However, somehow despite all parser theory and -engine progress, nobody still seems to be able to produce a C++ compiler with non cryptic errormessages:-)
From what I recall as both a Linux and BSD hack at the time, the boost to Linux was the earlier addition of an IDE interface, not licensing. BSD required a (more expensive and thus more rare) SCSI controller for much longer which limited the number of users (which then had some effect on developers).
Same for partitioning, to this day BSD still requires a primary partition, which means potentially more trouble for users to install.
"On a roof, such cells would require less than half the surface area to produce the same amount of power as today's standard solar panels, which have an efficiency of about 17%."
purely based on efficiency is dangerous. A lot of solar cells require a certain minimal light threshold before they start producing energy, and for reallife application, a lower threshold matters more than a few percent more of peak efficiency.
IOW efficiency is a function of among others light intensity
That's why I say libgcc, and not glibc. And libgcc IS in/usr/src/contrib/gcc, which is the non BSD licensed (read GNU tainted) directory of the BSD sources.
Check out this source directory so see exactly what is non BSD in the core BSD distribution of your choice.
First, BSD has some limitations (like carry notice and not misrepresent its origin).
Second, one must separate BSD licensed software from BSD based OSes. To my knowledge, all modern BSDs extensively use GNU software to create binaries, libgcc among others.
Third, one can link to GPL, LGPL code, as long as it is for own use (not redistributed).
Fourth, hardly any libraries come as GPL. Nearly all come with LGPL. Moreover, (L)GPL is a bit defanged if it comes with the distro. But static linking is still an issue.
When you transfer assets to your heirs you have to pay a fairly hefty tax, usually in the 10% magnitude. Same for other forms of property tax. And note that these taxes are on the value of the property, not on the revenue you generate with it.
It could be that equating copyright with property might benefit the IRS the most, and make that perpetual copyright might not be worthwhile because of tax pressure except for the 0.001% of all-time classics.
Yes. Note that the articles - mentions putting it on the barn, not an house. Barn is probably way larger in roof surface. - no analysis of cost/benefit as you say. - falls into the pseudo environmental category. If the production of something with a green principle is quite environmentally damaging, and the "green" benefit is low, the net result can still be _more_ polution. That's about the first thing they learn you in any engineering course about the environment, yet journalists seem to miss that en-masse.
The main "problem" then and now is that it is simply more expensive. So it only gets interesting if the oil is gone (or hard to exploit).
Of course it doesn't help for netto CO2 emissions. (or not much. It depends on conversion costs of both, but since C is further from C7Hx, I'd guess that converting coal is less efficient than crude oil)
It's also possible to convert it into hydrogen (water-shift reaction followed by an additional step to convert the CO (+H2O) into more H2), but I don't know if this is really worthwhile.
This is btw all standard Chemical Engineering introductionary stuff for decades
It can be worthwhile for performance reasons. They made a whole bunch of minor enhancements (like function inlining across units (D2005) and redoing all the assembler and memory manager (D2006). (and of course they merged some horrible Delphi.NET language abhorrations back into D2005/6 but you don't HAVE to use them).
If you do a lot of string processing, some more functions got standarised in strutils. Minor stuff, but hard to miss if you get used to it.
I was maintaining an realtime image recognition program, and it allowed me to squeeze one more camera using the same software per stock box that we deploy. (+/- 30% increase) However I still work with D7, I only compile releases with D2006. The new IDE is dog slow.
(btw portable versions of the C++ Turbovision is are available for GCC and portable Pascal ones (clones actually, the Pascal version was never PDed) via Free Pascal)
IMHO the QT designer stuff and IDE can't hold a candle to Delphi.
I ran the check tool from microsoft, and the machine passed all test except for the HD test.
It seems it requires all required space to be already _free_ on the machine (so that increases the requirement with the size of your current installation), and on the primary partition.
I move cursors and edit more than that I issue commands. Of course, with VI both also require extra keys.
More importantly, since you address typing, the shift-alt states are covered by a normal typing posture. The escape key is not. It is out there somewhere up left.
Agreed for clientside. The only true ones that I know are Borland's Kylix and Loki games and they failed, and the other commercial apps (e.g. Adobe Reader, Nvidia drivers ) are mostly vehicles for services or hardware, and their commercial angle (as in pay for software) is quite limited.
But the serverside is a different story. Still mostly services based, but way more money involved
There is some XCode integration for FPC btw. The last release was the initial XCode support, in the next release (within a month) it should work better.
Slashdot Mirror
Well, Delphi pays my bills, and occasionally I commercially sidestep with FPC.
I actually make money with GCC too. But only with a commercially supported one by some microprocessor vendor.
However Python? Please! Hobbyism.
However, somehow despite all parser theory and -engine progress, nobody still seems to be able to produce a C++ compiler with non cryptic errormessages
Good point. The release manifest indeed doesn't mention that Free Pascal also supports Mac like Object Pascal.
From what I recall as both a Linux and BSD hack at the time, the boost to Linux was the earlier addition of an IDE interface, not licensing. BSD required a (more expensive and thus more rare) SCSI controller for much longer which limited the number of users (which then had some effect on developers).
