The idea behind a test system, generally, is that "wrecking" it is meaningless. You were going to wipe it and image it from the production system (or from the backup of the production system, or however you handle that in your setup) again next time you needed to test anything, anyway.
That said, I don't have the luxury of a test sytem per se at work. We're small, and we don't have any computers we don't actively use. (We don't even have a dedicated firewall/NATbox; it has to share hardware with a cgi server. This is not ideal, but I can't justify two headless server boxes to my boss, when we have five-year-old systems that need replaced.) I do test things out myself before deploying them to other staff or to the public, if I can. (Some things I can't test very well myself, because they require the wrong hardware or OS or whatever for my workstation. In that case I do a quick test when I install and hope for the best.) So I sortof use my workstation as a test system, kindof, but not really entirely, because I do by necessity have some stuff on there that matters. (My data get backed up daily over the LAN on a cron job, but having to do a reinstall would be a real bummer, and by the time I got all the software I regularly use reinstalled and configured, I'd be out a couple of days at least.)
But if I had a test system, I'd handle it in a fashion such that wrecking it would be irrelevant, since it would just get imaged from another system or a backup or whatever before each testing session anyway.
These days, 17-inch monitors don't cost significantly more than 14-inch ones, and 19-inch ones are very reasonably priced. Consequently, it's possible to use resolutions higher than 640x480 without being forced to squint. As a result, you can fit more than 80 columns in an editor and read it comfortably.
There are obviously still limits on how wide your code can be without causing people problems viewing it, but I would say the limit is a bit more than 80 characters these days. Probably at least 100, if not 120. In any given workplace, you should get your programmers together and agree on a limit.
> Anyone who knows anything whatsoever about graphic design knows that putting > beige text on a light beige background is going to be hard on the eyes.
Depends how sensitive your eyes are to light.
What anyone who knows anything whatsoever about graphic design ought to know is that the color choices of 98.75% of all websites are horrible, and any sensible person who has to sit in fron of a browser for long hours first turns off page colors and forces his own color preferences on all sites. This doesn't help with images, so sites that are mostly graphics will still be hard on the eyes, but sites like/. that are mostly text are considerably better this way.
Personally, I favor #FFE6BC text on a #294D4A background, but you can set your browser to whatever colors you like.
Sure, and I expected them to send us stuff that wasn't selling well, and the judge probably figured too that they'd do that. But I didn't realize they'd send so many *duplicates*. If they'd sent us fifty *different* non-sellers, I'd figure hey, add some diversity to the collection, some stuff people might not otherwise have listened to. But this nonsense about being able to count the number of distinct titles on our fingers, that's plain wrong. We can't use multiple copies of this stuff.
> BTW: There is one big difference between music and books in public libraries. > It generally takes a person a week or two to read a book, while it only > takes an hour or so to copy a CD.
In theory, maybe. In practice, average checkout time for a CD is *very* similar to average checkout time for a book. CDs do get renewed marginally less often, but they usually stay out for most of the two weeks, and they come back late just as often as books, averaged out per item.
> the correct one would be to figure out your environment and build accordingly
No, the correct way is to build your software so that it makes no significant difference what platform it's running on. With modern languages and libraries and toolkits this is getting closer and closer to actually being possible. (Think: Parrot and wxWindows.)
> On the plus side, almost all of the women in the upper classes were smart
This is a well-known phenomenon. In strongly male-dominated fields (not fields where it's like 80%/20%, but more like 95%/5% or worse), the women you do find in those fields are some of the very best people in the field. This is an overgeneralization, but it's a *good* overgeneralization (i.e., it's not true every time, but it's true WAY more often than it's false).
I have a theory about why this is, and it goes like this: the women who don't have a strong affinity for the field strongly tend to leave the field due to the cultural pressures. The ones who are left are the ones whose inclination for the field was so strong that they couldn't make themselves give it up, even if all their friends and their parents and everyone think they ought to find something different to do.
One would think that the same would work in reverse for males in fields that are female-dominated, but if it does it seems to be to a lesser extent. Apparently either the psychology or the sociology of the situation is in some way different for men.
> Does this ratio of 9/125 represent the actual number of female worker Does > this ratio of 9/125 represent the actual number of female workers in the > IT industry, or is it lower?
