Perhaps it is because NoScript breaks FlashBlock (not sure about AdBlock). I don't know if this is still the case, since I removed it (FlashBlock) when I noticed that NoScript was preventing it from working.
Depending on your test environment (if you're testing through a dial-up-emulator proxy or just files across an intranet), you might simply be seeing the initial paint delay. It may or may not be this (or it could even be that it's really slower), but you can search for "delay" in about:config. By default, it waits about a half a second before painting the page (to avoid frequently flashing/rearranging the first page of contents for, e.g., people browsing over dial-up).
Google for initial paint delay to see more discussion of that. That's not to trivialize your experience--it could well be that FF is slower than some older browsers for intrinsic reasons instead of configuration reasons (examples include architectural decisions for compat/maintainability over speed, supporting more features/standards, etc.).
X86 machine code has also been "interpreted" for ages now. No hardware implements the x86 ISA as presented to programmers--they all convert it to internal opcodes using firmware. The importance of "interpreted" versus "compiled" stopped being relevant some time ago. The main issue now is just choosing the trade-off between speed and correctness (java VMs can be made reasonably faster if you turn off some of the safety checks; I'm not sure if that tuneability exists outside of the embedded VM implementations, though).
scp is pretty dumb about transferring files except in simple cases. My main peeve: it changes the mtime (to now), and has no equivalent of cp's -a (archive--preserves a bunch of stuff)--or in fact, any useful arguments pertaining to different ways to copy things.
rsync understands ssh, which makes it my preferred over-ssh transfer method. Where rsync is out of version sync or missing altogether, the tar method is a great standby (tar preserves all the original metadata by default, can be told to handle sparse files, and all sorts of other goodies). If I were a fan of cpio, that could probably work, too.
You're right, but what the grandparent describes really can't be hacked on easily. Authentication / authorization support must be designed into a system--there are many different aspects that need to hook into authorization services for them to be useful. Also, maintaining something that can implement company access policies (e.g., departmental politics) is a job in itself--especially if you're contradicting the goals the original software has ingrained into its design.
Just adding the usual friendly reminder that security (which authorization is a part of) can't be tacked on, but must be designed in early on.
Now when PHP et al gain aspect-oriented programming support, interesting possibilities may open themselves up in the "casual open-source/free-software hacking" front.
Perhaps it's part of the same effect where people do not really notice programs/operating systems crashing: they don't see there being any other way (or they don't care enough to seek out those ways). Computer geeks are aware of what software is capable of.
To apply this axiom to online ads, computer geeks are aware enough to know that there are ways to not see ads (like ad blockers, etc.), focus their abilities on achieving those goals, and thus get annoyed when they see ads (e.g., if it defeats their filtering or if they're using a public computer).
Non-computer-geeks, on the other hand, see ads as an uncontrollable reality and thus focus their abilities on ignoring them and setting up mental blocks.
To draw a strained analogy, the computer geeks are "clean" people who focus their efforts on cleaning and disinfecting their environment, so they never encounter diseases and thus never build immunities to them. Others, who experience "unclean" environments and occasionally get diseases, have stronger immune systems and thus the ability to cope with diseases when they get them (compare that to the "clean" people who will be much worsely affected if they ever do encounter a disease).
Another strained comparison: cultures living cities (with mostly domesticated animals) are easily aggravated by others' noisy pets, since they know that it's easy to silence those noises (by complaining to the owner, complaining to the city/law, etc). Compare that to living in a jungle environment, where the only way to handle the noises is to either build sound-proof houses or kill the noisy animals (not always possible).
I've noticed similar effects with TV compared to my own pre-DVR history: I get really annoyed when seeing ads on non-DVRed TV because I know the ads can theoretically be skipped, and I've gotten used to not having to act on my own ignore them (e.g., take bathroom breaks, make a sandwich, etc.).
As an added thought, computer geeks (especially the generation that got into the web in its early life in the 90s) remember that ads were not naturally a part of the web, but rather were added later in its life, so they know what a web without ads is like.
Anyway, those are just my thoughts and observations, based purely on speculation.
