Yeah, about 15 lines above indeed, it was set unconditionally to value "true", though it was inside an object (with only public members) which is probably what confuced the compiler ?
Only it wasn't, because I already though about this, and cached the value in a "final boolean" variable, to see if that would be the problem. It wasn't, still complains and refuces to compile. So it doesn't matter a bit if it's constant or not.
Well, it's not too bad a performance hit to create one extra object, but it's kinda bad for code readability, but what can I do.
I don't know, probably some of these defects might be actual problems, but unless the software is real good, it's always possible that certain cases never happen, although automatic software can find "defects".
As a rather "stupid" example, I had to initialize a Map to an empty HashMap just last week to get Sun's Java compiler accept my code, although the only two references to the Map where within two if-blocks, within the same function, both of which depended on the same boolean value, which wasn't changed in the whole function.
There's a difference between defect and a bug.
Tools that help in finding problems are great, but after all, they can only point possibly unsafe points. Ofcourse it's good to write code that doesn't trigger any such possibilities in the first place.
How about populating our sites with a community developed ActiveX control, that gives a popup, asking user to "install critical anti-virus update", or "enhance clock precision" or something, then when user clicks yes, it opens about 15 browsers all directed to goatse, and on top of the whole pile of shit, a message box "Are you really that stupid ??"
with yes and no buttons swapping themselves every time you put your mouse on the no button.
It's fine to just uninstall it after one time impact.. average user will repeat it a few times, until the lesson is learned.
I'm more than happy with my (now about one and half year old) Matrox G550. It runs Tux-racer just fine, so am I missing something here ?? Never had any driver problems, other than KT133 chipset having trouble with G550+Audigy a bit. Only drawback is that second screen only supports 1600x1200..
I think the worst thing that can happen with this SCO thing and software patents and all, is that the legal system will collapse.
I mean, I think we've avoided almost all patent issues by developing/distributing infringing software in Europe, but I don't think that it'll stop here if software patents become a law.
Instead it's probably only going to hurt the corporations, that now have to take closer watch at what they are using. As for individual developers, I hope there will be thousands and thousands of legal cases, if this becomes law..
Why ? Average case takes years even now. The legal system will have "scaling issues" so to say, if you double, maybe triple the cases. In addition it's not only the legal system. Open Source can always go underground. I think most people that develop OpenSource will do it what ever happens. What they might do is do it without their real names.
This is turn means that someone has to investigate who's behind that stuff. I can tell that cases with far more priority sometimes take years to investigate, at least here in Finland. By allowing software patents, the EU would allow something like a DoS attack on not only it's patent offices, but legal systems and police.
I might sound anarchist, but I really don't believe that EU could sustain as many lawers as US can, due to other differences in legal systems.
Personally, I won't be constraining my (personal, open source) code with limitations due to software patents. At most I'll stop putting my name in the copyright, and change license to something that doesn't have the condition 7 of GNU GPL in it, but instead something like "by using this software, you are intentionally infringing software patents, and take all responsability in possible legal cases".
My previous system before gentoo was a LFS so I believe I'm probably one of those that Gentoo is targeted to.
The point of my post is that emerge, rc-update and other handy tools in addition to the default gnome2 desktop would make it a good distribution for newbies as well.
After installing Gentoo, there isn't much that you need to be a Guru for, so too bad you have to choose between "easy to use" or "easy to install".
I've installed Gentoo for people that would not been able to do it, and they've been happy with it.. that that it's necessarily bad that someone has to get the system in a "startup condition" before letting the newbie in.. at least there won't be extra services before the newbie learns how to install them:)
Gentoo would be the absolute top of all distribution, if it wasn't for the installation. I mean, is there a nice graphical installer that configures X for you and let's you login to a working system and start learning from the desktop first..
Gentoo is superb if you have someone install it for you. The install.txt part of the install isn't the hard part actually, you get it to boot with not too much trouble if you have some idea of your system, but before you have a nice graphical desktop system, you have to tweak and emerge a lot of things.
One thing that would be nice if we had nice installer for gentoo would be to allow it to install/configure X and Gnome (or KDE if someone really prefers it) and then populate the menus with the most useful pieces of software.
Not really install them, but instead let them popup something like "Evolution is not yet installed on you system. Do you want me to install it ?" and then automatically emerge it if that's what the user wanted.
Actually, Mozilla seems to be the only browser, that has developers who actually understand the CSS box model (which shouldn't be that hard after all).
