The slashes go "the wrong way" (hmm, any explicit biases there?)
In how many programming languages is \ not a special character in strings? MS uses C++ quite a bit, where you can't type stuff like ``/usr/local/etc'' in your apps.
URLs use/
/ is generally found in the same place and is otherwise typed less frequently.
Hell, slash is easier to say than backslash.
Combine those, and you get the confusion I've seen many times of people typing ``http:\\blah'' or trying to read out some large path on a Windows system, usually starting with \\ to access a remote share and having to say ``backslash backslash data backslash stuff'' vs. ``slash slash data slash stuff.''
That it's even called ``backslash'' should be an indicator. When was the last time you saw a document use backslash as a separator for anything?
I had a micra on my keychain until those bastards at airport security took it away from me and put it in their big ``confiscated terrorist tools'' box. I wasn't even flying, just dropping someone off. They could've held it for a couple of minutes like they did when they took my phone...but no, I had to either have them throw it away or have someone miss a flight.
I used to always fly with my wave, and I felt pretty damned good about it. Nowadays, you have to be really careful where you take such a terrifying weapon, which greatly reduces its usability.
I don't really play games (well, I do sometimes on my playstation or game cube or whatever), but I don't really see where this is coming from. On my old 1.6 box, there's something like 170 games in/usr/pkgsrc (including stuff I might actually play like quake 3 arena). Any Linux based game should work as well on x86. I wouldn't really know, though.
2. It cannot be used by my grandma.
It has, in fact, been used by my in-laws who have no computer experience at all (which would make them my kids' grandma and grandpa). Login, click here for a web browser, go!
3. It lacks a GUI of any note.
Said web browser was running in X. You should probably consider looking at something before trying to tell other people about it.
4. There is no support available for it.
Yeah, there is. See wasabi systems.
5. It is an assortment of fragmented OSes.
Wait, are we talking about NetBSD, or Linux? NetBSD is a single, solid, clean OS that looks the same on many different hardware platforms. Linux is many different operating systems that have very little to do with each other.
6. It cannot be run on the x86 platform.
This is supreme ignorance. The laptop used by the in-laws was a thinkpad (which is x86 based).
7. You have to compile everything and know C.
I don't believe I compiled Mozilla (I never compiled the Netscapes I used before that, but if I did compile Mozilla, it certainly wasn't significant enough for me to remember it).
8. Support for the latest hardware is always poor.
This statement is true regardless of platform. The things that are important to people vary by platform. I seem to recall significant hardware support reaching NetBSD before Linux in the past as well.
9. It is incompatiable with GNU/Linux.
Well, BSD and other Unixes were here first, so of GNU/Linux has incompatibilities, it isn't the fault of the BSDs.
This, however, is my #1 reason for not using Linux in general. It doesn't work like any other system I use (on my LAN: MacOS X, Solaris 8, SunOS 4.1.4, NetBSD (various versions), IRIX 6.2, IRIX 6.5, probably some others). There are things I can expect to work a certain way in all of those places (i.e. netstat).
10. It is dying.
Wait a second, so is Apple, and this Apple's running a BSD, does that mean it's a double negative?
So an attacker who can gain access to your network -- over a wired connection or wirelessly -- can trick an affected system into trusting a rogue machine, and when the compromised machine reboots, take it over and even attack other systems on the network.
So, a guy has to get on my network, set up another machine as a trusted server, wait for me to reboot, and then...? Is this a fair comparison to email viruses, etc...?
My cube's been up for 90 days. I plan to take it down and upgrade it eventually. Does this mean I'm going to be vulnerable?
Programming can be a creative process, but for 99.9% of projects it's not. How is putting next weeks sales targets on an intranet site creative?
How is that programming?
Or allowing customers to check their bank balance online?
This is more like the kind of work that programmers do, however, it's a lot more complicated than it sounds. It has to be designed. It has to be designed securely, and so that it scales with the amount of customers real banks have when they're all checking their accounts around the same time, and it has to be managable so enhancements don't require starting over. It has to be well-tested (which is an art in itself).
Those are engineering problems, and can easily be solved using well-understood methods and techologies.
This statement seems to imply that such things exist. This is not the case. If it were, we wouldn't have so many contradictory schools of The Right Way to engineer software.
There's nothing creative involved, nor should there be.
There is creativity involved, but perhaps it shouldn't be, and maybe it won't at some point.