It's probably very typical, within one standard deviation of average.
> my field has a much higher representation from the fairer sex!
Yeah. So does my workplace. I'm the (entire) IT department and am one of two male employees, the other being the maintenance guy. All my other coworkers are female. (I work at a library.) One imagines this is some kind of really unusual coincidence. Or something.
> I asked my librarian at the school I go to, and she had thought that it > would be a bad idea to use it, because it's written by random people, > instead of scholars like in "traditional" encyclopedias.
The scholars who write traditional encyclopedias are pretty much random people. There's probably more editorial review, but still, traditional encyclopedias are nothing like authoritative. The librarian you asked is obviously not a reference librarian, or she'd have told you that you shouldn't use *any* encyclopedia for in-depth research, ever. Encyclopedias are a starting point. You consult them to get enough of an overview so that you know what to look for when you go to do your real research.
This is not to belittle the importance of encyclopedias. They're extremely useful and often good enough all by themselves for the merely curious. Often when you have to research a topic you don't know anything about, you don't know where to get started until you've gone through a couple of encyclopedia articles on the topic. Encyclopedias are great for that stuff -- and the Wikipedia is no exception. But if you're thinking of Britannica (or, worse, Worldbook) as a scholarly source to cite, you're out of your chair.
Most of the clues you can normally use to determine whether a message is legit have been stripped out. The hard-to-forge headers, such as the Received: headers, are absent for the test. The links have been altered so that viewing the source won't tell you anything about where the link in the actual message would have taken you. (The links in the test don't take you anywhere, but you can't even tell from the javascript source where the links were supposed to point originally; it's impossible for a spammer to do this in a real spam.) Thus, the test questions are all II (Insufficient Information to determine an answer) in my book. Their resulting failure figure (28%), then, is probably high.
After LOTR, we're going to shoot a James Bond flick starring Danny DeVito as 007, and then a musical costarring Leonard Nimoy, Clint Eastwood, and Britney Spears. (The plot of this musical will have to involve romance, obviously.)
Also, it's imperative that the Wizard of Oz be remade with current actors. I'm thinking John Goodman as the scarecrow, Tommy Lee Jones as the Cowardly Lion, Chris Kattan as the Tin Man, Jim Carey as Dorothy, Adam Sandler as the dog, Pee Wee Herman as the Wicked Witch, and Whoopee Goldberg as the Great Oz.
But to bring this back to topic: for LOTR, they definitely should have cast William Shattner as Tom Bombadil. It's simply impossible to think of Tom skipping rapidly down the forest path in a sky-blue jacket and yellow boots singing exuberantly about being a merry fellow without picturing William Shattner. It's just a question of who would be Goldberry.
It's important to test at different resolutions, test on systems where the default colors are set different from what you usually use, test with various browser settings changed, and so on. This is all at least as important as testing on multiple browsers.
Oh, and run it through validator.w3.org.
If you do all that, plus view it in MSIE, Gecko, and lynx, you should be doing pretty well.
Re:v6 could help solve some net problems
on
IPv6 is Here
·
· Score: 1
> "good netizens who want to be anonymous"
Being able to trace who owns the connection doesn't imply being able to trace who is using it. They can always go to the library and be anonymous. Most small public libraries don't even contemplate the possibility of asking to see ID when you use the internet.
> I'm just so glad to have an option other than DSL
Do you have any idea how long I've been *wishing* I had DSL as an option, *hoping* it comes to my area, *waiting*, *wondering* if they're ever going to get their act together and roll it out? If I could get DSL, the latency would be decent enough for X11-forwarding between home and work. That would *rock*. But it's not happening any time soon, apparently.
> Most patrons are barely capable of using existing public-access terminals > let alone a multi-tabbed browser.
Most patrons don't use the tabbed browsing feature, no. I have the tab bar configured to hide when only one tab is open, for just this reason. Some patrons do, however, appreciate the fact that closing the browser window automatically logs them out of everything. (This is because I configured cookies to have a limited lifetime of the current session, but the patrons are more interested in the result than the implementation.) If it's possible to do that with IE, I don't know how. As the computer guy, I appreciate something different about Mozilla: less maintenance.