You might be able to get away without the second two calls if you pick a key combo that has no normal interpretation.
I mention this for those who want to hack their own installation until similar functionality gets integrated into FireFox. This change does not require any recompiling--just unpack the jar file, edit the file I mentioned, and repack the jar file.
> Similarly, system widgets aren't used even for in-page items like radio buttons, > checkboxes, buttons, text fields. I cannot fathom the hubris that makes the Mozilla > developers feel that their application is so uniquely important that it > deserves to look different than every other application on my system.
FYI, I heard the explanation for this some time ago in #mozilla: no system widgets exist that would allow mozilla/firefox to implement the CSS specification(s).
> * Bookmarking groups of tabs and then opening them all with a single click - Mozilla yes, Firefox no.
Bookmarks/Bookmark this Page/"Bookmark all tabs in a folder". Been there since pre-0.8.
> * Opening new blank tabs with a button - Mozilla yes, Firefox no.
View/Toolbars/Customize..., find the "Open New Tab" button, and drag it somewhere. I like having both new-tab and new-window next to each other for quick access. This has also been there for a while.
Summary: do not stay logged into X as root (in fact, it'd be better to not log into X as root at all--instead, use gksu or xterm+su or xterm+sudo, etc.).
Public Domain -> any business logic I can reverse engineer I can use in my commercial project without mention or recompense. Public Domain is not the same as free-to-download-and-use (the latter is a subset of the former); you can still place restrictions on free-to-download-and-use (since you're accomplishing it with a license).
An open source license will beat release to Public Domain if only because it still imposes restrictions, and because they retain copyright of the original. Note one definite advantage to the GPL (from a business's perspective) is that even though your source is out there, none of your competitors can legally use (well, redistribute programs based on) that source without using the GPL themselves.
Also, they (Adobe) almost certainly license third party components/libraries to use in FrameMaker, and they probably have to pay royalties for some of them. That aspect would kill both the Public Domain idea and the free-to-download-and-use idea.
Regarding the idea that reversed binaries != source, you're technically correct but the distinction is lost for most people who'd go that far (the disassembly still reveals all the logic for the system, even if it's not pretty and commented; commercial entities who are disassembling competitors' products don't care about getting pretty, commented source).
Actually, atomic commits means something totally different from global revision numbers. Having atomic commits means that a software failure during repository-modifying activities leaves everything in a well-defined state. That is, if you are committing your changes and the network connection dies, your computer dies, the OS or source control client crashes, you kill the source control client, etc., then your repository should not be corrupted (and it will be as though you never committed at all). With CVS, some files may have gotten committed to, and others not, leaving the repository in an unknown/inconsistent state.
Check out the 'modules' program at modules.sourceforge.net. It makes it fairly easy to switch between versions of any program you like (and choose to set up with modules).
It's distribution-agnostic (works on non-Linux UN*X, too). The common usage is more or less module load gcc2.95 (now, 'gcc' will execute gcc v2.95).
Anyway, as someone else mentioned, Debian already offers this for some packages (the package maintainer has to do some things differently to make it work). For gcc, there's a gcc-2.95 package which installs/usr/bin/gcc-2.95 (CC=gcc-2.95./configure; make CC=gcc-2.95; etc.). Relatively few packages are available under this scheme, though.
Debian also has 'alternatives' (/etc/alternatives) which route common binaries to a system-wide configurable version of a package (or meta package). Example uses of this system in mainline Debian are 'editor' (vim, emacs, nano,...), 'sensible-browser' (galeon, mozilla, firefox, konq,...), and 'java' (kaffe, sun, blackdown,...).
Microsoft changed the language and made apps that wouldn't run on Sun's JVM. IBM wrote an IDE in (standard) Java which produces (standard) Java. How is what IBM did related to what MS did?
