OpenBSD Journal
Contributed by mbalmer on from the openbsd-is-the-better-linux dept.
"Please don't let this reverse engineering idiocy hinder wireless driver adoption, we're already falling far behind OpenBSD who are very successfully reverse engineering lots of wireless chipsets."
Read the full story on KernelTrap.org
(Comments are closed)
Copyright © - Daniel Hartmeier. All rights reserved. Articles and comments are copyright their respective authors, submission implies license to publish on this web site. Contents of the archive prior to as well as images and HTML templates were copied from the fabulous original deadly.org with Jose's and Jim's kind permission. This journal runs as CGI with httpd(8) on OpenBSD, the source code is BSD licensed. undeadly \Un*dead"ly\, a. Not subject to death; immortal. [Obs.]
By Venture37 (217.22.88.123) venture73 # hotmail com on www.geeklan.co.uk
Comments
By SH (82.182.103.172) on
>
Erm, OpenBSD did not stop developing new drivers after 3.7 nor stop improving the existing ones.
Comments
By Venture37 (217.22.88.123) venture37 # hotmail com on www.geeklan.co.uk
I didn't say they did!
Comments
By Paladdin (213.97.233.52) on
>
> I didn't say they did!
>
I catched the point at first time. But, you know, some comments are understood depending on the mood you woke up ;)
Regards!
By Anonymous Coward (156.34.219.26) on
Comments
By Anonymous OpenBSD and Debian user (83.17.211.222) on
> 'falling behind' in free drivers, perhaps
While they are failing in wifi drivers space, they are
far better in other areas. Think filesystem, performace,
and ease of use. You can't have everything.
> they will be motivated to address the shortcoming and
> less inclined to accept blobs and NDAs.
Linux kernel developers doesn't accept blobs. Period.
Many users doesn't accept them either.
While NDAs are bad thing, I choose to run hardware with
driver that was written under NDA than to drop it.
Why? Because if developer doesn't accept NDA then there
will probably be no driver at all. Rejecting NDAs
doesn't magically make hardware vendor provide "clean"
documentation.
> Influencing, scaring, embarrassing, or shaming
> (I'll take whatever I can get) the Linux people
> to be more vigilant about the source of their
Utter nonsense. I care about freenes of my drivers.
And what?
NOTHING HAPPENS.
> drivers -- especially if they would extend the
> concept beyond wireless -- would another shiny
> feather in the cap of the OpenBSD developers,
> and perhaps slow the 'cave to hardware vendors' trend
> that seems to be infliltrating opensource these days.
I don't use non-free drivers, I don't run non-free software
either. I sold my geforce card just to have one that can
be driven with free drivers.
And guess what? NVIDIA haven't opened their drivers since
then. ATI? No. Atheros. No.
Comments
By Anonymous Coward (70.27.15.123) on
> far better in other areas. Think filesystem, performace,
> and ease of use. You can't have everything.
Linux is certainly ahead in performance, but "filesystem"? What are they ahead in there? Having more filesystems that all do the same thing with varying degrees of data loss and corruption? That's not really a good thing is it? And I think you meant difficulty of use, because linux is sure as hell not easier. Its more difficult than openbsd for everything.
> Linux kernel developers doesn't accept blobs. Period.
> Many users doesn't accept them either.
Right, that's why they have done nothing to stop all the blobs, and have in fact gone out of their way to make changes to the kernel to better accomodate nvidia's shit.
> While NDAs are bad thing, I choose to run hardware with
> driver that was written under NDA than to drop it.
> Why? Because if developer doesn't accept NDA then there
> will probably be no driver at all. Rejecting NDAs
> doesn't magically make hardware vendor provide "clean"
> documentation.
It would if everyone did it. You are trading immediate personal convenience for everyone's long term freedom.
Comments
By Anonymous Coward (83.17.211.222) on
> > far better in other areas. Think filesystem, performace,
> > and ease of use. You can't have everything.
> Linux is certainly ahead in performance, but "filesystem"? What are they ahead in there?
Do:
$ time tar xzf
$ time cp -R large-tree large-tree-2
Compare.
> And I think you meant difficulty of use, because linux is sure as hell
> not easier. Its more difficult than openbsd for everything.
I understand that because of this rant someone modded you up. Great.
Do update on openbsd system. How simple is that?
Debian-based systems can be updated with one command.
> > Linux kernel developers doesn't accept blobs. Period.
> > Many users doesn't accept them either.
