Though for sufficiently serious problems, one can have recalls many years later.

I see no compelling need to move from my S2 running 2.3.6, and in fact I see some reason not to: the lack of a dedicated search button on the S4 (and S3). I like the idea of a consistent interface so that in every well-designed app where searching is relevant, I can search by pressing in exactly the same place. Plus, I like the fact that I can launch my favorite apps by holding down the search button for a second (though that does take an extra app). This is fairly minor, but from my point of view so are most of the improvements.

Plus I'd have to figure out how to legally root a new device. It took me months to figure out how to legally root my S2. (The standard ways of rooting of the S2 involved installing a custom ROM, but custom ROMs would contain copyrighted Samsung-copyrighted components, and hence downloading them is at least grayish.)

You might want to check the license as maybe the borrowing from work was already legit. IANAL, but the EULA with the downloadable CS2 says: "2.4 Portable or Home Computer Use. The primary user of the Computer on which the Software is installed may install a second copy of the Software for his or her exclusive use on either a portable Computer or a Computer located at his or her home, provided the Software on the portable or home Computer is not used at the same time as the Software on the primary Computer."

While most users most of the time don't need the extra features of Word, some users need them much of the time, and I expect many serious users would benefit from learning about them and using them on occasion.

In my own discipline (analytic philosophy), it is very common to have numbered propositions in articles, which are then referred to by number. The right way to do this in Word is to set up an automatically numbered list that allows for the list items to be noncontiguous but still numbered sequentially, and then to refer to the numbered propositions via the cross-referencing feature. This way, as you insert more numbered propositions while editing, the numbers and references get automatically updated. As far as I can see, Google Docs supports neither noncontiguous lists nor cross-referencing. That said, it is my feeling that while most people in my discipline use Word (a few use LaTeX, and I use it myself for symbol-heavy articles), most don't know about these features and just number and cross-reference manually.

For non-fiction writing where one needs to refer to other chapters, sections or footnotes (e.g., "See footnote 17, above" or "As we shall see in Section 4.3.1.a"), cross-referencing is pretty much essential if one wants to avoid error prone manual references, fixing which up can be nasty if the editor calls for revisions. Again, that's a reason to use LaTeX, but there is no need for LaTeX given that Word's alt-I,N does this quite well.

Assigning keys to special symbols is also very useful--sometimes I don't want to bother with LaTeX for something that just has some symbols from the Symbol font, but I certainly don't to be pressing alt-I,S and using the mouse to select the symbol (or alt-I,S,Tab,arrows,enter) each time, and a keyboard shortcut is just what the doctor ordered.

And while most of the time, macros aren't needed, there are times when they greatly reduce labor. For instance, recently I indexed one of my books. To do that, I used a perl script that uses in-text ASCII codes marking index entries and page breaks on the Word file back-generated from the galley PDF (the press refused to give me a copy of their Quark or InDesign file). But I had to insert all those ASCII codes to mark areas of text to reference, e.g., {{IndexEntryA:}}Text to which the index entry should point{{:IndexEntryA}}. Typing in the codes would be error prone. So I wrote some super-simple macros which greatly reduced labor. It still was a ton of work, but without macros it would have been very hard. How often do I index? Only once every couple of years when I have a book coming out. But when I do, I need all the help I can get from the software.

And there are little conveniences of macros. I was once writing something that didn't have much in the way of symbols, but kept on having italic variable letters with numerical subscripts in the text. I could have switched to LaTeX, but I liked writing in Word and that was about the only bit of technical symbolism I routinely needed. Without macros, a typical such sequence (T subscript 1, say) would be: ctrl-I, T, ctrl-I, ctrl-=, 1, ctrl-=. Easy to slip up, and a nuisance. With macros, I could just type T1, alt-S, and it would format it correctly. (A more sophisticated macro would check whether the subscript was numerical or a variable letter and italicize the latter but not the former. I can't remember if I had it do that.) The macro made it possible to more fluently, without the annoyance of slowing down to format subscripted variables whenever they occurred.

