If google starts a mail service, is it really to take on Yahoo?
Hey, I long ago started a mail service (for me and my wife) on my home machine. I've also installed qmail on a couple of work machines. But I don't see any articles about how I'm taking on Yahoo.
If your primary requirement is a cheap, fast way to get a lot of machines up and running, and and security isn't at the top of your list, then a "distribution" is the way to go. You still have to decide which one, and that depends on how your people will be using it.
If your primary requirement is a way to get a few machines (e.g., a server farm) up and running so that it is secure and reliable, then what you want to do is download all the source and build your own systems one piece at a time. You make sure you understand each package's configuration and security needs before you let it go live.
There is no best way for everyone.
Historically, linux has mostly been the favorite of people who want distributions. It tends to come with everything that compiles and passes the "make test" suite. It's no surprise that linux distributions should contain packages that are insecure. That's what happens when you let everyone throw their favorite apps into the bin.
Historically, the BSDs have been the favorite of people who have serious security concerns. It's no surprise that they should lack the full bag of bells and whistles of linux, and it's no surprise that they should have few security issues.
It does seem that, if you take the build-it-yourself approach with linux, you can get a well-tailored machine that's also fairly secure. But you will have to do the work and spend the time learning about the issues.
Manufacturers of windows for building just might be at risk soon. There are a number of companies now selling transparent LCD displays. There are several that can be used as real windows in walls. Google for "transparent LCD display" for information. (You'll have to wade through a lot of technical details to find pretty pictures.;-)
So arguing that windows in walls and Windows in a computer display are different subject areas might not work much longer. It's possible right now to have a window in your house that is a computer display. How practical this is, I'm not sure, but you could do it.
So shall we start a pool on the date of Microsoft's first C&D letter and/or lawsuit when someone does this?
Of course, it might have happened already, and we just haven't heard about it. And imagine the rage of MS's lawyers if such a window were controlled by a computer running Lindows...
BTW, if you think I was just joking, google for "the THE operating system". Right now, there are 418 hits. A few are typos, but most are about the THE OS itself.
But the attempt to register "THE" was, as I recall, reported as a bit of geek humor. Apparently the USPTO got the joke, laughed with them, and turned them down. But there's a serious question of whether their examiners would get it today.
You might also want to google for "English programming language". There are fewer hits, and most are just uses like "English programming language instruction". But a few are about the programming language, a dialect of SQL.
Is someone going to trademark the word "The" next,...
Actually, someone tried that, back in the 70's as I recall. It was the name of a new OS. The USPTO didn't accept it then. They just might today.
So why don't you give it a try? Let us know how it works out.
(There was also an attempt to register "English" as the name of a computer programming language. That wasn't accepted, either. It's worth another try, too.;-)
Yeah; I recall back in the 60's, when I was the undisputed chess champ in my high-school crowd. After a while that started to get old. I mean, you can either become a pro chess player, one of the worst jobs in the world, or you can move on to something else. I did the latter.
Actually, I started playing the piano a lot. The second-worst job in the world. Now, several instruments later, and with pianos transformed into a zillion descendant instruments, it still hasn't grown old.
But I also became a computer geek, so I can earn money. And I got into network programming, so I can take partial credit for the imminent destruction of the recording industry, which roughly a century ago took over music and made it nearly impossible to earn a living as a musician. We can all be happy about the revenge that we are starting to enact.
Along the way, I found the explanation for why people don't grow out of music:
Mother: Son, what do you want to do when you grow up:
Son: I want to be a musician.
Mother: Well, son, you have a choice. You can grow up, or you can be a musician.
It might be worth noting that linux was started in Finland, which is not exactly a 3rd-world country, but it's not what you'd call wealthy or powerful. Linus Torvalds has written about his motivation for for starting the project. One of his main reasons was that he wanted to work on the innards of a unix system, but he was a poor student who couldn't afford a commercial system. He could afford a PC clone, so he got that and a copy of the POSIX specs, and started coding. ("How hard can it be?";-)
This is likely a good model for the poorer parts of the world. Usable computer hardware is rapidly getting cheaper, but commercial software isn't. But we're talking about a part of the world with a lot of unemployed people, who have time on their hands and a strong desire to make a better life for themselves.
It doesn't take a great genius to imagine where this might lead. Especially if the information they need becomes available on the Net. And note that a number of schools are starting to put a lot of tech coursework online. MIT is no longer the only one with this sort of project underway.
(Hey, remember the article here recently about how MIT's Open Course Ware project was being implemented in India?;-)
But the controls don't seem to fit this model. On both machines, I only see a single "Printer Sharing" thingie. It's in the "Services" tab of the "Sharing" window. There is nothing to indicate which printer(s) should be shared.
This is significant because one of the PBs does have a second printer plugged into its USB port, and right now it sees both that printer and the airport printer. I did some experimenting, deleting all the printers for both PBs, and then adding them to PB2. Printer Sharing was unchecked. When I asked PB1's Print Center to find USB and Rendezvous printers, it failed. I checked "Printer Sharing" in PB1, and then PB2 was able to see both the Rendezvous (i.e., airport) printer and the USB printer on PB1.
I then closed PB1. PB2 lost sight of both printers within a few minutes. This makes sense for PB1's USB printer, but not for the airport printer.
So if I want PB1 to see the USB printer on PB2, PB2 must have printer sharing turned on. This makes sense. But then PB1 can't see the airport printer if PB2 is closed.
Note that this is the reverse of the original problem. Now PB2 has to be running for PB1 to see either printer. I'm guessing it's because I deleted all the printers and then had PB2 find both printers. So now PB2 "owns" both printers, although for the airport printer this makes no sense.
This has gotta be a bug, right? Did they actually design it to work this way?
I wonder if there's some place at Apple's tech support that would give a reasonable explanation of all this. But I'm afraid to call them up about it, because of the disaster the first time I tried. (I'm trying to get some work done, and I need the Internet connection to keep working.;-)
However, on linux, I do know what I'm doing.... This is relevant how, exactly?
Read the title of this thread. An answer of the form "I use X rather than Y because I know X and have had troubles learning Y" is about the most direct, relevant answer one could give.
Of course, this is usually used as a coverup for a refusal to learn. I was letting readers know that I am willing to learn, and I'm not using knowledge of linux as an excuse to ignore a new system that gets lots of good reviews. I put out some good money for a Powerbook. But, like any computer system, it's somewhat of a complex beast. I can't learn it overnight. If that's what's expected, then it's clearly not the right machine for a dummy like me. And maybe I should just stick to a system that was designed for a dummy like me.
... Apple tech support, who are obviously braindead morons.
Well, now; I think that's a bit harsh. It would be more accurate to say that they're the usual sort of harried, poorly-trained low-rank workers who barely have a handle on what they're expected to do, and are mostly working from scripts that handle only a list of pre-programmed cases.
This isn't a criticism of the support people themselves so much as a criticism of Apple (and the rest of the industry, for that matter).
