Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!


Forgot your password?

Comment Need to get better electric cars (Score 1) 332

I've read this far, and my main conclusion is that people are more concerned with looks than function. AFAIAC, a car must incorporate these eight important things:

1: It needs to be reliable
2: It needs to be affordable to buy
3: It needs to be affordable to run
4: It needs to meet the cargo / passenger requirements
5: It needs to be comfortable
6: It needs to provide sufficient practical amenities
7: It needs to be safe
8: It needs to pollute as little as possible

How it looks is wholly irrelevant.

Comment More in heaven than the app store (Score 1) 249

You do realise that all this announcement means is that...

It appears it may mean that some or all 32-bit APIs may be going away. So that's somewhat of a serious concern for me as it would break Qt's OS interface. Of course, there's always the VM route, which I expect to keep working until/unless VMWARE goes away. Oh, and Parallels, too. And the others.

you won't be able to sell the 32 bit version of your app through the Apple app store.

I don't sell it, I give it away, and I have (and plan to have) no dealings whatsoever with Apple's app store. So no, not really a realization for me. :)

And honestly, building for 64 bit only allows you to jettison so much ancient cruft, you should be pleased. ARC and the non fragile ABI are worth it by themselves.

ARC: I manage my own memory. I have no need of ARC. I'm glad it's there for others in some versions of the OS, though. The problem I can't get around is that the OS itself leaks like a sieve. Still true as of 10.12. I always hope it'll be better with each new release, but not so far. If you want to unmask the worst cases of it, just turn off VM for a week and watch your memory consumption. Pretty sad.

My kernal_task is currently consuming 2.87 GB. By far the biggest wastrel. 14.52 GB of memory in use; except, no, it isn't. Also 9.25 GB cached, which, meh, I'm not too worried about performance boost caching, but some of that stuff hasn't been used in weeks. I'll have to reboot to clean it up, or fling purge at the system, which cleans up waaaay too much.

ABI: The problem here is that it means abandoning support for earlier OS's, and there's no actual reason to do that at this time. Particularly in that there is operating hardware in the field that Apple won't let run the new OS levels. Quite arbitrarily, too.

I will confess though... if Apple and the Qt people actually made a serious effort to clean up all the reported bugs instead of only the ones that have X>N reporters, they might be able to tempt me into a new build using the latest and greatest.

Unfortunately, both of them are a lot more interested in "shiny" than they are in "works like we claimed it works." They both habitually leave broken software systems unrepaired, while leaping into the future with the New Shiny. And besides... if they did fix all those bugs, I expect I'd get the benefit without building anew. The reason I say that is that builds that are 100% under Windows are less so under OS X. It seems very likely indeed that the problem is specific to OS X, especially considering how badly the OS leaks for pretty much everything app it runs. Including itself.

Comment Say, Horatio... (Score 1) 249

nobody was going to target new applications to it anyway, regardless of Apple removing 32 bit support.

Wrong. I do. I build my SDR app - which is a very complex thing - for 10.6.8 and up, and XP and up, is tested on all subsequent releases of both OSX and Windows, and (among other CPUs) on core 2 duo machines prior to release level issuance. It takes awhile, but mostly its rote testing using only slightly updated test suites, since I am long past requiring new OS calls.

Unless you actually need something from a later OS or CPU, there's no particular reason to target said CPU specifically, or to artificially inconvenience or lock out users of earlier hardware/OS (as Apple has done... the "Sierra won't run on these machines" business has been shown to be an artificial limit by software patches around the "oh no you don't" code that enables it to run just fine.)

In my case, a C2D provides (just barely) enough horsepower to run my app. Since it can, I see to it that it will. Further, since I have a 17" C2D macbook pro that still works fine (albeit running 10.6.8), it is always possible that some of my users do, or a C2D mini, as well. Many are still running 10.6.8 for certain, as I get a lot of mail / forum posts thanking me for keeping them covered. My app doesn't have a huge number of users - a few tens of thousands - but there are more than enough that my dev process has to cover a lot of OS levels and CPU types to be sure I'm not kicking someone downhill without any good reason. And yeah, I still carry that laptop around, and frankly, it pleases me that my app runs adequately on it.

Apple's move to 64-bit code is interesting. But unless the CPUs involved can't run 32-bit code, I rate choosing to obsolete applications people have bought or otherwise obtained legitimately to run on previous versions of the OS as pretty high on the scale of "Hey! Let's be fuckheads!" It's not like Apple is short of resources. If they want to support those apps, again barring impossible to address CPU-specific issues, they can. Which boils down to "Apple didn't have to screw you; Apple chose to screw you."

