The best kittens, technology, and video games blog in the world.

Thursday, September 24, 2009

Google Everything and some alternatives

Search cat by zenera from flickr (CC-SA)
Do you feel a bit uncomfortable using countless web services all operated by a single company? I can feel your pain. Here's a list of things I use (or tried but didn't like), and some alternatives if I know any.

Google Search

It used to be the case that Google Search was vastly better than anything else. Not longer so.
Don't believe me? Try the Blind Search - it searches the same term in Google, Yahoo, and Bing, then you can vote for the winner. Most of the time the results are extremely similar, and the scores are Google - 41%, Bing - 31%, Yahoo - 28%. Google still leads but not overwhelmingly.

While with text searches there's no reason to switch away from Google, Bing Images and especially Bing Video is quite significantly better than Google equivalents. Bing is simply a better search engine for porn than Google, they know it, and even advertise this fact:


I recommend we start using the word "to bing" as "to search Internet for video porn", just like "to google" means "to search Internet for textual websites".

Google Mail

I tried plenty of other webmail providers, and plenty of offline email programs. Even some Intranet email solutions like Zimbra and Exchange. Google Mail is so overwhelmingly better than anything else it's not even funny.

Hotmail, and Yahoo Mail, which I tried as throwaway spam accounts, are just too painful to use. There are no alternatives, it's Google or pain.

Google Maps

Google Maps was quite revolutionary when it first appeared. Unlike with Google Mail, here the competition did catch up quite successfully. Yahoo Maps, Bing Maps, MultiMap, StreetMap (well, that one's UK-only), and probably some others, are all quite usable for directions, and all suck for searching local businesses.

I've heard some drivers having preferences for one or the other, but as a public transport user/occasional cyclist, I'm still waiting for the first to offer decent TFL integration.

YouTube

YouTube annoys me considerably - they don't want to host anything even remotely risky. They limit access to videos by geography, remove those that include some copyrighted music in the background to unrelated funny thing.

On the other hand they have no problem hosting this video of a brutal rape (NSFW).

There are some alternatives, none anywhere near the size of YouTube. Not counting the porn tube of course. I have limited experience with them, so no in-depth comparison here.

Google Chrome

It's nowhere near being the most popular browser. I'm also not sure what's its main selling point, other than tab sandboxing, which Firefox badly needs.

Other than that, Firefox is vastly better. It has countless useful plugins. It has Greasemonkey, and Greasefire. It has Firebug. It has AdBlockPlus. Chrome isn't even close.

Not like I'm complaining, if it can get even one user away from IE, it's a win.

Blogger

Let's face it - Blogger is a pretty crappy blogging platform. But then I haven't heard users of any other being happy about theirs, so perhaps all are equally bad. And effort of moving the blog somewhere else (plus losing the URL) seems too big for me to bother.

In particular the way they handle spam comments is simply atrociously bad. I get a notification of a comment, and then it's about ten clicks to remove it if it's spam, instead of just one or two.

Google Reader

It's basically an online RSS reader. What I hate about most desktop RSS readers is that they tend to push notifications onto you, and don't live inside the most natural place possible - the browser.

With Google Reader Watcher, Firefox tells me how many unread items there are, and which website they're for. Which is exactly as much notification as I need.

I think the only thing left to want is "open every new item in a fresh tab" function, but there's probably some Greasemonkey script for that, or if not it would be pretty simple to write it.

I don't see much need for alternatives.

Google Talk

The main problem with IM programs is that everyone is on a different network, so you need way too many programs to talk with everyone. Adium has most of them. Except Skype. And the most basic functionality only - not webcams, no microphones, not even all the icons.

I don't have particular problems with Google Talk, but I'm going to use this unique opportunity to rant about Skype a bit. OSX version of Skype doesn't seem to have any way of disabling text notifications of contacts becoming available or unavailable. So it constantly spams me with irrelevant garbage. It's not possible to use Skype network from other programs even for IM chats. Many networks try this restriction - and abuse their control over access to other users, but Skype seems to be the only one which did it successfully. Would you be happy if you needed an iPhone to call a person using iPhone, a Nokia phone to call a person using a Nokia phone, and basically 20 different phones to speak with everyone you want to? Me neither, so why are we allowing exactly the same thing in the IM world?

Google News

I don't see Google News as solving any real problem. My main problem with news is that they focus on events of the day, while the important things are usually long ongoing processes, with little need for daily updates. And if you're away for a week or two, there's no easy way to look back at this time and see if anything important happened.