I've found this as a useful answer to the suggestion that you should go with commercial vendors because you need their support. I counter with the suggestion that in fact linux (and *BSD) provides better support: public newsgroups and mailing lists. A few simple tests will show that this typically works a whole lot better than even the best of corporate tech support operations.
Actually, of course, both can be either good or bad, depending on how you ask the question, who you happen to get answers from, and the phase of the moon. But if you talk to people who routinely deal with both, you'll tend to find agreement that the random gaggle of "free" help on the internet is usually better than the paid help you get from corporate support.
Much of the reason is that the "free" help is usually from relative experts (at least in a narrow topic) who enjoy showing off their expertise. They may be insulting and grating, but they often give useful answers. The corporate help is usually low-paid, inexpert hired help who are just trying to get their job done without too much grief from customers or employers. They are friendly puppies, but often don't really know much.
If you're trying to get something to work, an arrogant jerk who solves your problem is usually better than a friendly dummy who doesn't solve your problem.
And back to the original question: "Why use Linux at all when there's Mac OS X?" I think that all this is a fairly good answer. My experiences with Apples tech support did convince my clients that OSX (and the Apple's airport) didn't look like a good bet for the company network infrastructure. They don't want warm, fuzzy feelings; they want to get answers fast. The arrogant, grating experts in the linux herd of cats are a much better bet if you want to solve problems fast.
This isn't necessarily a criticism of Apple or the Mac. Their market is pretty clearly people who want a computer that "just works", and who don't want to bother their pretty heads with the details. I would (and do) recommend a Mac to such people. I have one myself. But I'm frustrated by the difficulties in getting past the user-friendly interface. What I want is a programmer-friendly interface, and that's something rather different. If that's what you want, then you have a good reason to use linux even when OSX is available.
(And yes, I do know about the Mac Terminal. I have four of those on my PB's screen at the moment. But, unlike linux, and like MS Windows, there are a lot of things that don't seem to be available from a shell command. Or maybe they are, and I just can't find them.)
Well, it seems to me that you've given very good support to my comments.
Recall that the question was:
Why use Linux at all when there's Mac OS X?
Now, it's clear to nearly everyone here that I'm obviously an idiot who doesn't know the first thing about using OSX, and my problems were all because of my own stupid mistakes.
But anyone who is new to OSX is going to make lots of mistakes. There's a lot of stuff to learn, and no mere human can learn it all in the first hour (or month). If this means I shouldn't have been using OSX, well, OK; I shouldn't have been using OSX.
That seems a reasonable answer to the question. An admitted linux geek sticks his neck out and asks questions that make it clear that he's ignorant. So he gets the treatment he obviously deserves. Lesson: He should stick with linux and not bother with OSX. It's a simple, elegant answer to the question.
Myself, however, I like to learn things. That's why I bought the Powerbook. So I think I'll thumb my nose at the advice here, and continue to learn what I can. In my copious spare time. There are some good ideas in there.
When and if I ever become as familiar with OSX, I may start advising my clients to use it more. However, considering that this learning must be done in my spare time, while my day job involves linux development, this may not be real soon.
To put it all in some sort of perspective, I mentioned one apparent screwup within OSX to a Mac-using coworker, with a comment like "Hey, they f**ked it up even worse than in linux or solaris." His response was "Yeah, but they f**ked it up in a much more elegant and pretty fashion."
We boch chuckled and went on with the problem at hand.
We have two powerbooks. When PB1 is closed, PB2 loses contact with the printer plugged into the airport.
I contend that this makes no sense at all. PB2 should be able to talk to the printer without PB1 being present and powered up. PB1 is not in the data path PB2 -> airport -> printer.
I'm assuming that this is somehow because I first used the printer from PB1, so PB1 has some special relationship to either the airport or the printer. But I haven't found any evidence to support this. I've sat the two PBs next to each other, fired up the airport and printer admin tools on both, and compared the fields one by one. They are identical. There appears to be no difference in the ways the two PBs see the airport or the printer. But PB2 still loses the printer when PB1 is put to sleep.
This has gotta be a bug.
And yes, I did ask on a couple of newsgroups. I got "replies" that made the same mistake: They tried to explain why my PB can't talk to the printer. But that's not the problem at all. Both PBs can see the printer. The problem is if I turn PB1 off, PB2 loses contact with the printer.
What's worse is that it's inconsistent. The time between turning PB1 off and PB2 losing sight of the printer varies from minutes to hours.
if you're finding OS X slow, opaque and frustrating, it's because you don't know what you're doing, you just think you know what you're doing.
Heh. You're right there.
However, on linux, I do know what I'm doing.
Part of the reason for getting an OSX system (and so far refusing to bring up non-Mac things like X-Windows etc) is that I wanted to get familiar with it.
The problem here is that, as others have pointed out, when I did the obvious thing and asked Apple's CS for help, I got an idiot-level runaround. They didn't say "Hey, this guy knows how to run linux gateways and firewalls; maybe I'll bounce him to an expert level. They kept "helping" me with something that wasn't related to my original problem (getting airport + printer + powerbook to work), and bollixed up my network as a result. They didn't mention DHCP to me; I discovered it myself. And they told me quite clearly that I couldn't expect to use their equipment in a mixed-vendor network.
I'm starting to agree with them.
With linux, OTOH, when I ask CS (e.g., Red Hat's), I get intelligent answers. Or sometimes I get RTFM, and usually a pointer to the FM or some equivalent online docs. So I can learn what I don't know.
I freely admit that I don't know everything there is to know about linux, and I'm even farther from that with OSX. So far, my experience with linux is that finding answers sometimes takes time, but I can find them, and among the arrogant jerks there are a lot of helpful people. With OSX, my experience so far is that I get a lot of "... for dummies" sorts of answers, but very little else. Finding answers has been rather slow compared to linux. So I'd recommend linux for serious networking applications.
To put things in perspective, we also have a couple of MS windows machines (which my wife needs for her work), and even a cool wifi-enabled PalmOS machine. They're all talking to each other, though not always amicably. I'd say that the OSX system is far, far better than the MS boxes for nearly everything. I've recommended a Mac to lots of people who have expressed frustration with their Windows boxes. I'll continue to do so.
My summary at the moment is that I'd strongly advise linux and/or *BSD for any serious networking, server, or other infrastructure situations. I'd recommend a Mac for any computer non-geek who wants a machine that works and is fun to use. I'd recommend MS systems for masochists. (And that wireless PalmOS gadget is a really cool toy. Now if it had a cell-phone built in, too...)
Well, one recent example from personal experience:
I was having problems with our two powerbooks (OSX) talking to a printer that was plugged into the airport. Fine, I thought, I'll call up Apple's CS people and they'll figure it out.
Hah! The fellow I talked to had me go through the machine's internet connections. Why he did this isn't obvious. The printer should work even if the machines aren't connected to the internet. But never mind; by the end of the session, my previously-working internet connection was bollixed up so badly that nothing could communicate. The call ended with me starting an hour-long validation of the system, and then I was supposed to call back.
When it finished (not finding much, and the internet still f**ked up even after a reboot, I called back. The fellow I talked to that time went through all thee tthings from the first call. Then, when he discovered that the connection to the internet was through a linuux box, he demanded that I reboot it. I balked at that, and let him know in no uncertain terms that this was utterly unacceptable.
