I don't want to make or eat dog food (java or whatever the managers wanted for me). I want to make and eat crunchy and tasty nice cookies (Python or Ruby could be it).
I decided that, and understood what the term "dog food" really meant when reading this:
I see no reason to be closer to Java, as I'm more productive in C++ (with wxWidgets) than in Java, in the kind of applications where you can use either.
I guess you evaluated both toolkits but you have written real code only with QT. In my case I evaluated both and I have written real code only with wxWidgets, so take my bias into account.
I agree that MFC message maps are ugly, but they are not obligatory, in fact I use none of them in my new wxCode. I have code like this in the constructors (or wherever is needed):
I see no difference using one or the other except for this, so it's a non issue to me.
I can do someTextCtrl << "number: " << someintvar; if that's what you mean by nice modern stuff.
And yes, the wxWidgets license is a GREAT advantage (one of the best licenses ever).
And about your remark about a subset of C++ coming directly from the stone age: Templates are good and nice, when used like in the STL but they are not necessary, just like multiple inheritance is nice to use sometimes, but should not be overused just because C++ has it. And everything should not be a class (like in Java, ughh) when it's not necessary. Some classes don't need to be part of a class hierarchy, and some classes don't need getters and setters. And templates are much harder to debug, both for the programmer and for the IDE.
Using just a subset of C++ is the right way to use C++ IMHO. In fact, sometimes using it just as a better C is the right way. And you still have the full power of C++ when you need it, which is not all the time.
So, my point is (for the guy with the problem): wxWidgets is a very good option (IMO much better than java if you have a C++ codebase) an you should evaluate it without any bias.
Since fugu's poison can lead to instantaneous deaths of diners, only licensed cooks are allowed to prepare fugu. You must have special skills and knowledge about fugu to be licensed. Poisonous parts of fugu differ, depending on the kind of fugu.
However, managers usually go and get the cheaper/unlicensed cooks. And then complain about the project's failure.
I'm sorry but I can't agree with you. And in some sense I agree with you. We just have different definitions of what math is.
To me math is all about formal logical systems, even more than computer science, and this is what I've learnt studying pure mathematics.
Computer science on the other hand, deals with effective ways of performing computations. In retrospective, this search for effective ways have led me to math.
So, to me (and my math teachers, and their teachers, probably as well) math is a superset of logic, and of computer science.
Then again, the only problem I see with my teachers is that they want to be the computer as well (some are incredibly good at that) and you and me know that computers are far better than us for pushing bits around.
In math, you only express a relationship between things in numerical terms.
Modern maths are not about numbers. You're thinking about algebra there. Of course algebra is easier than programming.
In fact, what they teached me as maths when studing software engineering is very different to what they teached me when studing pure mathematics. In software engineering they teached me some pen and paper algorithms for doing some calculations. I met a lot of nice girls who couln't understand programming but were otherwise very good in calculus, so in general, yes, programming seems to be harder than maths.
But modern maths are very different to what they first teached me.
Modern Maths are about abstract things you define arbitrarily in terms of the relations between them. If you change a little of the definition you have something entirely different. Then you explore what those relationships can do inside a formal system. Sometimes you get completely unexpected results and unexpected relationships with other theories. Even if you use computers to automatize the most tedious part of a proof there is still a lot of work to do.
In programming you have very powerful tools that make most complex tasks easier. You don't need a parser in your head because compilers will tell you all syntactic errors. You don't need to guess about complexity that much, a profiler will tell you the botlenecks.
I stand in my assertion than someone who knows maths can easily learn to program, but that someone who can only program can hardly understand more advanced mathematical concepts. More difficult if they a very dependant on a particular tool.
Just look at the names of Turing, von Newman, Knuth, I probably missed most people, all mathematicians, and all of them made something significative for computer science.
Now is your turn, name someone who does a lot of programming and has no mathematical knowledge doing something really significative for computer science. "Web 2.0" reimplementations of 30+ years old concepts are not significative contributions.
I know some people who started in philosophy, got biten by the logic bug, and finished with a maths degree.
And then, a very smart female friend started in physics, and then went to continue with philosophy.
And then some of the people we study, Descartes is both a mathematician and a philosopher.
However, logic is very fundamental to mathematics. You do proofs using mathematical logic.
And then if you start with sets, and define some relationships like "belongs to" and after some more definitions suddenly you discover boolean algebra!
There are some ways to convert normal function calls to CPS.
And there is something called monads used to convert imperative algorithms to functional style.
And yes, continuations can be a very powerful technique.
However, CPS functional code is still coding an algorithm. Any way to compute something is an algorithm. May be you should name your critic "I dislike imperative algorithms, and I like CPS functional algorithms."