> [snip] The modern C spec says that integer variables default to 0 [/snip]
Are you sure? Even stack-allocated variables? And we're talking about C, not C++, right? I recall that stack-allocated storage is not initialized in any case (since it requires extra instructions to clear it). Static storage, on the other hand, is promised to be initialized, since that's recorded at compile time (i.e., is stored in the binary).
check out 'modules' (for switching btw. versions)
on
Debugging Configure
·
· Score: 1
Check out the 'modules' program at modules.sourceforge.net. It makes it fairly easy to switch between versions of any program you like (and choose to set up with modules).
Re:Newest ver. still lock up on 1st website visite
on
Mozilla 1.4b Loosed
·
· Score: 1
Try creating a new profile. Run mozilla with -P and create a new profile called "temp" (or whatever). If it works fine there, maybe there's something weird going on with your profile.
Also, try downloading mozilla from mozilla.org and unpacking it somewhere (/tmp, $HOME, wherever you can write to as a user). Run it with -P and create a new profile with it (so as not to risk your real profile, since profiles are not always backwards-compatible).
Also, orthogonal to this discussion, it never hurts to back up $HOME/.mozilla from time to time: cd && tar cjf _mozilla-bak-`date -I`.tar.gz.mozilla
This isn't even strictly an HTML exploit (as claimed in the summary): take out the CSS (style="position:fixed") and it loses its teeth. I know of at least one other related crasher (it's reported in bugzilla) involving certain objects and position:..., and this is probably related.
Slashdot Mirror
Perhaps it is because NoScript breaks FlashBlock (not sure about AdBlock). I don't know if this is still the case, since I removed it (FlashBlock) when I noticed that NoScript was preventing it from working.
Depending on your test environment (if you're testing through a dial-up-emulator proxy or just files across an intranet), you might simply be seeing the initial paint delay. It may or may not be this (or it could even be that it's really slower), but you can search for "delay" in about:config. By default, it waits about a half a second before painting the page (to avoid frequently flashing/rearranging the first page of contents for, e.g., people browsing over dial-up).
Google for initial paint delay to see more discussion of that. That's not to trivialize your experience--it could well be that FF is slower than some older browsers for intrinsic reasons instead of configuration reasons (examples include architectural decisions for compat/maintainability over speed, supporting more features/standards, etc.).
For getting CPU governing on (and useful), I had to make the following changes to Debian Sid (which more or less tracks Ubuntu nowadays):
/etc/modules (these aren't loaded automatically):
/etc/default/cpufrequtils
Add to
speedstep-centrino
cpufreq_conservative
cpufreq_ondemand
cpufreq_powersave
Install the package 'cpufrequtils' and edit
I use:
ENABLE="true"
GOVERNOR="conservative"
MAX_SPEED=1700000
MIN_SPEED=600000
(the conservative governor gives it "inertia" to keep from changing speeds for momentary blips in usage)
X86 machine code has also been "interpreted" for ages now. No hardware implements the x86 ISA as presented to programmers--they all convert it to internal opcodes using firmware. The importance of "interpreted" versus "compiled" stopped being relevant some time ago. The main issue now is just choosing the trade-off between speed and correctness (java VMs can be made reasonably faster if you turn off some of the safety checks; I'm not sure if that tuneability exists outside of the embedded VM implementations, though).
scp is pretty dumb about transferring files except in simple cases. My main peeve: it changes the mtime (to now), and has no equivalent of cp's -a (archive--preserves a bunch of stuff)--or in fact, any useful arguments pertaining to different ways to copy things.
rsync understands ssh, which makes it my preferred over-ssh transfer method. Where rsync is out of version sync or missing altogether, the tar method is a great standby (tar preserves all the original metadata by default, can be told to handle sparse files, and all sorts of other goodies). If I were a fan of cpio, that could probably work, too.
You're right, but what the grandparent describes really can't be hacked on easily. Authentication / authorization support must be designed into a system--there are many different aspects that need to hook into authorization services for them to be useful. Also, maintaining something that can implement company access policies (e.g., departmental politics) is a job in itself--especially if you're contradicting the goals the original software has ingrained into its design.
Just adding the usual friendly reminder that security (which authorization is a part of) can't be tacked on, but must be designed in early on.
