Piaw Na

You point out (correctly, imho) that the death of the development tools market is at least partially to blame for the sad state of affairs today. I think this is largely because developers just don't want to pay for tools. I understand this sentiment, but it's had the unintended consequence that the tools we have are often poor and fragmented. - Joel Webber

A symbolized stacktrace should be all it needs when it comes to debugging. And some strategic printfs. If something goes wrong, the code owner should be able to know these strategic places. Function-level stepping might be useful if you haven't written the code. So, this sounds like a smug remark, but it seems that the market shows that there is no need. - Henner Zeller

I think it shows the lack of respect for good tools in the programming community. Sure, you don't need a power tool when working on stuff, you could always use your hand tools. But the power tools go a long way towards making everything you do easier and faster. That many programmers don't see the point saddens me. The best programmers I know care about their tools, and work on them as well. - Piaw Na

The Java world is *all about* extensible, pluggable symbolic debuggers. (The C++ world, less so.) The fact that these debuggers can't cross RPC boundaries or work on servers has more to do with priorities and attitudes at your workplace than any general issue in the community at large. They do, for example, work very well on Android, mostly because the Android team came from the Java community and cared enough to make JDWP work seamlessly. - ⓞnor

In any case, *tools* are one thing; *symbolic breakpoint debuggers* are another. Many people would disagree stridently with your assertion that "caring about tools" requires people to be interested in stepping through code. Also, I don't think it has anything to do with paying -- open source delivers a tremendous amount of value. - ⓞnor

I'm happy to see other tools. I'm just objecting to the "print statements are the state of the art" crowd. - Piaw Na

The assertion that asymbolized stack trace is all that's necessary for debugging is useful only in the most pedantic sense. Technically I don't even need that, since my programs will generally be deterministic. But why on earth would I want to give up the ability to set breakpoints and step through code?I have always found it enormously useful to do so, and often do so just to make sure a piece of code behaves as I expect it to. - Joel Webber

What about convenient command line tools for RPC generation, database inspection, and library access? How about detailed status mini-websites displaying myriad statistics and snapshots about every running server? Or the ability to grab CPU or memory profiles of anything running anywhere at any time? These are some of the *tools* that people who are uninterested in stepping through code have built. It's all a question of priorities, but you can *not* say people "do not care about tools". - ⓞnor

Yes, lots of hours spent on post-hoc analysis, but many more hours were spent adding "logging" entries into systems that are effectively not much more than print statements. And yes, people seem very proud of it and consider it "state of the art." All this works only because people operating in that environment are very smart, and can work despite those handicaps. Geoworks died for the... more... - Piaw Na

Piaw, "m-x gdb" is not good enough for you? :) :) :) - Peng-Toh

Stack traces, breakpoints, etc. don't capture the state of a distributed system -- and for large distributed systems it's not even practical to capture all the state. They're rarely useful for finding performance problems. They're rarely useful for finding 1-in-a-million problems. So, fine, use them. But if I had to choose whether to give up logging+analysis or breakpoints+stacktraces, it's an easy choice. So looking to breakpoints+stacktraces as an indication of sophistication doesn't make sense to me. - Daniel Dulitz

Completely agree with Daniel, nicely phrased. - Henner Zeller

PT: if it was good enough for me, I wouldn't have described swat. - Piaw Na

Daniel: think about the length of a typical edit/debug/compile cycle, and then think about how much time is wasted doing nothing but adding more logging in an effort to correct problems. I don't think we can do away without logging for the reasons you described, but most problems are far easier to find, and a reasonable debugger and debugging environment will speed up development so you can focus your energies on the hard ones instead of spending more time on the easy ones. - Piaw Na

It's not that big either way if it's just a few retailers instead of Verizon itself... (which, I guess, also argues that the Amazon discounts on the Pre and Pixi aren't that big a deal.) - Andrew C

this must be with contract? when android 2 devices are sub-$200 unlocked, that's when ill be buying - Mike Chelen

With contract. Good luck getting them for sub $200 unlocked. - Piaw Na

The US market isn't really where the action is. Wait till Droid GSM launches worldwide across many carriers. - Peng-Toh

The U.S. is huge for Android 2.0 because many Americans drive everywhere. In Europe, the driving features aren't that necessary since many people take transit. Same goes for Asia. - Piaw Na

The world market for phones is far larger than the market for phones in the US. - Peng-Toh

Yes, but is it larger by revenue? Especially in Europe, where it's hard to dominate because the markets are so splintered, you could easily have to spend a lot of marketing dollars to sell half the phones. - Piaw Na

Any single phone, like Droid, will never dominate in the rest of the world. But the sum of all Android phones can. It will probably never be as profitable as iPhone is for Apple. It is also not clear that US carriers spend less than foreign carriers on marketing. I've almost never seen an ad for a specific phone except for the iPhone. It is always the carrier marketing itself, not the phones. Remember, phones are not locked to carriers. - Peng-Toh

I think he makes good points. - Andrew C from Android

I think his criteria apply to an "iPhone killer" but not for a successful handset platform. He makes one huge mistake when he says b/c Android is free, bottom-feeders will flock to the Android and Google brands. But those _brands_ are not free; Google and Android continue to control them and the experiences to which they attach. As far as "the app problem," he thinks that b/c there is a... more... - Daniel Dulitz

There were a flurry of app updates when the Hero launched (with 1.6) and even more when the Droid did (with 2.0)... - Andrew C

As for branding, my sense is that for now the carriers have most of the branding - T-Mobile's Android phones aren't widely considered Android so much as T-Mobile's, etc etc. - Andrew C