As a coder (you can't rule out C++ and lisp knowledge) and a mathematician (well, not really graduated, yet) I can say that programming and maths have a lot in common. In fact, math is harder, and after doing some higher maths, you will surely be a better coder. Maths expands your mind.
However, coding has (almost) nothing to do with what is normally seen as maths, it's not like algebra, it's not calculus, and is not linear algebra either (most of the time). In that sense you're totally right. The current emphasis in that math == calculus is, in my mind, just wrong, given current market trends.
I would like to see a more computer science oriented maths curriculum.
Maths are behind everything. Logic is a part of maths. The theory used to manage concurrent programming is, guess what? A mathematical model.
Regular grammars, and by extension, regular expressions are a part of mathematics. Don't tell me you don't use regular expressions in your code. At least, I do.
Most of the maths you need when coding are found in the book Concrete Mathematics by Knuth. And what is found in this book is also very different to what is normally teached as maths.
So, yes, if you know what maths really are (if math teachers knew about programming), coding has a lot to do with mathematics.
I would like to be able to click in some part of a comment to hide it and all its responses, so I can read the comments I like and not all the stuff that gets posted.
Please some of the CSS hackers submiting designs for the new/. take this into account.
Back in the day, when only Opera had tabs, I read lots of FF users saying that tabbed browsing was unnatural, idiotic, or just unneded. Now that FF has tabs, it is the best thing since sliced bread.
However, I vastty prefer to have technical discussions, and I admit I was wrong about aussie_a motives. You can read the rest of my conversation with him. And leave the stoning to the religious guys, I'm a proud atheist and I listen to arguments.
Now back to the article. You can say the autor is the biggest troll ever, but you can't deny the speed stuff. Or the CSS acid test. Or the memory leaks in FF. Even if the autor endorses a IE skin as his prefered browser, with you and I and ten thousand readers knowing that FF is better than anything IE related.
You just can't deny those points, so you went to attack the author of the article. Now who's the troll?
If Opera is not as used as its technical merits suggest (the only important ones to me as I don't care about political agendas as FF users do), is because the huge and ugly advertising banner Opera used to have, and which was removed in response to FF gaining popularity.
In fact I hope FF gains even more popularity, if just to force everyone to write standars compliant web pages. Believe me, banks are starting to notice. And then you and me and everyone else will use our favorite browser.
I think that the XML and the LaTeX source text would be processed separately, so the logic in one thing doesn't affect the logic in the other.
Something like this for the LaTeX stuff(using php), keeping the rest of your XML stuff intact:
<SCRIPT language="JavaScript" src="jsMath.js"></SCRIPT> <?php $number = random(); //something more fancy for rationals is needed here $ntype = gettype($number); ?> text and more text <DIV CLASS="math"> <!-- you see, this is just text: --> \phi(z) = <?php switch ($ntype) {
case "integer": //here goes more LaTeX, just printed in a string, no DOM processing
print "\sqrt{" . $number . "}";
break; // the other cases and other format for the numbers } ?> </DIV>
With that example you can see that it's not more taxing for your server, and the result can be seen with IE and Opera. The size of the fonts, could it be that you have other version of the fonts like the ones in SWP?
Anyway the whole point of jsMath is to shift LaTeX rendering to the browser. Server side LaTeX is much more expensive than server side XML and almost everything else.
I don't think Opera will implement MathML, neither IE, that's why I started looking for alternatives.
The bookmark bar is just called "personal toolbar" in Opera.
The buttons in the menu bar is just a big "uhh?" for me, as I see exactly the same space wasted in Opera as in Firefox, except for just a couple pixels in favour of FF. And I removed the bookmarks and status bars in FF.
The feeds feature, yes I can see why you use it that way, and indeed is a great idea. Opera should have a "server based session bookmark" feature, to compete here. Maybe a user.js bookmark button is the answer for that.
The find in page feature: you're just used to the other one, I see a non-issue there.
Slashdot Mirror
I don't want to make or eat dog food (java or whatever the managers wanted for me). I want to make and eat crunchy and tasty nice cookies (Python or Ruby could be it).
I decided that, and understood what the term "dog food" really meant when reading this:
I Have a Feeling We're Not In Emerald City Anymore by Henry G. Baker.
I guess you evaluated both toolkits but you have written real code only with QT. In my case I evaluated both and I have written real code only with wxWidgets, so take my bias into account.
I agree that MFC message maps are ugly, but they are not obligatory, in fact I use none of them in my new wxCode. I have code like this in the constructors (or wherever is needed):It's just convenient for some people used to the MFC maps.
And the fact that it uses no templates, is an advantage to me. (FAST compilation, smaller exes)is no harder to write thanI see no difference using one or the other except for this, so it's a non issue to me.
