My guess is it won't go down much. It might even go up. The two-way split is going to create two monopolies -- so if you buy MS now, you'll have stock in two very strongly positioned companies.
I think it's probably true, whatever kind of light it is. And actually, I don't think it's all that bad. The fact is, all of that information is out there -- the "really good source of technical support" actually exists."
Sadly, that isn't what it gets used for. Instead of, for example, handling bigger arrays easily, programmers just make their array code less efficient. Faster computer, slower code, same effect.
There's two factors you neglect, though: faster development time, and cheaper software.
That's a good point. But all of that is a significant amount of work -- there's an increasingly high barrier to entry. It'll happen, but not on near the scale of mp3.
--
Re:weirdest install concept i've ever heard of
on
Mandrake 7.1 Released
·
· Score: 2
Thanks for the info. I think that confirms my point -- namely, the importance of a package is arbitrarily selected by Mandrake.
This slider concept might make sense within highly focused areas ("C Development") or perhaps even within broader use-for-machine categories ("Graphics Workstation") but it's difficult to the point of impossible for a whole Linux distro. How important are the following packages: bind, gimp, emacs, lsof, apmd, pump/dhcpcd?
--
Re:weirdest install concept i've ever heard of
on
Mandrake 7.1 Released
·
· Score: 3
My feeling is that users who are able to pull out the CD and install specific apps (including resolving dependencies) are not the ones who would be intimidated by selecting packages (or broad groups thereof).
Another way when my analogy fails is that most people know enough about what should be in an automobile to notice if the windshield wipers are gone. But what if the selector decides not to install something less obvious -- a timing chip, or some sensor or another? ("It's just a tiny little sensor; it can't be very important.")
And it's not just about critical things. A new user who is too intimidated to go through lists of packages is likely to learn what's available by looking through the application menus, or maybe by typing (at the advice of a friend) "ls/usr/bin". The people whom this is supposed to help are going to be more lost later, because it'll be harder for them to know what they could have. I learned a lot about Linux during my first few installs by reading the descriptions of every package that I didn't immediately recognize. Sure, it took more time, but it was worth it -- ultimately reducing the learning curve.
Not knowing what's on the system also makes the job harder for the people who have to support the machine. This is part of why Red Hat's base component set is so large.
--
Re:weirdest install concept i've ever heard of
on
Mandrake 7.1 Released
·
· Score: 2
I don't think this is user-friendly at all. It might appear so on the surface, but it leaves the user with no concept of what might be installed on their system, leading to long term confusion.
Imagine if this feature were available when buying a car:
Sales system: "100%: Car costs $16,000". You: "Oh my, I can't afford that. Let's slide the bar down to 50%."
Later, when it starts to rain, you find that you've only got a windshield wiper on the passenger side. So, you pull over to the side of the road, and discover that there's no hazard lights either. Hey! There aren't any lights on this car at all! Luckily, brakes seem to have made the cut.... hope the airbags are there....
The beta had an extremely strange user interface concept in its installer. After you were done selecting packages, you were given a dialog box with a slider: "Installing 100% of the packages will take N MB". Changing the slider lowered the percentage of packages to be installed based on their (arbitrarily selected by Mandrake?) relative "importance". So if it were going to take 1.5GB to install everything, but you only had 750MB, you could drag the slider down to 50%.
It's a very interesting idea, but I'm pretty sure it's not a good one! Has this made it to the final release?
There is a fundamental difference between games and music/movies: the latter can be compressed acceptably with lossy formats, while presumably that won't work with software.
All those mp3s and VCDs are far smaller than the original.
Um, why should machines making up a search engine backend need access to Real Audio or Napster? Are you suggesting that the cluster is also used as people's workstations?
Perl is like English. It's got lots of words and concepts borrowed from other languages. Sometimes spellings are confusing and even irrational. There's irregularities everywhere you look. It's not particularly hard to learn the basics, but picking up all of the nuances takes real effort -- many people who use it daily are still confused by some of the grammar. It's also very flexible, and a favorite language of many for writing poetry.
Python is like Esperanto. It's designed from the beginning to make sense, and to be easy to learn. Spelling is phonetic, and "unnecessary complications" have been removed. It's still a fully-functional language in which any concept can be expressed -- being a planned language doesn't automatically make it Orwellian -- but there isn't the flexibility found elsewhere.