> Additionally, the majority of catalog lookups
This is irrelevant for us. Our catalog stations in the library are dumb terminals. We only use web browsers for actual web access. (We do have a web-based catalog, which patrons can access from home, but it's not used within the library, generally.) This will change when we migrate to a different automation system, but that's a couple of years out still.
That's gotta be a lie. IE wasn't even remotely close to usable, not even by the standards of the day, until at _least_ 5.0. There were a small handful of people using IE 4.0, but *nobody* used IE3, much less anything before that.
> Why does tabbed browsing keep rising to such prominence as a must-have
Because, before tabbed browsing, web-based fora (such as slashdot) were just about worthless, impossibly painful to use, _especially_ over dialup. We all used usenet back then because there were newsreaders with a decent interface. By letting you queue pages by middle-clicking (or ctrl-clicking, for the mouse-button-challenged) on links and continue to read uninterrupted in the meanwhile, tabbed browsing makes the web viable as a medium for many things it was just not suitable for previously. I spend a *lot* more time on the web than I did before. (It is arguable whether this is a good thing...)
> Does anyone know if it is possible to do similar thing with Mozilla > (ie. Default start page, proxy setttings, etc)?
To lock that stuff down, you're going to have to lock down the user's account at the OS level better than is possible with any version of Windows I've yet seen. If you think you have these things locked down, your users maybe just aren't imaginative enough to use Google to find out how to get around it.
Forcing a certain start page seems pointless. Forcing proxy settings is most easily done at the router, by blocking outgoing traffic. Otherwise, like I said, you're going to have to do some seriously heavy-duty stuff in terms of locking down the user's account, the kind of stuff that makes the system virtually worthless to the user because they can't even save a file.
> I bet that any library with public computers that uses mozilla would have > to hire an extra person to show people the advantages.
No, it's pretty much a complete non-issue. We use Mozilla.org browsers exclusively, and the biggest pain is remembering to install the Java plugin and Acrobat Reader every time we get a new system or have to reinstall Windows on an existing system. Vanishingly close to 100% of patron questions are website-specific, stuff like "Where's the link on this bank website to transfer money to my credit union account" and "Why won't Hotmail let me get this [malformed] attachment" and "Why does this website require me to give them an email address to sign up for this service?" (This last is really common. A lot of our patrons don't have email.)
Galion Public Library uses Mozilla.org browsers exclusively. (I'm the computer guy.)
However, we previously used mostly Communicator. We did have MSIE on *one* computer at one point, but that system was so much trouble that when Windows got cranky and needed to be reinstalled, we didn't bother. The librarians were offering to dig a hole in the flower gardens and bury it; they weren't interested in having it fixed; they wanted it replaced. Also, reinstalling would have been a problem since we didn't have the original driver disks (not my fault; we didn't have them when I was hired), and with its being a Compaq Deskpro (no model number _anywhere_, and there are dozens of models, and you have to know which one you have...), finding the correct drivers on the net was promising real pain. This was late 2000. I put TurboLinux on it and it served as a CGI server for a couple of years after that without incident.
None of the librarians has ever asked me why we don't use MSIE. (Some of them have asked me about the difference between Mozilla and Netscape, though.) No patron AFAIK has ever specifically asked for Internet Explorer either. I do get occasional complaints from patrons about certain plugins not being installed (most frequently Flash), but that's not nearly as many complaints as I get about the Yahoo! Mail and Hotmail interfaces (neither of which we endorse or recommend; we officially do not provide email: we merely provide access to the web).
I should note that our catalog stations within the library are not web-based. We have a web-based catalog so people can get to our catalog from home, but within the library the catalog stations are VT510 dumb terminals, connected only to the automation system via ports (on a DECServer) which are only privileged for OPAC (i.e., the catalog) and nothing else. For our older patrons, the dumb terminals are easier to use and less intimidating than a web-based system. (The OPAC literally tells you what buttons to push, and there's no need to know how to use a mouse, which is good because a lot of people around here aren't comfortable with computer mice yet.)
> Editing posts is evil! It may seem like a good idea for fixing that stupid > typo or the wrong link, but it usually ends in deleted/edited posts where > people just don't stand up to their words and nobody can follow what the > reason for the discussion/flamewar was.