I can do someTextCtrl << "number: " << someintvar; if that's what you mean by nice modern stuff.
And yes, the wxWidgets license is a GREAT advantage (one of the best licenses ever).
And about your remark about a subset of C++ coming directly from the stone age:
Templates are good and nice, when used like in the STL but they are not necessary, just like multiple inheritance is nice to use sometimes, but should not be overused just because C++ has it.
And everything should not be a class (like in Java, ughh) when it's not necessary. Some classes don't need to be part of a class hierarchy, and some classes don't need getters and setters.
And templates are much harder to debug, both for the programmer and for the IDE.
Using just a subset of C++ is the right way to use C++ IMHO.
In fact, sometimes using it just as a better C is the right way.
And you still have the full power of C++ when you need it, which is not all the time.
So, my point is (for the guy with the problem): wxWidgets is a very good option (IMO much better than java if you have a C++ codebase) an you should evaluate it without any bias.
Although I vastly prefer Opera, anything that can help decrease market share of IE and its broken everything is good.
It's by far the most flexible language I've ever used. It allows programmers to modify the most fundmental aspects of the language.
I guess you will love LISP macros.
Software is like cooking Blow-Fish / FUGU.
Since fugu's poison can lead to instantaneous deaths of diners, only licensed cooks are allowed to prepare fugu. You must have special skills and knowledge about fugu to be licensed. Poisonous parts of fugu differ, depending on the kind of fugu.
However, managers usually go and get the cheaper/unlicensed cooks. And then complain about the project's failure.
I'm sorry but I can't agree with you. And in some sense I agree with you. We just have different definitions of what math is.
To me math is all about formal logical systems, even more than computer science, and this is what I've learnt studying pure mathematics.
Computer science on the other hand, deals with effective ways of performing computations. In retrospective, this search for effective ways have led me to math.
So, to me (and my math teachers, and their teachers, probably as well) math is a superset of logic, and of computer science.
Then again, the only problem I see with my teachers is that they want to be the computer as well (some are incredibly good at that) and you and me know that computers are far better than us for pushing bits around.
Modern maths are not about numbers. You're thinking about algebra there. Of course algebra is easier than programming.
In fact, what they teached me as maths when studing software engineering is very different to what they teached me when studing pure mathematics. In software engineering they teached me some pen and paper algorithms for doing some calculations. I met a lot of nice girls who couln't understand programming but were otherwise very good in calculus, so in general, yes, programming seems to be harder than maths.
But modern maths are very different to what they first teached me.
Modern Maths are about abstract things you define arbitrarily in terms of the relations between them. If you change a little of the definition you have something entirely different. Then you explore what those relationships can do inside a formal system. Sometimes you get completely unexpected results and unexpected relationships with other theories. Even if you use computers to automatize the most tedious part of a proof there is still a lot of work to do.
In programming you have very powerful tools that make most complex tasks easier. You don't need a parser in your head because compilers will tell you all syntactic errors. You don't need to guess about complexity that much, a profiler will tell you the botlenecks.
And you can be very formal in computer science.
I stand in my assertion than someone who knows maths can easily learn to program, but that someone who can only program can hardly understand more advanced mathematical concepts. More difficult if they a very dependant on a particular tool.
Just look at the names of Turing, von Newman, Knuth, I probably missed most people, all mathematicians, and all of them made something significative for computer science.
Now is your turn, name someone who does a lot of programming and has no mathematical knowledge doing something really significative for computer science. "Web 2.0" reimplementations of 30+ years old concepts are not significative contributions.
I think the controversy will still be open.
I know some people who started in philosophy, got biten by the logic bug, and finished with a maths degree.
And then, a very smart female friend started in physics, and then went to continue with philosophy.
And then some of the people we study, Descartes is both a mathematician and a philosopher.
However, logic is very fundamental to mathematics. You do proofs using mathematical logic.
And then if you start with sets, and define some relationships like "belongs to" and after some more definitions suddenly you discover boolean algebra!
All this is very circular, in a way.
What you're talking about is very close to CPS.
There are some ways to convert normal function calls to CPS.
And there is something called monads used to convert imperative algorithms to functional style.
And yes, continuations can be a very powerful technique.
However, CPS functional code is still coding an algorithm. Any way to compute something is an algorithm. May be you should name your critic "I dislike imperative algorithms, and I like CPS functional algorithms."
As a coder (you can't rule out C++ and lisp knowledge) and a mathematician (well, not really graduated, yet) I can say that programming and maths have a lot in common. In fact, math is harder, and after doing some higher maths, you will surely be a better coder. Maths expands your mind.