As far as I can tell, it's equally true for all alternatives. If you know a good way to follow important news, tell me about it in comments.

Google Docs

Google Docs is a pretty cool webapp for editing documents together. I have irrational hatred towards WYSIWYGs, especially web WYSIWYGs, so don't ask me about fancy advanced features, but it solves the basic need well enough, and I don't see what could be the big selling point for its alternatives.

Google Code

Google Code is a nice hosting service for Open Source projects. It supports Subversion (obviously), and Mercurial, but not git.

Some of the alternatives are SourceForge (both Subversion, and git, also some other VCSs I don't care about), github (only git, duh). I'm not sure how bad github's limitation of 300MB per open source repository is - my personal code repository's .git is easily bigger than that, but then I freely store binaries and stuff in it (like font files for jrpg).

Overall I don't really see any major advantage for Google Code over it's countless competitors, and it's not even particularly big.

Google Analytics

I'm going to be pretty clueless about it, as Google Analytics is the only one I tried.

Good things - it gives you pretty decent overview of traffic, and is really easy to install.

Bad things - no decent API (screen scrapping possible but quite painful), no access to raw data, I'm not a big fan of their interface, which seems to be focused on all the wrong things.

If I started a website these days, I would definitely try something else, if only to see if it's any better. Switching an existing site would probably not be worth the effort, as all historical data would get lost.

Summary

These are just some of Google services and programs I tried. It's scary in two ways. First, a single company has way too much control over the Internet, and over people's often private data. Google has been relatively well-behaved as far as huge soulless corporations go.

Second, and perhaps more important problem is - what the fuck is the entire competition doing? The only reason people use Google Absolutely Everything is because most of everything else sucks so much. Google didn't succeed just once or twice, it had long stream of spectacular successes where everyone else failed (search, maps), and often even failed to catch up years later (webmail, text ads). Seriously, how difficult is it to innovate a bit online? Difficult enough that a single company does most of it? Hardly anything good on the web ever came from Yahoo, Microsoft, Apple, Sun, and countless other huge Internet companies in ages. The only other major company that seems to be doing anything interesting is Amazon. A damn bookstore. A damn bookstore is doing more web innovation than all the software and web giants put together.

Yes, there are still startups doing the cool stuff, but most of them stop as soon as they get bought. Like del.icio.us, which was revolutionary, but then got bought by Yahoo, and is stagnant ever since.

In search for Just Wars

Allied World War I soldiers by Dunechaser from flickr (CC-NC-SA)
I know people don't really read my political posts. If I wanted to be read, I'd be writing angry rants about video games, programming, and Harry Potter fan-fiction.

Here's a quick post about politics, namely about just wars, and other violence. Right wing nuts are generally pro-war, no matter who's fighting whom. What's surprising is that even left wing nuts are pro-war. Here's some Chomsky, it doesn't get much left-wing-nuttier than that:

ROBERT B. SILVERS: ... Under what conditions, if any, can violent action be said to be "legitimate"? ...

NOAM CHOMSKY: My general feeling is that this kind of question can't be answered in a meaningful way when it's abstracted from the context of particular historical concrete circumstances. Any rational person would agree that violence is not legitimate unless the consequences of such action are to eliminate a still greater evil. Now there are people of course who go much further and say that one must oppose violence in general, quite apart from any possible consequences. I think that such a person is asserting one of two things. Either he's saying that the resort to violence is illegitimate even if the consequences are to eliminate a greater evil; or he's saying that under no conceivable circumstances will the consequences ever be such as to eliminate a greater evil. The second of these is a factual assumption and it's almost certainly false. One can easily imagine and find circumstances in which violence does eliminate a greater evil. As to the first, it's a kind of irreducible moral judgment that one should not resort to violence even if it would eliminate a greater evil. And these judgments are very hard to argue. I can only say that to me it seems like an immoral judgment.


Oh, really? So, could you please point me to some examples of wars that successfully eliminated greater evils? The criteria are, you can agree that they are sane:
  1. Without the war, very large number of people would have died or otherwise suffered.
  2. There was no other plausible way of saving them, and all alternatives have been tried.
  3. Thanks to the war, most of those people were saved.
  4. Few people died or otherwise suffered in the war, especially civilians, but also participants. Some suffering is allowed, but only if it's vastly overwhelmed by suffering prevented.
  5. It must have been possible to predict it advance that this success was very likely. Lucky wins don't count. And of course, it must have worked. Failed good intentions don't count.
I have really difficult time finding out wars that fit these criteria, even with very liberal grading.

