinery to convert them to a product (the cooked burger patty). Then someone else in the assembly line combines that with other components to build the delivered product. Sounds like manufacturing, right?
With only a small stretch, we could also reclassify marketing and sales as "manufacturing". After all, they are building public images (sometimes out of whole cloth) and assembling them in the marks' - uh, I mean valued customers' -- minds.
Of course, this is being pushed by politicians who talk of "manufacturing a concensus".
Before long, we'll all be in manufacturing.
Heh, heh; if I had any mod points, I'd give you a "funny".
But, like most jokes, there's a grain of truth at the core.
I use google a fair amount, but probably a lot less than many non-geeks. I know how to bookmark things and organize my bookmarks. Periodically I move many of my bookmarks to more permanent places in several "Links.html" files. So I don't have to keep looking things up; my computers remember the good ones for me.
And, of course, I do remember a lot of URLs. I've been known to type in search URLs for google and a few other sites, eliminating a download of their main search page.
And another thing that confuses the browser statistics: It's quite common now for browsers to include a selection of ID strings to mimic other browsers. Mostly, this is used to mimic IE, so that you can see a site that only works for IE. There were a few stories recently about opera's facility for this, and the stories explained about the msn.com sites that actively looked for opera and sent damaged pages. Recently I got a BlackBerry 7280, installed its real browser, and one of the first things it did was ask me if I wanted it to pretend to be IE. Guess why they include this feature?
So we should expect that the statistics count lots of other browsers as IE. It takes some rather sophisticated programming to avoid this, and in some cases, you can't determine the true nature of a web client at all. Many of the counters are commercial operations doing the counting as part of a marketing operation, and they aren't well-motivated to try to uncover all the browser ID-string deceptions.
Of course, this use of fake ID strings was started years ago, when the first IE versions told servers that they were Mozilla. They still do this, actually, though they now include "IE" in a later part of the ID string. Anyone trying to extract such data from server logs is familiar with this.
In any case, anyone who takes the browser statistics seriously is either seriously ignorant of the problems, or knows the problems but wants to push some PR agenda.