Paddings don't work right in IE or in Opera.
Other than that, if you sum the support of IE and Opera, you pretty much get Mozilla.
One interesting concept with this is that if it can be made so that it doesn't need any read-write files it's very painful to exploit even if there was a minor security hole.
You could be sure there are no trojan's atleast, if you make sure it's impossible to mount even ramdisks as read-write. It's OK to have write-only devices for logging (forward to another host/printer/???) as long as the host is unable to read (or at least execute) them.
I'd love to have such a firewall. CD would act like immutable firmware while in the system, while you could still update it.
Maybe use initrd and remount it (permanently, maybe special kernel patch) read-only on boot.
Re:I had a feeliing it would get posted to slashdo
on
Linux Router Project Dead
·
· Score: 0, Offtopic
You could try running Win95 under an emulator like VMWare/plex86, or if you have the CPU to burn, bochs.
The first at least should be able to run Win95 just fine (it emulates ne2k net IIRC and has it's own VGA drivers) but costs some money. There's a free evaluation version to play with though.
Once the GUI become OUI (Object User Interface) or BUI (Bubble Use Interface) the Windows GUI paradigm will be replaced with something more powerful.
Now, this is already happening. I see the GUI frontends to commandline tools as part of separation of GUI from component, which is basicly the only logical way to improve current situation (not limited to CLI vs GUI ofcourse).
Another thing is the "WEB" interface which is most interesting because the application doesn't need to live on the users machine, which allows for things like slashdot to be implemented quite easily in a cross-platform way. Again, OS doesn't matter.
My second point however is, that there's nothing that keeps one from changing the GUI of Windows to whatever one finds appropriate. Explorer.exe can be replaced by some other system (alternative exist) almost as easily as a GNU/Linux user switches a Window Manager (just there's no GUI to do it for you).
I try to use only the visual and sorting tools in Outlook for differnt categories, hierarchies, but otherwise make sure that all my data is also available in standard format. If Outlook were to disappear from the world tomorrow I would miss it, but would have nothing really "tied" into it.
Personally I use GNU/Linux with Evolution (which btw acts/looks basicly like Outlook without as many problems and with few improvements in GUI, Exchange connectivity is available for a fee) as my mail client, which happens to have local folders in ~/evolution/Local (inbox is ~/evolution/Local/Inbox) and as far as I can say the main mailbox file is standard mbox. There's few binary files too, but they are for unimportant things like indexes and such.
Unfortunately, if most of people choose Windows® in the Present, Open Source solutions will have less users, which means that the statistical probablities of those people having everything on Windows® and running into problems with GNU/Linux and other Open Source easily increases, which would mean that in the Future the choice is even easier.
Whether there actually are such problems for a given domain of "things I want to do" is unfortunately irrelevant. These days it already takes me minutes to find a category of software that has no (acceptable) Open Source solutions.
With my new Gentoo installation I haven't had a single problem (installation is somewhat tricky but after that it's like a dream).
The reality is that it's not even a question of operating system. Most popular Open Source applications work just fine on Windows.
After slowly changing the applications into Open Source the operating system question starts losing relevance.
The question should not be whether to use closed or open software, instead, it should be Could some of the products we use could be replaced with Open Source alternatives ?
More important than Open Source are Open Dataformats.
You know, flame all you want but we all know everyone started in this programming language.
No, not everyone started with this one. Actually, I personally haven't done a single serious application with it.
Instead, I started with GWBASIC, which I used until my apps became too large for it to handle and I started looking for alternatives. I found Pascal, didn't like it at all, then found C, and liked it, only few years later I actually saw QBASIC for the first time.
At some point I also looked at COBOL (before QBASIC, that is) but it didn't make much sense to a 11 years old me back then..
If I could go back and select my first language again, I'd go for LISP though.
Please remember Linux(tm) is a trademark of Linus Torvalds and properly indicate it when talking about Linux(tm) or he might be interested in sueing you.
IIRC some old IBM Fortran compiler used to produce very small by using almost nothing but subroutine calls in the actual program producing something kinda like P-code. Maybe they are doing something similar.
Like some others have pointed out, printing isn't really that slow.
I can write in cursive, at least in two different styles, and I've always been told my writing is very beautiful and easy to read. I still use it quite often when I want my text to look beautiful.