However, coding has (almost) nothing to do with what is normally seen as maths, it's not like algebra, it's not calculus, and is not linear algebra either (most of the time). In that sense you're totally right. The current emphasis in that math == calculus is, in my mind, just wrong, given current market trends.
I would like to see a more computer science oriented maths curriculum.
Maths are behind everything. Logic is a part of maths. The theory used to manage concurrent programming is, guess what? A mathematical model.
Regular grammars, and by extension, regular expressions are a part of mathematics. Don't tell me you don't use regular expressions in your code. At least, I do.
Most of the maths you need when coding are found in the book Concrete Mathematics by Knuth. And what is found in this book is also very different to what is normally teached as maths.
So, yes, if you know what maths really are (if math teachers knew about programming), coding has a lot to do with mathematics.
That's very good news indeed.
I think that's brilliant, or at least the only workable solution.
Gmail (and others) have a "Mark as SPAM" button, and now that button should be extended to the entire web.
But it can only be in the google toolbar, and similar toolbars, and not many users have them installed, or do they?
And then is the whole "I will report as SPAM my competition's home page" issue.
I didn't knew about that.
And what about other browsers besides Firefox? Or when I'm in places where I can't install software?
That's why I want this to be by default in slashdot.
You have just proved to me that it should be here by default.
Anyway, I'm trying them ASAP, there is anyone that you prefer?
It seems one is deprecated and other is untested.
I would like to be able to click in some part of a comment to hide it and all its responses, so I can read the comments I like and not all the stuff that gets posted.
/. take this into account.
Please some of the CSS hackers submiting designs for the new
It's not silly, it's also based on experience, mostly experience listening to priests, ministers and so on.
However, what you said is very logical, I wish every believer had that attitude.
It's better to have an interactive read-eval-print loop and program using it.
Sorry that your language doesn't seem to provide one.
Why I feel like a cockroach after a meeting. :P
Back in the day, when only Opera had tabs, I read lots of FF users saying that tabbed browsing was unnatural, idiotic, or just unneded. Now that FF has tabs, it is the best thing since sliced bread.
= 18185, with time I have learned to distrust the technical arguments of lots of FF users.
This is another of those pearls (beware, link in spanish): http://www.cristalab.com/v3/foros/viewtopic.php?t
However, I vastty prefer to have technical discussions, and I admit I was wrong about aussie_a motives. You can read the rest of my conversation with him. And leave the stoning to the religious guys, I'm a proud atheist and I listen to arguments.
Now back to the article. You can say the autor is the biggest troll ever, but you can't deny the speed stuff. Or the CSS acid test. Or the memory leaks in FF. Even if the autor endorses a IE skin as his prefered browser, with you and I and ten thousand readers knowing that FF is better than anything IE related.
You just can't deny those points, so you went to attack the author of the article. Now who's the troll?
If Opera is not as used as its technical merits suggest (the only important ones to me as I don't care about political agendas as FF users do), is because the huge and ugly advertising banner Opera used to have, and which was removed in response to FF gaining popularity.
In fact I hope FF gains even more popularity, if just to force everyone to write standars compliant web pages. Believe me, banks are starting to notice. And then you and me and everyone else will use our favorite browser.
So, what's your record in seconds?
I got a lucky one once and made 13 seconds in the small board.
http://mywebpages.comcast.net/SupportCD/FirefoxMyt hs.html
Then people will sell the combos.
"PS3 plus 50+ games for only US$500!!!" on eBay and related sites.
The console dying thing, well that's a lot of legal ground for a class action suit.
Something like this for the LaTeX stuff(using php), keeping the rest of your XML stuff intact:With that example you can see that it's not more taxing for your server, and the result can be seen with IE and Opera. The size of the fonts, could it be that you have other version of the fonts like the ones in SWP?
Anyway the whole point of jsMath is to shift LaTeX rendering to the browser. Server side LaTeX is much more expensive than server side XML and almost everything else.
I don't think Opera will implement MathML, neither IE, that's why I started looking for alternatives.
Just to clarify my other comment.
You print LaTeX source code in the page. It's just more text.
The JS makes all the hard work of converting that source code to a nice math output. Pure Web client based stuff.
It has nothing to do with DVI, PDF, PS, or any usual LaTeX toolchain.
The bookmark bar is just called "personal toolbar" in Opera.
The buttons in the menu bar is just a big "uhh?" for me, as I see exactly the same space wasted in Opera as in Firefox, except for just a couple pixels in favour of FF. And I removed the bookmarks and status bars in FF.
The feeds feature, yes I can see why you use it that way, and indeed is a great idea. Opera should have a "server based session bookmark" feature, to compete here. Maybe a user.js bookmark button is the answer for that.
The find in page feature: you're just used to the other one, I see a non-issue there.