I just tried 1.4 and a recent fix that was in the nightly that I've been using for the last two weeks didn't make it in. The Mozilla developers finally fixed it so that the new mail notification can play sounds! Before, no matter what sound you specified, it would always beep the system beeper. Too bad the fix didn't make it into 1.4, I was looking forward to using an actual release for a while...
Have you ever written a parser? I have written several. I start with a tokenizer (lexer) that is built using regular expressions. My favorite tool for generating lexers is JFlex. The job of the lexer is to break the stream into tokens. Regular expressions are very useful for this.
Once the lexer is complete, a parser can be built. The tokens from the lexer are assembled into a parse tree. My favorite tool for generating parsers is CUP. The grammar for the parser is usually specified in BNF form.
As you can see, the parser depends on the lexer and the lexer depends on regular expressions. If you really want to be picky about it, there are some parser generators (javacc) that let you stick regular expressions right on the leaves of the grammar and handle both the parsing and the lexing. So yes, you need more than regular expressions to do parsing, but parsers are still almost always based on regular expressions.
I was just looking over the documentation for JFlex and non-greedy regular expression matching isn't mentioned. I'll be that means that it isn't implemented, but I haven't tested it. Since I write parsers using JFlex, its a good thing I can come up with the non-greedy syntax when I need to.
In some regular expression packages (not all) [\x00-\x7f] can be written as [^]. That is, all the characters that are not in the empty set of characters. Very nice shorthand. I also like the [A]|[^A] does the same thing.
Ah yes, if you have that feature of regular expressions, that is true. However, non-greedy matching is a regex feature that I found is not reliably implemented everywhere that I need it.
Hmmm, I just tried it in my text editor and it worked. Maybe its more widely implemented than I thought. I could have sworn this didn't work last time I tried it there. Maybe it was added in a recent version.:-)
You make an excellent point. The regular expression I came up with would not do the right thing in that situation when finding comments in your text editor.
Parsers are, however, based on regular expressions. I orginally wrote this regular expression when I was writing a lexer (using JFlex) for Java. The examples that I saw used a state machine and I wanted to do it with a regex. When combined with regular expression to find sting literals (and all the regular expressions for other junk), it does the right thing.
I should put your example on the page somewhere.:-)
I know from my own experiences that writing a regular expression to describe something is not always as easy as it would seem at first glance. I found it difficult to write a regular expression to define a c-style comment:/* comment */ Well, not impossible, just more difficult that I thought it would be. I posted my thought process about how I constructed a regular expression to pick out a c-style comment on my website. It's the kind of thing I like to ask interview candidates.
USB has allowed me to give up an expensive KVM. All the computers in my rack now have a couple USB ports.
Get a USB keyboard with a built in hub (~$50)
Get a USB mouse with a scroll wheel (~$30)
Get a USB switch (~$20)
Plug Each of your computers into the switch's outputs
Plug your keyboard into the switch input
Plug your mouse into your keyboard
For less that $100 you won't have to by an expensive (up to $500) KVM.
There are a couple drawbacks:
I don't know of a USB monitor, so you will need a different monitor switch.
Because this actually disconnects your mouse and keyboard from the computer, they may take a few seconds be be recognized by the computer and start working when you switch
You can get monitor/usb KVM switches, I don't know how much they cost, but the scroll wheel will probably work just fine though them as well.
Gnome 1 had lots of applets but no system tray. Gnome 2 has a system tray so that many applets weren't needed. Many applets were not ported to Gnome 2. The version of Gnome 2 that came with rh 8 didn't have the system tray. It didn't have the applets. Oops.
I'd have to agree in man respects. For a Gnome desktop user, 7.3 --> 8.0 is a huge step backwards. 8.0 uses Gnome 2, but it didn't have many bells and whistles yet. The version in 8.0 didn't have any panel appelets, nor did it have the system tray that was designed to eliminate the need for many panel applets. The metacity window manager in 8.0 is very basic compared to sawfish in 7.3. Also, Apache 2.0 doesn't work with my favoriet apache mod: jserve.
If you haven't tried rh 9.0 yet, try it. It has applets again (not as many, but it has the important ones). It also has the system tray. It still uses metacity, but you can switch back to sawfish by killing metacity and then starting sawfish before metacity restarts. I guess I should switch to Apache Tomcat to replace mod_jserve.
Now I'm starting to think that my 7.3 box is looking a bit shabby and that I should upgrade to 9.0.
They are not an original work of authorship, but instead an application of a mathematical algorithm to "fingerprint" a file; they're just a list of numbers. That would be like copyrighting the output of "ls -l".
The numbers are not just any list of numbers, they are a list of numbers mathematically derived from copyright material. You could certainly prevent somobody from distributing information gleaned from your files including the file names, sizes, and permissions. You created that information, you hold copyright over it. Yes the output of ls -l falls under copyright rules.
the hashes would certainly fall under the "fair use" exception to the copyright on the SCO files, as they are a form of commentary on them.
I don't see how fair use would apply here. Fair use has exceptions to copyright for commentary, but only for short excerpts such as a quote or a video clip. It also has exceptions for parody, which clearly does not apply here. The hashes basically contain the entire work in a different (albeit unrecoverable) form. As far as I know fair use has no provisions for such a situation.