I did agree to disconnecting the linux box from the chain, and hooking the airport up to the cable modem. He still couldn't get it to work. And still, nothing at all had been done with the printer. He left me with yet another system validation task running, and by the time it finished, it was well past their closing time. And he'd made it clear that they wouldn't be able to help us as long as there was a linux box on the LAN.
I went into personal search-and-recover mode, and got the internet hookup working (through my linuux gateway, of course). I couldn't have done it through the OSX net tools that I found; I mostly used the diagnostic tools on the linux box to tell me what was happening.
One thing I learned from the linux tools was that the airport was running a DHCP server. I hadn't yet seen anything in the Apple tools that told me this. Its address range overlapped with the range used by the linux box's DHCP server. Those who understand this paragraph will understand the problem.
I found the DHCP controls in the Apple software and moved the airport's DHCP server to a different subnet. A bit more judicious configging, and I got my internet connection back again. I even got the printer to work, sorta, though we do still have a problem that when I close my powerbook, my wife's loses contact with the printer. (I've asked about this on some newsgroups, with no answer. I'm afraid to call Apple about it.)
The major effect of all this was that I wrote it up for the people I'm consulting for. They had been looking into trying some OSX boxes as part of their corporate network infrastructure. After my report, they dropped that idea. A couple of other people did a bit of judicious inquiring, and gave independent reports that I hadn't exaggerated a bit.
Nonetheless, I have a powerbook in addition to two linux boxes. I think that OSX has a lot of good things going for it. But I wouldn't recommend it for some things that are routine on linux. One is handling networks with a mixture of different machines.
Much of the problem is the "Don't worry your little head about it" attitude of the Mac community and Apple. That's fine if you're a non-nerd. But if you know what you're doing, and you want tools that give you access to all the innards, you'll find OSX slow, opaque and frustrating. With linux, you'll find a community that is willing and able to help you (if not always in a friendly manner;-).
The hardware is nice. If I could get a 17" powerbook with linux (with drivers for all the hardware), I'd find it really useful. I think I'll start bookmarking reports like this one, and maybe invest in one of these machines in the not-too-distant future.
[Microsoft] have 95% of the browsers and webpages are coded to whatever crap IE renders whenever necessary.
You do have to be a bit careful with repeating this sort of claim, because many of the statistics you'll read fall into the "87% of all statistics are just made up" category. It's very easy to interpret web-server logs and sales figures in radically different ways.
Thus, I recently installed the latest Opera on my Powerbook, as part of my collection of browsers for testing web pages. I checked with a nearby server log, and its default id string is:
"Mozilla/4.0 (compatible; MSIE 5.0; Mac_PowerPC) Opera 6.0 [en]"
It's easy for an unscrupulous marketer or a sloppy programmer to interpret this as Mozilla, Netscape, IE or Opera. It can be (and is) counted as any of them, depending on what you want to "prove".
As for sales figures, I like to mention the Dell box sitting in the row of computers on the shelf next to my desk. It's running RH linux 8.0 at the moment. There's no trace of Microsoft software on it. But it was delivered with MS windows, so both Dell and Microsoft count me as a Windows customer. And, since I haven't used their customer support for it, I'm obviously a happy, satisfied customer. This box is, of course, also counted in any industry figures as having an IE browser. Wrong again.
We really don't have any good numbers on the scale of this sort of misrepresentation. Any numbers you see are probably in that 87% of statistics that are just made up.
One of my more fun examples of padded browser statistics: I also have an old W98 box on my shelf. It's used for web testing, of course, and is usually turned off. But when I do use it, one very real problem is that it rarely survives more than a dozen web pages before it hangs and I have to reboot it. When I restart the browser, I usually have to re-fetch a number of the pages that I was working on. Those fetches go into the statistics, of course. I'd claim that counting reloads caused by browser crashes is totally bogus. This is padding your numbers in the worst possible way.
We have no way of knowing how often Windows users have to do this, because there's no reliable way to distinguish such downloads from others in a server log. And it's not just IE; the Firebird browser on my linux and OSX boxes hangs regularly and has to be killed and restarted. Those reloads are also bogus numbers in the statistics.
Any figures you see on this topic are to be taken with a large grain of salt. They are mostly PR, not facts.
There's also the theory that the extra filters are to produce pictures that look right to the eyes of the visiting aliens that are helping NASA. I mean, how would you like to have pictures that cut out the red portion of the spectrum?
For that matter, there are plenty of earth species that can see frequencies outside the human visual range. We are always getting complaints from our conure and cockatiels about the poor color quality of our computer screens. The ultraviolet is almost entirely missing.
I suppose the cockroaches would complain, too, but they only use the computers after we've gone to bed, and who knows what sort of mods they've made to the displays?
In any case, it's not easy to get a really accurate color display from the current primitive crop of manufacturers on this benighted planet. Not to mention the difficulty of getting good UV and IR pigments from the inkjet-printer manufacturers. For the forseeable future, those visiting aliens will just have to order the inks from home, and you know how slow delivery in this part of the galaxy can be.
Why, no, not at all. If you look closely at/. stories in the past few months, you'll find others from Al Jazeera. Also, if you go to new.google.com, you'll find them well represented in the top stories. In particular, they've had good summaries of a lot of technical and scientific stories.
You'd think that the Al Jazeera folks are trying to be a respected news source or something.
(I was trying hard not to say "fair and balanced.;-)
Their Middle-East reporting makes for interesting reading, too. They often give you a somewhat different slant than Western news services.
think the point, us nerds would like to be able to hack our phones like we hack on our computer systems.
Close, but I'd express it differently.
I don't really want to hack my phone. I want to replace it. What I want to replace it with is a PDA-like gadget that will fit into my pocket, and be able to talk to both the phone system and the wireless Internet. And I want to be able to use it like a computer, i.e., it must be programmable.
An important part of this is the "will fit into my pocket" phrase. Most PDAs flunk this test.
I have in my pocket what looked like a good start a couple of years ago: a Kyocera 6035 "smartphone". It has a lot of problems, though. One is that the web browser works over IP that's PPP over the phone system. It's sloooooow, and you get charged full air time for the connection, even when no packets are being passed. This is far too expensive to use it routinely.
My wife has a new Tungsten, that comes with wi-fi networking built in. But it doesn't do phone calls. And it's too big for my pocket (though it does fit into her purse).
Also, these are both PalmOS. After a couple of years of exploring their development stuff, I find that it's really not worth the effort. Doing even the smallest thing takes forever, because you just can't debug the stuff. The slightest error freezes everything, you have to reboot, and you have no clues as to what went wrong. There's nothing at all like gdb available. And most of the internal working are invisible and undocumented to outsiders like me.
To be credible, I'd want something that I can actually program. This means that the innards should be documented, and there should be places to ask dumb questions. PalmOS doesn't even come close. I haven't tried Symbian, and I do wonder if it's better.