It seems to work fine on perlmonks.org. (Okay, so perlmonks also has a more advanced moderation system than slashdot, which probably helps. Still.)
> If cancer was cured, the pharmaceutical companies would be out of business.
Hardly. Doctors didn't go out of business when smallpox was wiped out, or when antibiotics were discovered that could cure bacterial infections, were they? There's always another kind of illness, one that we barely noticed before because the one we just cured was worse or more common. Cure cancer, and something else will be the big medical problem.
> Emacs is... bloated enough that it needs three different entries You say "bloated", I say "featureful". I remember when Lotus 123 for DOS was called bloated, because it used over 100 kilobytes of RAM and over a megabyte of disk space.
> What you are answering is something developer - not admin
Most of my job is admin stuff. I do some development, but most of it is site-specific custom development for admin purposes. I'd call that admin. I write Perl scripts that do stuff like make backups of important things and copy them across the network. Call that developer if you want, I guess.
> I do not know any admin who use emacs, for example It's difficult to imagine what life as an admin would be like without Emacs.
> And shell + awk + sed very often is better than perl Now you're just trolling. shell+awk+sed+grep is barely in the same category as Perl4 and not nearly as useful (for admin stuff, I mean) as Perl5.
> A is for awk, which runs like a snail Yeah, these days we use Perl for that stuff. > B is for biff, which reads all your mail Emacs/Gnus takes care of this, thanks. (Yeah, I know that's not really what biff does, but hey, it's what the poem says.) > C is for cc, as hackers recall C is for nursing along old legacy code that hasn't had the good graces to be rewritten in a real language yet; for new stuff we use high-level languages. > D is for dd, the command that does all I must admit that this one is still handy from time to time... > E is for emacs, which rebinds your keys Can't live without that... > F is for fsck, which rebuilds your trees One word: journaling. > G is for grep, a clever detective I admit I still occasionally use this, but Perl's regexen are more powerful. > H is for halt, which may seem defective I usually use shutdown or init when I need to do hardware maintenance. > I is for indent, which rarely amuses Emacs does this automatically, of course. Has for years. > J is for join, which nobody uses Indeed, what does it even do? From a quick look at the man page, it looks at first glance like a Perl one-liner, give or take a dozen strokes. > K is for kill, which makes you the boss I do still use this sometimes. > L is for lex, which is missing from DOS Isn't that one of those C things? Its days are numbered. > M is for more, from which less was begot With eshell (the Emacs shell), there's no need for a pager any longer. This letter should now go to man, IMO. > N is for nice, which really is not Here's another one I must admit to using occasionally. Also renice. > O is for od, which prints out things nice I had to use man to even find out what this does, and then it turns out to be something I'd almost never use, but if I did need that functionality, I could do it in a Perl one-liner faster than look up the od manpage again. > P is for passwd, which reads in strings twice > Q is for quota, a Berkeley-type fable Okay, those might still be relevant. > R is for ranlib, for sorting a table Definitely a Perl job, that. > S is for spell, which attempts to belittle Emacs has ispell/aspell integration these days. > T is for true, which does very little Indeed. > U is for uniq, which is used after sort Another thing we use Perl for in the modern era. > V is for vi, which is hard to abort If you really want vi (*WHY*?), Emacs has a version of it built in. > W is for whoami, which tells you your name If you need a program for that, commands aren't going to solve your problems. > X is, well, X, of dubious fame Emacs has better colors if you use X. 24-bit. Without it, 16 colors. > Y is for yes, which makes an impression, and I suppose... > Z is for zcat, which handles compression Most of us use either zip (or Archive::Zip) or gzip (or Zlib) for that now. We'll say zip, because it starts with z and so won't screw up the poem any worse than I already have;-)
Slashdot Mirror
> Yeah, but why rush to wreck them
The idea behind a test system, generally, is that "wrecking" it is meaningless.
You were going to wipe it and image it from the production system (or from the
backup of the production system, or however you handle that in your setup)
again next time you needed to test anything, anyway.
That said, I don't have the luxury of a test sytem per se at work. We're
small, and we don't have any computers we don't actively use. (We don't
even have a dedicated firewall/NATbox; it has to share hardware with a
cgi server. This is not ideal, but I can't justify two headless server
boxes to my boss, when we have five-year-old systems that need replaced.)