Apple Newton - nice idea, bad implementation. Palm Pilot - same idea (copied), nice implementation.
PalmOS is the worst operating system I've ever used. It's extremely difficult to develop for, has horrible limitations, no standard ways to access any stored data, is just now coming around with ways to extend the storage (but only for very specific things and using an API unlike any other storage mechanism, etc...
NewtonOS was incredibly easy to develop for, applications were very consistent (it took work to make an application not support beaming, faxing, etc...), storage was consistent.
Actually, the storage was amazing...you could have an individual phone record stored on an external card device, and the ability to move it was consistent across all apps (unless someone worked to make it inconsistent). The database virtualization layer was such that a single query would return records from records from multiple sources simultaneously, and if one of those sources were to be removed, or a new one inserted, the UI would update with the new records.
I got palm primarily for interoperability. It was frustrating. It still is. Palms gained popularity because they were cheaper and less functional, not because they were more elegant. People weren't ready for PDAs, but they were cool with having a small electronic phone book thing to carry around.
The really amazing thing is the large amount software that's been developed for PalmOS. It's incredibly hard to write even trivial apps for palm, and there's almost no opportunity for code reuse, so things have to be rewritten from scratch.
Evidence: take a look at the memo pad application. It's a huge amount of code. Now, what if you want to write an app that reads/writes the memo DB? You have to reproduce or copy and paste a large chunk of it.
or just don't run rsync as a server. There's no need to for most uses anyway - just install the client at both ends and connect with the "-e ssh" flag and you're laughing
What if I don't want system users for every rsync user? What if I need to run my connections through an http proxy server (yes, I really, really do)? What if I want standard mechanisms for listing available modules? What if I want to limit the number of simultaneous connections for a specific area? What if I want to limit the files available in a specific area? What if I want to transfer sensitive files on a system periodically from cron, but I don't want to have an ssh key that grants access to do this without a password on the recipient machine?
I think that pretty much sums up the ways I most commonly use rsync around the house. I do use it with the -e ssh option for one-off things sometimes as well, but not running a server is certainly no workaround for me.
For some reason I've been under the impression that not all browsers supported single quotes, but that probably shows my age more than anything else.
It was valid in HTML 2.0, and I assume it still is. You're getting closer to the point, though...
I think what annoys me is the Rule Nazis basically say THEIR WAY is the One True HTML
HTML is a specification available to anyone who wants to read it, implement something that reads HTML, or write HTML him/herself. Ambiguity is *not* helpful in standards. I believe HTML 2.0 did allow you to sometimes omit quotes in attributes, but not always. There were specific characters that might cause your attribute to break. This gets even worse in auto-generated stuff where you might just end up with one of those things in a tag somewhere and suddenly your page doesn't render.
So things got easier. If an element has an attribute, you quote it. No exceptions. Parsers are easier and more reliable and people can worry about the harder problems.
If you break the rules, and do things that are not valid according to the specifications, you should expect things to break.
According to various references, the <center> tag is immoral. Instead, they want you to say <div align = "center">. They tell us that, in some abstract universe, this is easier to understand.
(do you really space out your elements like that? I've never seen that done)
``align'' is no longer a valid attribute of div for the same reason that ``center'' is no longer a valid element of HTML. It's layout. HTML didn't start out as a fancy layout language, just another markup language. Once people started using it for layout, it was realized that it's *really* bad at it, and documents could no longer be maintained. It's now broken into two parts:
HTML is for document structure (this is a header for this paragraph, etc...).
CSS is for layout (headers look like this, paragraphs look like that, etc...)
Why? What's wrong with keeping HTML a language that's easy for humans to read, maintain and understand?
Exactly! I've got *pages* of HTML in my editor that produce a single, small bit of content just in the right spot on the screen. Nested tables in nested tables with every other row and column stretching a 1pixel image to give *just* the right layout. We go in there and try to add a column and suddenly it's not even on the screen. I've replaced a lot of this kind of thing with the most simple possible HTML and CSS that can rearrange it *outside* of the document itself.
This is really important for things like the product search engine internal to my company. I can perform a search and get the results on my screen in pretty much plain HTML format, then pop open the CSS editor in firebird and manipulate it until it looks exactly right. Then I just put that css in the file this thing includes and nobody can tell the difference...
except the programmer/html guy who has to go in there next time we need to add a column or something.
It makes it consistent. To XML, it's just an attribute with a name and a value. The values can have spaces. What's the point of making specific HTML tags a special case?