The latest war in Iraq

The latest war in Iraq is easily not anywhere close to a just war. Saddam wasn't a nice guy, but amount of suffering he was causing recently was quite insignificant as for third world dictator standards, so it fails on count one, and automatically two and three.

According to Lancet estimates there were 1.3 million excess deaths due to it. Many other Iraqis suffered in other ways. So fail on count four.

Only fools would have expected an easy win, and being greeted with flowers, it's the Middle East after all. All claims about WMD, Saddam's connections to Al-Qaeda and so on don't count, as they were known lies, and even if Bush believed them, they would fail by count five.

Score: 0/5

The latest war in Afghanistan

This war is slightly better, as number of casualties on all sides is closer to 60 thousand, not counting another 10 thousands or so in the spill-off to Pakistan, what's most likely an underestimate in a poorly researched place like Afghanistan, but still much lower than Iraq's. And Bin Laden was in Afghanistan.

Still, it fails on all counts. There were people who died in Al-Qaeda terrorism, but other than a single well-known case, repetition of which was fairly easy to prevent, their numbers were really low. Even with 9/11 much lower than number of people who could be expected to die in the war, so count one fails.

There were plenty of plausible alternatives for getting or killing Bin Laden, and dealing with terrorism, so fail at criterion two. See IRA and ETA for good examples. Or even Clinton-style targeted killings.

Terrorism increased greatly after the war, so it fails criterion three. Count four doesn't fail that badly, assuming figures are not overly underestimated (they usually are, counting only direct deaths), but it's still pretty large number just to catch one guy. Let's be generous and count it as a point.

Anyone who expected an easy win in Afghanistan must have slept through history classes. And Afghanistan is still in chaos, terrorists still terrorize, and Bin Laden was never caught or killed, so criterion five spectacularly fails.

Score: 1/5

The Second World War

Let's move back in time to World War II, widely proclaimed to be the most just war in history, without any basis in fact.

So first, suffering. This is very likely - if everyone just gave Hitler what he wanted (mostly Central Europe to share with Stalin), millions of people would be reduced to status of second class citizens, and depending on how it would go, hundreds of thousands or even millions would die. It's not certain if large scale death camps would be implemented without the war - it seems Nazis were fine with Jews and everyone else they didn't like just getting out of their country (500,000 or so emigrated from Germany before the war, also see the famous Madagascar Plan), but let's assume they really wanted to perform massive genocide. In any case, the first criterion is quite plausible, so full point.

The second criterion fails. The obvious plausible alternative of letting people persecuted by Nazis emigrate to other countries, most were not terribly happy about letting them in, even through Nazis were quite glad about just kicking them out (first confiscating most of their property). Forced resettlement would suck, but it still beats getting gassed.

Now, the third one fails miserably - Nazis killed pretty much everyone they wanted to kill. Most people who would have suffered living under Nazi dictatorship, were instead forced to live under Soviet dictatorship. Approximately 70 million people have died, most of them outside territories that Nazis would have if the world decided against going to war with them. So it's spectacular failure on the counts three and four.

There could be no expectation of an easy victory - spectacular fall of France might have been unexpected, but World War I style trench warfare with millions of casualties was the primary alternative. It's not even obvious if Allies really cared about defeating Nazis - if they were serious about winning, going after them in September 1939 while they were busy in Poland seemed like the best moment, but they didn't do anything until they were attacked.

Score: 1/5

American Civil War

Here's another "glorious" war - American Civil War. It's really historically inaccurate, but let's grant the common myth that the cause of war was abolition of slavery. There were four million slaves in USA in 1860, to be freed. So a somewhat mythical point, but granted.

As for alternatives being untried, most of the world abolished already or soon after 1860, without getting into any major wars. Even countries highly dependent on slave labour, which quite smoothly transitioned into poor waged labour, with or without compensations for former slave owners. There's no reason why it wouldn't work in United States, sooner or later. No points here.

As for point three, yes, all slaves were freed, so entire point here. Sure, there was a century of discrimination ahead, but let's not get too nitpicking.

Point four utterly fails. 620 thousand soldiers were killed, plus some unspecified number of civilians. It was quite clean as far as wars go, so I wouldn't expect a huge value for civilian excess deaths. Still, that's only 6 slaves freed for one death. I'm sure most abolitionist activists had far better than this number - by collecting money, and buying off slaves' freedom, or by nagging slave-owners' conscience until they agreed to free slaves one person could surely free far more than 6 slaves with lifetime's effort.