Now when PHP et al gain aspect-oriented programming support, interesting possibilities may open themselves up in the "casual open-source/free-software hacking" front.
WFM, same set-up, but unstable instead of testing.
Maybe try a new profile and/or with no extensions installed?
Or a "brilliant" marketroid. "Well, if it changes, they have to buy a new unit!"
https://bugzilla.mozilla.org/show_bug.cgi?id=55181
and
https://bugzilla.mozilla.org/show_bug.cgi?id=26018 6
seem to talk about the issue. (note Bugzilla doesn't permit slashdot links, so either copy-and-paste or open-link-in-new-window.)
This will bring Windows into the realm of normal property: now you can get pre-pwned Windows. (ba-dum ching)
Thank you, I'll be here all week.
Perhaps it's part of the same effect where people do not really notice programs/operating systems crashing: they don't see there being any other way (or they don't care enough to seek out those ways). Computer geeks are aware of what software is capable of.
To apply this axiom to online ads, computer geeks are aware enough to know that there are ways to not see ads (like ad blockers, etc.), focus their abilities on achieving those goals, and thus get annoyed when they see ads (e.g., if it defeats their filtering or if they're using a public computer).
Non-computer-geeks, on the other hand, see ads as an uncontrollable reality and thus focus their abilities on ignoring them and setting up mental blocks.
To draw a strained analogy, the computer geeks are "clean" people who focus their efforts on cleaning and disinfecting their environment, so they never encounter diseases and thus never build immunities to them. Others, who experience "unclean" environments and occasionally get diseases, have stronger immune systems and thus the ability to cope with diseases when they get them (compare that to the "clean" people who will be much worsely affected if they ever do encounter a disease).
Another strained comparison: cultures living cities (with mostly domesticated animals) are easily aggravated by others' noisy pets, since they know that it's easy to silence those noises (by complaining to the owner, complaining to the city/law, etc). Compare that to living in a jungle environment, where the only way to handle the noises is to either build sound-proof houses or kill the noisy animals (not always possible).
I've noticed similar effects with TV compared to my own pre-DVR history: I get really annoyed when seeing ads on non-DVRed TV because I know the ads can theoretically be skipped, and I've gotten used to not having to act on my own ignore them (e.g., take bathroom breaks, make a sandwich, etc.).
As an added thought, computer geeks (especially the generation that got into the web in its early life in the 90s) remember that ads were not naturally a part of the web, but rather were added later in its life, so they know what a web without ads is like.
Anyway, those are just my thoughts and observations, based purely on speculation.
This seems to work (this goes at the end of the section in browser.jar!/content/browser/search.xml:
<handler event="keypress" keycode="VK_RIGHT" modifiers="control" phase="capturing">
this.firstChild.click();
event.preventDefault();
event.stopPropagation();
</handler>
You might be able to get away without the second two calls if you pick a key combo that has no normal interpretation.
I mention this for those who want to hack their own installation until similar functionality gets integrated into FireFox. This change does not require any recompiling--just unpack the jar file, edit the file I mentioned, and repack the jar file.
> Similarly, system widgets aren't used even for in-page items like radio buttons,
> checkboxes, buttons, text fields. I cannot fathom the hubris that makes the Mozilla
> developers feel that their application is so uniquely important that it
> deserves to look different than every other application on my system.
FYI, I heard the explanation for this some time ago in #mozilla: no system widgets exist that would allow mozilla/firefox to implement the CSS specification(s).
"Open in tabs" - in every folder, including the ones you create using the method mentioned above.
> * Bookmarking groups of tabs and then opening them all with a single click - Mozilla yes, Firefox no.
Bookmarks/Bookmark this Page/"Bookmark all tabs in a folder". Been there since pre-0.8.
> * Opening new blank tabs with a button - Mozilla yes, Firefox no.
View/Toolbars/Customize..., find the "Open New Tab" button, and drag it somewhere. I like having both new-tab and new-window next to each other for quick access. This has also been there for a while.
Hope that helps.
See this xscreensaver FAQ.
