But I don't really like the thought of creating and killing millions/billions of things that are/maybe/might be/could have been/sorta/etc humans to get there.
I think it's important to stress that Bush isn't interested so much in protecting human life. Even if his motives are genuine and straightforward, he's only interested in protecting human life at an arbitrary age cutoff.
Bush has sent a significant number of humans, some of them my family members, in harms way without, it turns out, verifying the validity of his research before acting. What gets me is the forked tongue speeches about the sanctity of human life on one hand, and the wholesale slaughter on the other.
This does however put me in mind of one of my favorite lawyer jokes.
Why are lab scientists increasingly turning to experimentation on lawyers instead of white mice?
Population predictions indicate that there will probably never be a shortage of lawyers.
Studies have shown that research scientists often develop emotional attachment to white mice.
There are some things you just can't get a white mouse to do.
I realize it's useless to reply to this, but I'm annoyed by the constant references to "Do No Evil" that miss the point.
They actually have a page where they spell out what they mean by Do No Evil. It doesn't mean, do no evil by the standards of every human in America. Their guidelines very clearly indicate that doing no evil means acting in good faith rather than trying to dupe users. It has more to do with returning honestly ranked search results, not installing spyware, and making programs you install easy to uninstall.
There's nothing evil about complying with the local laws, whatever you may think of them. If the local law happens to be communist, Google should respect that, and they do. It should just make you grateful you live in a country where you can get full, unadulturated Google.
A pivot table is a primitive data cube. (I'm sure that helped). Basically, given a table of data with different dimensions (factoids), you can come up with aggregate groupings to analyze different segments of your population.
So say you have a group of customers with these dimensions:
accountnumber
gender
state
activity
countOfPurchases
totalPurchasesInCents
latestPurchase
tenureOfMembership
averageNumberOfItemsPerBasket
Each customer gets one row in the table. The pivot allows you to cross section this data. You could, for example, put gender across the top and tenure of membership along the left side with countOfAccountnumber (aggregate of accountnumber) to see whether males or females tend to stay with you longer. You could change the count to a percentage to see this normalized across different genders.
You could put in both a sum of items and an average of items per order to see if your customers tend to buy more all at once, or in smaller chunks. You could pop in latest purchase and see if this trend is increasing or decreasing.
You could do all of this with SQL, but the pivot table makes it really fast and convenient. Even PHB's can use them in our company, and often find interesting pivots.
The most common uses that I've seen are using pivots to isolate:
Large churn segments (customers leaving in a big group)
Seasonal buying patterns (patterns grouped by month to see that X happens every December
Campaign analysis (Have isControl be a dimension and compare data based on whether or not they were control)
I use pivot tables in my job extensively, and I've spend a good deal of time messing with data pilot. The lastest iteration is much better than the previous one. You can now drag around dimensions, double clicking on dimension names. I just found subtotalling.
Here are my annoyances:
1. You have to right click and go into Start to edit too many things. Double clicking on the headers in the actual data pilot table should bring up the options like it does in the Start view. 2. Arbitrary filters should be available from the data pilot table. The filter thing is actually more intuitive than Excel's, and less prone to errors, but a list of items dropped down from the header should be able to build a condition for you. 3. Double clicking on the intersection of fields should open a new sheet with only those records that fall into that intersection. This is incredibly useful (and one of the main reasons we use pivots). Once you've identified an intersting segment, you want to see the items in that segment to look for further detail. 4. Fields in the data area should have sticky 'data types' assignable. This could actually be done better than Excel. Once a field is given a particular type and format (ie, comma separated number with no decimals), it should stay that way for the duration of the pivot. If you drag the field in and out, it should remain in the same format and data type. This would be much better than excel's current behavior and would possibly tip my boss' opinion. 5. The data pilot wizard should give an option to create the pilot in a new sheet. 6. One of the more powerful featurs of pivot tables is pivot charts. You can actually pivot the chart in chart mode and see the visualization change. I know you can do this with the numbers in the pivot, but sometimes you notice something in a graph that you wouldn't see otherwise. 7. Date grouping is very advanced in excel. You can right click on a date and tell it which elements of the date to group by. I can change a monthly report to a quarterly report in about 4 clicks. This saves needless reproduction of date fields.