> Right, that's why they have done nothing to stop all the blobs,
> and have in fact gone out of their way to make changes to the
> kernel to better accomodate nvidia's shit.
Just plain lies. Converting funtions to EXPORT_SYMBOL_GPL
was for what? Tell greg k-h that he helps nvidia!
> > While NDAs are bad thing, I choose to run hardware with
> > driver that was written under NDA than to drop it.
> > Why? Because if developer doesn't accept NDA then there
> > will probably be no driver at all. Rejecting NDAs
> > doesn't magically make hardware vendor provide "clean"
> > documentation.
> It would if everyone did it. You are trading immediate personal
> convenience for everyone's long term freedom.
More power for you!
The sad part is that posters like you who (I think) work hard
to make OpenBSD best system on this planet typically[*] go home,
and use Windows (or install it to your less-geeky persons)
because OpenBSD simply doesn't do the job.
[*] I have meet to many BSD-zealots behaving just like that.
Strangely they forgot about freedom when they
recomend windows to their friends.
Comments
By takahide (84.217.47.96) on
>
> Do:
>
> $ time tar xzf
> $ time cp -R large-tree large-tree-2
>
> Compare.
Read the entire message before replying. When talking about filesystems parent said: "Having more filesystems that all do the same thing with varying degrees of data loss and corruption?".
> Do update on openbsd system. How simple is that?
> Debian-based systems can be updated with one command.
I've had Debian systems go boo-boo after using apt-get and aptitude. That was quite some time ago though, but it was enough to encourage me to switch to a more stable system that in the long-term was easier to maintain and update.
> The sad part is that posters like you who (I think) work hard
> to make OpenBSD best system on this planet typically[*] go home,
> and use Windows (or install it to your less-geeky persons)
> because OpenBSD simply doesn't do the job.
Crybaby, no connection to reality. OpenBSD isn't the solution to everything, neither is Linux nor Windows.
Comments
By Anonymous Coward (83.17.211.222) on
> filesystems parent said: "Having more filesystems that all do
> the same thing with varying degrees of data loss and corruption?".
No, parent said:
"Linux is certainly ahead in performance, but "filesystem"?"
*and*:
"What are they ahead in there? Having more filesystems that all
do the same thing with varying degrees of data loss and corruption?
That's not really a good thing is it?"
Read on what I've been commenting on, please.
I've pointed out in what Linux is better on.
> I've had Debian systems go boo-boo after using apt-get and aptitude.
> That was quite some time ago though, but it was enough to encourage
> me to switch to a more stable system that in the long-term was easier to
> maintain and update.
Yes, you can do bad things with aptitude, like any tool used incorrectly.
I found that managing large number of computers (that really *are*
updated) is lot easier with automatic tools that Debian provide.
I also have found that many administrators never update BSD boxes
because, well "they are that secure, that it isn't needed".
When I told them that w^x wont help them with sql-injection attacks
just ignored the issue.
> > The sad part is that posters like you who (I think) work hard
> > to make OpenBSD best system on this planet typically[*] go home,
> > and use Windows (or install it to your less-geeky persons)
> > because OpenBSD simply doesn't do the job.
>
> Crybaby, no connection to reality.
Great to know. Maybe I was totaly drunk when I last used Blender
under Linux. Maybe it's now possible to do under OpenBSD too?
(With hardware acceleration.)
..or maybe I should simply use Windows because I want to do some
3d-graphics design?
Comments
By takahide (84.217.47.96) on
> > filesystems parent said: "Having more filesystems that all do
> > the same thing with varying degrees of data loss and corruption?".
>
> No, parent said:
>
> "Linux is certainly ahead in performance, but "filesystem"?"
>
> *and*:
>
> "What are they ahead in there? Having more filesystems that all
> do the same thing with varying degrees of data loss and corruption?
> That's not really a good thing is it?"
>
Yes parent said that, but with filesystems that are easily corrupted and suffer from data loss performance means nothing.
> Read on what I've been commenting on, please.
> I've pointed out in what Linux is better on.
Will do.
>
> > I've had Debian systems go boo-boo after using apt-get and aptitude.
> > That was quite some time ago though, but it was enough to encourage
> > me to switch to a more stable system that in the long-term was easier to
> > maintain and update.
>
> Yes, you can do bad things with aptitude, like any tool used incorrectly.
> I found that managing large number of computers (that really *are*
> updated) is lot easier with automatic tools that Debian provide.
