FIRST--I will recommend my gBasic project (gbasic.sourceforge.com) because it will offer Rapid Application Development, terminal independant interfaces (including curses, X, and HTML/CGI), and extensive data processing / string manipulation capabilities. Also aiming for some degree of VBScript compatibility. What I've done already allows Grammar morphing such as this script formt he gbasic.conf file: NCONST SCONST expr "+" expr expr "-" expr expr "*" expr expr "/" expr expr "=" expr expr ">" expr expr " "if" cond "then" stmt var "=" expr "print" expr "input" var However, different languages have different strengths in different areas and I, as a programmer of several, do not feel we are anywhere close to creating any one to take over for all. This question largely depends upon the application you are creating with CGI. But obviously it should have good string manipulation capabilities and in most cases good database and/or data processing capabilities, and should in most cases be quick to develop in...as the nature of the web itself is quick.. Perl meets all of these points very well from a functional stand point (in a way that no other is worthy of comparison--except maybe Python)--but suffers miserably when considering ease of programming and readability. The Perl syntax is rather difficult to get used to as where in some places even one misplaced space can make it not work...and it's syntax is highly un-intuitive. It's a mix of shell scripting, C, and other similar languages that makes certain parts of it easier for those who understand those languages before learning Perl. Chipmunk BASIC -- On the other hand is, in my opinion, the next best candidate (may come as a surprise to most programmers). It's VERY EASY and FAST to program in...has decent string manipulations and acceptible array/file I/O capabilities.. But it has no DB support, no specific CGI functions (although that's very easy to make), is primarilly developed for the Macintosh (Linux/UNIX second (but decent), and the Windows version sucks badly), and is not GPL'd. --Slicker
Seems to me it'd be better and faster to use an Expert System for technology patents. Although I've always objected to calling them part of Artificial Intelligence (since the're just a bunch of embedded if..then..else..clauses), humans can't possibly do as well as a machine in researching the mountains of patent descriptions that exist. Patent law DEFINITELY needs an overhaul and vast liberalization as a whole, but I feel there are still some legitimate uses of patenting. I can't quite place the line in terms of what "technologies" really should be patentable when it comes to software....but vaguelly perhaps a person could say, "the sole use of a complex algorythm not previously used by anyone else." Even this could be abused.. Copyrights on software are ok...I guess patents really don't make any sense at all. --Matthew
Re:Motif "ugly" while GTK "beautiful"?? I LOVE QT.
on
Death of CDE & Motif?
·
· Score: 1
I'm nuts about QT--programming-wise. MFC and OWL are also OO, but QT is the only toolkit I'm aware of that actually fully employs the philosophy. It's vastly abstracted and yet more flexible. I don't need to have it event driven, but I can. I can reduce my programming work becaues it works like the concept of small tools....everything is directly inter-operable. QT truelly is a revolutionariy concept..and it's fast....but I agree that GTK and Motif are both vastly easy to code than MFC or OWL. Plus...QT is portable... I like that, too. GTK is just another Motif almost. QT is fundamentally new. email: matthew@tedder.com --Matthew
Standards need to be developed across desktops but Linux and UNIX are operating systems of choice and I don't think any of them should die. 1. Cut and Paste should be extended across all of them to handle objects and stuff...hopefully better than OLE. 2. A desktop's API should have some standard developed for basic things...just like POSIX covers as much as possible across UNIXes. I think that any compatibility sacrifices with old software to produce such standards in the future would even be worthwhile sacrifices. The point is that we MUST GET STANDARDS NOW! Even a semi-bad standard is better than none. email: matthew@tedder.com --Matthew
First, Linux man pages are only good for experts who really don't even need them as much. Second, why do they have to be so boring? AIX man pages MUCH easier for people who are learning and provides command and useful examples at the end of each. Linux could learn A LOT from AIX--simplicity without sacrificing flexibility. Next, we need all those HOWTOs that reference each other every other sentence to link to at least link to each other instead. But I have a much better idea--EXPANDING HYPERLINKS. My wife is a doctoral candidate at the College of Journalism and Mass Communications at the University of Florida. She's concentrated her various studies on the effectiveness of learning via computer-mediated communications. Some of her findings include that traditional linear text has better recall and comprehension scores for all types of readers than does paged hypertext such as HTML, SGML, etc.. It now seems clear that, while they can get to specific information faster, flipping through pages tends to moderate disorient readers causing poorer learning. My suggested solution is EXPANDING HYPERTEXT which she will be doing an experiment with in January / Febuary for her dissertation (Ph.D final research paper). My program (which I'm working on little bit little) uses a slightly extended version of Rich Text Format (RTF) for paged and expanding links. The expanding link simply insert the text from the linked page right into a designated spot (typically the end of a sentence). As an example, I make make a list of chapter headings as the first document. Each chapter heading serves as a link to subsections titles. Each subsection title serves as a link to paragraph topic sentences. Each topic sentence serves as a link to the paragraph. Words within the paragraph link to anything anywhere... By clicking on the links, you can expand and deflate greater detail where desired. Some of these links may even refer to previous chapters or paragraphs--which is good. Unstructured links are better for recall and comprehension. As a programmer I have to admit this is nice also because 90% of learning anything from tutorials is sifting through the BS and repetition for useful content. Examples are therefore often the best content to be found. Explanations and theory are secondary. --Matthew