But it's pretty obvious that a pocket-sized linux gadget with both wi-fi and cell-phone hardware would do the job quite nicely. Nothing hidden there, and lots of places to ask dumb questions (and get RTFM answers, for which I can ask "So where's the FM for that?";-)
I'm not dogmatic about linux, though. FreeBSD would be nice, too, and OSX would be pretty good (though parts of its innards are blocked by brick walls).
I also wonder about iTron. Is there any way for a US resident (with little Japanese) to get meaningful experience with it?
I do think it might be easier to use one browser at a time. Unless, of course, Firebird does some horrible and evil Windows-like "only one instance of this program can be open" thing,...
But that's exactly how mozilla (and the original netscape) have always worked. Actually, on my linux box, there are always exactly four mozilla-bin processes, regardless of how many windows I have open, but only one of them is ever using the cpu at any given time. On my Mac (OSX), top shows one "mozilla-bi" process and one "MozillaFir" process. When I do something with either, their cpu usage goes up. But it doesn't matter how many windows I have open or which window I'm using; there is only one active process for each browser.
It's good that Firebird split off the mail/news task into a separate program. At least, when the browser windows all hang, you can switch to email or news. But if a browser window freezes, it doesn't work to switch to a different browser window, because that one's frozen, too.
The problem that I see in firebird isn't a total hang. You can still scroll around in web pages. But clicking on links merely gets you the "busy" icon, and no new pages can be fetched in any browser window. This is further evidence that all the windows belong to one big process.
I suspect that this won't be fixed in firebird, because it's not a bug, it's an intentional design. But not a very good design, for a lot of reasons. It certainly does interfere with productivity, especially when you have to kill all your browser windows and restart them all.
Funny thing: My wife just replaced her aging Visor with a Tungsten. She likes some things about it. But some things are just flakey. The browser sometimes works great; sometimes it gets a fatal error when you start typing a URL and you have to reboot. It found our home wireless network without problems and (after I overrode its choice of DNS server with one that works) works there fine. A work, she can red her email, but attempts to reply or send get incomprehensible error messages.
As far as I can tell, there's no chance whatsoever that I could fix these problems. Yeah, I've poked around in it, and I've used PalmOS enough to basically know what's going on. But so much of it is hidden, with no way I can learn to discover what's happening internally. And my attempts to learn enough to program the suckers have gotten so many RTFM (with no FM mentioned) and other such insults rather than answers that I despair of ever being able to do even the smallest programming job on such systems.
Rather than "It just works", I'd summarize it as "The things that work are good; the rest you just have to learn to live with."
Myself, I experimented with a PDA phone for a while. Then one day, it went berserk. The support people helpfully talked me through a total re-initialization, wiping everything out and reconfiguring it. I'd done backups, of course, so I thought I could just do a restore. But when I tried that, it first backed up its current (empty) files over the backup files on my disk, thus wiping out my entire appointment book. I was never able to find what I'd done wrong.
So I went back to a paper pocket calendar. It doesn't fail and lose everything, and it doesn't need to be backed up.
If the PalmOS backup files had been in a format that I could decode, I could have handled much of the job myself. I'd looked at them, and sorta made sense of them, but not entirely. So, until I find useful documentation about the internal representation of those files, I won't be relying on a PalmOS system for anything important.
If I can use a linux-based PDA phone, I'll jump onto it right away. I'd trust it to not lose my important stuff, because I know I can get at the data and write my own program to read and write the files. (If not, well, I'll just use a different app whose file formats are available. Or I'll write my own, so I can get a calendar that can handle events that last past midnight.;-)
Something that keeps my data in proprietary formats can never qualify as "reliable", in my mind.
Mine has even less. Cheers; it and Firebird make the web fun.
Yeah, except that lately my firebird has only lasted a day or two at a time. Then suddenly it starts hanging when attempting to download a new URL. Things like scrolling and switching between tabs work, but no new pages can be downloaded (and the current pages can't be updated). You can't get it to quit with anything short of "kill -9". This isn't fun.
This morning, it happened repeatedly. I tracked it down, by starting up windows or tabs one at a time and checking to see when the freeze-up happened. It was on the/. page from yesterday about the latest SCO news! Something in that page hangs firebird's downloader, at least on my machines (one linux and one OSX).
There are a lot of nice things about firebird. But it still implements the worst misfeature of mozilla (and netscape and IE): It is one monolithic process. So if that process hangs, all your browser windows are dead.
If they want a real user-friendly browser, they'll make it fork for each new window. Then when one of them hangs, you don't have to kill off everything you're working on and restart them all.
It's really annoying seeing this Microsoft-like monolithic program worm its way into unixoid systems. We knew better than that 20 years ago.
But at least I have 5 or 6 other browsers on each machine, so I can get the benefits of multiprogramming by simply starting different browsers.
Giving governments control of the net is the worst possible idea... apart from all the other ideas which are worse.
Good point. And we might note that there is an unstated presupposition at work here: The idea that the Internet should be controlled by some organization.
We should be pointing out an alternative: Freedom of the Internet's users from control of their speech (with the qualification that we need ways of preventing people like marketers and politiciant from imposing their "free speech" on unwilling listeners).
If we must have a single organization controlling the Internet, in much of the world that organization probably should be the government. In some parts of the world (the US, Canada, most of Europe, etc.), there are laws in place that protect people from the government. These laws include the right to speak and publish, the right to due process if charged with a crime, etc. Such laws aren't always recognized by the current ruling gang, true, but the courts generally do recognize and enforce them, when they can.
At present, such protections don't apply in areas controlled by corporations. If you say something that offends a manager, you're out. You have no right to call home during work hours. You have no right to keep personal items in your desk. If charged with an offense, you have no right to a fair trial. You have no rights at all, except maybe the right to walk out.
A year or so back, we saw reported here the case of an ISP in Arizona that was bought out by msn.com, and one of the things they did was to cut off email to anyone not running Microsoft software. And if you read Microsoft EULAs, you often find a clause stating that you can't publish anything critical of them or their software. These are the sorts of things that corporations have the legal right to do. Many governments don't have such rights, and you can challenge them in court if they try to force you to kowtow to a chosen corporation.
I suppose we all understand that most governments can't be trusted very far, either. Even the best are not exactly known to be supportive of citizens who publicly criticise the government. But if we're on government property, at least we have some rights, and we can fight their attempts to control us. On corporate property, we have no rights whatsoever.
Still, the best situation would be to prevent total control by any organization, government or corporate.
Well, yeah; we all thought it was funny, too. In a sick sort of way.
I mean, think of the implications. The people who wrote the software that monitors things while you're on the operating table or sitting in a seat on an airliner are highly likely to think that fast code is more important than correct results. They refuse to switch to languages that do overflow or bounds checking, because code in those languages doesn't run as fast as Fortran or C.
Slashdot Mirror
If google starts a mail service, is it really to take on Yahoo?
Hey, I long ago started a mail service (for me and my wife) on my home machine. I've also installed qmail on a couple of work machines. But I don't see any articles about how I'm taking on Yahoo.
What does it take to get such fawning coverage?
Depends on how you define "better".