When data pilot gets these features, it'll be a viable competitor with Excel's pivot tables.
I was actually thinking about this, blowing bubbles with my daughter. It occured to me that I could see a very detailed reflection in the surface of the bubble. It also occured to me that the bubble lasted a surprizingly long time so long as it wasn't disturbed. I wondered if the study of bubbles might not lead to a new translucent holographic display. I don't have any experience with the optics or chemistry to know if it's possible, but bubbles have some interesting properties
they can be spherical, domes or flat
They are translucent
They are extremely cheap
Probably just me daydreaming, but I thought it worth noting.
You could just assume, like I did, that it was a reference to "The wall-like Aias" (Anglified to Ajax) from the Illiad. He was the biggest and the strongest of the Achaians who won the trojan war.
I prefer Ajax to RIA because it's easier to remember than another TLA.
Actually, this kind of thing generally backfires in other ways as well. The more specific and comprehensive the rating system, the easier it is for the (children savvy > parent savvy) population to get the raunchy stuff.
I can remember being a kid scanning for late night Showtime movies that were R rated. I can remember, with delight, when TV guide started putting (nudity/language/violence) tags on them, so I could filter out the stupid movies and go straight for the breasts.
This isn't censorship. This is making it harder for people to say that they got to a site 'accidentally' because it's quite easy to filter out the XXX content.
Now if only they could force anything in a popup to have a.pop extension, we could 'censor' those too.
That's interesting because I have my IM client (Gaim, but trillian does it also) notify me when my mail comes in. Sadly Google doesn't have an IM service yet, but I use gmail notifier for that. The upshot is, my e-mail is easily as quick about notifying me when there are new messages as IM.
I find it amusing that instant gratification in communication is used pejoratively. Do you consider talking on the phone and in person also distasteful? Often, conversations are more efficient in IM because you can have question and answer sessions without large gaps of time passing between comments.
In principal, IM is just e-mail without the overhead of headers and subject lines. I have threads at work where it took days to have a conversation that could have taken minutes because we were using email instead of IM. All of the people I with whom I work closely use IM to share code snippets, ask questions and hash out details.
Most IM clients now have searchable history and offline history. This takes away e-mail's permanence as a differentiator. We do use e-mail at work, both because not everyone has IM and because some of the execs don't use it, but I find that it's much easier to locate a detail in a continuous IM conversation than trying to search through subject lines guessing at which message holds the right info.
I imagine a lot of objection to cookies, particularly those storing information, is not so much privacy from the web administrators, but privacy from others using the terminal.
Cookies can indicate that you've been visiting porn, personals, job search, etc websites that indicate to your boss or SO that you've been up to things you oughtnt. Given the enormous mass of sites that put cookies on your browser, it's faster to eliminate all cookies once a month than to sort through them and eliminate only those suspect ones.
Given that people obviously don't want to admit this is their motivation, they must complain loudly about privacy issues and Big Brother looking over their uniquely identified shoulder.
I think it would be interesting to develop three classes of cookies.
The interesting thing to me is that the correctness of grammar is not boolean. It's not [Perfect, Meaningless], it's a scale of meaningfullness. I believe that erring on the side of not learning language skills at all is unfortunate because one may have excellent ideas that people ignore because they're badly expressed.
On the other hand, the cost/gain of triple proofing every bit of written and spoken content begins to waste valuable time that could be spend communicating on dotting i's and crossing t's that could easily be taken as given. (That was probably a run on)
I find that it is useful to learn to write well to the point where most grammatic rules are automatic, so that I can be understood. I find that it is counter-productive to allow yourself to be so focused on the correctness of grammar that you demean others or lose access to their thoughts because of a trivial technicality.
As a postscript: Part of the reason programmers do better with computer languages (with the exception of maybe Perl) is that they are generally developed with a limited vocabulary, several predictable and consistant rules and a well documented structure over a period of a few years.
English is the bastard child of at least 3 languages which developed over centuries from disparate gramatical structures because of weird sociological pressures. In most languages, it's not even possible to split an infinitive.
