Posted on 05/26/2009 11:04:51 AM PDT by Gomez
The University of Virgina collects and compiles information about students' computer use. The information has been collected over the past decade by student employees of ITC, known as Student Consultants (SCs) and formerly, Computing Advisors (CAs), a group of first-year students hired to advise and assist their peers with computing. The data is based on SC and CA surveys of first-year residence halls each fall.
Basically the same thing as someone changing the output of an API you depend on in an object-oriented application.
Which are you more likely to use? Why?
It would depend on whether it is ever supposed to be human-readable or exchanged with other applications (although you can export). But of course I'd prefer a chess game object that had pieces objects with properties for their own color and position, then save it to a database or serialize it to a binary XML file. But that doesn't have much to do with the file layout of the underlying system.
>But of course I’d prefer a chess game object that had pieces objects with properties for their own color and position, then save it to a database or serialize it to a binary XML file. But that doesn’t have much to do with the file layout of the underlying system.
LOL - And you prove my point; you want an object when you’re working with it and something storable to store, a file.
The method of using an intermediary file as the bridge between input & output is common, and sometimes necessary; but I would argue that it is a [very] poor choice for IPC (which is what your script file is, essentially, doing in the above-mentioned case).
RE: “A lot of what you are seeing is simply that Apple offers every-hour free training classes on various Mac topics and software and people bring their laptops and even their iMacs in to participate in those classes. There are also paid ($100 for a year) one hour one-on-one weekly training sessions that require you to bring your computer.
Finally, as to reliability. You could not be more wrong. Apple Macs have consistently been rated either #1 or (infrequently) #2 in reliability among computer makers.
As to customer satisfaction, Apple’s beat all comers:
.......... (many paragraphs and charts and graphs followed.....”
************
Good grief! What am I ‘wrong’ about? Why so defensive? I merely asked why so many people would be hauling computers back thru the mall to the Apple store. I have no problem with Mac; in fact know very little about it. I looked at them once when in a mall — they looked fine to me.
“That is what makes Microsoft’s new campaign so interesting. They are attacking Apple.
As a market leader, they should NOT attack the underdog - it’s the wrong strategy for them. Fools.
Two different things: a quick command or short script to get something done, and writing an application.
I'd rather use a command with awk, sed, grep, etc., for some on-the-fly task (or even a scheduled one) than design and write an object-oriented application. For example, if you were throwing those text files on my server constantly, wanting them to be parsed out for some task, a one-line command on a chron job could probably do the job very well. I don't even need to recompile when you change the parsing criteria.
>Two different things: a quick command or short script to get something done, and writing an application.
Are they? I mean isn’t programming very broadly defined as the implementation of a solution to a problem/task? If that is the case, then shell-scripting is, obviously, just another case of programming. (IE Batch programming and application programming are still both programming.)
>I’d rather use a command with awk, sed, grep, etc., for some on-the-fly task (or even a scheduled one) than design and write an object-oriented application.
...but you yourself said that that was no different from using APIs and being SOL when they change. I’m of the belief that a good scripting system is something good/desirable; take Scheme or LISP for example, as they are scripting/interpreted languages... also note that LISP _was_ the equivalent of the batch-file/script for the “Lisp Machines” { Search that term here: http://www.gigamonkeys.com/book/introduction-why-lisp.html }
>For example, if you were throwing those text files on my server constantly, wanting them to be parsed out for some task, a one-line command on a chron job could probably do the job very well. I don’t even need to recompile when you change the parsing criteria.
This is true. However it seems to be the poor-man’s not-quite “hot code deployment”. ( Search that term here: http://www.defmacro.org/ramblings/fp.html ) In all, it sounds very much like you are advocating a needlessly componotized/isolated version of the Lisp Machine’s capabilities wherein the scripting-language LISP was the glue that held the system together (Obviously you could run other programs, compilers, interpreters, apps, etc under such a system.).
>Or how about an engineer willing - anxious even - to trade endless frustration, breakdowns, slowdowns, maddening interface inconsistencies, and mile-deep trash heaps of menu heirarchies for simple, reliable, controllable, and - yes, FLEXIBLE computing?
Are you advocating making a new generation of Lisp Machines? {Perhaps with Scheme instead...}
>>It’s a computer, for Pete’s sake. Not a “lifestyle” or “religion”.
>
>And exactly how much have you used a modern Mac?
What does using a Mac have to do with noting that, as with all else in life, there are priorities? I don’t see him as bashing Macs per se, but rather the MacMania that some of their users have.
It is, after all, a computer. A tool to be used and evaluated, improved or discarded as necessary. Though sometimes “the people” don’t choose their tools wisely; Commodore is a case-in-point, their computers were superior, technically speaking, to pretty much all of their competitors.