If your primary requirement is a cheap, fast way to get a lot of machines up and running, and and security isn't at the top of your list, then a "distribution" is the way to go. You still have to decide which one, and that depends on how your people will be using it.
If your primary requirement is a way to get a few machines (e.g., a server farm) up and running so that it is secure and reliable, then what you want to do is download all the source and build your own systems one piece at a time. You make sure you understand each package's configuration and security needs before you let it go live.
There is no best way for everyone.
Historically, linux has mostly been the favorite of people who want distributions. It tends to come with everything that compiles and passes the "make test" suite. It's no surprise that linux distributions should contain packages that are insecure. That's what happens when you let everyone throw their favorite apps into the bin.
Historically, the BSDs have been the favorite of people who have serious security concerns. It's no surprise that they should lack the full bag of bells and whistles of linux, and it's no surprise that they should have few security issues.
It does seem that, if you take the build-it-yourself approach with linux, you can get a well-tailored machine that's also fairly secure. But you will have to do the work and spend the time learning about the issues.
Manufacturers of windows for building just might be at risk soon. There are a number of companies now selling transparent LCD displays. There are several that can be used as real windows in walls. Google for "transparent LCD display" for information. (You'll have to wade through a lot of technical details to find pretty pictures. ;-)
...
So arguing that windows in walls and Windows in a computer display are different subject areas might not work much longer. It's possible right now to have a window in your house that is a computer display. How practical this is, I'm not sure, but you could do it.
So shall we start a pool on the date of Microsoft's first C&D letter and/or lawsuit when someone does this?
Of course, it might have happened already, and we just haven't heard about it. And imagine the rage of MS's lawyers if such a window were controlled by a computer running Lindows
BTW, if you think I was just joking, google for "the THE operating system". Right now, there are 418 hits. A few are typos, but most are about the THE OS itself.
But the attempt to register "THE" was, as I recall, reported as a bit of geek humor. Apparently the USPTO got the joke, laughed with them, and turned them down. But there's a serious question of whether their examiners would get it today.
You might also want to google for "English programming language". There are fewer hits, and most are just uses like "English programming language instruction". But a few are about the programming language, a dialect of SQL.
Is someone going to trademark the word "The" next, ...
;-)
Actually, someone tried that, back in the 70's as I recall. It was the name of a new OS. The USPTO didn't accept it then. They just might today.
So why don't you give it a try? Let us know how it works out.
(There was also an attempt to register "English" as the name of a computer programming language. That wasn't accepted, either. It's worth another try, too.
Yeah; I recall back in the 60's, when I was the undisputed chess champ in my high-school crowd. After a while that started to get old. I mean, you can either become a pro chess player, one of the worst jobs in the world, or you can move on to something else. I did the latter.
Actually, I started playing the piano a lot. The second-worst job in the world. Now, several instruments later, and with pianos transformed into a zillion descendant instruments, it still hasn't grown old.
But I also became a computer geek, so I can earn money. And I got into network programming, so I can take partial credit for the imminent destruction of the recording industry, which roughly a century ago took over music and made it nearly impossible to earn a living as a musician. We can all be happy about the revenge that we are starting to enact.
Along the way, I found the explanation for why people don't grow out of music:
Mother: Son, what do you want to do when you grow up:
Son: I want to be a musician.
Mother: Well, son, you have a choice. You can grow up, or you can be a musician.
It might be worth noting that linux was started in Finland, which is not exactly a 3rd-world country, but it's not what you'd call wealthy or powerful. Linus Torvalds has written about his motivation for for starting the project. One of his main reasons was that he wanted to work on the innards of a unix system, but he was a poor student who couldn't afford a commercial system. He could afford a PC clone, so he got that and a copy of the POSIX specs, and started coding. ("How hard can it be?" ;-)
;-)
This is likely a good model for the poorer parts of the world. Usable computer hardware is rapidly getting cheaper, but commercial software isn't. But we're talking about a part of the world with a lot of unemployed people, who have time on their hands and a strong desire to make a better life for themselves.
It doesn't take a great genius to imagine where this might lead. Especially if the information they need becomes available on the Net. And note that a number of schools are starting to put a lot of tech coursework online. MIT is no longer the only one with this sort of project underway.
(Hey, remember the article here recently about how MIT's Open Course Ware project was being implemented in India?
Hmmm ... That might have something to do with it.
;-)
But the controls don't seem to fit this model. On both machines, I only see a single "Printer Sharing" thingie. It's in the "Services" tab of the "Sharing" window. There is nothing to indicate which printer(s) should be shared.
This is significant because one of the PBs does have a second printer plugged into its USB port, and right now it sees both that printer and the airport printer. I did some experimenting, deleting all the printers for both PBs, and then adding them to PB2. Printer Sharing was unchecked. When I asked PB1's Print Center to find USB and Rendezvous printers, it failed. I checked "Printer Sharing" in PB1, and then PB2 was able to see both the Rendezvous (i.e., airport) printer and the USB printer on PB1.
I then closed PB1. PB2 lost sight of both printers within a few minutes. This makes sense for PB1's USB printer, but not for the airport printer.
So if I want PB1 to see the USB printer on PB2, PB2 must have printer sharing turned on. This makes sense. But then PB1 can't see the airport printer if PB2 is closed.
Note that this is the reverse of the original problem. Now PB2 has to be running for PB1 to see either printer. I'm guessing it's because I deleted all the printers and then had PB2 find both printers. So now PB2 "owns" both printers, although for the airport printer this makes no sense.
This has gotta be a bug, right? Did they actually design it to work this way?
I wonder if there's some place at Apple's tech support that would give a reasonable explanation of all this. But I'm afraid to call them up about it, because of the disaster the first time I tried. (I'm trying to get some work done, and I need the Internet connection to keep working.
However, on linux, I do know what I'm doing. ...
This is relevant how, exactly?
Read the title of this thread. An answer of the form "I use X rather than Y because I know X and have had troubles learning Y" is about the most direct, relevant answer one could give.
Of course, this is usually used as a coverup for a refusal to learn. I was letting readers know that I am willing to learn, and I'm not using knowledge of linux as an excuse to ignore a new system that gets lots of good reviews. I put out some good money for a Powerbook. But, like any computer system, it's somewhat of a complex beast. I can't learn it overnight. If that's what's expected, then it's clearly not the right machine for a dummy like me. And maybe I should just stick to a system that was designed for a dummy like me.
Which brings us back to the title of the thread.
... Apple tech support, who are obviously braindead morons.
Well, now; I think that's a bit harsh. It would be more accurate to say that they're the usual sort of harried, poorly-trained low-rank workers who barely have a handle on what they're expected to do, and are mostly working from scripts that handle only a list of pre-programmed cases.
This isn't a criticism of the support people themselves so much as a criticism of Apple (and the rest of the industry, for that matter).
I've found this as a useful answer to the suggestion that you should go with commercial vendors because you need their support. I counter with the suggestion that in fact linux (and *BSD) provides better support: public newsgroups and mailing lists. A few simple tests will show that this typically works a whole lot better than even the best of corporate tech support operations.