Summary: do not stay logged into X as root (in fact, it'd be better to not log into X as root at all--instead, use gksu or xterm+su or xterm+sudo, etc.).
The abstract doesn't even come into play in any patent-related arguments. Claims are king, so check them before jumping to a conclusion.
Public Domain -> any business logic I can reverse engineer I can use in my commercial project without mention or recompense. Public Domain is not the same as free-to-download-and-use (the latter is a subset of the former); you can still place restrictions on free-to-download-and-use (since you're accomplishing it with a license).
An open source license will beat release to Public Domain if only because it still imposes restrictions, and because they retain copyright of the original. Note one definite advantage to the GPL (from a business's perspective) is that even though your source is out there, none of your competitors can legally use (well, redistribute programs based on) that source without using the GPL themselves.
Also, they (Adobe) almost certainly license third party components/libraries to use in FrameMaker, and they probably have to pay royalties for some of them. That aspect would kill both the Public Domain idea and the free-to-download-and-use idea.
Regarding the idea that reversed binaries != source, you're technically correct but the distinction is lost for most people who'd go that far (the disassembly still reveals all the logic for the system, even if it's not pretty and commented; commercial entities who are disassembling competitors' products don't care about getting pretty, commented source).
Such ends my rambling; have a good day.
Actually, atomic commits means something totally different from global revision numbers. Having atomic commits means that a software failure during repository-modifying activities leaves everything in a well-defined state. That is, if you are committing your changes and the network connection dies, your computer dies, the OS or source control client crashes, you kill the source control client, etc., then your repository should not be corrupted (and it will be as though you never committed at all). With CVS, some files may have gotten committed to, and others not, leaving the repository in an unknown/inconsistent state.
Check out the 'modules' program at modules.sourceforge.net. It makes it fairly easy to switch between versions of any program you like (and choose to set up with modules).
It's distribution-agnostic (works on non-Linux UN*X, too). The common usage is more or less module load gcc2.95 (now, 'gcc' will execute gcc v2.95).
Anyway, as someone else mentioned, Debian already offers this for some packages (the package maintainer has to do some things differently to make it work). For gcc, there's a gcc-2.95 package which installs /usr/bin/gcc-2.95 (CC=gcc-2.95 ./configure; make CC=gcc-2.95; etc.). Relatively few packages are available under this scheme, though.
Debian also has 'alternatives' (/etc/alternatives) which route common binaries to a system-wide configurable version of a package (or meta package). Example uses of this system in mainline Debian are 'editor' (vim, emacs, nano, ...), 'sensible-browser' (galeon, mozilla, firefox, konq, ...), and 'java' (kaffe, sun, blackdown, ...).
Microsoft changed the language and made apps that wouldn't run on Sun's JVM. IBM wrote an IDE in (standard) Java which produces (standard) Java. How is what IBM did related to what MS did?
> [snip] The modern C spec says that integer variables default to 0 [/snip]
Are you sure? Even stack-allocated variables? And we're talking about C, not C++, right? I recall that stack-allocated storage is not initialized in any case (since it requires extra instructions to clear it). Static storage, on the other hand, is promised to be initialized, since that's recorded at compile time (i.e., is stored in the binary).
Check out the 'modules' program at modules.sourceforge.net. It makes it fairly easy to switch between versions of any program you like (and choose to set up with modules).
Try creating a new profile. Run mozilla with -P and create a new profile called "temp" (or whatever). If it works fine there, maybe there's something weird going on with your profile.
.mozilla
Also, try downloading mozilla from mozilla.org and unpacking it somewhere (/tmp, $HOME, wherever you can write to as a user). Run it with -P and create a new profile with it (so as not to risk your real profile, since profiles are not always backwards-compatible).
Also, orthogonal to this discussion, it never hurts to back up $HOME/.mozilla from time to time: cd && tar cjf _mozilla-bak-`date -I`.tar.gz
This isn't even strictly an HTML exploit (as claimed in the summary): take out the CSS (style="position:fixed") and it loses its teeth. I know of at least one other related crasher (it's reported in bugzilla) involving certain objects and position:..., and this is probably related.