Some language books like
- Turbo Pascal (preferably something from Jeff Duntemann). A lot used to learn programming still.
- (introduction to) Modula 2 (Niklaus Wirth)
- Algorithms + Datastructures=programs (Niklaus Wirth)
- Delphi books, preferable by Marco Cantu' Not necessarily need the latest one.
- Math: Handbook for Automation series (half of the volumes seems to be German though)
E.g. when I'm abroard, reading the home news (-headlines) is the most important to me, but probably I won't be logged in via my own ISP.
And a e.g. US provider isn't going to pay for Swedish/Dutch content.
I can't decide what bundle of content I want, the ISP will buy what is cheapest. So probably I'll end up with some popular but worthless crap (Big Brother shows and relatives), because the usefull things are a niche. (according to the ISPs)
Paying is not bad. Paying and not being able to select yourself is bad, since it ends up to be the lowest common denomitor again.
Because it is purely OOP. I like to be able to switch between OOP and procedural programming depending on the job (or the part of the program I'm working on), without switching languages.
It is also said to be slow (not an own experience)
Why do you believe that if you borrow enough features from languages the language actually improves?
What you DON'T implement, is just as important as what you implement. Why?
Because sooner or later somebody stuff an obfuscated piece of code under your nose, and you wish that the ugly tricks in that code were never dubbed "features"
Also I never understood the reason why languages that stuff everything on one line, using nearly the entire ASCII set are more "powerfull"
Powerfull can be a lot of things (quality of code, easyness of tackeling new problems, clear syntax that saves on debugging time), but the amount of chars you have to type is never the limiting factor in programming.
I think they are afraid that if
they publish their use of BSD licensed software too much, the public opinion is going to demand the same favour from them, forced by licenses or not.
So Microsoft doesn't want to have publically funded GPL projects?
Pro:
Microsoft payed taxes too, and those publically funded projects could be indirectly funded by M$.
Con-to-Pro:
So what? Microsoft can still use those projects, it just can't sell them. It could runHotmail on Linux if it wanted. It just had to donate changes back:-)
Con-to-Pro-2: GPL project owners can still sign a commercial use license, which could be done to allow government developped technology in both commercial apps and Open Source protected by GPL.
But Microsoft wants it cheap, and for themselves alone. I'm member of a GPL project, and Microsoft hasn't approached us for a commercial license yet.
(which we would be happy to honor)
1.
Do you see possibilities on Linux for the small software developer (e.g. the shareware writer) ?
Or will Open Source be a means for the big corps to effectively kill commercial small software developper?
2.
Do you see a market for software without services (e.g. like Borland seems to try with Kylix) ?
Which is perfectly normal. However NOT over the backs of hobby programmers.
At least Microsoft applies its horrible licenses to stuff it actually owns.
- That would make the kernel *source* platform dependant (and i386 only probably).
- shared linked to a dozen libraries
- or static and make the kernel even more bigger.
If the work is solely yours, you can submit multiple licenses.
A lot of people (and specially anti-GPL zealots) often forget that.
So you can GPL it, give the appopriate license to the uni (and you implicitely seem to that),
write your cousin a commercial license, and if somebody from the BSDs bothers you about the GPL, you can sign him a BSD license etc, etc.
Re:But can you debug your programs?
on
GCC 3.0 Released
·
· Score: 1
Try the CVS. I also managed to get (Free) Pascal support running with post 5.0 GDB CVS on FreeBSD with some patches.
When I last checked (6 months ago), the CVS was in a pretty usable form.
Pascal,M2 ARE strongly typed, stronger than C in most respects. (Borland dialects are a bit more lenient btw)
Learning general programming skills is about learning fundamental concepts, and most important GET A CLUE WHAT YOU ARE DOING.
Strong typing forces that.
Same about autocleaning. If you can't deal with it, the study is a level too high and abstract for you. Might I suggest the Law department?:-)
It is not that working with an autocleaning language is bad, as long as it has benefits for the loss of control (like faster development cycles). But AFTER you graduate,
Generating CS majors that are unaware what their own code does is plain and simple bad.
During a lot of studies (I'm still working on my Chemistry Masters, but I did a CS bachelors at the side) the theory in classes is much heavier, and often never used when you start working. (Quantum Physics?). But they teach you a professional way of thinking, increase your level, and make you more flexible and all-round.
Slashdot Mirror
Kazaa also has some annoying ad stuff that is
installed if you install default.
One of them is new.net that ruined a lot of people's
machines on the ISP that I work.
New.net is a TLD "extender", but it is problematic.
Euuh, I don't want to ruin a good article, but the
"leadership" at Microsoft works because the leader
controls the pay check.
How do you enforce leadership on volunteers? Or even a group of companies that are practically independant of eachother?
Some language books like
- Turbo Pascal (preferably something from Jeff Duntemann). A lot used to learn programming still.
- (introduction to) Modula 2 (Niklaus Wirth)
- Algorithms + Datastructures=programs (Niklaus Wirth)
- Delphi books, preferable by Marco Cantu' Not necessarily need the latest one.
- Math: Handbook for Automation series (half of the volumes seems to be German though)
If you stack all papers discussing life on Mars on top of eachother, you can probably reach Mars
E.g. when I'm abroard, reading the home news (-headlines) is the most important to me, but probably I won't be logged in via my own ISP.