>Show me a ten year old PC capable of running Vista usefully.
That’s unfair... vista is a GIANT step _backward_ in terms of usability and you know it! ;)
But on the other hand, it’s VERY surprising what modern hardware Win98SE can run on!
I really just cruise these threads to see Mac-bashers and Apple haters come out of the closet. ;’)
And different things are suited to different tasks. I'm not going to design and build a whole OO program when a simple shell script will suffice. Conversely, scripts are too often used when a full application would be better, which is one reason I don't like Perl. It's often used as a sledgehammer for any problem, writing huge applications. I also think it's ugly as a full programming language, but that's a personal thing.
but you yourself said that that was no different from using APIs and being SOL when they change.
That is a problem with any programming, so assigning it only to scripts was unfair.
In all, it sounds very much like you are advocating a needlessly componotized/isolated version of the Lisp Machine’s capabilities
I hate LISP. It was the worst language I had to deal with in grad school, and that included FORTRAN and COBOL. For this kind of work I like small components that do their job and only their job well, and then chaining them together for the complete task -- shell scripts.
For example, a few weeks ago I could have written or purchased an app, or done LISP, to get what I needed out of a couple gigabytes of http logs. But a one-liner (admittedly one rather long line) using cat, grep and sed with some pipes and redirects gave me what I wanted in about a minute.
On the other hand, I had some freaky stuff that needed to be done with a database regularly to fix a deficiency in a purchased app, and the help desk rep needed to be able to hand-hold the process and make decisions along the way, plus get the info necessary for him to make such decisions. Yes, you guessed it, an OO application for that one.
I read the title as, “Mac explodes at University of Virginia...”
:-)
In other words, Carter, your experience with Macs is woefully outdated. You are spouting off from a base of no experience at all with modern Macs. Frankly, we already knew that.
>>but you yourself said that that was no different from using APIs and being SOL when they change.
>
>That is a problem with any programming, so assigning it only to scripts was unfair.
I know, but you brought it up so I thought it best to bring it back as a counterpoint; you did well to recognize it as a general-programming problem. (This is true of any state-based storage; if you are reading a bunch of comma-formatted names, as in military nomenclature, and suddenly come on a semi-colon delimited name you’ll blink in surprise, figure it out, and continue on... unfortunately, programming the analog to such a simple, generalized error-recovery procedure is astonishingly complex.)
>>I mean isn’t programming very broadly defined as the implementation of a solution to a problem/task?
>
>And different things are suited to different tasks.
We are in agreement then. The example of polar- vs. Cartesian-coordinates as your ‘workspace’ was intended to highlight that. Likewise, Imperative (procedural or OO), Logical, Functional, and Rule-based programming are ALL particularly suited to different problems.
>>In all, it sounds very much like you are advocating a needlessly componotized/isolated version of the Lisp Machine’s capabilities
>
>I hate LISP. It was the worst language I had to deal with in grad school, and that included FORTRAN and COBOL. For this kind of work I like small components that do their job and only their job well, and then chaining them together for the complete task — shell scripts.
I think you’re missing my point here. I understand about your feelings bout LISP; I had my first introduction to it (well, Scheme actually) this last semester. It was fun and interesting in a way, and maddeningly stubborn in others. {I think it appealed to the part of me that is fascinated by different languages, especially expressing the same idea/concept in them. It’s defiantly not the way I’m used to thinking about programming.}
Ah, but I digress. I brought up the Lisp Machine not because of LISP, but because of the high level of integration that they used; the programming/scripting language WAS the OS, WAS the script-file language, and WAS eminently usable. You can see glimmers of how beneficial such a high-level of integration can be in modern RAD environments... they take a LOT more planning, design, and work to implement (and moreso to implement correctly) but the result can be VERY good. (IMO)
>On the other hand, I had some freaky stuff that needed to be done with a database regularly to fix a deficiency in a purchased app, and the help desk rep needed to be able to hand-hold the process and make decisions along the way, plus get the info necessary for him to make such decisions. Yes, you guessed it, an OO application for that one.
Hm, that sounds like it might have been fairly easy to express in a rule-based language...
I’m kinda-working on an OS right now, and it seems to me that it would be good to make it not high-level, but very high-level; think what HLLs are to assembly... that is what it should be, on the programmer-level, to current OSes. Take ADTs for example, it seems rather wasteful to have programmers roll their own, or the language (and it’s libraries) provide the common ones like stack and queue... that can be done by the OS/OE with an Interface (in the OOP/Delphi/Java sense). Then, when the OS updates its ADT every application [that’s using the interface, obviously] enjoies the same optimization benefit. {Imagine working on a system with a Sort interface, but the implementing object was programmed up using bubble-sort, then replaced by a Quicksort or Shell-sort... all the searching system-wide would be improved.} That is a small example, of course, but the idea has been around since DLLs and before even... it just seems that they stopped the integration & abstraction too short and on too low a level.