As for point five, it was successful, and due to North massively overwhelming South in industrial capacity, military strength, and population, it could be expected to be an easy win. It wasn't all that easy, but point granted.

Score: 3/5

American War of Independence

American Revolutionary War is another big one, but I really don't see why.

First, suffering - there were some extra taxes white protestant elites weren't terribly fond of. That's zero points, or else Washington, D.C. has as much right to rebel right now.

Second, alternatives. It's automatic fail due to lack of significant suffering, but other countries like Canada successfully got as much autonomy as they wanted without armed rebellion.

Third, prevention of suffering. Also automatic fail, but notice how United States kept slavery long after the rest of British Empire abolished it. It's not at all obvious that it was more representative, or more democratic than other colonies.

Four, casualties. There were about 52 thousand soldiers killed, plus unknown number of civilians (but not too many, as it was one of the rare clean wars). Vast majority of those were from disease, most commonly scurvy, a disease that was already known to be easily preventable. Battle deaths were barely over 10 thousand. That's small by modern standards, but not trivial relative to two million people living in rebelling States back then - just about 2%. I'll grant a point, but it's just being generous.

As for point five, expecting an easy victory against British Empire at peak of its power was foolish. It was a lucky win. No points.

Score: 1/5

Cambodian–Vietnamese War

That covers the usual suspects, which I all mentioned mostly to point out how they fail at being just wars. Now let's look at some obscure wars that might actually succeed.

Cambodian–Vietnamese War is my best candidate. Khmer Rouge killed approximately 1.5 million people out of population of 7.1 million, and weren't stopping. They didn't even limit themselves to their borders, but also raided some towns in Vietnam and massacred civilians there. Overall, spectacular victory at point one.

It's hard to tell what alternatives could be tried, but none are obvious, and due to massive urgent suffering we cannot blame the Vietnamese for not looking too hard. A second point.

Pol Pot was successfully overthrown. Guerrilla lasted for another decade, but it was on much smaller scale. Full point.

Point four is not entirely successful, as China attacked Vietnam to stop their Cambodian operation, with wide variety of estimated, but something of order of 100,000 soldiers and civilians killed. Still, that's far less than number of Khmer Rouge casualties, so point granted.

War could be expected to be easy - huge, well equipped, and experienced Vietnamese army could and did easily overwhelm Khmer army, and it did so successfully. Full point.

Score: 5/5

Most other wars of independence

Cambodian-Vietnamese War proves a just war is possible, even if really rare. Let's go back to other wars for a moment. One very popular kind are all independence/liberation wars, and uprisings against dictatorships.

They usually fail on suffering front - most colonial powers don't engage in massive killing, so the suffering is mostly widespread discrimination, and lack of representative democracy.

They fail on the second front, as countless countries succeeded in getting significant autonomy, or independence, by political action. It worked even against the Soviets, so it doesn't depend on the colonial power being "nice".

As for the third point, most governments that get into power after such wars are dominated by military, or otherwise very far from democratic ideal, and hardly much better than previous colonial powers. You might be lucky sometimes, but I wouldn't count on it.

Fourth - there's usually plenty of civilian casualties. Independence wars, unless the colonial power simply decides to give up early, tend to be really dirty. No points.

Fifth - by balance of power, it's really naive to expect a victory here. Yes, sometimes it happens, not that rarely even, but lucky wins don't count. And there's even lower chance of the new government being much better than the one being replaced. No points.

Related reading: From Dictatorship To Democracy.

Score: usually 0/5, occasionally higher score.

Terrorism

If we go even further from guerrilla uprisings to terrorism, it gets even worse.

Now occasionally like in Palestine the suffering is real, so point granted. In many places like Basque Country it's fairly insignificant.

As for alternatives, political methods not involving violence have been quite successful in countless cases. So zero points.

One interesting thing about terrorism, is that is pretty much always fails. Of very long list of organizations employing terrorism, the only totally successful case is Hezbollah, which kicked Israelis out of Lebanon. Tamil Tigers - who had semi-independent state in North Sri Lanka for quite a long time before being defeated recently, and Hamas and other Palestinian groups, which fought out some very limited autonomy for Palestinians, can be counted as partially successful. Everything else was a spectacular failure. Notice how the successful cases are really guerrillas employing some limited terrorism against mostly military targets as a small side activity, so they shouldn't even count as primarily terrorist organizations. Zero points here.