There is no programming language that I know of that requires, or even allows, quoted numbers as numeric values.
There are a lot of programming languages out there (HTML isn't one of them). Many of them allow quoted numbers. sh, perl and tcl off the top of my head (not that I particularly like programming in any of those, but they meet your requirements). As far as requiring it, there's probably one of those out there, too.
I find the religious attitude of strict HTML purists a little silly. If it's no harder to code a browser so omitting quotes works, why force them on people, when in many ways the old style was far more readable?
Because the standards say it shouldn't work that way. If we all decide which parts of the standards we feel we should honor, we end up with the HTML of 1996 or so. Things really are better now (even though there still are decaying browsers around that people who use that don't support CSS properly just yet).
Sure, but I run this stuff on NetBSD, Solaris, FreeBSD, Mac OS X, IRIX, and SunOS 4.1.4 at home. It works for anything in any directories I want, and it automatically revision controls the files so they can be rolled back if you realize you made a mistake after you wake up the next morning.
as for arch, i've been looking into it, but it's still a hack onto cvs.
WTF are you talking about? Arch is only related to CVS in that it's in the same application category.
i'm spoiled and i've been trying to find a free (just as in beer, free as in speech would be nice, but i just want to get my work done quickly and without hassle) source control client that can mimic the functionality and stability of perforce, and i've yet to find one.
We use perforce at work and I'm fed up with the lack of functionality compared to what I get from arch (which I use for most of my home projects and home dir and stuff).
We have a few integrations with other software that makes perforce tolerable (bug tracking mostly). These could be done with arch as well, but it's not worth it to me yet.
You point it at a dir and run it from cron nightly. It also gives you a handy nightly mail telling you what changed. Excellent for those late night changes to systems where you don't remember what you did...or if someone else made some late night changes that you'd like to undo.
I'm not American, but I thought the issue was that he lied to a Grand Jury, not simply that he lied.
Sure, but that's not the argument I typically get from people. It's just that he lied. If a democrat lies in the US, it's over.
My personal feeling is that it's OK to lie about questions that shouldn't have been asked in the first place. We wasted millions of dollars worrying about some guy's sex life, and people still use him as an example of a horrible President because he lied about his sex life.
I don't really care about that. When will people start talking the same way about Bush lying about motivations for this war?
The issue doesn't start with the fact that he lied trying to protect his family from a scandal. The issue is that he was cheating on his wife in the first place!
I'm not sure where to begin on this...what does any of this have to do with being POTUS? There's a job he's hired to do. Whatever he does that does not affect this job is his own business.
Until we find something better than humans (or politicians or whatever) to run governments, we're going to have deal with them having human characteristics.
Honestly, unless you're Hillary Clinton, what difference does it make to you?
Bill Clinton also got caught lying to the entire USA and that don't matter becuase over half the population still loves him and his adorable wife.
Clinton got caught lying about something that wasn't anyone's business. It's his personal life and you've gotta be some odd kind of sheep to hate his wife for it.
MS lies about things that directly affect people in my industry. They do so to destroy competition (technology, jobs, etc...).
So, to review:
Clinton lies to protect his personal life and family from his own mistakes at the potential cost of his family.
MS lies to grow its business at the cost of jobs, technology, and freedom of people in the computer industry.
Bush lies to protect his investments at the cost of lives and governments.
Well, it's no fair that you have voice recognition or something on your computer so you can start the application without clicking.
It might be different in the Windows version, but on the Mac, I can place an audio CD in my slot loading drive, and it will automatically rip with my settings and spit the CD back out. It makes the process about as easy as it can be without creating a robot to change the CDs.
*None* of the MP3 players he's hawking resolves all 5 points
This is what I was looking for someone to point out. After each point, a player was shown that was better at that point. I didn't see two pictures of the same player. Does this suggest that no player is even good at two of them?
How do you cross a classloader boundary ? Just curious, I thought that was not possible.
Have a cache in a top level classloader, cache something in a lower classloader, and then pull it out from a peer classloader.
This has bitten me a few times when I've got a utility set of classes that include a class and are in the classpath of a servlet container. Each webapp trying to use the cache sees the others namespace, but can't make any sense of the objects in the cache.
Because if a user gets a message like "Traceback (most recent call last): File "", line 1, in ? TypeError: cannot concatenate 'str' and 'int' objects" then you might as well have said "5", it'd be about as meaningful.