Incorrectly? I've had apt-get and aptitude _overwrite_ settings in /etc without warning (no extra flags used). The X maintainer once decided to enable a feature that caused lots of people some serious grief. No warning during update, nothing on the website, no nothing.
> I also have found that many administrators never update BSD boxes
> because, well "they are that secure, that it isn't needed".
> When I told them that w^x wont help them with sql-injection attacks
> just ignored the issue.
I have the same experience with Linux, Solaris, OS X and Windows administrators.
> > Crybaby, no connection to reality.
>
> Great to know. Maybe I was totaly drunk when I last used Blender
> under Linux. Maybe it's now possible to do under OpenBSD too?
> (With hardware acceleration.)
> ..or maybe I should simply use Windows because I want to do some
> 3d-graphics design?
Now I definately know you have no connection to reality. I've read your other replies and you seem to have problems with the Linux community as well. In your case, yes you should use Windows. Preferrably you should also get another hobby.
What a waste of time reading your comments. Nowhere did you point out where Linux did better.
Comments
By Anonymous Coward (83.17.211.222) on
> and suffer from data loss performance means nothing.
And for sure you dealt with many corrupted ext2/3 filesystems?
> > Yes, you can do bad things with aptitude, like any tool used incorrectly.
> > I found that managing large number of computers (that really *are*
> > updated) is lot easier with automatic tools that Debian provide.
> Incorrectly? I've had apt-get and aptitude _overwrite_ settings in /etc
> without warning (no extra flags used). The X maintainer once decided
> to enable a feature that caused lots of people some serious grief.
> No warning during update, nothing on the website, no nothing.
And this of course happened during old-stable to stable upgrade?
Maybe you used -testing or -unstable?
> > I also have found that many administrators never update BSD boxes
> > because, well "they are that secure, that it isn't needed".
> > When I told them that w^x wont help them with sql-injection attacks
> > just ignored the issue.
> I have the same experience with Linux, Solaris, OS X and Windows
> administrators.
On the other hand majority administrators _I_ know run Linux only.
> Now I definately know you have no connection to reality. I've read your
> other replies and you seem to have problems with the Linux community
> as well. In your case, yes you should use Windows.
Windows? You must be kidding me. I used Linux/BSD exclusively since
6 years or more. Before that time I used proprietary UNIX and Windows.
I've never needed to use Windows since I've switched to Linux.
Partially because I've accepted *free* 3d drivers written under NDA.
You may as pure as RMS. I wish you that, but sadly I think
that this "has no connection with reality."
> Preferrably you should also get another hobby. What a waste of
> time reading your comments.
Sorry for wasting your time, then.
> Nowhere did you point out where Linux did better.
This was never my intention. I don't come to *BSD forums and cry that
Linux is better. All I wantend is to point out to that while NDAs are
evil (I haven't said otherwise) they sometimes allow _me_ to do job
on otherwise *free* system. You will probably rightly point that system
isn't free anymore. Yes, it isn't but it's a whole lot better than switching
to Windows.
Comments
By Anonymous Coward (83.17.211.222) on
> > administrators.
>
> On the other hand majority administrators _I_ know run Linux only.
...and I know that they update their machines.
Better to forget about this altogether.
By takahide (84.217.47.96) on
Yes.
> And this of course happened during old-stable to stable upgrade?
> Maybe you used -testing or -unstable?
Stable.
> On the other hand majority administrators _I_ know run Linux only.
OK, so out of the minority that run *BSD, only a few of them don't update their systems because one or more scripts have an SQL injection vulnerability? Here's a tip, don't talk so much or you might reveal what you really are.
> Windows? You must be kidding me. I used Linux/BSD exclusively since
> 6 years or more. Before that time I used proprietary UNIX and Windows.
I began using Linux exclusively back in '95 and later switched to OpenBSD as my OS of choice.
> You may as pure as RMS. I wish you that, but sadly I think
> that this "has no connection with reality."
I think you have me confused with someone else or you just assume too much. I do use OpenBSD at home and at work, but if you had read my reply more carefully you should know that I'm not under the delusion that any OS will let you easily solve all problems thrown at it.
> > Nowhere did you point out where Linux did better.
>
> This was never my intention.
Cut'n paste from your previous reply:
> Read on what I've been commenting on, please.
> I've pointed out in what Linux is better on.
Take it easy with the alcohol.
By Anonymous Coward (66.11.66.41) on
> because, well "they are that secure, that it isn't needed".