I do test things out myself before deploying them to other staff or to
the public, if I can. (Some things I can't test very well myself, because
they require the wrong hardware or OS or whatever for my workstation. In
that case I do a quick test when I install and hope for the best.) So I
sortof use my workstation as a test system, kindof, but not really entirely,
because I do by necessity have some stuff on there that matters. (My data
get backed up daily over the LAN on a cron job, but having to do a reinstall
would be a real bummer, and by the time I got all the software I regularly
use reinstalled and configured, I'd be out a couple of days at least.)
But if I had a test system, I'd handle it in a fashion such that wrecking
it would be irrelevant, since it would just get imaged from another system
or a backup or whatever before each testing session anyway.
These days, 17-inch monitors don't cost significantly more than 14-inch ones,
and 19-inch ones are very reasonably priced. Consequently, it's possible to
use resolutions higher than 640x480 without being forced to squint. As a
result, you can fit more than 80 columns in an editor and read it comfortably.
There are obviously still limits on how wide your code can be without causing
people problems viewing it, but I would say the limit is a bit more than 80
characters these days. Probably at least 100, if not 120. In any given
workplace, you should get your programmers together and agree on a limit.
> Anyone who knows anything whatsoever about graphic design knows that putting
/. that are mostly text are considerably
> beige text on a light beige background is going to be hard on the eyes.
Depends how sensitive your eyes are to light.
What anyone who knows anything whatsoever about graphic design ought to know
is that the color choices of 98.75% of all websites are horrible, and any
sensible person who has to sit in fron of a browser for long hours first
turns off page colors and forces his own color preferences on all sites.
This doesn't help with images, so sites that are mostly graphics will still
be hard on the eyes, but sites like
better this way.
Personally, I favor #FFE6BC text on a #294D4A background, but you can set
your browser to whatever colors you like.
> The RIAA is using libraries to dump overstock.
Sure, and I expected them to send us stuff that wasn't selling well, and the
judge probably figured too that they'd do that. But I didn't realize they'd
send so many *duplicates*. If they'd sent us fifty *different* non-sellers,
I'd figure hey, add some diversity to the collection, some stuff people might
not otherwise have listened to. But this nonsense about being able to count
the number of distinct titles on our fingers, that's plain wrong. We can't
use multiple copies of this stuff.
> BTW: There is one big difference between music and books in public libraries.
> It generally takes a person a week or two to read a book, while it only
> takes an hour or so to copy a CD.
In theory, maybe. In practice, average checkout time for a CD is *very*
similar to average checkout time for a book. CDs do get renewed marginally
less often, but they usually stay out for most of the two weeks, and they
come back late just as often as books, averaged out per item.
Now, videos, those have very different numbers.
> the correct one would be to figure out your environment and build accordingly
No, the correct way is to build your software so that it makes no significant
difference what platform it's running on. With modern languages and libraries
and toolkits this is getting closer and closer to actually being possible.
(Think: Parrot and wxWindows.)
> On the plus side, almost all of the women in the upper classes were smart
This is a well-known phenomenon. In strongly male-dominated fields (not
fields where it's like 80%/20%, but more like 95%/5% or worse), the women
you do find in those fields are some of the very best people in the field.
This is an overgeneralization, but it's a *good* overgeneralization (i.e.,
it's not true every time, but it's true WAY more often than it's false).
I have a theory about why this is, and it goes like this: the women who
don't have a strong affinity for the field strongly tend to leave the field
due to the cultural pressures. The ones who are left are the ones whose
inclination for the field was so strong that they couldn't make themselves
give it up, even if all their friends and their parents and everyone think
they ought to find something different to do.
One would think that the same would work in reverse for males in fields
that are female-dominated, but if it does it seems to be to a lesser extent.
Apparently either the psychology or the sociology of the situation is in some
way different for men.
> Does this ratio of 9/125 represent the actual number of female worker Does
> this ratio of 9/125 represent the actual number of female workers in the
> IT industry, or is it lower?
It's probably very typical, within one standard deviation of average.
> my field has a much higher representation from the fairer sex!