As so many people have noted, you shouldn't look to Slashdot for legal advice. I would certainly ask an intellectual property lawyer before distributing the hash sums from SCO code. A copyright lawyer may tell you that it won't be a problem, but I doubt it.
When I read the headline, I thought how could comparing MD5 sums of files work? I'm sure the files have been modified at both ends. Even if they were from the same source orginally, there is no way they would match now.
The scheme instead computes an MD5 sum for each line (actually each five lines together) and publishes the hashed versions of the files. Then anybody can do the line by line compare without ever seeing a readable version of the source code.
The theory is that SCO can't complain about somebody distributing these hashes because you can't get the source code from it.
The only problem I see is that the hashes are still derived from SCO's intellectual property and are therefore still covered by copyright. SCO could still put up a stink about it. Especially since they have stated that their goal is to sue IBM for money rather than to identify the peices and rectify the situation. SCO has said that they are afraid that if the lines are known, the problem will be fixed and they won't be able to sue any more. (Poor babies.)
The previous article you point to shows recommendations from a group of companies that argue that bug reports should not be made public. Bugtraq does not follow this recommendation, and I doubt that it ever will. Bugtraq fully discloses bugs to the general public and I don't see that changing any time soon.
Unless bugtraq is falling down on the job, why do we need another one?
Just 40 some odd volts?
on
42-Volt Autos
·
· Score: 1
My car already has a 144 volt electrical system. (Enough to power the car up a hill.) With the battery banks in the trunk, I really don't have to worry about leaving the dome light on either.
Re:The Java.net creator is on to something
on
Sun Opens Java.net
·
· Score: 4, Insightful
Java could really benifit from something along the lines of what CPAN has done for Perl. There is a heck of a lot of Java software out there, but I know of no central repository from which you can download, compile, and test, in a standard way.
The CPAN commmunity makes all this happen. If you have never used it, it is quite slick. The process goes something like this:
# install somenew:package
downloading |----------------100%|
compiling |----------------100%|
testing |--------------OK|
The Giant Java Tree is a nice collection of Java software, but it doesn't have the automatic downloads or tests. It also isn't as comprehensive as CPAN, but that is mostly because of the mountains of Java software available.
I can see myself using generics when designing APIs. Lets say that I currently have a method that returns a list of Strings. Internally I store those Strings in a Vector so that I can add to it easily. I want users to be able to access these Strings so I create a getStrings() method. Since want the user to know that they are Strings and not have to do any casting I currently do something like this:
public String[] getStrings(){ String[] s = new String[vector.size()]; return vector.toArray(s); }
Maybe in the future I could just store them in a list and return that list because it is a string list specifically:
> I can't quite get my head round what combination of
> VM, compiler and/or language changes would iron out
> the following incompability between arrays and
> generics:
>
> Cat[] cats = new Cat[10];
> Animal[] animals = cats;// legal
>
> List cats = new List(10);
> List animals = cats;// illegal
> according to JSR-14
This problem is addressed in the "variance" extension to
the generic type system, which will be included in the
imminent JSR14 prototype 2.0. Details are enclosed in
the prototype.
Placement of the task bar top vs bottom is hardly analogous to round vs square wheels. Maybe snow tires vs chains or wide vs narrow. Other than you tires have to be able to turn (you task bar has to do its job) there doesn't seem to be One Right Way for tires either.
A task bar at the top is a horrible idea if you like to have them auto hidden. I prefer my task bar to auto-hide because it takes up much less space. However, when its at the top it pops up when you overshoot the close button, and then obscures the close button. Similar for using the scroll bar if you put it on the right. I have one bar on the bottom and one on the left, and it works well for me.
Bottom line: There is no One Right Way, it needs to be configurable.
Slashdot Mirror
I think it was only a Linux bug. I wouldn't know if it ever worked on Windows.
I just tried 1.4 and a recent fix that was in the nightly that I've been using for the last two weeks didn't make it in. The Mozilla developers finally fixed it so that the new mail notification can play sounds! Before, no matter what sound you specified, it would always beep the system beeper. Too bad the fix didn't make it into 1.4, I was looking forward to using an actual release for a while...
Once the lexer is complete, a parser can be built. The tokens from the lexer are assembled into a parse tree. My favorite tool for generating parsers is CUP. The grammar for the parser is usually specified in BNF form.
As you can see, the parser depends on the lexer and the lexer depends on regular expressions. If you really want to be picky about it, there are some parser generators (javacc) that let you stick regular expressions right on the leaves of the grammar and handle both the parsing and the lexing. So yes, you need more than regular expressions to do parsing, but parsers are still almost always based on regular expressions.
Parsers are based on lexers which are based on regular expressions. Is that clear then?
I was just looking over the documentation for JFlex and non-greedy regular expression matching isn't mentioned. I'll be that means that it isn't implemented, but I haven't tested it. Since I write parsers using JFlex, its a good thing I can come up with the non-greedy syntax when I need to.