> When I told them that w^x wont help them with sql-injection attacks
> just ignored the issue.
You don't need to update OpenBSD to solve SQL injection attacks. OpenBSD does not contain anything that uses SQL in any way. If you install packages, running pkg_add -r is very easy. If you install 3rd party software from source, its not openbsd's job to make it easy for you to update that software.
By Anonymous Coward (128.171.90.200) on
> because, well "they are that secure, that it isn't needed".
> When I told them that w^x wont help them with sql-injection attacks
> just ignored the issue.
Updating OpenBSD would not fix the issue, as it is not an OpenBSD issue.
If you use user-level software that is susceptible to sql-injection then that is where the problem is, not the operating system.
By Anonymous Coward (66.11.66.41) on
>
> Do:
>
> $ time tar xzf
> $ time cp -R large-tree large-tree-2
Ok, here you go.
OpenBSD:
14.32s - 36.81s
Linux:
19.22s - 40.73s
> Compare.
Learn to read. Linux has half a dozen filesystems, none of which appear to actually be stable and reliable. The fact that some of those filesystems perform better than FFS for certain tasks doesn't make linux ahead, it just means if you really need performance for a certain corner case scenario, then openbsd is not for you. Nobody wants openbsd to have lots of filesystems that all do the same thing, only with different problems.
> Do update on openbsd system. How simple is that?
for r in `ls /path/to/openbsd/distfiles/*`; do tar zxpf $r; done
cp /path/to/kernel/bsd /
run mergeslave or mergemaster to update /etc
reboot
I'd have to say, pretty fucking simple.
> Debian-based systems can be updated with one command.
It sure can. And because it is just one command, it can trash your system in horrible ways without any warning. OpenBSD already has plenty of "just one command"s that can break your system. Try rm /usr/lib/* or dd if=/dev/random of=/dev/rwd0c if you would like to break your system.
> Just plain lies. Converting funtions to EXPORT_SYMBOL_GPL
> was for what? Tell greg k-h that he helps nvidia!
I didn't say they all help nvidia. But if you pay attention to commits, you'd notice that Linus does. Since he controls the whole thing, it matters more wether he is sucking nvidia's cock then if the developers do.
Comments
By Matthias Kilian (84.134.17.58) on
>
> for r in `ls /path/to/openbsd/distfiles/*`; do tar zxpf $r; done
Congratulations! If /path/to/openbsd/distfiles is a typical installation path of a typical OpenBSD disk (or some RELEASEDIR filled by applying release(8)), you've just overwritten /etc/passwd and /etc/master.passwd (and probably other important files).
> cp /path/to/kernel/bsd /
> run mergeslave or mergemaster to update /etc
Too late, and the wrong order (the kernel always comes first). See the fine upgrade FAQ. And the very fun part (updating ports) is missing, too. pkg_tools have seen massive improvements in the last months (thanks to espie@ for this). And best of all, it just works (with the exception of some strange non-error error messages).
FWIW, in case someone asks how an OpenBSD installation looks like, you can point him to a ttymation by telling him to run telnet dead-parrot.de ;-)
If I find the time, I'll add an update show (3.8 -> 3.9). And don't comment on the sudoers(5) configuration -- I'm aware that it's insane, and I hope to fix it next weekend.
Comments
By Anonymous Coward (66.11.66.41) on
> >
> > for r in `ls /path/to/openbsd/distfiles/*`; do tar zxpf $r; done
>
> Congratulations! If /path/to/openbsd/distfiles is a typical installation path of a typical OpenBSD disk (or some RELEASEDIR filled by applying release(8)), you've just overwritten /etc/passwd and /etc/master.passwd (and probably other important files).
Congratulations! I didn't say it was a path to a cd. Its the path to the tarballs you want to extract. Perhaps I should have been more clear, knowing in advance that some nitpicking twat would come along and bitch about it. But it wasn't a guide to upgrading, it was a quick example of how simple it is to upgrade. Yes, you have to be bright enough to only download the tarballs you want. You also have to be bright enough to not have the source tarball there, or any other random tarball that will overwrite stuff. You even have to be in / and not some random directory. Use of your brain is still required for this process. But then again, if you do mess up and clobber your /etc, its not terribly hard to grab the copies from /var/backups/, although that would also require use of that brain thingy again.