Comment This is a complex, ingrained problem (Score 1) 608

If it's unwanted then it's been delivered badly

That doesn't follow. Unwanted can have many root causes. And the "calculation" involved in an advance is generally "I would like to know this person." Wait too long, you may lose to someone quicker on the draw. Wait, and you miss opportunities, even when you eventually succeed; every day, once gone, is gone forever.

If you are a desirable person, then you're going to be desired.

Your point about power dynamic vs. advances is very (probably most) important and will remain relevant even if women learn to shoulder an equal load in the "making advances" category. This is a potential problem in any structured authority system. Generally speaking, the most effective solutions involve only starting relationships at one's own level of authority.

However, this is not a panacea. My SO's father married his own secretary. My SO was one of my students. Both relationships are very long term (many decades) and have proven to be quite healthy and strong. Authority mismatch is not always an impediment or a bad thing; but the potential for it being a bad thing is much higher than with an equal level person / peer.

Best thing, IMHO, is that women learn to be initiators so that men don't have to do it the majority of the time. Then we can expect to be approached if we're desired, and we won't have to always be approaching women who may not want to be approached.

And yes, of course, politeness, consideration, and a willingness to accept whatever answer is given are important for all involved.

No one should support rude approaches. But when they occur, and they will, don't escalate. Be polite and withdraw as quickly and effectively as possible.

Comment I'm not talking about just ads (Score 1) 257

How in the exact crap it helps is that the links to paywalled sites stop working - they fail instantly (I actually generate a "this URL is disabled in hosts" page... OSX has a built-in webserver) and you're not deluged with crap like "pay me." Hit back (or whatever key combo your browser uses for back), you're back at google. No screwing with the junk sites, basically an immediate response, doesn't even have to hit the network.

Not only does this keep the offending site from sullying your browser, it doesn't even give them the courtesy of an Internet reach-around: the click never reaches their servers.

If you don't want paywalled sites, the best move is to vote with your wallet and you clicks. So don't go to them, even accidentally. This accomplishes that. Sites that misbehave are dead to you.

Comment Quality can mean other things (Score 1) 257

The other "sources" of news read and summarize those original articles, often with a much lower quality level.

When Sanders was running for the Democratic nomination, the NYT, Washington Post, and Guardian - which are all more or less liberal operations - "failed to adequately report" (by which I mean intentionally downplayed) coverage of his campaign.

I don't need "news" that uses bias to push its own agenda. That's not news. That's propaganda. "Quality level" isn't just about writing well. It's also about reporting without bias. The more bias there is, the less the actual value of the source. It really doesn't matter if they have reporters on the ground if what gets in the news source is all triaged into viewpoint-addled-mush.

I do want to know what's happening. Informing me of that in a completely even-handed manner, IMHO, is the news media's only legitimate job. I'll form my own opinions on it. Since the news media is doing a very, very poor job of avoiding exactly that kind of bias, I'll keep looking elsewhere. Even a poor-grammar, poor-spelling summary of one of those articles is better if the MSM slant is edited out.

Comment Choice (Score 1) 257

You always have a choice. My choice is to add problem sites to my hosts file as;

This simple move serves to solve the paywall (and autoplay video, and spammy page covering ad, and junklink [try this one weird trick!], and we-don't-allow-commenting-sites, and similar) problems.

As far as the news goes, many non-paywalled sources remain. I use them.

Perhaps someday the news will move from a for-profit model to a for-the-people model. That'd make it much more worth supporting. Unless/until then... pffft.

Comment Reading Tips (Score 1) 179

If you want to complain no one has answered your take on things, you should bring reading skills and capacities above the third grade level (and I may be being unkind to many third graders there) in case someone actually, you know, answers you.

It is the height of rudeness to press for a response, and then fail to bother to read it. I'm absolutely unwounded – this is a profoundly obvious case of consider the source – but I certainly know better to continue engagement with demonstrated social cripples.

That's /thread. Anything further is just aimless scratching of itches.

Comment Re:Going in seems so pointless (Score 2) 226

OK, you get free food at work. You have a comfortable lounge.
  You are also within walking distance. I could see a lot of people actually enjoying working in the office with those two perks.

Imagine all the poor bastards who have to either pack their lunch or waste a chunk of their paycheck on take-out food, burn 30-60 minutes and a few gallons of gas in traffic, only to go to a cube farm with bad coffee and no place to relax.

So it is notable that you STILL prefer working from home!

Comment No, lets confuse people with facts. (Score 4, Informative) 79

Facts like people were writing and widely sharing open source code well before any of this nix-specific event came along in 1984. Go look at early issues of Byte, Kilobaud, Dr. Dobbs and so forth and so on.

I had source code of my own published and shared nationally in November 1977, and I wasn't anywhere near the first.