So many holes in his arguments. For example, does the Windows brand get diluted because hundreds of manufacturers make crap PCs or use their own brand on top of the Windows brand? Or should Microsoft rename Windows 7 to Microsoft 7? Based on his arguments, MSFT should have died ages ago. Android right now is like Windows 1.0. You don't have to be vertically integrated to win. The tide... more... - Peng-Toh

Andrew, with every new iPhone there is a new SDK and a flurry of new apps. When the platform changes (1.5, 1.6, 2.0), apps will need to respond. One question is whether non-boundary-pushing phone introductions cause platform changes, and it appears not. Another question is whether it takes a lot of effort to track platform changes, and it also appears not. - Daniel Dulitz

The incompatibilities in the Windows world -have- given it a bad name, Peng-Toh... - Andrew C

That didn't stop Windows from becoming dominant. - Piaw Na

Fair enough. But without the advantage of a major price difference, and with more corporate support for Apple this time, I don't think this will be a straight replay of Win vs Mac. - Andrew C

Andrew, I'm not sure that Windows' incompatibilities have given it a bad name. There may have been grumbling about the platform changes of Windows 3, or Windows 95, but it was quickly drowned out by happiness that the new platform was great. The early days of Windows had enormous incompatibility... versus Mac OS where everything just worked. But that didn't give Windows a bad name; to... more... - Daniel Dulitz

More corporate support for Apple? First I've heard of it. :-) And the price difference is coming. Apple has left a big huge price umbrella below it. - Piaw Na

I haven't seen a successful* high-margin phone come down in price quickly since the Razr... (* word chosen to exclude the Pre.) - Andrew C

And yes, there's room to cut, but that's what Apple will also do with the next revision - bring it in for $200 and cut the 3GS price to be the new lower priced iPhone. - Andrew C

$13k/year is still staggering. And I would suggest that your comparison seems too 'anchored', to use Ariely's term. - Andrew C from Android

No, I went to school at Berkeley in the late 80s/early 90s. My tuition was $4000 a year, except for the first year, which was darn near $20k for non-in-state students. So I'm not anchored to anything except what I consider a bargain. UC Berkeley is a bargain. - Piaw Na

One of the most unique virtues of UC Berkeley is that it's an engine of social mobility. Unfortunately thanks to the dysfunctional California government, that seems to be going by the wayside. :-( - Ruchira S. Datta

Ruchira: this makes it more important to fix California government, which was something already important. The Supermajority rule for budgets have got to go. - Piaw Na

It may be a bargain compared to USC or Harvard, but it's still outrageously expensive IMO. Far more so than $4k/year. - Andrew C from Android

Well, $4K a year was 18 years ago. I don't think you have an understanding of how higher education inflation works . :-) - Piaw Na

Yeah, it really does feel like they're putting the nails in the coffin of the UC system with this. - Victor Ganata

On the contrary, I think raising the price is the right thing to do. The alternative is to layoff faculty! - Piaw Na

I guess what I mean specifically is the death of the Master Plan. - Victor Ganata

It's like regular inflation except much much much faster. - Andrew C from Android

I think if Californians have the will to get rid of the supermajority rule, we can fund the UC system again, and this will only be a slight delay. If Californians don't have the will to get rid of it, then worrying about this will be like worrying about your smoking lung cancer when you're falling off a cliff without a parachute. - Piaw Na

Like Ruchira points out, the Master Plan was a great engine of social mobility in California. Going through the UC system certainly inspired a lot of people I went to school with to try and give back to their communities. I'm also thinking back to all those high school outreaches in underrepresented communities, inspiring kids who might not have thought it was an option into going into... more... - Victor Ganata

The alternative, Piaw, would be to cut out some of the considerable administrative fat at UC. It's hard times for everyone. They should have gone to every department in the system and asked them to identify where waste was happening. Every little dept knows its own bailiwick better than others do. And individual workers wanting to save jobs will be motivated to identify places where waste and redundancy are happening. - Spidra Webster

UC is supposed to provide higher education to the children of the taxpayers of CA. That was its primary purpose. And it has strayed from that purpose for many years, providing cheap R&D to for-profit corporations and having the undergraduates be merely grist for the grad & research mills. - Spidra Webster

Spidra, just this fall they began an initiative euphemistically titled "Operational Excellence" to do just that. I'm going to wait and see on this one. I fully expect that they will find a lot of places where processes can be streamlined, but I don't know whether the will exists to follow through on implementing the recommendations they end up making. On the other hand, there are no easy choices at this point. - Ruchira S. Datta

I think it's really crazy to blame the UC. If anyone's to blame, it's the feckless voters of California who've never been willing to make tough decisions (i.e., hike taxes on themselves) for the sake of long term success. While I think that there's some administrative fat, I don't think it would be even sufficient to bring the tuition hike down by even 1%. The reality is, we're not... more... - Piaw Na

Piaw, I'm not one of those people who votes down every tax. I'm in favor of taxing ourselves to provide services to ourselves. But it's not crazy to blame UC. UC has done a LOT of bone-headed shit and the bone-headedness exists on many levels. I'm a graduate of UCLA and I worked for UCB for a while so I'm basing this on my experience as well as what I read in the papers. http://www.google.com/search... - Spidra Webster

Sounds like we didn't fund enforcement/auditing either. Talk about penny-wise pound foolish. - Piaw Na

Haha... the first comment: "Wow, a sample size of three, and you made up two of the numbers. Even by the standards of half-assed hand waving arguments, that's pretty impressive." - Peng-Toh

That comment does kick ass. - Andrew C