I would think this had more weight if Google actually served up porn ads. They don't. In fact, I've never had any embarassing ads remotely related to my midget-viking-horse porn fetish.
I was just going to say that. Go is the natural extension of a strategy game. It removes all questions of luck, terrain, or unit quality. It assumes that all such distinctions distract from the true quality of a good general.
What I love about Go, and turn-based strategy vs. rts, is that you have to take full responsibility for every move. There's no concept of rushing or using cover to advantage. That way, when you pull off a subterfuge, or are victimized by it, you feel the full weight.
Slashdot Mirror
I think it's important to stress that Bush isn't interested so much in protecting human life. Even if his motives are genuine and straightforward, he's only interested in protecting human life at an arbitrary age cutoff.
Bush has sent a significant number of humans, some of them my family members, in harms way without, it turns out, verifying the validity of his research before acting. What gets me is the forked tongue speeches about the sanctity of human life on one hand, and the wholesale slaughter on the other.
This does however put me in mind of one of my favorite lawyer jokes.
Why are lab scientists increasingly turning to experimentation on lawyers instead of white mice?
I realize it's useless to reply to this, but I'm annoyed by the constant references to "Do No Evil" that miss the point.
They actually have a page where they spell out what they mean by Do No Evil. It doesn't mean, do no evil by the standards of every human in America. Their guidelines very clearly indicate that doing no evil means acting in good faith rather than trying to dupe users. It has more to do with returning honestly ranked search results, not installing spyware, and making programs you install easy to uninstall.
There's nothing evil about complying with the local laws, whatever you may think of them. If the local law happens to be communist, Google should respect that, and they do. It should just make you grateful you live in a country where you can get full, unadulturated Google.
So say you have a group of customers with these dimensions:
Each customer gets one row in the table. The pivot allows you to cross section this data. You could, for example, put gender across the top and tenure of membership along the left side with countOfAccountnumber (aggregate of accountnumber) to see whether males or females tend to stay with you longer. You could change the count to a percentage to see this normalized across different genders.
You could put in both a sum of items and an average of items per order to see if your customers tend to buy more all at once, or in smaller chunks. You could pop in latest purchase and see if this trend is increasing or decreasing.
You could do all of this with SQL, but the pivot table makes it really fast and convenient. Even PHB's can use them in our company, and often find interesting pivots.
The most common uses that I've seen are using pivots to isolate:
Hope that helped.
I use pivot tables in my job extensively, and I've spend a good deal of time messing with data pilot. The lastest iteration is much better than the previous one. You can now drag around dimensions, double clicking on dimension names. I just found subtotalling.
Here are my annoyances:
1. You have to right click and go into Start to edit too many things. Double clicking on the headers in the actual data pilot table should bring up the options like it does in the Start view.
2. Arbitrary filters should be available from the data pilot table. The filter thing is actually more intuitive than Excel's, and less prone to errors, but a list of items dropped down from the header should be able to build a condition for you.
3. Double clicking on the intersection of fields should open a new sheet with only those records that fall into that intersection. This is incredibly useful (and one of the main reasons we use pivots). Once you've identified an intersting segment, you want to see the items in that segment to look for further detail.
4. Fields in the data area should have sticky 'data types' assignable. This could actually be done better than Excel. Once a field is given a particular type and format (ie, comma separated number with no decimals), it should stay that way for the duration of the pivot. If you drag the field in and out, it should remain in the same format and data type. This would be much better than excel's current behavior and would possibly tip my boss' opinion.
5. The data pilot wizard should give an option to create the pilot in a new sheet.
6. One of the more powerful featurs of pivot tables is pivot charts. You can actually pivot the chart in chart mode and see the visualization change. I know you can do this with the numbers in the pivot, but sometimes you notice something in a graph that you wouldn't see otherwise.
7. Date grouping is very advanced in excel. You can right click on a date and tell it which elements of the date to group by. I can change a monthly report to a quarterly report in about 4 clicks. This saves needless reproduction of date fields.
When data pilot gets these features, it'll be a viable competitor with Excel's pivot tables.