> > cp /path/to/kernel/bsd /
> > run mergeslave or mergemaster to update /etc
>
> Too late, and the wrong order (the kernel always comes first). See the fine upgrade FAQ. And the very fun part (updating ports) is missing, too. pkg_tools have seen massive improvements in the last months (thanks to espie@ for this). And best of all, it just works (with the exception of some strange non-error error messages).
It rarely matters if you do the kernel first or not. As someone who does this to several machines at the same time, I test it first on the test machine, and see if I need the extra reboot or not. So I just tossed the commands out there to show how simple it is. Feel free to play it safe and do the kernel first, and then reboot and do the rest. Feel free to copy the kernel safely even if you like, or if you don't keep a /bsd.rd laying around. That brain thing might come in handy here.
Comments
By Matthias Kilian (84.134.54.59) on
Comments
By Anonymous Coward (70.27.15.123) on
It is your duty to laugh derisively at idiots who blindly copy and paste stuff without turning their brains on. It is not your duty to try to help those people. You cannot help those who do not wish to learn.
By Martin Schröder (87.193.63.70) martin@oneiros.de on http://www.oneiros.de
This statement doesn't become true by just repeating it.
Please demonstrate that xfs and ext3 are unstable and unreliable.
Comments
By pkplex (60.234.135.124) on http://127.0.0.1
>
> This statement doesn't become true by just repeating it.
>
> Please demonstrate that xfs and ext3 are unstable and unreliable.
Well just recently I had problems with a fedora box, where for no good reason I was getting all sorts of fsck errors on a 60Gig /home ext3 partition after normal reboots... it happened serveral times, im trying centos now, seeing how that goes. Whats more they were errors for which I actually had to be at the console for, whereas bsd just fixes the stuff and continues booting. It certainly does not seem reliable to me.
Comments
By pkplex (60.234.135.124) on http://127.0.0.1
> >
> > This statement doesn't become true by just repeating it.
> >
> > Please demonstrate that xfs and ext3 are unstable and unreliable.
>
> Well just recently I had problems with a fedora box, where for no good reason I was getting all sorts of fsck errors on a 60Gig /home ext3 partition after normal reboots... it happened serveral times, im trying centos now, seeing how that goes. Whats more they were errors for which I actually had to be at the console for, whereas bsd just fixes the stuff and continues booting. It certainly does not seem reliable to me.
>
>
I forgot to mention I ran segates diagnostic tools, which check the drives for bad blocks and what not. The drive was fine.
By tedu (69.12.168.114) on
>
> This statement doesn't become true by just repeating it.
>
> Please demonstrate that xfs and ext3 are unstable and unreliable.
you mean like they still get patched regularly?
like http://www.kernel.org/pub/linux/kernel/v2.6/testing/ChangeLog-2.6.17-rc4
has
[XFS] Fix a possible metadata buffer (AGFL) refcount leak when fixing an
AG freelist.
and
[XFS] Fix a project quota space accounting leak on rename.
and
[XFS] Fix a possible forced shutdown due to mishandling write barriers
with remount,ro.
or like http://www.kernel.org/pub/linux/kernel/v2.6/testing/ChangeLog-2.6.17-rc6
has
[PATCH] ext3 resize: fix double unlock_super()
?
[look kids, i can use the internet!]
Comments
By Martin Schröder (87.193.39.172) martin@oneiros.de on http://www.oneiros.de
In contrast to the ffs code in OpenBSD, which is error-free since 2000?
Get real. XFS has been used for years with multi-TB-filesystems in commercial environments. Of course there are still bugs in it, but most likely fewer then in OpenBSD-ffs, because XFS has been used more and longer.
> [look kids, i can use the internet!]
Oh, really? Then read this: http://en.wikipedia.org/wiki/XFS and this: http://en.wikipedia.org/wiki/Comparison_of_file_systems
Comments
By Anonymous Coward (70.27.15.123) on
>
> In contrast to the ffs code in OpenBSD, which is error-free since 2000?
How many serious data corruption and/or data loss bugs has openbsd's FFS had in that same time? Of those, how many only existed for a period of a few hours to a few days, and only in -current?
> Get real. XFS has been used for years with multi-TB-filesystems in commercial environments. Of course there are still bugs in it, but most likely fewer then in OpenBSD-ffs, because XFS has been used more and longer.
That's quite an interesting theory. But FFS goes back quite a ways too you know. Simply being old doesn't make something good. You can tell because XFS still contains major data corruption issues (at least under linux).