This kind of hype reminds me of Apple/IBM/Radioshack/Commodore and fans variously claiming they were the "roots" of the PC market. Look into the history of Altair, Imsai and SWTPC, for instance. Not saying who was first, I'm not sure by any means -- but I am sure who wasn't.

Comment Re:And so... (Score 1) 68

One more thing: Bugs abound in Amazon's Echo. As well as some fairly annoying designed-in, unfixable shortcomings. MyCroft is brand new. A little late, but not horribly so. I'm perfectly willing to give it some time and effort. Seeing as how I can change it to address any shortcomings that really concern me; I can't change the Echo, nor can I get it to do the things I really want it to, nor the way I want it to. And yes, I own the Echo.

Comment Not desirable (Score 1) 68

But it will never be as good as the competitors that learn by users sharing data

It'll never be as good at sharing data with people you don't know and shouldn't trust, anyway. But the fact that you, a human, can do excellent speech-to-text, means that local software and hardware can get there eventually. And it'll very likely get there faster if that's the goal we preferentially chase.

And we should.

Comment Some GPL things (Score 2) 179

I'm not the AC, but I have some commentary on the issue.

My position is that the GPL is inherently toxic to the economy, specifically to the portion of the economy that consists of opportunity to earn from creating software. That very economy is important to those of us who can program at a level where we can create commercial products. My tiny sector of said economy shall not be poisoned thus.

And before someone says "yes, but you can charge for support", the way I see it, the optimum path is to write software that doesn't need support. Why?

First, because that's best for the consumer: good docs, good performance, high reliability, minimum bugs, secure. I prefer to charge for the product, fix it for free if it breaks (after all, I sold it with the idea that it did X under conditions Y...Z; I maintain that's an obligation I must address if I possibly can), and charge for actual new features V...W. Likewise, if I write proper documentation (which I also maintain is my obligation), I don't need to be telling people how to do X, because I already told them how to do X. Not that I don't end up pointing people to the docs on how to do X, but that's not a failing of mine I have to be concerned about having foisted off on the consumer.

Second, because it's very good for me. If my users come to think of my products as reliable, well documented, fixed for free quickly if broken, and overall functional as advertised, they're a lot more likely to come back to me than if I continually charge them for doing the above things poorly. I'm not guessing here: This has worked extremely well for me.

What isn't good for the consumer is when Joe(anne) programmer looks at an idea, runs into the GPL, and goes, "no, not going to give my ideas to my competition" and wanders off elsewhere.

Licensing code under a non-copyleft free license means you are fine with someone taking your code and building a commercial product with it, and never giving it back in any way or form.

You contradict yourself. The opportunity to get a return is absolutely there. If Joe(anne) builds a commercial product, I have the opportunity to purchase that result; to encourage and benefit the programmer(s) who built the product in direct return for the benefit they provide me, and perhaps to benefit yet again further down the road as they consider the opportunity to earn more. Likewise, if it's me doing the programming, I'm one heck of a lot more likely to consider it if it means my family and I get to eat because I created something useful to someone than I am if my earnings are now compromised by having handed my work to the competition.

If you're okay with that, how come you're not okay with someone who does give it back, except in a way you can't use?

Because it represents an opportunity to earn; because it represents an incentive for the earner to make something for me, right along with a reason for them to do it. And I can use their work. Commercial products are usable. Writing code isn't the only use of code. Using the programs the code is incorporated in is a significant benefit. I can't be writing every program I use. Commercial products have been a huge boon to my career and my family.

Now, none of this means that I object to someone else's decision to license their stuff under the GPL; that's their code and they are absolutely free to license it any way they see fit. What it does mean, though, is that spotting a GPL license means I'm going to walk away from whatever it is. It won't benefit me in my commercial applications unless I give away my own inventions, and it has the potential to harm me by giving my work to my competition, so: not having any.

Personally, when I make code public (which I've done a bit of), I make it public, as in, you want to use it, copy it, learn from it, incorporate it in your commercial product, you just go right ahead, and I am happy for you. Because it pleases me that someone would benefit from code I publish. Financially, learning (and so hopefully financially in the long run), whatever path they want to take.

Rewards "as fame": You can't eat fame. 15 minutes or otherwise. I would encourage programmers everywhere to think through the financial implications of giving away their inventions. Implications that land on others as well as you. If you want to do the most good, as I see it, the most likely way is to open that door as wide as possible, so as to not hamstring the recipient(s) of your gift(s).

TL;DR: I won't use GPL source code. You can use GPL source code, and I don't care. The reason I won't use GPL is because it poisons commercial code with a toxic requirement to publish your own inventions.

Slashdot Top Deals

Have you reconsidered a computer career?