There is a *huge* difference between getting the wrong answer (5) and getting an exception. As frustrating as an error may be for a user, it's far, far better than misleading the user.
Lets face it: the only languages suitable for a production environment are statically typed.
Static typing isn't enough, it must also be strongly typed. I write a lot of java (semi-statically typed, semi-strongly typed sometimes-ish). I can still write code and lie to the compiler about a type and have it break at runtime. Or worse, I can cross a classloader boundary and get ClassCastExceptions on objects that are the right type. Ugh.
I agree with your basic point, though. Ocaml wins here. Types are strong, static, and inferred. Runtime is insanely fast. Exceptions work reliably where I want them.
My lazily written ocaml is as reliable as my carefully written java, but much smaller and prettier.:)
If you use '-w' with Perl, as is recommended by virtually every Perl guru in existence, you get:
bash$./test.pl Argument "some string" isn't numeric in addition (+) at./test.pl line 6. 5
That seems pretty non-mysterious, doesn't it?
You still got 5. Your program still broke. However, if you look around through wherever that log goes when it made it to ``production,'' you might have a clue as to what the condition was that caused it to break.
Slashdot Mirror
The slashes go "the wrong way" (hmm, any explicit biases there?)
/
In how many programming languages is \ not a special character in strings? MS uses C++ quite a bit, where you can't type stuff like ``/usr/local/etc'' in your apps.
URLs use
/ is generally found in the same place and is otherwise typed less frequently.
Hell, slash is easier to say than backslash.
Combine those, and you get the confusion I've seen many times of people typing ``http:\\blah'' or trying to read out some large path on a Windows system, usually starting with \\ to access a remote share and having to say ``backslash backslash data backslash stuff'' vs. ``slash slash data slash stuff.''
That it's even called ``backslash'' should be an indicator. When was the last time you saw a document use backslash as a separator for anything?
The original point was about / being writable. The problem is that if / is writable (but not sticky), then it'd be possible to do this:
./
cp -r etc myetc; mv etc etc.old ; mv myetc etc
And then you control etc.
However, due to the sticky bit:
dustin2wti:/tmp/test 520% ls -ld . etc
drwxrwxr-t 3 root admin 102 15 Dec 14:10
drwxr-xr-x 2 root wheel 68 15 Dec 14:10 etc/
dustin2wti:/tmp/test 521% mv etc newetc
mv: rename etc to newetc: Operation not permitted
(because of the sticky bit and my lack of ownership over etc)
Remember, renames are *directory* modifications, not file modifications. The sticky bit fills in the difference.
I have a Leatherman Micra
I had a micra on my keychain until those bastards at airport security took it away from me and put it in their big ``confiscated terrorist tools'' box. I wasn't even flying, just dropping someone off. They could've held it for a couple of minutes like they did when they took my phone...but no, I had to either have them throw it away or have someone miss a flight.
I used to always fly with my wave, and I felt pretty damned good about it. Nowadays, you have to be really careful where you take such a terrifying weapon, which greatly reduces its usability.
But hey, we all feel safer, right?
Just in case someone's reading this crap:
/usr/pkgsrc (including stuff I might actually play like quake 3 arena). Any Linux based game should work as well on x86. I wouldn't really know, though.
1. You can not play games on it.
I don't really play games (well, I do sometimes on my playstation or game cube or whatever), but I don't really see where this is coming from. On my old 1.6 box, there's something like 170 games in
2. It cannot be used by my grandma.
It has, in fact, been used by my in-laws who have no computer experience at all (which would make them my kids' grandma and grandpa). Login, click here for a web browser, go!
3. It lacks a GUI of any note.
Said web browser was running in X. You should probably consider looking at something before trying to tell other people about it.
4. There is no support available for it.
Yeah, there is. See wasabi systems.
5. It is an assortment of fragmented OSes.
Wait, are we talking about NetBSD, or Linux? NetBSD is a single, solid, clean OS that looks the same on many different hardware platforms. Linux is many different operating systems that have very little to do with each other.
6. It cannot be run on the x86 platform.
This is supreme ignorance. The laptop used by the in-laws was a thinkpad (which is x86 based).
7. You have to compile everything and know C.
I don't believe I compiled Mozilla (I never compiled the Netscapes I used before that, but if I did compile Mozilla, it certainly wasn't significant enough for me to remember it).
