Linked by Thom Holwerda on Mon 15th Oct 2007 21:06 UTC, submitted by Valour
OpenBSD "A few weeks ago, the OpenBSD Project announced that the Portable C Compiler had been added to the OpenBSD source tree. There has already been some explanation of why the traditional GNU Compiler Collection is troublesome and why a new compiler is needed, but there are still some details left uncovered. In this interview, Theo de Raadt and Otto Moerbeek of the OpenBSD Project offer more information about PCC and GCC and where they are headed within the project."
Thread beginning with comment 278451
To view parent comment, click here.
To read all comments associated with this story, please click here.
RE: sight...
by Janizary on Mon 15th Oct 2007 23:41 UTC in reply to "sight..."
Janizary
Member since:
2006-03-12

If the GCC was more community oriented, it wouldn't be dropping hardware support that the community wants, but older platforms are being dropped - why do you think OpenBSD has to use multiple GCCs?

Reply Parent Score: 3

RE[2]: sight...
by diegocg on Mon 15th Oct 2007 23:53 in reply to "RE: sight..."
diegocg Member since:
2005-07-08

GCC drops platforms because nobody maintains them. In other words, the community does not care so much about them as you say, because they'd support it if they really cared.

You have two choices: The dictadure of the minority or the dictadure of the majority. You're free to choose the dictadure of the minority, but don't expect the majority to care about you.

Reply Parent Score: 17

RE[3]: sight...
by edwdig on Tue 16th Oct 2007 01:52 in reply to "RE[2]: sight..."
edwdig Member since:
2005-08-22

GCC drops platforms because nobody maintains them. In other words, the community does not care so much about them as you say, because they'd support it if they really cared.

I don't think you can realistically say that. Compilers are one of the most complex types of software projects there are. Very, very, few people are capable of contributing to a modern compiler.

Also, ones of the major reasons for this PCC movement is that the GCC code is difficult to work with by design.

Things have to be REALLY bad to reach the point where people think it's easier to take a primative compiler an get it up to speed than it is to work with an existing one.

Reply Parent Score: 5

v RE[3]: sight...
by Duffman on Tue 16th Oct 2007 07:00 in reply to "RE[2]: sight..."
RE[3]: sight...
by Oliver on Tue 16th Oct 2007 16:33 in reply to "RE[2]: sight..."
Oliver Member since:
2006-07-15

>free to choose the dictadure of the minority

Or some people call this some respect for the minority to some extent. This big community you're speaking of, is out of laziness almost instantly ready to axe a project.

Reply Parent Score: 1

RE[2]: sight...
by butters on Tue 16th Oct 2007 07:44 in reply to "RE: sight..."
butters Member since:
2005-07-08

why do you think OpenBSD has to use multiple GCCs?


Because the OpenBSD project didn't have either the desire or the means to maintain GCC's niche architecture support? I suppose it must be a lack of desire, since a lack of means would be a problem if they intend to implement all of these architectures for an alternative compiler on their own.

GCC clearly isn't where it needs to be as the keystone of the free software toolchain. Hopefully some good ideas will come out of PCC. In the free software world, you must never be afraid of competition. However, for practical reasons I have a hard time envisioning PCC becoming a successful project in its own right. It seems the odds are stacked against them. It also seems that OpenBSD has more important challenges that they could be focusing on. I feel that PCC could become a distraction.

On the whole, I think that PCC is an interesting project whose progress should be watched closely. I'm glad that somebody is undertaking this challenge, as you have to be a glutton for punishment to reinvent the compiler suite. The legendary Fred Brooks remarked that his troubled project to develop an Algol compiler for OS/360 taught him that compilers generally take three times longer than expected to develop.

Reply Parent Score: 6

RE[3]: sight...
by Soulbender on Tue 16th Oct 2007 09:39 in reply to "RE[2]: sight..."
Soulbender Member since:
2005-08-18

However, for practical reasons I have a hard time envisioning PCC becoming a successful project in its own right


Well, it does primarily aim to be a C compiler, not an "everything-and-the-kitchen-sink" compiler that GCC is. That cuts down the scope quite a bit.
(Fortran? News flash, it's not 1975 anymore...)

It also seems that OpenBSD has more important challenges that they could be focusing on.


Like what? Being ignored by the GCC people? ;)
Obviously they've grown weary of battling GCC and took a good look at the alternatives. These decisions arent made overnight, although a lot of people here (most certainly almost all non-developers) seems to think so.

Either way, OpenBSD's goal for PCC is to have it build the base system, not all of ports, so it does not have to deal with all those "esoteric" GCC features. GCC will still be in ports and will most likely be required to build many ports for a long time to come.

Reply Parent Score: 2

RE[3]: sight...
by theine on Tue 16th Oct 2007 10:02 in reply to "RE[2]: sight..."
theine Member since:
2005-09-29

GCC clearly isn't where it needs to be as the keystone of the free software toolchain.

It seems everybody enjoys bitching about GCC, but I have yet to hear a compelling argument for why the above statement is true.

Reply Parent Score: 2

RE[3]: sight...
by Oliver on Tue 16th Oct 2007 16:44 in reply to "RE[2]: sight..."
Oliver Member since:
2006-07-15

>I suppose it must be a lack of desire, since a lack of means would be a problem if they intend to implement all of these architectures for an alternative compiler on their own.

It's always nice to read OSNews, where people are full of FUD because of massive lack of any knowledge.

1.) it's primary a NetBSD project
2.) OpenBSD will contribute to it (and some FreeBSD developers too)
3.) if you read the interview, it's somewhat an idea for future development

Last not least, they have a lack of developers. And don't forget they're eager to use such a compiler because it's easier to maintain in the long run.

Reply Parent Score: 1

RE[2]: sight...
by NxStY on Tue 16th Oct 2007 07:52 in reply to "RE: sight..."
NxStY Member since:
2005-11-12

GCC drops an architechture only if it's
1. unmaintained.
2. Has been marked as deprecated for two releases (two years).
3. Has issues that nobody is fixining due to the lack of a maintainer.

So if the community relly wants this arch, why doesn't anyone just send a mail to the gcc list during those two years and complain? Or preferably submit some patches? And if the arch has been dropped they can always fix it and resubmit it.

Reply Parent Score: 8