Actually, of course, both can be either good or bad, depending on how you ask the question, who you happen to get answers from, and the phase of the moon. But if you talk to people who routinely deal with both, you'll tend to find agreement that the random gaggle of "free" help on the internet is usually better than the paid help you get from corporate support.
Much of the reason is that the "free" help is usually from relative experts (at least in a narrow topic) who enjoy showing off their expertise. They may be insulting and grating, but they often give useful answers. The corporate help is usually low-paid, inexpert hired help who are just trying to get their job done without too much grief from customers or employers. They are friendly puppies, but often don't really know much.
If you're trying to get something to work, an arrogant jerk who solves your problem is usually better than a friendly dummy who doesn't solve your problem.
And back to the original question: "Why use Linux at all when there's Mac OS X?" I think that all this is a fairly good answer. My experiences with Apples tech support did convince my clients that OSX (and the Apple's airport) didn't look like a good bet for the company network infrastructure. They don't want warm, fuzzy feelings; they want to get answers fast. The arrogant, grating experts in the linux herd of cats are a much better bet if you want to solve problems fast.
This isn't necessarily a criticism of Apple or the Mac. Their market is pretty clearly people who want a computer that "just works", and who don't want to bother their pretty heads with the details. I would (and do) recommend a Mac to such people. I have one myself. But I'm frustrated by the difficulties in getting past the user-friendly interface. What I want is a programmer-friendly interface, and that's something rather different. If that's what you want, then you have a good reason to use linux even when OSX is available.
(And yes, I do know about the Mac Terminal. I have four of those on my PB's screen at the moment. But, unlike linux, and like MS Windows, there are a lot of things that don't seem to be available from a shell command. Or maybe they are, and I just can't find them.)
Well, it seems to me that you've given very good support to my comments.
Recall that the question was:
Why use Linux at all when there's Mac OS X?
Now, it's clear to nearly everyone here that I'm obviously an idiot who doesn't know the first thing about using OSX, and my problems were all because of my own stupid mistakes.
But anyone who is new to OSX is going to make lots of mistakes. There's a lot of stuff to learn, and no mere human can learn it all in the first hour (or month). If this means I shouldn't have been using OSX, well, OK; I shouldn't have been using OSX.
That seems a reasonable answer to the question. An admitted linux geek sticks his neck out and asks questions that make it clear that he's ignorant. So he gets the treatment he obviously deserves. Lesson: He should stick with linux and not bother with OSX. It's a simple, elegant answer to the question.
Myself, however, I like to learn things. That's why I bought the Powerbook. So I think I'll thumb my nose at the advice here, and continue to learn what I can. In my copious spare time. There are some good ideas in there.
When and if I ever become as familiar with OSX, I may start advising my clients to use it more. However, considering that this learning must be done in my spare time, while my day job involves linux development, this may not be real soon.
To put it all in some sort of perspective, I mentioned one apparent screwup within OSX to a Mac-using coworker, with a comment like "Hey, they f**ked it up even worse than in linux or solaris." His response was "Yeah, but they f**ked it up in a much more elegant and pretty fashion."
We boch chuckled and went on with the problem at hand.
Um, you didn't quite read it right.
We have two powerbooks. When PB1 is closed, PB2 loses contact with the printer plugged into the airport.
I contend that this makes no sense at all. PB2 should be able to talk to the printer without PB1 being present and powered up. PB1 is not in the data path PB2 -> airport -> printer.
I'm assuming that this is somehow because I first used the printer from PB1, so PB1 has some special relationship to either the airport or the printer. But I haven't found any evidence to support this. I've sat the two PBs next to each other, fired up the airport and printer admin tools on both, and compared the fields one by one. They are identical. There appears to be no difference in the ways the two PBs see the airport or the printer. But PB2 still loses the printer when PB1 is put to sleep.
This has gotta be a bug.
And yes, I did ask on a couple of newsgroups. I got "replies" that made the same mistake: They tried to explain why my PB can't talk to the printer. But that's not the problem at all. Both PBs can see the printer. The problem is if I turn PB1 off, PB2 loses contact with the printer.
What's worse is that it's inconsistent. The time between turning PB1 off and PB2 losing sight of the printer varies from minutes to hours.
if you're finding OS X slow, opaque and frustrating, it's because you don't know what you're doing, you just think you know what you're doing.
...)
Heh. You're right there.
However, on linux, I do know what I'm doing.
Part of the reason for getting an OSX system (and so far refusing to bring up non-Mac things like X-Windows etc) is that I wanted to get familiar with it.
The problem here is that, as others have pointed out, when I did the obvious thing and asked Apple's CS for help, I got an idiot-level runaround. They didn't say "Hey, this guy knows how to run linux gateways and firewalls; maybe I'll bounce him to an expert level. They kept "helping" me with something that wasn't related to my original problem (getting airport + printer + powerbook to work), and bollixed up my network as a result. They didn't mention DHCP to me; I discovered it myself. And they told me quite clearly that I couldn't expect to use their equipment in a mixed-vendor network.
I'm starting to agree with them.
With linux, OTOH, when I ask CS (e.g., Red Hat's), I get intelligent answers. Or sometimes I get RTFM, and usually a pointer to the FM or some equivalent online docs. So I can learn what I don't know.
I freely admit that I don't know everything there is to know about linux, and I'm even farther from that with OSX. So far, my experience with linux is that finding answers sometimes takes time, but I can find them, and among the arrogant jerks there are a lot of helpful people. With OSX, my experience so far is that I get a lot of "... for dummies" sorts of answers, but very little else. Finding answers has been rather slow compared to linux. So I'd recommend linux for serious networking applications.
To put things in perspective, we also have a couple of MS windows machines (which my wife needs for her work), and even a cool wifi-enabled PalmOS machine. They're all talking to each other, though not always amicably. I'd say that the OSX system is far, far better than the MS boxes for nearly everything. I've recommended a Mac to lots of people who have expressed frustration with their Windows boxes. I'll continue to do so.
My summary at the moment is that I'd strongly advise linux and/or *BSD for any serious networking, server, or other infrastructure situations. I'd recommend a Mac for any computer non-geek who wants a machine that works and is fun to use. I'd recommend MS systems for masochists. (And that wireless PalmOS gadget is a really cool toy. Now if it had a cell-phone built in, too
Well, one recent example from personal experience:
;-).
I was having problems with our two powerbooks (OSX) talking to a printer that was plugged into the airport. Fine, I thought, I'll call up Apple's CS people and they'll figure it out.
Hah! The fellow I talked to had me go through the machine's internet connections. Why he did this isn't obvious. The printer should work even if the machines aren't connected to the internet. But never mind; by the end of the session, my previously-working internet connection was bollixed up so badly that nothing could communicate. The call ended with me starting an hour-long validation of the system, and then I was supposed to call back.
When it finished (not finding much, and the internet still f**ked up even after a reboot, I called back. The fellow I talked to that time went through all thee tthings from the first call. Then, when he discovered that the connection to the internet was through a linuux box, he demanded that I reboot it. I balked at that, and let him know in no uncertain terms that this was utterly unacceptable.