Two Questions: 1. Can I expect KDE 2.0 to be significantly faster and less resource hungery that previous versions? To everyone I've demonstrated KDE to (unless on P-450s), those problems have been immediately noticeable...especially when compared with Win95 on the same machine. 2. Has there been any re-work to resize dialogs to accomodate at least 800x600 resolutions? Many people still prefer that resolution, because they hate tiny fonts and icons... I don't like the large dialogs even on 1024x768, cause I'm trying to do as much as I can on my desktop without having to flip screens constantly.. Or are these two issues unmentioned or ignored? --Matthew
I'm in a Stinger Unit and we've shot down all kinds of missile just for target practice all the time....right up their butts...even when the burners went out....missile to missile hits is not hard at all. And I mean direct hits to small, fast moving missiles.. If fact, it's much harder to miss.. The problem with the scuds was that they were fired upon as they began their descent, speeding up which broke them apart... The Patriots then had two and three targets instead of one. This is also the primary difficulty of a Nuclear missile defense...such things are done intentionally. Also...when the U.S.S.R broke apart, Yeltsin proposed it and the parliament voted and decided who would keep the nukes and maintain the various treaties. But I think that in the diplomatic environment, treaties on paper are merely a manifestation of what they countries would do anyway--remember there is no international police that's going to lock up a country which breaches a treaty... Even UN sanctions and the world court rulings only work when countries decide to support them. Its right of the mightiest--and that's all. Treaties just make things look official. But I believe strongly in the need for such development, regardless.. The DPRK's Rodong and Taepodong missiles are of much improved quality with long ranges and high accuracy..plus, they're far more willing to use them and openly export the technology to "anyone willing to pay." Which has been several African countries (for resale to various Arab countries), Iran, and possibly Pakastan and India. AMM (Anti-missile missiles) need to be cheaper and more numerous. Hell, an F16 could shoot down a nuke if it were in the right place at the right time. --Matthew
Slashdot Mirror
<expr> NCONST </expr>
<expr> SCONST </expr>
<expr=plus> expr "+" expr </expr>
<expr=minus> expr "-" expr </expr>
<expr=multiply> expr "*" expr </expr>
<expr=divide> expr "/" expr </expr>
<cond=equals> expr "=" expr </cond>
<cond=greater> expr ">" expr </cond>
<cond=less> expr "<" expr </cond>
<stmt=ifthen> "if" cond "then" stmt </stmt>
<stmt=assign> var "=" expr </stmt>
<stmt=print> "print" expr </stmt>
<stmt=input> "input" var </stmt>
FIRST--I will recommend my gBasic project (gbasic.sourceforge.com) because it will offer Rapid Application Development, terminal independant interfaces (including curses, X, and HTML/CGI), and extensive data processing / string manipulation capabilities. Also aiming for some degree of VBScript compatibility. What I've done already allows Grammar morphing such as this script formt he gbasic.conf file: NCONST SCONST expr "+" expr expr "-" expr expr "*" expr expr "/" expr expr "=" expr expr ">" expr expr " "if" cond "then" stmt var "=" expr "print" expr "input" var However, different languages have different strengths in different areas and I, as a programmer of several, do not feel we are anywhere close to creating any one to take over for all. This question largely depends upon the application you are creating with CGI. But obviously it should have good string manipulation capabilities and in most cases good database and/or data processing capabilities, and should in most cases be quick to develop in...as the nature of the web itself is quick.. Perl meets all of these points very well from a functional stand point (in a way that no other is worthy of comparison--except maybe Python)--but suffers miserably when considering ease of programming and readability. The Perl syntax is rather difficult to get used to as where in some places even one misplaced space can make it not work...and it's syntax is highly un-intuitive. It's a mix of shell scripting, C, and other similar languages that makes certain parts of it easier for those who understand those languages before learning Perl. Chipmunk BASIC -- On the other hand is, in my opinion, the next best candidate (may come as a surprise to most programmers). It's VERY EASY and FAST to program in...has decent string manipulations and acceptible array/file I/O capabilities.. But it has no DB support, no specific CGI functions (although that's very easy to make), is primarilly developed for the Macintosh (Linux/UNIX second (but decent), and the Windows version sucks badly), and is not GPL'd. --Slicker
Seems to me it'd be better and faster to use an Expert System for technology patents. Although I've always objected to calling them part of Artificial Intelligence (since the're just a bunch of embedded if..then..else..clauses), humans can't possibly do as well as a machine in researching the mountains of patent descriptions that exist. Patent law DEFINITELY needs an overhaul and vast liberalization as a whole, but I feel there are still some legitimate uses of patenting. I can't quite place the line in terms of what "technologies" really should be patentable when it comes to software....but vaguelly perhaps a person could say, "the sole use of a complex algorythm not previously used by anyone else." Even this could be abused.. Copyrights on software are ok...I guess patents really don't make any sense at all. --Matthew