The Esperanto approach has its advantages, especially for programming. (Let's hope Python has more success in the real world!) In an alternate, more sensible universe, international documents are written in Esperanto. Likewise, it makes sense to design big collaborative projects in Python. On the other hand, I really enjoy the flexiblity and room for creativity provided by the English language -- and by Perl.
That's one way of looking at things, and it's obviously very popular -- but there is a price to pay. I highly recommend reading this excellent essay at Salon.com:
I think I subscribe to the "malice" theory on this one. Sure, it's not terribly difficult to get strong encryption -- but it's extra steps, and if there's one thing we've learned from MS IE's standings in the browser wars, it's that most consumers aren't going to bother.
Being proficient in one distro doesn't necessarily make you unable to handle others. There's no particular reason you can't be an expert in several different areas -- especially because the differences aren't really that great.
As for the LPI certification -- again, it's got potential, but it's just a multiple choice test. Until they add some lab-based stuff, it's not going to be the "best possible cert in the world".
Slashdot Mirror
--
--
There's two factors you neglect, though: faster development time, and cheaper software.
--
--
This slider concept might make sense within highly focused areas ("C Development") or perhaps even within broader use-for-machine categories ("Graphics Workstation") but it's difficult to the point of impossible for a whole Linux distro. How important are the following packages: bind, gimp, emacs, lsof, apmd, pump/dhcpcd?
--
Another way when my analogy fails is that most people know enough about what should be in an automobile to notice if the windshield wipers are gone. But what if the selector decides not to install something less obvious -- a timing chip, or some sensor or another? ("It's just a tiny little sensor; it can't be very important.")
And it's not just about critical things. A new user who is too intimidated to go through lists of packages is likely to learn what's available by looking through the application menus, or maybe by typing (at the advice of a friend) "ls
Not knowing what's on the system also makes the job harder for the people who have to support the machine. This is part of why Red Hat's base component set is so large.
--
Imagine if this feature were available when buying a car:
Sales system: "100%: Car costs $16,000".
You: "Oh my, I can't afford that. Let's slide
the bar down to 50%."
Later, when it starts to rain, you find that you've only got a windshield wiper on the passenger side. So, you pull over to the side of the road, and discover that there's no hazard lights either. Hey! There aren't any lights on this car at all! Luckily, brakes seem to have made the cut.... hope the airbags are there....
--
It's a very interesting idea, but I'm pretty sure it's not a good one! Has this made it to the final release?
--
All those mp3s and VCDs are far smaller than the original.
--
--
--
Perl is like English. It's got lots of words and concepts borrowed from other languages. Sometimes spellings are confusing and even irrational. There's irregularities everywhere you look. It's not particularly hard to learn the basics, but picking up all of the nuances takes real effort -- many people who use it daily are still confused by some of the grammar. It's also very flexible, and a favorite language of many for writing poetry.
Python is like Esperanto. It's designed from the beginning to make sense, and to be easy to learn. Spelling is phonetic, and "unnecessary complications" have been removed. It's still a fully-functional language in which any concept can be expressed -- being a planned language doesn't automatically make it Orwellian -- but there isn't the flexibility found elsewhere.
The Esperanto approach has its advantages, especially for programming. (Let's hope Python has more success in the real world!) In an alternate, more sensible universe, international documents are written in Esperanto. Likewise, it makes sense to design big collaborative projects in Python. On the other hand, I really enjoy the flexiblity and room for creativity provided by the English language -- and by Perl.
--
--
http://partners.nytimes.com/ etc. (article has since moved)
The new and supposedly brilliant URL is
http://partners.nytimes.com
Now, since you are so wise, you want to explain why that's an improvement? Cripes, indeed.
--
The dumbing-down of programming.
--
--
including
Piranha
--
This has been a message from someone who works with VMS.
--
--
btw: you seem a bit bitter. what's the problem?
--
--
--
Real-world experience is more important than degrees or certifications.
--
As for the LPI certification -- again, it's got potential, but it's just a multiple choice test. Until they add some lab-based stuff, it's not going to be the "best possible cert in the world".
--
--