I did agree to disconnecting the linux box from the chain, and hooking the airport up to the cable modem. He still couldn't get it to work. And still, nothing at all had been done with the printer. He left me with yet another system validation task running, and by the time it finished, it was well past their closing time. And he'd made it clear that they wouldn't be able to help us as long as there was a linux box on the LAN.
I went into personal search-and-recover mode, and got the internet hookup working (through my linuux gateway, of course). I couldn't have done it through the OSX net tools that I found; I mostly used the diagnostic tools on the linux box to tell me what was happening.
One thing I learned from the linux tools was that the airport was running a DHCP server. I hadn't yet seen anything in the Apple tools that told me this. Its address range overlapped with the range used by the linux box's DHCP server. Those who understand this paragraph will understand the problem.
I found the DHCP controls in the Apple software and moved the airport's DHCP server to a different subnet. A bit more judicious configging, and I got my internet connection back again. I even got the printer to work, sorta, though we do still have a problem that when I close my powerbook, my wife's loses contact with the printer. (I've asked about this on some newsgroups, with no answer. I'm afraid to call Apple about it.)
The major effect of all this was that I wrote it up for the people I'm consulting for. They had been looking into trying some OSX boxes as part of their corporate network infrastructure. After my report, they dropped that idea. A couple of other people did a bit of judicious inquiring, and gave independent reports that I hadn't exaggerated a bit.
Nonetheless, I have a powerbook in addition to two linux boxes. I think that OSX has a lot of good things going for it. But I wouldn't recommend it for some things that are routine on linux. One is handling networks with a mixture of different machines.
Much of the problem is the "Don't worry your little head about it" attitude of the Mac community and Apple. That's fine if you're a non-nerd. But if you know what you're doing, and you want tools that give you access to all the innards, you'll find OSX slow, opaque and frustrating. With linux, you'll find a community that is willing and able to help you (if not always in a friendly manner
The hardware is nice. If I could get a 17" powerbook with linux (with drivers for all the hardware), I'd find it really useful. I think I'll start bookmarking reports like this one, and maybe invest in one of these machines in the not-too-distant future.
Yeah, well, I was expecting an exciting Obfuscated Eiffel Contest, to really give the C and perl folks something to admire in eiffel.
Such a disappointment.
[Microsoft] have 95% of the browsers and webpages are coded to whatever crap IE renders whenever necessary.
You do have to be a bit careful with repeating this sort of claim, because many of the statistics you'll read fall into the "87% of all statistics are just made up" category. It's very easy to interpret web-server logs and sales figures in radically different ways.
Thus, I recently installed the latest Opera on my Powerbook, as part of my collection of browsers for testing web pages. I checked with a nearby server log, and its default id string is:
"Mozilla/4.0 (compatible; MSIE 5.0; Mac_PowerPC) Opera 6.0 [en]"
It's easy for an unscrupulous marketer or a sloppy programmer to interpret this as Mozilla, Netscape, IE or Opera. It can be (and is) counted as any of them, depending on what you want to "prove".
As for sales figures, I like to mention the Dell box sitting in the row of computers on the shelf next to my desk. It's running RH linux 8.0 at the moment. There's no trace of Microsoft software on it. But it was delivered with MS windows, so both Dell and Microsoft count me as a Windows customer. And, since I haven't used their customer support for it, I'm obviously a happy, satisfied customer. This box is, of course, also counted in any industry figures as having an IE browser. Wrong again.
We really don't have any good numbers on the scale of this sort of misrepresentation. Any numbers you see are probably in that 87% of statistics that are just made up.
One of my more fun examples of padded browser statistics: I also have an old W98 box on my shelf. It's used for web testing, of course, and is usually turned off. But when I do use it, one very real problem is that it rarely survives more than a dozen web pages before it hangs and I have to reboot it. When I restart the browser, I usually have to re-fetch a number of the pages that I was working on. Those fetches go into the statistics, of course. I'd claim that counting reloads caused by browser crashes is totally bogus. This is padding your numbers in the worst possible way.
We have no way of knowing how often Windows users have to do this, because there's no reliable way to distinguish such downloads from others in a server log. And it's not just IE; the Firebird browser on my linux and OSX boxes hangs regularly and has to be killed and restarted. Those reloads are also bogus numbers in the statistics.
Any figures you see on this topic are to be taken with a large grain of salt. They are mostly PR, not facts.
$12K a pop!
Yeah, and it doesn't even include cell-phone capability. What were they thinking?
Now GPS + WIFI + PCS would be useful. Especially if I could develop software for it.
There's also the theory that the extra filters are to produce pictures that look right to the eyes of the visiting aliens that are helping NASA. I mean, how would you like to have pictures that cut out the red portion of the spectrum?
For that matter, there are plenty of earth species that can see frequencies outside the human visual range. We are always getting complaints from our conure and cockatiels about the poor color quality of our computer screens. The ultraviolet is almost entirely missing.
I suppose the cockroaches would complain, too, but they only use the computers after we've gone to bed, and who knows what sort of mods they've made to the displays?
In any case, it's not easy to get a really accurate color display from the current primitive crop of manufacturers on this benighted planet. Not to mention the difficulty of getting good UV and IR pigments from the inkjet-printer manufacturers. For the forseeable future, those visiting aliens will just have to order the inks from home, and you know how slow delivery in this part of the galaxy can be.
Why, no, not at all. If you look closely at /. stories in the past few months, you'll find others from Al Jazeera. Also, if you go to new.google.com, you'll find them well represented in the top stories. In particular, they've had good summaries of a lot of technical and scientific stories.
;-)
You'd think that the Al Jazeera folks are trying to be a respected news source or something.
(I was trying hard not to say "fair and balanced.
Their Middle-East reporting makes for interesting reading, too. They often give you a somewhat different slant than Western news services.
think the point, us nerds would like to be able to hack our phones like we hack on our computer systems.
;-)
Close, but I'd express it differently.
I don't really want to hack my phone. I want to replace it. What I want to replace it with is a PDA-like gadget that will fit into my pocket, and be able to talk to both the phone system and the wireless Internet. And I want to be able to use it like a computer, i.e., it must be programmable.
An important part of this is the "will fit into my pocket" phrase. Most PDAs flunk this test.
I have in my pocket what looked like a good start a couple of years ago: a Kyocera 6035 "smartphone". It has a lot of problems, though. One is that the web browser works over IP that's PPP over the phone system. It's sloooooow, and you get charged full air time for the connection, even when no packets are being passed. This is far too expensive to use it routinely.
My wife has a new Tungsten, that comes with wi-fi networking built in. But it doesn't do phone calls. And it's too big for my pocket (though it does fit into her purse).
Also, these are both PalmOS. After a couple of years of exploring their development stuff, I find that it's really not worth the effort. Doing even the smallest thing takes forever, because you just can't debug the stuff. The slightest error freezes everything, you have to reboot, and you have no clues as to what went wrong. There's nothing at all like gdb available. And most of the internal working are invisible and undocumented to outsiders like me.
