Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!

 



Forgot your password?
typodupeerror

Comment Go get a real education from a real college... (Score 1) 213 213

Go learn what you are actually doing rather than learning how to tick a box on a goddamn test. Learn what NP-hard and NP-complete mean so you don't go around trying to solve intractable problems. Software development is NOT a trade you can learn by going to some trade school and getting certificates. It is a profession and, some would say, a science. Go take a few classes in Computer Science and learn how to really do it.

Certifications are JUST a way for companies to make money. That is all. It is a way for them to extract money from the developers who learn the skills to deal with their software.

GC

Comment Re:He did this to GNUstep as well.... (Score 1) 551 551

You're correct. I don't have to and, in fact, will not comply with his request. The reason I consider it unfair is because it is an absurd thing to ask. GNUstep is already a fringe project. We need all of the users and developers we can get. Asking us to deliberately cut out an option which benefits our developers for political ends is not a sacrifice I'm willing to make. I would fork the project before doing any such thing.

GC

Comment Re:Try understanding Free Software goals first (Score 1) 551 551

I believe you're making a gross assumption by picking apart a minor error in my post. Most people use IP as shorthand. Normally I would have said copyright and patent law. I, in fact, have consider the problem very deeply. Many years ago I wrote one of the first petitions against software patents:

https://web.archive.org/web/20011214055725/http://www.petitiononline.com/pasp01/petition.html
http://www.justice.gov/atr/public/hearings/ip/221258.htm (mentioned here in a FTC report on the effect of patents)

PetitionOnline.com has since been closed down, so I used the wayback machine to find it. :) I wrote it along with RMS and a few others. So, I have more than a passing understanding on the subject.

The issue with the FSF is that it is a political organization and doesn't give a crap about the quality of the software they release. This is the main reason why they haven't been able to release a decent kernel in the last 20 years and the reason why so many FSF projects are either very fringe or are utter failures. Not considering the command line tools such as ls and sed etc, the only projects which are truly successful in the GNU-land are gcc, gdb, emacs and very few others. Other than that the FSF's work is largely shunned by the community at large... see GNUstep and many others.

What RMS doesn't want is a threat to what he sees as his privileged place in the community. Compilers are hard. When gcc was first written it was a monumental achievement.... but gcc was also crippled as described in my earlier post.

I am not ascribing any motivations or priorities on RMS that he hasn't expressed all by himself. I have been part of the FSF on the GNUstep project for almost 17 years. I have spoken to him on many occasions and met him a few times in person. While I don't pretend to know him very well, I do know enough regarding his motivations not to have to make assumptions regarding them. Richard has said in the past that the FSF is political organization, not a technical one. Thus the quality of software can be sacrificed in the name of freedom. While in my youth I may have agreed with this point, I find it harder and harder to do so as I get older. I see software as an essential part of life. It should be open and free, but there also needs to be a balance.

The fact of the matter is that RMS should spend his time fighting the real enemy and that is proprietary software and companies like Oracle that like to think that they have the right to own the world. What RMS should NOT be doing is fighting against a compiler which is under a recognized free software license (Modified BSD https://www.gnu.org/philosophy/license-list.html). I can't help but hear a note of jealousy in RMS' tone when he fights against tools which are already free for the sake of preserving the FSF's position. The FSF has, quite frankly, sewn the seeds of this for many years by taking the policy that it did with GCC and other projects and making them inferior in order to keep them free. This position BEGS another group to come in and do it right. Some people may argue that he has a problem with the license.... I suspect that even if it were LGPLv3+ or GPLv3+ he would still have an issue since it is NOT gcc. If you make a more useful tool... people will flock to it. This is a lesson that I hope the FSF learns well and takes away from this experience.

Please don't think for a moment that I don't understand the importance of software freedom. If I didn't I wouldn't have devoted much of my life to GNUstep and to the cause for Free Software. Before you start making value judgements about someone you should, quite honestly, do your research. I have fought for software freedom for most of my adult life and I challenge anyone else here to make a similar claim.

Sincerely, Gregory Casamento

Comment He did this to GNUstep as well.... (Score 2, Insightful) 551 551

RMS did the very same thing to GNUstep. GNUstep currently supports both GCC and LLVM/Clang. The project does this for good reason: because Objective-C is better supported in clang than it is in gcc. GCC doesn't even consider ObjC as a release critical compiler and LLVM/Clang looks on it as central. Additionally clang supports many modern features of ObjC that gcc lacks and shows no signs of ever attaining.

RMS specifically indicated that supporting LLVM/Clang by mentioning it on our wiki page (http://wiki.gnustep.org/index.php/ObjC2_FAQ) was harming the GNU project in an important place. Our response was swift and unanimous against remove it since all we are doing is providing user choice and, given that GCC is inferior to LLVM/Clang for ObjC, we MUST support LLVM/Clang. To date we have gotten no response from RMS.

I think it's grossly unfair of RMS to request this. By supporting Clang and LLVM and LLDB we are not impacting user freedom. All we are doing is offering users a choice which, last time I checked was completely okay. What we have here is a problem where RMS sees his role in the FLOSS community diminishing because someone has come up with a faster, more useful and better support compiler.

If anyone has damaged the FSF it is not the folks at Clang/LLVM it is RMS and the FSF itself. They have systematically impacted developer freedom by doing the following to GCC:

https://gcc.gnu.org/ml/gcc/2005-01/msg00008.html

"One of our main goals for GCC is to prevent any parts of it from being
used together with non-free software. Thus, we have deliberately
avoided many things that might possibly have the effect of
facilitating such usage, even if that consequence wasn't a certainty.

We're looking for new methods now to try to prevent this, and the outcome
of this search would be very important in our decision of what to do." -- RMS

This is terrible! Why would you do this?! RMS is trying to achieve through technical means what proprietary software companies try to do via copyright and IP law.

RMS is risking an all out rebellion of pretty much all of the FSF/GNU projects if he keeps this up. My advice to the FSF and to RMS is to allow developer freedom and stop viewing LLVM/Clang as a threat or a setback for it is neither.

GC

Comment The real question is... (Score 1) 626 626

Should governments be profiting and depending on revenue which requires citizens to break the law in order to maintain it. If so, is not the local government contributing to illegal activity and accidental property damage, injury, and deaths by not allowing driverless cars on the road since they would be inherently safer?

Food for thought.

Comment Re:2.4% duh (Score 1) 426 426

You know, I think it's those in the field who don't have CS degrees who always spout that they are not necessary out of some feel of inadequacy. It's like saying a you don't need to have an MD in order to operate on someone. You can, of course, but you're very likely to screw things up in ways that you're not immediately aware of. I can't tell you how many times I've had to clean up the messes of people who are NOT computer science graduates because they decided to solve a common problem poorly due to ignorance of the field.

Greg C.

Comment Meritocracy.... (Score 2) 263 263

The web is not a US colony it is a meritocracy. In this case the US has most of the services which are essential for NOW. But this may not always be the case. So, Sweden and other countries that are upset should start working on the technological things that might give them the edge in the future.

In this world it is now all about the information and the technology. There is no fair and balanced here. Whoever knows the most... wins.

GC

Comment Re:Hmmm (Score 1) 131 131

There's ARC in Objective-C for example, which is a huge difference. Not sure if that's implemented in GNUstep.

Yes, it's implemented. If you're using CLANG. GNUstep's libobjc2 runtime supports it fully. GCC, unfortunately, has a long way to go to catch up on some features since, for some reason, they don't consider ObjC to be release critical.

A programming language is low level when its programs require attention to the irrelevant.

Working...