As for random unrelated casualties, they're present by design. Usually they're not terribly numerous compared to a proper war, so actually we could grant half a point here, but that would be stretching it. And that would only count if it worked, what not a single case of terrorism so far did.

As for chances of success - well, so far they're 0 successes in far too many tries, so no points.

Score: 0/5

Bombing of Bosnia and Herzegovina

Let's look at operations like bombing of Bosnia and Herzegovina.

First, suffering. Over 100,000 people died in the war, about half of them civilians, so it was a pretty good cause. Plus millions displaced, usual looting, rape, and what you usually see in dirty wars. Full point.

Second, alternatives. Some effort was taken, and out of three fighting parties, UN managed to get two to agree to stop. There were also attempts to limit civilian casualties by creating safe zones (openly abused for military advantage, but that's a completely different issues). You could argue about lack of fairness by UN, and if really everything that was possible was tried, but I'll grant the full point. Also military action was quite limited, aimed at getting quick political solution, not full annihilation of Serbian military capacity.

Third, prevention of further suffering - that's an easy point, operation was fully successful.

Fourth, no extra casualties - bombing is estimated to kill over 100 Serbian soldiers, a trivial number, and many more of just Serbian soldiers would have died if the fighting continued. Full point.

As for estimating chances of victory... well, wars are rarely won by just dropping a few bombs, but NATO was overwhelmingly stronger than any of the fighting parties in Bosnia, and it was successful, so full point.

Score: 5/5

Kosovo and other operations

While bombing of Bosnia was really successful, other peace enforcement campaigns like one in Kosovo much less so.

There were around 2000 casualties of the war between Serbian military and police and KLA, but no massive genocide ongoing. The best example of war crimes found was Račak massacre of about 40-45 people, and even that isn't really terribly solid. There was killing due to fighting, and large scale displacement, but that's rather underwhelming, so just half point this time.

Alternatives were under-explored. There was some attempt at negotiating, but it's not really obvious if NATO seriously wanted a solution that would be acceptable to both parties, as opposed to just getting some cover. No pressure was put on KLA, even though they would be fairly easy to press. And during the negotiations all sides issued a joint statement agreeing to "substantial autonomy for Kosovo", and "invited international civilian and military presence in Kosovo", so it's not really obvious what the subsequent fighting was about. No points.

Prevention of suffering was also quite unsuccessful, the war erupted even more strongly after bombing started, with some proper massacres this time. 80% of people killed, and 90% of people displaced happened after the bombing started. 250 thousand Serbian refugees never returned to Kosovo. No points.

Bombing as such didn't kill terribly many people, around 500 civilians, but due to massive increase in fighting on the ground huge number of people died. There was also a later spill-over of fighting into Macedonia, but it wasn't that severe. No points again.

I'm undecided if the last point should be granted or not - previous successes like Bosnia suggested that it would be an easy victory, but it worked only so-so, and large scale ground fighting caused a lot more suffering that was prevented. Half point.

Score: 1/5

Summary

According to criteria of successfully and reliably preventing significant suffering, the only good wars I could find where overwhelming force is used to stop already happening war or other large scale atrocities, and they limit themselves to minimalist goals instead of trying major political restructuring.

All wars where one side is not overwhelming, like WW2, and American Civil War, tend to cause massive suffering, comparable or worse than whatever can be prevented, even if one of the sides is clearly doing a lot of evil.

"Wars to End All Wars" are as old as known history, and include not only WW1, and WW2, but events as ancient as civil wars of late Roman Empire, and probably many before it. Success rate - 0%. Using this argument to support any war is just ridiculous.

All fighting from position of weakness - anywhere from terrorism to full scare insurgency - tend to cause massive suffering, and have even few successes at preventing major suffering.

If you know of any other wars that score very high according to my criteria, or have problems with my rating system (ratings are quite subjective, but it's hard to come up with anything much better), post it in comments.

Chomsky is a total idiot

Which all means that Chomsky is a total idiot, as he is a big fan of the kind of unsuccessful wars from position of weakness like Tamils in Sri Lanka (80 thousand or more people killed with nothing to show for it), and a big critic of successful wars like Bosnia (apparently Srebrenica was "not really a massacre" with only 8,000 or so people killed).

Yes, American government engages in many stupid wars, and causes a lot of suffering. But that doesn't mean that everyone who happens to criticize American government is automatically right and entitled to the high moral ground. Reversed Stupidity Is Not Intelligence, as my favourite philosopher Eliezer Yudkowsky says.