> Oh, really? Then read this: http://en.wikipedia.org/wiki/XFS and this: http://en.wikipedia.org/wiki/Comparison_of_file_systems
Very nice. Which part of those have anything to do with all of linux's filesystem options being broken to various degrees?
By takahide (84.217.47.96) on
>
> In contrast to the ffs code in OpenBSD, which is error-free since 2000?
>
> Get real. XFS has been used for years with multi-TB-filesystems in commercial environments. Of course there are still bugs in it, but most likely fewer then in OpenBSD-ffs, because XFS has been used more and longer.
>
> > [look kids, i can use the internet!]
>
> Oh, really? Then read this: http://en.wikipedia.org/wiki/XFS and this: http://en.wikipedia.org/wiki/Comparison_of_file_systems
XFS is a bad example on a filesystem that is reliable and fast. It's extremely sensitive to bad RAM and it doesn't handle power-loss or hard-resets very well (user data not fsync'd in the last 5-10 seconds are gone). IIRC XFS has no badblock handling and relies on the firmware to fix a failed write or read. This does pose a problem since many(?) drives only check for failed writes.
XFS was made for performance, not data integrity.
By Anonymous Coward (67.64.89.177) on
By Anonymous Coward (84.134.34.59) on
So if you think open source is a good thing, better don't count on Linux. Even if the Linux kernel developers start to fight BLOBs in a way similar to OpenBSD, the big distributions won't be stopped from doing the wrong thing.
Comments
By Anonymous OpenBSD and Debian user (83.17.211.222) on
> BLOBs that will be included in Ubuntu.
Sad, if it's true. Could you provide the link?
> Similarly, Novell/SuSE also plans to simplify development
> and inclusion of BLOBs.
Please don't spread FUD. See: http://lwn.net/Articles/184118/
While it might be used to simplify keeping propriteary drivers
*up to date* it might be used for free (linux out-of-tree)
drivers too.
> So if you think open source is a good thing, better don't
> count on Linux. Even if the Linux kernel developers start
> to fight BLOBs in a way similar to OpenBSD, the big
> distributions won't be stopped from doing the wrong thing.
Because it's about *freedom*. Isn't that the same freedom that
you, *BSD-only people like to bring when GPL vs. BSD battle begin?
Comments
By SH (82.182.103.172) on
> > BLOBs that will be included in Ubuntu.
>
> Sad, if it's true. Could you provide the link?
>
Here is one, and some googling is sure to bring more links:
http://undeadly.org/cgi?action=article&sid=20060508055853&pid=14
Comments
By Anonymous OpenBSD and Debian user (83.17.211.222) on
>
> http://undeadly.org/cgi?action=article&sid=20060508055853&pid=14
This is really sad... Mark was Debian person, I wouldn't expect
this from him.
On the other hand Ubuntu/SUSE is targeting in very differeng group
than OpenBSD (and less Debian). While I love OpenBSD's consistency,
using it on desktop isn't nice experience. (I mean OpenOffice.org,
why I need that bloody emulation? Yes, different target audience.)
If you (or we, if you count me :) want to make OpenBSD
valiable alternative to desktop-linux there is really lot to do.
DesktopBSD/PC-BSD are nice projects but as you know for sure
users' freedom isn't their priority. (ath_hal comes to mind.)
Comments
By Anonymous Coward (65.94.99.88) on
> than OpenBSD (and less Debian). While I love OpenBSD's consistency,
> using it on desktop isn't nice experience. (I mean OpenOffice.org,
> why I need that bloody emulation? Yes, different target audience.)
You stupid or something? The bloody emulation is needed because OpenOffice developers don't want a portable office suite, they want a Solaris/Linux/Windows office suite. This isn't an OpenBSD problem, it's an OpenOffice one.
Feel free to become an OpenOffice developer and keep it working natively on OpenBSD if you want it to run native.
Comments
By Anonymous Coward (83.17.211.222) on
> OpenOffice developers don't want a portable office suite, they
> want a Solaris/Linux/Windows office suite. This isn't an OpenBSD
> problem, it's an OpenOffice one.
It's not OpenOffice devs that show up on different forums and say
that OpenBSD is viable Linux alternative. OO.org devs probably
doesn't run OpenBSD so how they can fix OpenBSD-related problems?
Isn't OpenOffice working on FreeBSD natively? It is.
So why it can't on OpenBSD? Userspace differences between FreeBSD
and OpenBSD are relatively minor, I think.