To be credible, I'd want something that I can actually program. This means that the innards should be documented, and there should be places to ask dumb questions. PalmOS doesn't even come close. I haven't tried Symbian, and I do wonder if it's better.
But it's pretty obvious that a pocket-sized linux gadget with both wi-fi and cell-phone hardware would do the job quite nicely. Nothing hidden there, and lots of places to ask dumb questions (and get RTFM answers, for which I can ask "So where's the FM for that?"
I'm not dogmatic about linux, though. FreeBSD would be nice, too, and OSX would be pretty good (though parts of its innards are blocked by brick walls).
I also wonder about iTron. Is there any way for a US resident (with little Japanese) to get meaningful experience with it?
I do think it might be easier to use one browser at a time. Unless, of course, Firebird does some horrible and evil Windows-like "only one instance of this program can be open" thing, ...
But that's exactly how mozilla (and the original netscape) have always worked. Actually, on my linux box, there are always exactly four mozilla-bin processes, regardless of how many windows I have open, but only one of them is ever using the cpu at any given time. On my Mac (OSX), top shows one "mozilla-bi" process and one "MozillaFir" process. When I do something with either, their cpu usage goes up. But it doesn't matter how many windows I have open or which window I'm using; there is only one active process for each browser.
It's good that Firebird split off the mail/news task into a separate program. At least, when the browser windows all hang, you can switch to email or news. But if a browser window freezes, it doesn't work to switch to a different browser window, because that one's frozen, too.
The problem that I see in firebird isn't a total hang. You can still scroll around in web pages. But clicking on links merely gets you the "busy" icon, and no new pages can be fetched in any browser window. This is further evidence that all the windows belong to one big process.
I suspect that this won't be fixed in firebird, because it's not a bug, it's an intentional design. But not a very good design, for a lot of reasons. It certainly does interfere with productivity, especially when you have to kill all your browser windows and restart them all.
Funny thing: My wife just replaced her aging Visor with a Tungsten. She likes some things about it. But some things are just flakey. The browser sometimes works great; sometimes it gets a fatal error when you start typing a URL and you have to reboot. It found our home wireless network without problems and (after I overrode its choice of DNS server with one that works) works there fine. A work, she can red her email, but attempts to reply or send get incomprehensible error messages.
;-)
As far as I can tell, there's no chance whatsoever that I could fix these problems. Yeah, I've poked around in it, and I've used PalmOS enough to basically know what's going on. But so much of it is hidden, with no way I can learn to discover what's happening internally. And my attempts to learn enough to program the suckers have gotten so many RTFM (with no FM mentioned) and other such insults rather than answers that I despair of ever being able to do even the smallest programming job on such systems.
Rather than "It just works", I'd summarize it as "The things that work are good; the rest you just have to learn to live with."
Myself, I experimented with a PDA phone for a while. Then one day, it went berserk. The support people helpfully talked me through a total re-initialization, wiping everything out and reconfiguring it. I'd done backups, of course, so I thought I could just do a restore. But when I tried that, it first backed up its current (empty) files over the backup files on my disk, thus wiping out my entire appointment book. I was never able to find what I'd done wrong.
So I went back to a paper pocket calendar. It doesn't fail and lose everything, and it doesn't need to be backed up.
If the PalmOS backup files had been in a format that I could decode, I could have handled much of the job myself. I'd looked at them, and sorta made sense of them, but not entirely. So, until I find useful documentation about the internal representation of those files, I won't be relying on a PalmOS system for anything important.
If I can use a linux-based PDA phone, I'll jump onto it right away. I'd trust it to not lose my important stuff, because I know I can get at the data and write my own program to read and write the files. (If not, well, I'll just use a different app whose file formats are available. Or I'll write my own, so I can get a calendar that can handle events that last past midnight.
Something that keeps my data in proprietary formats can never qualify as "reliable", in my mind.
But then, I'm a programmer. YMMV.
Mine has even less. Cheers; it and Firebird make the web fun.
/. page from yesterday about the latest SCO news! Something in that page hangs firebird's downloader, at least on my machines (one linux and one OSX).
Yeah, except that lately my firebird has only lasted a day or two at a time. Then suddenly it starts hanging when attempting to download a new URL. Things like scrolling and switching between tabs work, but no new pages can be downloaded (and the current pages can't be updated). You can't get it to quit with anything short of "kill -9". This isn't fun.
This morning, it happened repeatedly. I tracked it down, by starting up windows or tabs one at a time and checking to see when the freeze-up happened. It was on the
There are a lot of nice things about firebird. But it still implements the worst misfeature of mozilla (and netscape and IE): It is one monolithic process. So if that process hangs, all your browser windows are dead.
If they want a real user-friendly browser, they'll make it fork for each new window. Then when one of them hangs, you don't have to kill off everything you're working on and restart them all.
It's really annoying seeing this Microsoft-like monolithic program worm its way into unixoid systems. We knew better than that 20 years ago.
But at least I have 5 or 6 other browsers on each machine, so I can get the benefits of multiprogramming by simply starting different browsers.
Giving governments control of the net is the worst possible idea... apart from all the other ideas which are worse.
Good point. And we might note that there is an unstated presupposition at work here: The idea that the Internet should be controlled by some organization.
We should be pointing out an alternative: Freedom of the Internet's users from control of their speech (with the qualification that we need ways of preventing people like marketers and politiciant from imposing their "free speech" on unwilling listeners).
If we must have a single organization controlling the Internet, in much of the world that organization probably should be the government. In some parts of the world (the US, Canada, most of Europe, etc.), there are laws in place that protect people from the government. These laws include the right to speak and publish, the right to due process if charged with a crime, etc. Such laws aren't always recognized by the current ruling gang, true, but the courts generally do recognize and enforce them, when they can.
At present, such protections don't apply in areas controlled by corporations. If you say something that offends a manager, you're out. You have no right to call home during work hours. You have no right to keep personal items in your desk. If charged with an offense, you have no right to a fair trial. You have no rights at all, except maybe the right to walk out.
A year or so back, we saw reported here the case of an ISP in Arizona that was bought out by msn.com, and one of the things they did was to cut off email to anyone not running Microsoft software. And if you read Microsoft EULAs, you often find a clause stating that you can't publish anything critical of them or their software. These are the sorts of things that corporations have the legal right to do. Many governments don't have such rights, and you can challenge them in court if they try to force you to kowtow to a chosen corporation.
I suppose we all understand that most governments can't be trusted very far, either. Even the best are not exactly known to be supportive of citizens who publicly criticise the government. But if we're on government property, at least we have some rights, and we can fight their attempts to control us. On corporate property, we have no rights whatsoever.
Still, the best situation would be to prevent total control by any organization, government or corporate.
Funny? FUNNY?
Well, yeah; we all thought it was funny, too. In a sick sort of way.
I mean, think of the implications. The people who wrote the software that monitors things while you're on the operating table or sitting in a seat on an airliner are highly likely to think that fast code is more important than correct results. They refuse to switch to languages that do overflow or bounds checking, because code in those languages doesn't run as fast as Fortran or C.
That's real funny.