When I need to write something fast, I use printed letters, since I can print about 5 times as fast, which is almost as fast as anybody can write, and about 10 times easier to read than average person's cursive, plus has the added benefit of being less straining to my wrist.
No offence hopefully, but I still remember not-so-long-ago when people used to say that "Linux is ideal for small webservers where little complex software beyond Apache or Perl is required, meaning any OS can conceivably be used. What sets...
Strange I've been using GNU/Linux at home since something like Â95 and always had more problem
with other peoples Windows. Go figure.
Slashdot Mirror
It was about like this.. void foo() { java.util.Map map; final boolean buz = true; if(buz) { map = new java.util.HashMap(); map.put("foo", "bar"); } /* something else */
if(!buz) {
System.out.println(
"doing something else");
} else {
System.out.println(map.get("foo"));
}
}
Only it wasn't, because I already though about this, and cached the value in a "final boolean" variable, to see if that would be the problem. It wasn't, still complains and refuces to compile. So it doesn't matter a bit if it's constant or not. Well, it's not too bad a performance hit to create one extra object, but it's kinda bad for code readability, but what can I do.
As a rather "stupid" example, I had to initialize a Map to an empty HashMap just last week to get Sun's Java compiler accept my code, although the only two references to the Map where within two if-blocks, within the same function, both of which depended on the same boolean value, which wasn't changed in the whole function.
There's a difference between defect and a bug. Tools that help in finding problems are great, but after all, they can only point possibly unsafe points. Ofcourse it's good to write code that doesn't trigger any such possibilities in the first place.
How about populating our sites with a community developed ActiveX control, that gives a popup, asking user to "install critical anti-virus update", or "enhance clock precision" or something, then when user clicks yes, it opens about 15 browsers all directed to goatse, and on top of the whole pile of shit, a message box "Are you really that stupid ??" with yes and no buttons swapping themselves every time you put your mouse on the no button.
It's fine to just uninstall it after one time impact.. average user will repeat it a few times, until the lesson is learned.
And please, don't mod me funny, I'm serious.
I mean, I think we've avoided almost all patent issues by developing/distributing infringing software in Europe, but I don't think that it'll stop here if software patents become a law.
Instead it's probably only going to hurt the corporations, that now have to take closer watch at what they are using. As for individual developers, I hope there will be thousands and thousands of legal cases, if this becomes law..
Why ? Average case takes years even now. The legal system will have "scaling issues" so to say, if you double, maybe triple the cases. In addition it's not only the legal system. Open Source can always go underground. I think most people that develop OpenSource will do it what ever happens. What they might do is do it without their real names.
This is turn means that someone has to investigate who's behind that stuff. I can tell that cases with far more priority sometimes take years to investigate, at least here in Finland. By allowing software patents, the EU would allow something like a DoS attack on not only it's patent offices, but legal systems and police.
I might sound anarchist, but I really don't believe that EU could sustain as many lawers as US can, due to other differences in legal systems. Personally, I won't be constraining my (personal, open source) code with limitations due to software patents. At most I'll stop putting my name in the copyright, and change license to something that doesn't have the condition 7 of GNU GPL in it, but instead something like "by using this software, you are intentionally infringing software patents, and take all responsability in possible legal cases".
How about comments within the source or the README that also tell you to RTFM ?
The point of my post is that emerge, rc-update and other handy tools in addition to the default gnome2 desktop would make it a good distribution for newbies as well.
After installing Gentoo, there isn't much that you need to be a Guru for, so too bad you have to choose between "easy to use" or "easy to install".
I've installed Gentoo for people that would not been able to do it, and they've been happy with it.. that that it's necessarily bad that someone has to get the system in a "startup condition" before letting the newbie in.. at least there won't be extra services before the newbie learns how to install them :)
Gentoo is superb if you have someone install it for you. The install.txt part of the install isn't the hard part actually, you get it to boot with not too much trouble if you have some idea of your system, but before you have a nice graphical desktop system, you have to tweak and emerge a lot of things.
One thing that would be nice if we had nice installer for gentoo would be to allow it to install/configure X and Gnome (or KDE if someone really prefers it) and then populate the menus with the most useful pieces of software. Not really install them, but instead let them popup something like "Evolution is not yet installed on you system. Do you want me to install it ?" and then automatically emerge it if that's what the user wanted.
Biogas (which is mainly methane) is used in Finland. The Finnish Biogas Association has some reports and links.