- they can be spherical, domes or flat
- They are translucent
- They are extremely cheap
Probably just me daydreaming, but I thought it worth noting.You seem to have left out reliable.
You could just assume, like I did, that it was a reference to "The wall-like Aias" (Anglified to Ajax) from the Illiad. He was the biggest and the strongest of the Achaians who won the trojan war. I prefer Ajax to RIA because it's easier to remember than another TLA.
Actually, this kind of thing generally backfires in other ways as well. The more specific and comprehensive the rating system, the easier it is for the (children savvy > parent savvy) population to get the raunchy stuff.
I can remember being a kid scanning for late night Showtime movies that were R rated. I can remember, with delight, when TV guide started putting (nudity/language/violence) tags on them, so I could filter out the stupid movies and go straight for the breasts.
This isn't censorship. This is making it harder for people to say that they got to a site 'accidentally' because it's quite easy to filter out the XXX content.
Now if only they could force anything in a popup to have a .pop extension, we could 'censor' those too.
That's interesting because I have my IM client (Gaim, but trillian does it also) notify me when my mail comes in. Sadly Google doesn't have an IM service yet, but I use gmail notifier for that. The upshot is, my e-mail is easily as quick about notifying me when there are new messages as IM.
I find it amusing that instant gratification in communication is used pejoratively. Do you consider talking on the phone and in person also distasteful? Often, conversations are more efficient in IM because you can have question and answer sessions without large gaps of time passing between comments. In principal, IM is just e-mail without the overhead of headers and subject lines. I have threads at work where it took days to have a conversation that could have taken minutes because we were using email instead of IM. All of the people I with whom I work closely use IM to share code snippets, ask questions and hash out details. Most IM clients now have searchable history and offline history. This takes away e-mail's permanence as a differentiator. We do use e-mail at work, both because not everyone has IM and because some of the execs don't use it, but I find that it's much easier to locate a detail in a continuous IM conversation than trying to search through subject lines guessing at which message holds the right info.
I imagine a lot of objection to cookies, particularly those storing information, is not so much privacy from the web administrators, but privacy from others using the terminal.
Cookies can indicate that you've been visiting porn, personals, job search, etc websites that indicate to your boss or SO that you've been up to things you oughtnt. Given the enormous mass of sites that put cookies on your browser, it's faster to eliminate all cookies once a month than to sort through them and eliminate only those suspect ones.
Given that people obviously don't want to admit this is their motivation, they must complain loudly about privacy issues and Big Brother looking over their uniquely identified shoulder.
I think it would be interesting to develop three classes of cookies.
The interesting thing to me is that the correctness of grammar is not boolean. It's not [Perfect, Meaningless], it's a scale of meaningfullness. I believe that erring on the side of not learning language skills at all is unfortunate because one may have excellent ideas that people ignore because they're badly expressed.
On the other hand, the cost/gain of triple proofing every bit of written and spoken content begins to waste valuable time that could be spend communicating on dotting i's and crossing t's that could easily be taken as given. (That was probably a run on)
I find that it is useful to learn to write well to the point where most grammatic rules are automatic, so that I can be understood. I find that it is counter-productive to allow yourself to be so focused on the correctness of grammar that you demean others or lose access to their thoughts because of a trivial technicality.
As a postscript: Part of the reason programmers do better with computer languages (with the exception of maybe Perl) is that they are generally developed with a limited vocabulary, several predictable and consistant rules and a well documented structure over a period of a few years.
English is the bastard child of at least 3 languages which developed over centuries from disparate gramatical structures because of weird sociological pressures. In most languages, it's not even possible to split an infinitive.
I would think this had more weight if Google actually served up porn ads. They don't. In fact, I've never had any embarassing ads remotely related to my midget-viking-horse porn fetish.
I was just going to say that. Go is the natural extension of a strategy game. It removes all questions of luck, terrain, or unit quality. It assumes that all such distinctions distract from the true quality of a good general. What I love about Go, and turn-based strategy vs. rts, is that you have to take full responsibility for every move. There's no concept of rushing or using cover to advantage. That way, when you pull off a subterfuge, or are victimized by it, you feel the full weight.