And a e.g. US provider isn't going to pay for Swedish/Dutch content.
I can't decide what bundle of content I want, the ISP will buy what is cheapest. So probably I'll end up with some popular but worthless crap (Big Brother shows and relatives), because the usefull things are a niche. (according to the ISPs)
Paying is not bad. Paying and not being able to select yourself is bad, since it ends up to be the lowest common denomitor again.
I don't call that procedural. I can create static container classes in Java too.
It is not just the function namespace itself, it also means having straight types which aren't OOP contained.
One of the major uses of this is that you can write gluecode to external code/data in the language itself.
Don't forget to weigh the values by population!
C++ is not ideal to standarise on. It carries too much legacy.
A redesigned cleaned up C++ might do it (though it would pretty much look like Delphi's Object Pascal I think)
Because it is purely OOP. I like to be able to switch between OOP and procedural programming depending on the job (or the part of the program I'm working on), without switching languages.
It is also said to be slow (not an own experience)
Why do you believe that if you borrow enough features from languages the language actually improves?
What you DON'T implement, is just as important as what you implement. Why?
Because sooner or later somebody stuff an obfuscated piece of code under your nose, and you wish that the ugly tricks in that code were never dubbed "features"
Also I never understood the reason why languages that stuff everything on one line, using nearly the entire ASCII set are more "powerfull"
Powerfull can be a lot of things (quality of code, easyness of tackeling new problems, clear syntax that saves on debugging time), but the amount of chars you have to type is never the limiting factor in programming.
I think they are afraid that if
they publish their use of BSD licensed software too much, the public opinion is going to demand the same favour from them, forced by licenses or not.
Only the "integrated" IE does that.
:-)
If you kill it with 98lite/ IEradicator, and install back stand alone, it doesn't
I did that at the job to save some mem for Mozilla
No spectacular breakthrough. Interfaces are still defined as C only.
Even Microsoft specifies quite a lot of interfaces in IDL nowadays.
Fully agree, you can also see this on *BSD, where the Ports tree is beloved for exactly the same reasons
So Microsoft doesn't want to have publically funded GPL projects?
Pro:
Microsoft payed taxes too, and those publically funded projects could be indirectly funded by M$.
Con-to-Pro:
So what? Microsoft can still use those projects, it just can't sell them. It could runHotmail on Linux if it wanted. It just had to donate changes back
Con-to-Pro-2: GPL project owners can still sign a commercial use license, which could be done to allow government developped technology in both commercial apps and Open Source protected by GPL.
But Microsoft wants it cheap, and for themselves alone. I'm member of a GPL project, and Microsoft hasn't approached us for a commercial license yet.
(which we would be happy to honor)
to companies that market with Linux also actually deliver, and also in a workable way (GPL).
Now I just hope that the GPL won't be too much of a problem to push this into FreeBSD
1.
Do you see possibilities on Linux for the small software developer (e.g. the shareware writer) ?
Or will Open Source be a means for the big corps to effectively kill commercial small software developper?
2.
Do you see a market for software without services (e.g. like Borland seems to try with Kylix) ?
Which is perfectly normal. However NOT over the backs of hobby programmers. At least Microsoft applies its horrible licenses to stuff it actually owns.
How could you do that?
- That would make the kernel *source* platform dependant (and i386 only probably).
- shared linked to a dozen libraries
- or static and make the kernel even more bigger.
If the work is solely yours, you can submit multiple licenses.
A lot of people (and specially anti-GPL zealots) often forget that.
So you can GPL it, give the appopriate license to the uni (and you implicitely seem to that),
write your cousin a commercial license, and if somebody from the BSDs bothers you about the GPL, you can sign him a BSD license etc, etc.
Try the CVS. I also managed to get (Free) Pascal support running with post 5.0 GDB CVS on FreeBSD with some patches.
When I last checked (6 months ago), the CVS was in a pretty usable form.
I see language changes (Java integrated, Fortran improvements), but I kind of miss the years old Gnu Pascal project?
Still, I think you should be more concerned about their basic programming skills, then their ability to write (more advanced) C++ code.
IOW if a proper introductionary programming costs a semester C++ programming, I think that is a good tradeoff.
Pascal,M2 ARE strongly typed, stronger than C in most respects. (Borland dialects are a bit more lenient btw)
Learning general programming skills is about learning fundamental concepts, and most important GET A CLUE WHAT YOU ARE DOING.
Strong typing forces that.
Same about autocleaning. If you can't deal with it, the study is a level too high and abstract for you. Might I suggest the Law department?
It is not that working with an autocleaning language is bad, as long as it has benefits for the loss of control (like faster development cycles). But AFTER you graduate,
Generating CS majors that are unaware what their own code does is plain and simple bad.
During a lot of studies (I'm still working on my Chemistry Masters, but I did a CS bachelors at the side) the theory in classes is much heavier, and often never used when you start working. (Quantum Physics?). But they teach you a professional way of thinking, increase your level, and make you more flexible and all-round.
6510 and a 6502 in a 9 node cluster (A c64 + 8 drives (8 trough 15). The connecting bus is a little bit slower than normal though :-)