Whenever one is doing anything repetitively, a macro will help. I suppose for simpler cases when it's just a keyboard shortcut, one could use Google Docs and a keyboard macro program on one's desktop. But sometimes the macros need to aware of what's in the word processor text.

I wonder if they're going to enforce this no-financial-interest-in-competing-product rule with regard to academic books. In disciplines where book writing is expected of scholars (most of the humanities, but much less so in the sciences), most of the best qualified reviewers will themselves be scholars who have a book on a related topic or be working towards having a book on a related topic.

Almost certainly you're right. But there did used to be vector monitors. It seems rather unlikely to me that we'd go back to them, but who knows?

The EU press release says that people can still sue for civil remedies, as these aren't preempted by the fine, and that the EU decision can be used as a proof of fact in court. I assume there will be something like class action suits now, or does Europe not have these?

You may be right about most cases, but not all cases.

Some people might prefer older versions because Google from time to time disables an API thereby breaking some app important to them. For instance, in JB, Google made it impossible for non-root non-system apps to access system logs, which kills many apps that monitor the logs to check for system conditions like app launch (e.g., I have an open source app that lets you set CPU, orientation and other settings differently for different apps; on JB, alas, it needs root as there is no non-root way to detect app launches). In at least some of the 4.2 builds, it seems that apps can no longer insert global sound effects, which kills global equalizer or audio boost apps, which again may be quite important to a user. Google may have good security reasons for introducing such changes (though I think they should let the user decide), but a user may also have good reasons for sticking with the older OS version.

There may also be low-level apps that do undocumented stuff and that are important to a user that are broken more innocently by newer OS versions. For instance, last I checked, ChainFire3D's night mode, which turns all screen colors to red-and-black on rooted devices, doesn't work on ICS. Yet if one does amateur astronomy, something like ChainFire3D's night mode is really important. (I wrote an open source alternative that works on rooted ICS, but it's only for some Samsung phones.)

There will also be minor convenience preferences. For instance, I hate what Google did with the recent apps list in ICS--I don't see the point of oversize previews of apps that you have to scroll through, as it's much faster to have a single window of icons like in earlier versions--one tap instead of slide and tap. No doubt many users like the recent apps list.

Finally, if an OS version is close to rock-solid on one's phone, and does almost everything one wants, it may not be worth the risk of new instabilities to upgrade.

In my own case, I am deliberately sticking to GB on my Galaxy S2. (My wife has ICS on hers. Her S2 is far less stable than mine, even though I have tons of apps on mine doing all sorts of custom stuff. It could be a hardware issue, though.)

Whether a statement is a lie does not depend on whether the statement is true or not, but on what the speaker believes.

I suppose there is no way to make Tor work only for users in repressive countries?

I don't think that committing rape and murder correlates with high intelligence.

There are tons of movies and TV series that just aren't available on Netflix or Amazon for rental.

That one can get away with breaking a law isn't an excuse for doing so.

There was a time when I wanted to use an open source MPEG-4 player on our PalmOS devices at home. I actually got a patent license from MPEG LA, since there are no fees for under 100,000 units or so. I expressly told them that I'd just be doing this for personal use. They sent me the license agreements overnight. Twice a year, I've had had to report the number of units, so I duly reported one per installed build, each time I installed a build. I don't think they realized how absurdly costly this was for them. Too bad for them: I kind of hoped they'd say that with such a small number of units, they're fine with me doing it without a license.

Strictly speaking, going by the linked article, he never said that embryology (or evolution or the Big Bang, for the matter) was a lie. He said that all that he was taught about embryology, evolution or the Big Bang was a lie. In the case of evolution and the Big Bang that's probably not much of a distinction. In the case of embryology, it might be a bigger distinction.

IANAL, but I skimmed through the judge's decision, and it's not just for his downloading the stuff, but also for redistribution. The judge seems to think that one way to evaluate the costs of the redistribution would be to think about how much an unlimited license for all and sundry on the Internet for the songs would be worth.

That said, it sure seems excessive, but not as ridiculously excessive as it would be just for downloading.