Yeah. So does my workplace. I'm the (entire) IT department and am one of
two male employees, the other being the maintenance guy. All my other
coworkers are female. (I work at a library.) One imagines this is some
kind of really unusual coincidence. Or something.
> I asked my librarian at the school I go to, and she had thought that it
> would be a bad idea to use it, because it's written by random people,
> instead of scholars like in "traditional" encyclopedias.
The scholars who write traditional encyclopedias are pretty much random people.
There's probably more editorial review, but still, traditional encyclopedias
are nothing like authoritative. The librarian you asked is obviously not a
reference librarian, or she'd have told you that you shouldn't use *any*
encyclopedia for in-depth research, ever. Encyclopedias are a starting point.
You consult them to get enough of an overview so that you know what to look
for when you go to do your real research.
This is not to belittle the importance of encyclopedias. They're extremely
useful and often good enough all by themselves for the merely curious. Often
when you have to research a topic you don't know anything about, you don't
know where to get started until you've gone through a couple of encyclopedia
articles on the topic. Encyclopedias are great for that stuff -- and the
Wikipedia is no exception. But if you're thinking of Britannica (or, worse,
Worldbook) as a scholarly source to cite, you're out of your chair.
Most of the clues you can normally use to determine whether a message is legit
have been stripped out. The hard-to-forge headers, such as the Received:
headers, are absent for the test. The links have been altered so that
viewing the source won't tell you anything about where the link in the
actual message would have taken you. (The links in the test don't take
you anywhere, but you can't even tell from the javascript source where the
links were supposed to point originally; it's impossible for a spammer to
do this in a real spam.) Thus, the test questions are all II (Insufficient
Information to determine an answer) in my book. Their resulting failure
figure (28%), then, is probably high.
After LOTR, we're going to shoot a James Bond flick starring Danny DeVito as
007, and then a musical costarring Leonard Nimoy, Clint Eastwood, and Britney
Spears. (The plot of this musical will have to involve romance, obviously.)
Also, it's imperative that the Wizard of Oz be remade with current actors.
I'm thinking John Goodman as the scarecrow, Tommy Lee Jones as the Cowardly
Lion, Chris Kattan as the Tin Man, Jim Carey as Dorothy, Adam Sandler as the
dog, Pee Wee Herman as the Wicked Witch, and Whoopee Goldberg as the Great Oz.
But to bring this back to topic: for LOTR, they definitely should have
cast William Shattner as Tom Bombadil. It's simply impossible to think of
Tom skipping rapidly down the forest path in a sky-blue jacket and yellow
boots singing exuberantly about being a merry fellow without picturing
William Shattner. It's just a question of who would be Goldberry.
It's important to test at different resolutions, test on systems where the
default colors are set different from what you usually use, test with various
browser settings changed, and so on. This is all at least as important as
testing on multiple browsers.
Oh, and run it through validator.w3.org.
If you do all that, plus view it in MSIE, Gecko, and lynx, you should
be doing pretty well.
> "good netizens who want to be anonymous"
Being able to trace who owns the connection doesn't imply being able to
trace who is using it. They can always go to the library and be anonymous.
Most small public libraries don't even contemplate the possibility of asking
to see ID when you use the internet.
> I'm just so glad to have an option other than DSL
Do you have any idea how long I've been *wishing* I had DSL as an option,
*hoping* it comes to my area, *waiting*, *wondering* if they're ever going
to get their act together and roll it out? If I could get DSL, the latency
would be decent enough for X11-forwarding between home and work. That would
*rock*. But it's not happening any time soon, apparently.
> Most patrons are barely capable of using existing public-access terminals
> let alone a multi-tabbed browser.
Most patrons don't use the tabbed browsing feature, no. I have the tab bar
configured to hide when only one tab is open, for just this reason. Some
patrons do, however, appreciate the fact that closing the browser window
automatically logs them out of everything. (This is because I configured
cookies to have a limited lifetime of the current session, but the patrons
are more interested in the result than the implementation.) If it's possible
to do that with IE, I don't know how. As the computer guy, I appreciate
something different about Mozilla: less maintenance.
> Additionally, the majority of catalog lookups
This is irrelevant for us. Our catalog stations in the library are dumb
terminals. We only use web browsers for actual web access. (We do have a
web-based catalog, which patrons can access from home, but it's not used
within the library, generally.) This will change when we migrate to a
different automation system, but that's a couple of years out still.