Same for partitioning, to this day BSD still requires a primary partition, which means potentially more trouble for users to install.
I'm not a solar panel expert, but the statement
"On a roof, such cells would require less than half the surface area to produce the same amount of power as today's standard solar panels, which have an efficiency of about 17%."
purely based on efficiency is dangerous. A lot of solar cells require a certain minimal light threshold before they start producing energy, and for reallife application, a lower threshold matters more than a few percent more of peak efficiency.
IOW efficiency is a function of among others light intensity
(obligatory 1984 reference)
Who is going to police it? The newly set up Ministry of Truth?
That's why I say libgcc, and not glibc. And libgcc IS in
Check out this source directory so see exactly what is non BSD in the core BSD distribution of your choice.
First, BSD has some limitations (like carry notice and not misrepresent its origin).
Second, one must separate BSD licensed software from BSD based OSes. To my knowledge, all modern BSDs extensively use GNU software to create binaries, libgcc among others.
Third, one can link to GPL, LGPL code, as long as it is for own use (not redistributed).
Fourth, hardly any libraries come as GPL. Nearly all come with LGPL. Moreover, (L)GPL is a bit defanged if it comes with the distro. But static linking is still an issue.
> You can do whatever you want with BSD code
You can not misrepresent its origin.
Note also that *BSD aren't 100% BSD licensed. Most notably, you will probably link to a LGPL licensed libgcc.
When you transfer assets to your heirs you have to pay a fairly hefty tax, usually in the 10% magnitude. Same for other forms of property tax. And note that these taxes are on the value of the property, not on the revenue you generate with it.
It could be that equating copyright with property might benefit the IRS the most, and make that perpetual copyright might not be worthwhile because of tax pressure except for the 0.001% of all-time classics.
But of course then I didn't care either....
Yes. Note that the articles
- mentions putting it on the barn, not an house. Barn is probably way larger in roof surface.
- no analysis of cost/benefit as you say.
- falls into the pseudo environmental category. If the production of something with a green principle is quite environmentally damaging, and the "green" benefit is low, the net result can still be _more_ polution. That's about the first thing they learn you in any engineering course about the environment, yet journalists seem to miss that en-masse.
Trying to follow that gives me a headache.
I thought you weren't allowed to use those anymore in the UK ? :-)
The main "problem" then and now is that it is simply more expensive. So it only gets interesting if the oil is gone (or hard to exploit).
Of course it doesn't help for netto CO2 emissions. (or not much. It depends on conversion costs of both, but since C is further from C7Hx, I'd guess that converting coal is less efficient than crude oil)
It's also possible to convert it into hydrogen (water-shift reaction followed by an additional step to convert the CO (+H2O) into more H2), but I don't know if this is really worthwhile.
This is btw all standard Chemical Engineering introductionary stuff for decades
Isn't the Fisher-Trops 65 years old already? Germans used it in WWII for aux fuel, and so did South Africa during the boycott (SASOL).
The Club of Rome also named this as possibility in 1980 (I never read the first report, only the revised one)
> It's possible that using part of OpenOffice.org in KOffice might allow closed-source derivatives of KOffice.
Excuse me sir, but that is nonsense.
It can be worthwhile for performance reasons. They made a whole bunch of minor enhancements (like function inlining across units (D2005) and redoing all the assembler and memory manager (D2006). (and of course they merged some horrible Delphi.NET language abhorrations back into D2005/6 but you don't HAVE to use them).
If you do a lot of string processing, some more functions got standarised in strutils. Minor stuff, but hard to miss if you get used to it.
I was maintaining an realtime image recognition program, and it allowed me to squeeze one more camera using the same software per stock box that we deploy. (+/- 30% increase) However I still work with D7, I only compile releases with D2006. The new IDE is dog slow.
(btw portable versions of the C++ Turbovision is are available for GCC and portable Pascal ones (clones actually, the Pascal version was never PDed) via Free Pascal)
IMHO the QT designer stuff and IDE can't hold a candle to Delphi.
I ran the check tool from microsoft, and the machine passed all test except for the HD test.
It seems it requires all required space to be already _free_ on the machine (so that increases the requirement with the size of your current installation), and on the primary partition.
Afaik Bill Joy himself stated this.
I move cursors and edit more than that I issue commands.
Of course, with VI both also require extra keys.
More importantly, since you address typing, the shift-alt states are covered by a normal typing posture. The escape key is not. It is out there somewhere up left.
Keyboards have cursor movement keys nowadays. No need anymore for an escape mode. (even though some people idealise that out of principle now)
Agreed for clientside. The only true ones that I know are Borland's Kylix and Loki games and they failed, and the other commercial apps (e.g. Adobe Reader, Nvidia drivers ) are mostly vehicles for services or hardware, and their commercial angle (as in pay for software) is quite limited.
But the serverside is a different story. Still mostly services based, but way more money involved
There is some XCode integration for FPC btw. The last release was the initial XCode support, in the next release (within a month) it should work better.