FreeBSD does have people working on it to work natively.
(I heard that OpenBSD does too have some people working on it.)
OpenOffice haven't started to magically work on fbsd, it wouldn't
start on obsd either.
Enlight me -- isn't lack of people interested main problem?
> Feel free to become an OpenOffice developer and keep it working
> natively on OpenBSD if you want it to run native.
...or OpenBSD ports' tree maintaiter. Porting apps is their task,
isn't it?
Comments
By Anonymous Coward (84.135.77.54) on
> isn't it?
No, it isn't.
Comments
By Anonymous Coward (83.17.211.222) on
> > isn't it?
>
> No, it isn't.
Thanks for clarification.
So how would you call providing patches for various
apps to compile and work properly on OpenBSD system?
Maybe adapting? So, why not adapt OpenOffice.org?
Comments
By Anonymous Coward (216.209.146.192) on
> apps to compile and work properly on OpenBSD system?
>
> Maybe adapting? So, why not adapt OpenOffice.org?
>
So, where's your patch? You want it? You do it.
Comments
By Anonymous Coward (83.17.211.222) on
> So, where's your patch? You want it? You do it.
So there is no problem, because I'm unable to fix this myself?
Exactly the same response like in linux community...
Comments
By Chris (24.76.100.162) on
>
> > So, where's your patch? You want it? You do it.
>
> So there is no problem, because I'm unable to fix this myself?
>
> Exactly the same response like in linux community...
>
>
It seems this needs to be pointed out time and time again.
There *is* a problem -- OpenOffice doesn't work on OpeBSD (or whatever, I don't really care about the specifics). But it's *your* problem. If you're lucky, someone will have the same problem and fix it, and you get to benefit for free.
But if you can't or don't want to fix the problem, you have no justifications for bitching and whining on lists or message boards because someone won't do free work that you want done, when you want it done.
These people are not your employees. They don't owe you anything.
Comments
By Anonymous Coward (83.17.211.222) on
> (or whatever, I don't really care about the specifics). But it's *your*
> problem. If you're lucky, someone will have the same problem
> and fix it, and you get to benefit for free.
>
> But if you can't or don't want to fix the problem, you have no justifications
> for bitching and whining on lists or message boards because someone
> won't do free work that you want done, when you want it done.
>
> These people are not your employees. They don't owe you anything.
Very well said. I agree completly. I'm sorry for sugesting otherwise.
(I must misundersood you before.)
By Anonymous Coward (69.70.207.240) on
> > > isn't it?
> >
> > No, it isn't.
>
> Thanks for clarification.
>
> So how would you call providing patches for various
> apps to compile and work properly on OpenBSD system?
>
> Maybe adapting? So, why not adapt OpenOffice.org?
>
http://www.undeadly.org/cgi?action=article&sid=20060508184046
By couderc (212.234.204.97) on
> > OpenOffice developers don't want a portable office suite, they
> > want a Solaris/Linux/Windows office suite. This isn't an OpenBSD
> > problem, it's an OpenOffice one.
>
> It's not OpenOffice devs that show up on different forums and say
> that OpenBSD is viable Linux alternative. OO.org devs probably
> doesn't run OpenBSD so how they can fix OpenBSD-related problems?
There is a thing called "writing portable code".
There are also things called standards like ISO C or POSIX.
Maybe we should make OO devs aware of this ?
> Isn't OpenOffice working on FreeBSD natively? It is.
> So why it can't on OpenBSD? Userspace differences between FreeBSD
> and OpenBSD are relatively minor, I think.
FreeBSD ? You mean the Linux like distribution with a BSD license ?
> FreeBSD does have people working on it to work natively.
> (I heard that OpenBSD does too have some people working on it.)
>
> OpenOffice haven't started to magically work on fbsd, it wouldn't
> start on obsd either.
That's the point, it should have.
> > Feel free to become an OpenOffice developer and keep it working
> > natively on OpenBSD if you want it to run native.
>
> ...or OpenBSD ports' tree maintaiter. Porting apps is their task,
> isn't it?
Not everybody would have fun in porting such bullshit.
</rant>
By Anonymous Coward (24.81.14.122) on
> > OpenOffice developers don't want a portable office suite, they
> > want a Solaris/Linux/Windows office suite. This isn't an OpenBSD
> > problem, it's an OpenOffice one.
>
> It's not OpenOffice devs that show up on different forums and say
> that OpenBSD is viable Linux alternative. OO.org devs probably
> doesn't run OpenBSD so how they can fix OpenBSD-related problems?