> I've been a faithful IE user since 2.0
That's gotta be a lie. IE wasn't even remotely close to usable, not even by
the standards of the day, until at _least_ 5.0. There were a small handful
of people using IE 4.0, but *nobody* used IE3, much less anything before that.
> Why does tabbed browsing keep rising to such prominence as a must-have
Because, before tabbed browsing, web-based fora (such as slashdot) were just
about worthless, impossibly painful to use, _especially_ over dialup. We all
used usenet back then because there were newsreaders with a decent interface.
By letting you queue pages by middle-clicking (or ctrl-clicking, for the
mouse-button-challenged) on links and continue to read uninterrupted in the
meanwhile, tabbed browsing makes the web viable as a medium for many things
it was just not suitable for previously. I spend a *lot* more time on the
web than I did before. (It is arguable whether this is a good thing...)
> IE can be locked down
IE can be *what*? +1, Funny.
> Does anyone know if it is possible to do similar thing with Mozilla
> (ie. Default start page, proxy setttings, etc)?
To lock that stuff down, you're going to have to lock down the user's account
at the OS level better than is possible with any version of Windows I've yet
seen. If you think you have these things locked down, your users maybe just
aren't imaginative enough to use Google to find out how to get around it.
Forcing a certain start page seems pointless. Forcing proxy settings is
most easily done at the router, by blocking outgoing traffic. Otherwise,
like I said, you're going to have to do some seriously heavy-duty stuff in
terms of locking down the user's account, the kind of stuff that makes the
system virtually worthless to the user because they can't even save a file.
> I bet that any library with public computers that uses mozilla would have
> to hire an extra person to show people the advantages.
No, it's pretty much a complete non-issue. We use Mozilla.org browsers
exclusively, and the biggest pain is remembering to install the Java plugin
and Acrobat Reader every time we get a new system or have to reinstall
Windows on an existing system. Vanishingly close to 100% of patron
questions are website-specific, stuff like "Where's the link on this
bank website to transfer money to my credit union account" and "Why won't
Hotmail let me get this [malformed] attachment" and "Why does this website
require me to give them an email address to sign up for this service?"
(This last is really common. A lot of our patrons don't have email.)
Galion Public Library uses Mozilla.org browsers exclusively. (I'm the
computer guy.)
However, we previously used mostly Communicator. We did have MSIE on *one*
computer at one point, but that system was so much trouble that when Windows
got cranky and needed to be reinstalled, we didn't bother. The librarians
were offering to dig a hole in the flower gardens and bury it; they weren't
interested in having it fixed; they wanted it replaced. Also, reinstalling
would have been a problem since we didn't have the original driver disks
(not my fault; we didn't have them when I was hired), and with its being a
Compaq Deskpro (no model number _anywhere_, and there are dozens of models,
and you have to know which one you have...), finding the correct drivers on
the net was promising real pain. This was late 2000. I put TurboLinux on
it and it served as a CGI server for a couple of years after that without
incident.
None of the librarians has ever asked me why we don't use MSIE. (Some of
them have asked me about the difference between Mozilla and Netscape, though.)
No patron AFAIK has ever specifically asked for Internet Explorer either. I
do get occasional complaints from patrons about certain plugins not being
installed (most frequently Flash), but that's not nearly as many complaints
as I get about the Yahoo! Mail and Hotmail interfaces (neither of which we
endorse or recommend; we officially do not provide email: we merely provide
access to the web).
I should note that our catalog stations within the library are not web-based.
We have a web-based catalog so people can get to our catalog from home, but
within the library the catalog stations are VT510 dumb terminals, connected
only to the automation system via ports (on a DECServer) which are only
privileged for OPAC (i.e., the catalog) and nothing else. For our older
patrons, the dumb terminals are easier to use and less intimidating than
a web-based system. (The OPAC literally tells you what buttons to push,
and there's no need to know how to use a mouse, which is good because a
lot of people around here aren't comfortable with computer mice yet.)
> Editing posts is evil! It may seem like a good idea for fixing that stupid
> typo or the wrong link, but it usually ends in deleted/edited posts where
> people just don't stand up to their words and nobody can follow what the
> reason for the discussion/flamewar was.