Paddings don't work right in IE or in Opera. Other than that, if you sum the support of IE and Opera, you pretty much get Mozilla.
You could be sure there are no trojan's atleast, if you make sure it's impossible to mount even ramdisks as read-write. It's OK to have write-only devices for logging (forward to another host/printer/???) as long as the host is unable to read (or at least execute) them.
I'd love to have such a firewall. CD would act like immutable firmware while in the system, while you could still update it.
Maybe use initrd and remount it (permanently, maybe special kernel patch) read-only on boot.
just to say i'd mod you up if I had any points :)
Personally, I think it's great that software is Open Source by OSI's definition, but 9 times of 10 I prefer Free Software over Open Source.
The first at least should be able to run Win95 just fine (it emulates ne2k net IIRC and has it's own VGA drivers) but costs some money. There's a free evaluation version to play with though.
wasteland:~# emerge -u gnome gnome-office evolution
Once the GUI become OUI (Object User Interface) or BUI (Bubble Use Interface) the Windows GUI paradigm will be replaced with something more powerful.
Now, this is already happening. I see the GUI frontends to commandline tools as part of separation of GUI from component, which is basicly the only logical way to improve current situation (not limited to CLI vs GUI ofcourse).
Another thing is the "WEB" interface which is most interesting because the application doesn't need to live on the users machine, which allows for things like slashdot to be implemented quite easily in a cross-platform way. Again, OS doesn't matter.
My second point however is, that there's nothing that keeps one from changing the GUI of Windows to whatever one finds appropriate. Explorer.exe can be replaced by some other system (alternative exist) almost as easily as a GNU/Linux user switches a Window Manager (just there's no GUI to do it for you).
I try to use only the visual and sorting tools in Outlook for differnt categories, hierarchies, but otherwise make sure that all my data is also available in standard format. If Outlook were to disappear from the world tomorrow I would miss it, but would have nothing really "tied" into it.
Personally I use GNU/Linux with Evolution (which btw acts/looks basicly like Outlook without as many problems and with few improvements in GUI, Exchange connectivity is available for a fee) as my mail client, which happens to have local folders in ~/evolution/Local (inbox is ~/evolution/Local/Inbox) and as far as I can say the main mailbox file is standard mbox. There's few binary files too, but they are for unimportant things like indexes and such.
Whether there actually are such problems for a given domain of "things I want to do" is unfortunately irrelevant. These days it already takes me minutes to find a category of software that has no (acceptable) Open Source solutions.
With my new Gentoo installation I haven't had a single problem (installation is somewhat tricky but after that it's like a dream).
The reality is that it's not even a question of operating system. Most popular Open Source applications work just fine on Windows. After slowly changing the applications into Open Source the operating system question starts losing relevance.
The question should not be whether to use closed or open software, instead, it should be Could some of the products we use could be replaced with Open Source alternatives ? More important than Open Source are Open Dataformats.
No, not everyone started with this one. Actually, I personally haven't done a single serious application with it.
Instead, I started with GWBASIC, which I used until my apps became too large for it to handle and I started looking for alternatives. I found Pascal, didn't like it at all, then found C, and liked it, only few years later I actually saw QBASIC for the first time.
At some point I also looked at COBOL (before QBASIC, that is) but it didn't make much sense to a 11 years old me back then..
If I could go back and select my first language again, I'd go for LISP though.
Can't http:// be forced in slashcode ?
Please remember Linux(tm) is a trademark of Linus Torvalds and properly indicate it when talking about Linux(tm) or he might be interested in sueing you.
Other than that ? Probably just marketing.
Time to test GPL in court ?
I've tried few editors, but VIM (with scripting for folding and automatic adding of end-tags) is best I've found for both my XML and XHTML needs.
I can write in cursive, at least in two different styles, and I've always been told my writing is very beautiful and easy to read. I still use it quite often when I want my text to look beautiful.
When I need to write something fast, I use printed letters, since I can print about 5 times as fast, which is almost as fast as anybody can write, and about 10 times easier to read than average person's cursive, plus has the added benefit of being less straining to my wrist.
No offence hopefully, but I still remember not-so-long-ago when people used to say that "Linux is ideal for small webservers where little complex software beyond Apache or Perl is required, meaning any OS can conceivably be used. What sets...
Strange I've been using GNU/Linux at home since something like Â95 and always had more problem with other peoples Windows. Go figure.