8. Support for the latest hardware is always poor.
This statement is true regardless of platform. The things that are important to people vary by platform. I seem to recall significant hardware support reaching NetBSD before Linux in the past as well.
9. It is incompatiable with GNU/Linux.
Well, BSD and other Unixes were here first, so of GNU/Linux has incompatibilities, it isn't the fault of the BSDs.
This, however, is my #1 reason for not using Linux in general. It doesn't work like any other system I use (on my LAN: MacOS X, Solaris 8, SunOS 4.1.4, NetBSD (various versions), IRIX 6.2, IRIX 6.5, probably some others). There are things I can expect to work a certain way in all of those places (i.e. netstat).
10. It is dying.
Wait a second, so is Apple, and this Apple's running a BSD, does that mean it's a double negative?
So an attacker who can gain access to your network -- over a wired connection or wirelessly -- can trick an affected system into trusting a rogue machine, and when the compromised machine reboots, take it over and even attack other systems on the network.
So, a guy has to get on my network, set up another machine as a trusted server, wait for me to reboot, and then...? Is this a fair comparison to email viruses, etc...?
My cube's been up for 90 days. I plan to take it down and upgrade it eventually. Does this mean I'm going to be vulnerable?
Whatever.
Programming can be a creative process, but for 99.9% of projects it's not. How is putting next weeks sales targets on an intranet site creative?
How is that programming?
Or allowing customers to check their bank balance online?
This is more like the kind of work that programmers do, however, it's a lot more complicated than it sounds. It has to be designed. It has to be designed securely, and so that it scales with the amount of customers real banks have when they're all checking their accounts around the same time, and it has to be managable so enhancements don't require starting over. It has to be well-tested (which is an art in itself).
Those are engineering problems, and can easily be solved using well-understood methods and techologies.
This statement seems to imply that such things exist. This is not the case. If it were, we wouldn't have so many contradictory schools of The Right Way to engineer software.
There's nothing creative involved, nor should there be.
There is creativity involved, but perhaps it shouldn't be, and maybe it won't at some point.
Apple Newton - nice idea, bad implementation.
Palm Pilot - same idea (copied), nice implementation.
PalmOS is the worst operating system I've ever used. It's extremely difficult to develop for, has horrible limitations, no standard ways to access any stored data, is just now coming around with ways to extend the storage (but only for very specific things and using an API unlike any other storage mechanism, etc...
NewtonOS was incredibly easy to develop for, applications were very consistent (it took work to make an application not support beaming, faxing, etc...), storage was consistent.
Actually, the storage was amazing...you could have an individual phone record stored on an external card device, and the ability to move it was consistent across all apps (unless someone worked to make it inconsistent). The database virtualization layer was such that a single query would return records from records from multiple sources simultaneously, and if one of those sources were to be removed, or a new one inserted, the UI would update with the new records.
I got palm primarily for interoperability. It was frustrating. It still is. Palms gained popularity because they were cheaper and less functional, not because they were more elegant. People weren't ready for PDAs, but they were cool with having a small electronic phone book thing to carry around.
The really amazing thing is the large amount software that's been developed for PalmOS. It's incredibly hard to write even trivial apps for palm, and there's almost no opportunity for code reuse, so things have to be rewritten from scratch.
Evidence: take a look at the memo pad application. It's a huge amount of code. Now, what if you want to write an app that reads/writes the memo DB? You have to reproduce or copy and paste a large chunk of it.
or just don't run rsync as a server. There's no need to for most uses anyway - just install the client at both ends and connect with the "-e ssh" flag and you're laughing
What if I don't want system users for every rsync user? What if I need to run my connections through an http proxy server (yes, I really, really do)? What if I want standard mechanisms for listing available modules? What if I want to limit the number of simultaneous connections for a specific area? What if I want to limit the files available in a specific area? What if I want to transfer sensitive files on a system periodically from cron, but I don't want to have an ssh key that grants access to do this without a password on the recipient machine?
I think that pretty much sums up the ways I most commonly use rsync around the house. I do use it with the -e ssh option for one-off things sometimes as well, but not running a server is certainly no workaround for me.
I found the opposite to be true. Does that mean we cancel each other out?
For some reason I've been under the impression that not all browsers supported single quotes, but that probably shows my age more than anything else.
It was valid in HTML 2.0, and I assume it still is. You're getting closer to the point, though...