It seems to work fine on perlmonks.org. (Okay, so perlmonks also has a more
advanced moderation system than slashdot, which probably helps. Still.)
> If cancer was cured, the pharmaceutical companies would be out of business.
Hardly. Doctors didn't go out of business when smallpox was wiped out, or
when antibiotics were discovered that could cure bacterial infections, were
they? There's always another kind of illness, one that we barely noticed
before because the one we just cured was worse or more common. Cure cancer,
and something else will be the big medical problem.
Details change. The big picture stays the same.
> Emacs is ... bloated enough that it needs three different entries
You say "bloated", I say "featureful". I remember when Lotus 123 for DOS was
called bloated, because it used over 100 kilobytes of RAM and over a megabyte
of disk space.
> What you are answering is something developer - not admin
Most of my job is admin stuff. I do some development, but most of it is
site-specific custom development for admin purposes. I'd call that admin.
I write Perl scripts that do stuff like make backups of important things
and copy them across the network. Call that developer if you want, I guess.
> I do not know any admin who use emacs, for example
It's difficult to imagine what life as an admin would be like without Emacs.
> And shell + awk + sed very often is better than perl
Now you're just trolling. shell+awk+sed+grep is barely in the same category
as Perl4 and not nearly as useful (for admin stuff, I mean) as Perl5.
> A is for awk, which runs like a snail ;-)
Yeah, these days we use Perl for that stuff.
> B is for biff, which reads all your mail
Emacs/Gnus takes care of this, thanks. (Yeah, I know that's not really
what biff does, but hey, it's what the poem says.)
> C is for cc, as hackers recall
C is for nursing along old legacy code that hasn't had the good graces to be
rewritten in a real language yet; for new stuff we use high-level languages.
> D is for dd, the command that does all
I must admit that this one is still handy from time to time...
> E is for emacs, which rebinds your keys
Can't live without that...
> F is for fsck, which rebuilds your trees
One word: journaling.
> G is for grep, a clever detective
I admit I still occasionally use this, but Perl's regexen are more powerful.
> H is for halt, which may seem defective
I usually use shutdown or init when I need to do hardware maintenance.
> I is for indent, which rarely amuses
Emacs does this automatically, of course. Has for years.
> J is for join, which nobody uses
Indeed, what does it even do? From a quick look at the man page, it looks
at first glance like a Perl one-liner, give or take a dozen strokes.
> K is for kill, which makes you the boss
I do still use this sometimes.
> L is for lex, which is missing from DOS
Isn't that one of those C things? Its days are numbered.
> M is for more, from which less was begot
With eshell (the Emacs shell), there's no need for a pager any longer.
This letter should now go to man, IMO.
> N is for nice, which really is not
Here's another one I must admit to using occasionally. Also renice.
> O is for od, which prints out things nice
I had to use man to even find out what this does, and then it turns out
to be something I'd almost never use, but if I did need that functionality,
I could do it in a Perl one-liner faster than look up the od manpage again.
> P is for passwd, which reads in strings twice
> Q is for quota, a Berkeley-type fable
Okay, those might still be relevant.
> R is for ranlib, for sorting a table
Definitely a Perl job, that.
> S is for spell, which attempts to belittle
Emacs has ispell/aspell integration these days.
> T is for true, which does very little
Indeed.
> U is for uniq, which is used after sort
Another thing we use Perl for in the modern era.
> V is for vi, which is hard to abort
If you really want vi (*WHY*?), Emacs has a version of it built in.
> W is for whoami, which tells you your name
If you need a program for that, commands aren't going to solve your problems.
> X is, well, X, of dubious fame
Emacs has better colors if you use X. 24-bit. Without it, 16 colors.
> Y is for yes, which makes an impression, and
I suppose...
> Z is for zcat, which handles compression
Most of us use either zip (or Archive::Zip) or gzip (or Zlib) for that now.
We'll say zip, because it starts with z and so won't screw up the poem any
worse than I already have
1. Emacs
2. Perl
3. ssh/scp
4. Emacs
5. CPAN (and search.cpan.org)
6. screen
7. Perl
8. Emacs
9. Google Groups
10. cperl-mode
HTH.HAND.