I hope that makes sense to you... because I’m trying to put my design philosophy for this OS I’m working on down in words that others can understand. (It’s a big project.)
You are indeed, "one-of-a-kind".
Then what you know is somewhat wrong and somewhat right. If, as you say the success rate of virus writers in attacking any target has to do with the prevalence of the target. Macs are now over 35,000,000 (probably now north of 38 million) in the wild. In the US the percentage of Macs in the installed base is between 10% and 20% depending on the study being reported. In the rest of the world, Macs represent about 4-5% of all computers installed.
If, as you claim, prevalence of the target is important in "success" of a malware writer, then we should see that 4-5% of all malware should be Mac malware. It isn't. The number of viable self-propagating, self-installing, self-duplicating Mac OS X viruses in the wild, after eight years, is still ZERO. The number of Spyware in the wild is still ZERO. There are about 15 well documented and known Trojan horse applications, but they require extraordinary steps to get a Mac user to install them. Where is the Mac's fair share, its proportionate share, of the over 1 million malware now estimated to be in the Wild. It just does not exist.
Perhaps the reason it doesn't exist has something to do with the second part of what you know... the level of difficulty in creating attacks. It is extremely difficult to find a vector for any Mac malware to use to spread.
The actual fact is that the new science of Networking has shown that any computer placed on the internet is just as targetable as any other regardless of the prevalence of that computer. In 2006, the Witty Worm—written by ONE, single person—infected all 12,000 vulnerable Windows computers in the world (through their firewalls and anti-virus protection) in under 45 minutes (some reports say a half hour), regardless of where they were, or what they were doing on the Internet. They were not prevalent... they were, by any definition, extremely obscure; but their obscurity did not protect them from the Witty Worm. It found them, installed itself, and infected them.
What has prevented ONE, single person from writing a similarly effective and virulent Mac OS X virus to infect the 35 million Macs running bare naked of any protection at all (many without firewalls)? Obscurity? No way. Your second thing you know is the reason.
Neither of those examples is pertinent.
There’s no question Mac’s closed architecture does lend itself a greater level of protection than open environments like the PC world.
The core OS of OS X is UNIX, an open source architecture OS. Apple even publishes the source code of their kernel and many of their applications including Safari.
"Safari uses Apple's WebKit for rendering web pages and running JavaScript. WebKit consists of WebCore (based on Konqueror's KHTML engine) and JavaScriptCore (based on KDE's JavaScript engine named KJS). Like KHTML and KJS, WebCore and JavaScriptCore are free software and are released under the terms of the GNU Lesser General Public License. Some Apple improvements to the KHTML code are merged back into the Konqueror project. Apple also releases additional code under an open source 2-clause BSD-like license."
A lot of other components of OS X have been in the Open Source and GNU licensing for decades and are available for anyone to inspect and even change. Apple, which adhere to open standards, is far more open than Microsoft. Hundreds of thousands of programers have inspected and even added to the open source parts Mac OS, finding and fixing vulnerabilities.
You assumption of closed architecture is much more descriptive of Microsoft where only hundreds or thousands have inspected and/or revised the Windows source code to close vulnerabilities.
Since I know what I’m doing and know how to avoid most threats, I’d rather branch out, have a computer with a greater software compatibility list and more change in my pocket.
Currently, the Mac can legally run more software than can Windows computers.
Cal, sorry but your post reeked of FUD... with the implication being that Mac users were so dissatisfied with their Macs that they had to carry them back to the store for repair and servicing far in excess of what would be considered normal. I have seen the exact same statement/implication used by FUD spreaders. My apologies since the FUD shoe doesn't seem to fit... or is it the glove? ;^)>
Those who claim that there is no qualitative or productive difference to using a Mac are always those who have never really used one. There world view of computers is completely formed from their experience of using the crippled OS world of Windows and all of its issues and assuming that is the way it has to be. Most Mac users are intimately familiar with both platforms having to use Windows in their work. Many of us are IT professionals.
Commodore is a case-in-point, their computers were superior, technically speaking, to pretty much all of their competitors.
I finally threw my Amiga 3000 out this winter, so I have to agree. I recall showing a Mac Head my Amiga running Mac OS 8.6 in emulation faster than his same speed processor, $10,000 Mac II could.
I also had an Amiga 1000, 500, 2000, 2500, and the 3000. I never bought the 4000. I was never tempted by the Amiga 1200. I had an Orange Micro PC AT card installed in my 3000 as well.
Guilty!
But on the other hand, it’s VERY surprising what modern hardware Win98SE can run on!
That probably says more about Intel's and AMD's commitment to legacy hardware code than anything Microsoft may have done ...
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.