I think what annoys me is the Rule Nazis basically say THEIR WAY is the One True HTML
HTML is a specification available to anyone who wants to read it, implement something that reads HTML, or write HTML him/herself. Ambiguity is *not* helpful in standards. I believe HTML 2.0 did allow you to sometimes omit quotes in attributes, but not always. There were specific characters that might cause your attribute to break. This gets even worse in auto-generated stuff where you might just end up with one of those things in a tag somewhere and suddenly your page doesn't render.
So things got easier. If an element has an attribute, you quote it. No exceptions. Parsers are easier and more reliable and people can worry about the harder problems.
If you break the rules, and do things that are not valid according to the specifications, you should expect things to break.
According to various references, the <center> tag is immoral. Instead, they want you to say <div align = "center">. They tell us that, in some abstract universe, this is easier to understand.
(do you really space out your elements like that? I've never seen that done)
``align'' is no longer a valid attribute of div for the same reason that ``center'' is no longer a valid element of HTML. It's layout. HTML didn't start out as a fancy layout language, just another markup language. Once people started using it for layout, it was realized that it's *really* bad at it, and documents could no longer be maintained. It's now broken into two parts:
HTML is for document structure (this is a header for this paragraph, etc...).
CSS is for layout (headers look like this, paragraphs look like that, etc...)
Why? What's wrong with keeping HTML a language that's easy for humans to read, maintain and understand?
Exactly! I've got *pages* of HTML in my editor that produce a single, small bit of content just in the right spot on the screen. Nested tables in nested tables with every other row and column stretching a 1pixel image to give *just* the right layout. We go in there and try to add a column and suddenly it's not even on the screen. I've replaced a lot of this kind of thing with the most simple possible HTML and CSS that can rearrange it *outside* of the document itself.
This is really important for things like the product search engine internal to my company. I can perform a search and get the results on my screen in pretty much plain HTML format, then pop open the CSS editor in firebird and manipulate it until it looks exactly right. Then I just put that css in the file this thing includes and nobody can tell the difference...
except the programmer/html guy who has to go in there next time we need to add a column or something.
Fry's is also in other reputable locations:
Can you be specific as to which of those are reputable?
<a href = http://www.amazing.com/>amazing.com<a>
Are you sure that was ever correct?
print("<img src = \"foo.gif\" height = \"100\" width = \"50\">\n");
That's still wrong (unless you didn't mean for that spacing to be in there), how about this:
print("<img src='foo.gif' height='100' width='50'>\n")
What earthly purpose does quoting numbers serve?
It makes it consistent. To XML, it's just an attribute with a name and a value. The values can have spaces. What's the point of making specific HTML tags a special case?
There is no programming language that I know of that requires, or even allows, quoted numbers as numeric values.
There are a lot of programming languages out there (HTML isn't one of them). Many of them allow quoted numbers. sh, perl and tcl off the top of my head (not that I particularly like programming in any of those, but they meet your requirements). As far as requiring it, there's probably one of those out there, too.
I find the religious attitude of strict HTML purists a little silly. If it's no harder to code a browser so omitting quotes works, why force them on people, when in many ways the old style was far more readable?
Because the standards say it shouldn't work that way. If we all decide which parts of the standards we feel we should honor, we end up with the HTML of 1996 or so. Things really are better now (even though there still are decaying browsers around that people who use that don't support CSS properly just yet).
So, do tell, how do you use CSS to replace tables and frames?
You use tables where tabular data should be presented, and not where it shouldn't.
For example, the following page contains no tables (but it used to, and it was really difficult to manage changes):
http://bleu.west.spy.net/~dustin/
Sure, but I run this stuff on NetBSD, Solaris, FreeBSD, Mac OS X, IRIX, and SunOS 4.1.4 at home. It works for anything in any directories I want, and it automatically revision controls the files so they can be rolled back if you realize you made a mistake after you wake up the next morning.
as for arch, i've been looking into it, but it's still a hack onto cvs.
WTF are you talking about? Arch is only related to CVS in that it's in the same application category.
i'm spoiled and i've been trying to find a free (just as in beer, free as in speech would be nice, but i just want to get my work done quickly and without hassle) source control client that can mimic the functionality and stability of perforce, and i've yet to find one.
We use perforce at work and I'm fed up with the lack of functionality compared to what I get from arch (which I use for most of my home projects and home dir and stuff).
We have a few integrations with other software that makes perforce tolerable (bug tracking mostly). These could be done with arch as well, but it's not worth it to me yet.