I'm nuts about QT--programming-wise. MFC and OWL are also OO, but QT is the only toolkit I'm aware of that actually fully employs the philosophy. It's vastly abstracted and yet more flexible. I don't need to have it event driven, but I can. I can reduce my programming work becaues it works like the concept of small tools....everything is directly inter-operable. QT truelly is a revolutionariy concept..and it's fast....but I agree that GTK and Motif are both vastly easy to code than MFC or OWL. Plus...QT is portable... I like that, too. GTK is just another Motif almost. QT is fundamentally new. email: matthew@tedder.com --Matthew
Standards need to be developed across desktops but Linux and UNIX are operating systems of choice and I don't think any of them should die. 1. Cut and Paste should be extended across all of them to handle objects and stuff...hopefully better than OLE. 2. A desktop's API should have some standard developed for basic things...just like POSIX covers as much as possible across UNIXes. I think that any compatibility sacrifices with old software to produce such standards in the future would even be worthwhile sacrifices. The point is that we MUST GET STANDARDS NOW! Even a semi-bad standard is better than none. email: matthew@tedder.com --Matthew
First, Linux man pages are only good for experts who really don't even need them as much. Second, why do they have to be so boring? AIX man pages MUCH easier for people who are learning and provides command and useful examples at the end of each. Linux could learn A LOT from AIX--simplicity without sacrificing flexibility. Next, we need all those HOWTOs that reference each other every other sentence to link to at least link to each other instead. But I have a much better idea--EXPANDING HYPERLINKS. My wife is a doctoral candidate at the College of Journalism and Mass Communications at the University of Florida. She's concentrated her various studies on the effectiveness of learning via computer-mediated communications. Some of her findings include that traditional linear text has better recall and comprehension scores for all types of readers than does paged hypertext such as HTML, SGML, etc.. It now seems clear that, while they can get to specific information faster, flipping through pages tends to moderate disorient readers causing poorer learning. My suggested solution is EXPANDING HYPERTEXT which she will be doing an experiment with in January / Febuary for her dissertation (Ph.D final research paper). My program (which I'm working on little bit little) uses a slightly extended version of Rich Text Format (RTF) for paged and expanding links. The expanding link simply insert the text from the linked page right into a designated spot (typically the end of a sentence). As an example, I make make a list of chapter headings as the first document. Each chapter heading serves as a link to subsections titles. Each subsection title serves as a link to paragraph topic sentences. Each topic sentence serves as a link to the paragraph. Words within the paragraph link to anything anywhere... By clicking on the links, you can expand and deflate greater detail where desired. Some of these links may even refer to previous chapters or paragraphs--which is good. Unstructured links are better for recall and comprehension. As a programmer I have to admit this is nice also because 90% of learning anything from tutorials is sifting through the BS and repetition for useful content. Examples are therefore often the best content to be found. Explanations and theory are secondary. --Matthew
Two Questions: 1. Can I expect KDE 2.0 to be significantly faster and less resource hungery that previous versions? To everyone I've demonstrated KDE to (unless on P-450s), those problems have been immediately noticeable...especially when compared with Win95 on the same machine. 2. Has there been any re-work to resize dialogs to accomodate at least 800x600 resolutions? Many people still prefer that resolution, because they hate tiny fonts and icons... I don't like the large dialogs even on 1024x768, cause I'm trying to do as much as I can on my desktop without having to flip screens constantly.. Or are these two issues unmentioned or ignored? --Matthew
I'm in a Stinger Unit and we've shot down all kinds of missile just for target practice all the time....right up their butts...even when the burners went out....missile to missile hits is not hard at all. And I mean direct hits to small, fast moving missiles.. If fact, it's much harder to miss.. The problem with the scuds was that they were fired upon as they began their descent, speeding up which broke them apart... The Patriots then had two and three targets instead of one. This is also the primary difficulty of a Nuclear missile defense...such things are done intentionally. Also...when the U.S.S.R broke apart, Yeltsin proposed it and the parliament voted and decided who would keep the nukes and maintain the various treaties. But I think that in the diplomatic environment, treaties on paper are merely a manifestation of what they countries would do anyway--remember there is no international police that's going to lock up a country which breaches a treaty... Even UN sanctions and the world court rulings only work when countries decide to support them. Its right of the mightiest--and that's all. Treaties just make things look official. But I believe strongly in the need for such development, regardless.. The DPRK's Rodong and Taepodong missiles are of much improved quality with long ranges and high accuracy..plus, they're far more willing to use them and openly export the technology to "anyone willing to pay." Which has been several African countries (for resale to various Arab countries), Iran, and possibly Pakastan and India. AMM (Anti-missile missiles) need to be cheaper and more numerous. Hell, an F16 could shoot down a nuke if it were in the right place at the right time. --Matthew