In some regular expression packages (not all) [\x00-\x7f] can be written as [^]. That is, all the characters that are not in the empty set of characters. Very nice shorthand. I also like the [A]|[^A] does the same thing.
You are very welcome. I used regular expressions to build the parser, so you arn't too far offtopic. ;-)
Hmmm, I just tried it in my text editor and it worked. Maybe its more widely implemented than I thought. I could have sworn this didn't work last time I tried it there. Maybe it was added in a recent version. :-)
Parsers are, however, based on regular expressions. I orginally wrote this regular expression when I was writing a lexer (using JFlex) for Java. The examples that I saw used a state machine and I wanted to do it with a regex. When combined with regular expression to find sting literals (and all the regular expressions for other junk), it does the right thing.
I should put your example on the page somewhere. :-)
I know from my own experiences that writing a regular expression to describe something is not always as easy as it would seem at first glance. I found it difficult to write a regular expression to define a c-style comment: /* comment */ Well, not impossible, just more difficult that I thought it would be. I posted my thought process about how I constructed a regular expression to pick out a c-style comment on my website. It's the kind of thing I like to ask interview candidates.
- Get a USB keyboard with a built in hub (~$50)
- Get a USB mouse with a scroll wheel (~$30)
- Get a USB switch (~$20)
- Plug Each of your computers into the switch's outputs
- Plug your keyboard into the switch input
- Plug your mouse into your keyboard
For less that $100 you won't have to by an expensive (up to $500) KVM.There are a couple drawbacks:
- I don't know of a USB monitor, so you will need a different monitor switch.
- Because this actually disconnects your mouse and keyboard from the computer, they may take a few seconds be be recognized by the computer and start working when you switch
You can get monitor/usb KVM switches, I don't know how much they cost, but the scroll wheel will probably work just fine though them as well.Gnome 1 had lots of applets but no system tray. Gnome 2 has a system tray so that many applets weren't needed. Many applets were not ported to Gnome 2. The version of Gnome 2 that came with rh 8 didn't have the system tray. It didn't have the applets. Oops.
If you haven't tried rh 9.0 yet, try it. It has applets again (not as many, but it has the important ones). It also has the system tray. It still uses metacity, but you can switch back to sawfish by killing metacity and then starting sawfish before metacity restarts. I guess I should switch to Apache Tomcat to replace mod_jserve.
Now I'm starting to think that my 7.3 box is looking a bit shabby and that I should upgrade to 9.0.
As so many people have noted, you shouldn't look to Slashdot for legal advice. I would certainly ask an intellectual property lawyer before distributing the hash sums from SCO code. A copyright lawyer may tell you that it won't be a problem, but I doubt it.
The scheme instead computes an MD5 sum for each line (actually each five lines together) and publishes the hashed versions of the files. Then anybody can do the line by line compare without ever seeing a readable version of the source code.
The theory is that SCO can't complain about somebody distributing these hashes because you can't get the source code from it.
The only problem I see is that the hashes are still derived from SCO's intellectual property and are therefore still covered by copyright. SCO could still put up a stink about it. Especially since they have stated that their goal is to sue IBM for money rather than to identify the peices and rectify the situation. SCO has said that they are afraid that if the lines are known, the problem will be fixed and they won't be able to sue any more. (Poor babies.)
The bug finding, reporting, fixing, and patching process should minimize the potential damage. If your goal is to minimize damage then neither full immediate discloser or no disclosure is a good answer. Bruce Schneier has written a good article about full disclosure in his Crypto-Gram newsletter.
Unless bugtraq is falling down on the job, why do we need another one?
My car already has a 144 volt electrical system. (Enough to power the car up a hill.) With the battery banks in the trunk, I really don't have to worry about leaving the dome light on either.
The CPAN commmunity makes all this happen. If you have never used it, it is quite slick. The process goes something like this:
# install somenew:package
downloading |----------------100%|
compiling |----------------100%|
testing |--------------OK|
The Giant Java Tree is a nice collection of Java software, but it doesn't have the automatic downloads or tests. It also isn't as comprehensive as CPAN, but that is mostly because of the mountains of Java software available.
GCC Printers Elite Color 16 DN $2099
That should be plenty for a church.
Maybe in the future I could just store them in a list and return that list because it is a string list specifically:
From http://forum.java.sun.com/thread.jsp?forum=316&thr ead=389987&start=15&range=15&tstart=0&trange=1 5:
> I can't quite get my head round what combination of // legal // illegal
> VM, compiler and/or language changes would iron out
> the following incompability between arrays and
> generics:
>
> Cat[] cats = new Cat[10];
> Animal[] animals = cats;
>
> List cats = new List(10);
> List animals = cats;
> according to JSR-14
This problem is addressed in the "variance" extension to
the generic type system, which will be included in the
imminent JSR14 prototype 2.0. Details are enclosed in
the prototype.
Also a link to a MIT research paper on variance from that thread.
Placement of the task bar top vs bottom is hardly analogous to round vs square wheels. Maybe snow tires vs chains or wide vs narrow. Other than you tires have to be able to turn (you task bar has to do its job) there doesn't seem to be One Right Way for tires either.
Bottom line: There is no One Right Way, it needs to be configurable.