I have a script that does all of this for me:
http://bleu.west.spy.net/~dustin/soft/filemonitor
You point it at a dir and run it from cron nightly. It also gives you a handy nightly mail telling you what changed. Excellent for those late night changes to systems where you don't remember what you did...or if someone else made some late night changes that you'd like to undo.
I'm not American, but I thought the issue was that he lied to a Grand Jury, not simply that he lied.
Sure, but that's not the argument I typically get from people. It's just that he lied. If a democrat lies in the US, it's over.
My personal feeling is that it's OK to lie about questions that shouldn't have been asked in the first place. We wasted millions of dollars worrying about some guy's sex life, and people still use him as an example of a horrible President because he lied about his sex life.
I don't really care about that. When will people start talking the same way about Bush lying about motivations for this war?
The issue doesn't start with the fact that he lied trying to protect his family from a scandal. The issue is that he was cheating on his wife in the first place!
I'm not sure where to begin on this...what does any of this have to do with being POTUS? There's a job he's hired to do. Whatever he does that does not affect this job is his own business.
Until we find something better than humans (or politicians or whatever) to run governments, we're going to have deal with them having human characteristics.
Honestly, unless you're Hillary Clinton, what difference does it make to you?
Bill Clinton also got caught lying to the entire USA and that don't matter becuase over half the population still loves him and his adorable wife.
Clinton got caught lying about something that wasn't anyone's business. It's his personal life and you've gotta be some odd kind of sheep to hate his wife for it.
MS lies about things that directly affect people in my industry. They do so to destroy competition (technology, jobs, etc...).
So, to review:
Clinton lies to protect his personal life and family from his own mistakes at the potential cost of his family.
MS lies to grow its business at the cost of jobs, technology, and freedom of people in the computer industry.
Bush lies to protect his investments at the cost of lives and governments.
Everybody lies. Why do you lie?
Sounds just like Windows XP's CD ripper.
Will it rip to any standard formats? I can configure iTunes to rip to AIFF, MP3 or AAC.
Not that ``rip and spit'' is anything novel, I've written shell scripts to do the same thing, but that's not the original point.
Well, it's no fair that you have voice recognition or something on your computer so you can start the application without clicking.
It might be different in the Windows version, but on the Mac, I can place an audio CD in my slot loading drive, and it will automatically rip with my settings and spit the CD back out. It makes the process about as easy as it can be without creating a robot to change the CDs.
*None* of the MP3 players he's hawking resolves all 5 points
This is what I was looking for someone to point out. After each point, a player was shown that was better at that point. I didn't see two pictures of the same player. Does this suggest that no player is even good at two of them?
How do you cross a classloader boundary ? Just curious, I thought that was not possible.
Have a cache in a top level classloader, cache something in a lower classloader, and then pull it out from a peer classloader.
This has bitten me a few times when I've got a utility set of classes that include a class and are in the classpath of a servlet container. Each webapp trying to use the cache sees the others namespace, but can't make any sense of the objects in the cache.
Yes, I'm doing so right now, in fact.
:)
Because if a user gets a message like "Traceback (most recent call last): File "", line 1, in ? TypeError: cannot concatenate 'str' and 'int' objects" then you might as well have said "5", it'd be about as meaningful.
There is a *huge* difference between getting the wrong answer (5) and getting an exception. As frustrating as an error may be for a user, it's far, far better than misleading the user.
Lets face it: the only languages suitable for a production environment are statically typed.
Static typing isn't enough, it must also be strongly typed. I write a lot of java (semi-statically typed, semi-strongly typed sometimes-ish). I can still write code and lie to the compiler about a type and have it break at runtime. Or worse, I can cross a classloader boundary and get ClassCastExceptions on objects that are the right type. Ugh.
I agree with your basic point, though. Ocaml wins here. Types are strong, static, and inferred. Runtime is insanely fast. Exceptions work reliably where I want them.
My lazily written ocaml is as reliable as my carefully written java, but much smaller and prettier.
If you use '-w' with Perl, as is recommended by virtually every Perl guru in existence, you get:
./test.pl ./test.pl line 6.
bash$
Argument "some string" isn't numeric in addition (+) at
5
That seems pretty non-mysterious, doesn't it?
You still got 5. Your program still broke. However, if you look around through wherever that log goes when it made it to ``production,'' you might have a clue as to what the condition was that caused it to break.