>
> Isn't OpenOffice working on FreeBSD natively? It is.
> So why it can't on OpenBSD? Userspace differences between FreeBSD
> and OpenBSD are relatively minor, I think.
>
> FreeBSD does have people working on it to work natively.
> (I heard that OpenBSD does too have some people working on it.)
>
I tried to make a port about a year ago. OOO is huge. version 1.x is over 200 megs of compressed source code. They include tons possibly modified libraries. It uses a distributed build system called dmake which segfaults a lot on OpenBSD.
> OpenOffice haven't started to magically work on fbsd, it wouldn't
> start on obsd either.
>
> Enlight me -- isn't lack of people interested main problem?
>
I patched a dozen things in it and I was still nowhere near getting it working. OOO is a bloated piece of shit that is probably not any better than MS word code quality wise.
> > Feel free to become an OpenOffice developer and keep it working
> > natively on OpenBSD if you want it to run native.
>
> ...or OpenBSD ports' tree maintaiter. Porting apps is their task,
> isn't it?
>
Ports are maintained by those who care about having an easy way to deploy certain tools. If you're going to use OpenBSD as a platform for an app with many dependant packages, like a complex web site, be prepared to build and update ports. At times I am tempted to use Linux or FreeBSD because there are more users and someone else is already maintaining packages for what I use, but I don't because I like the simplicity of OpenBSD.
By Matthias Kilian (84.134.34.59) on
> using it on desktop isn't nice experience. (I mean OpenOffice.org,
> why I need that bloody emulation? Yes, different target audience.)
Who needs OpenOffice? Really?
I mean, most people I know use OpenOffice when some retard sent them an email containing just a MS Office document. So do I. Just to be able to read that shmuck.
But using it for writing documents from scratch? Sorry, there're much better tools like tex, lout or even roff. Or txt2tags, if you're really lazy (like me).
Don't let the world drown in eye-candy. Content is more important than layout; form follows function.
> If you (or we, if you count me :) want to make OpenBSD
> valiable alternative to desktop-linux there is really lot to do.
"Desktop" isn't the primary goal of OpenBSD; "Desktop" even isn't well-defined. There're people using a VT as their desktop. (Not me, since I need some colourful browser for my famous online-game ;-))
Comments
By Anonymous Coward (83.17.211.222) on
Rather large group of people, I think.
There is too many .docs, .ppts and .xls floating around
to ignore them.
And while I use gnuplot for charts less advanced users
like OO.org's spreadsheed's functionality.
> I mean, most people I know use OpenOffice when some retard sent
> them an email containing just a MS Office document. So do I.
> Just to be able to read that shmuck.
>
> But using it for writing documents from scratch? Sorry,
> there're much better tools like tex, lout or even roff.
> Or txt2tags, if you're really lazy (like me).
I use plain tex/latex personally. I'm just too stupid to use
OO.org -- I can't even manage to write nested enumerated list
properly...
> Don't let the world drown in eye-candy. Content is more
> important than layout; form follows function.
...but there are lots of people who can't manage to use
tex/latex. LyX is nice but in "Office"-type enviroments you
can't expect people will start using it. (I tried.)
There is no program for everyone.
> "Desktop" isn't the primary goal of OpenBSD; "Desktop"
> even isn't well-defined. There're people using a VT as
> their desktop. (Not me, since I need some colourful
> browser for my famous online-game ;-))
Very true. Even sadly true, because if anyone wants/expects
people switch to OpenBSD because it's better, they must
know what people expect from that mythical desktop-thing.
And currently it's not what system is providing.
Comments
By Menace 3 Society (71.126.109.162) meance3society@gmail.com on
>
> > "Desktop" isn't the primary goal of OpenBSD; "Desktop"
> > even isn't well-defined. There're people using a VT as
> > their desktop. (Not me, since I need some colourful
> > browser for my famous online-game ;-))
>
> Very true. Even sadly true, because if anyone wants/expects
> people switch to OpenBSD because it's better, they must
> know what people expect from that mythical desktop-thing.
> And currently it's not what system is providing.
Here's the thing. There's this one free operating system, that has chosen to define its success based on the number of users it has (actually, I shouldn't limit it to that operating system--there are a large number of projects that care only about tracking the number of unique users). Therefore, making a few enough changes to a program to get an extra two or three hundred users is time well spent, in their opinion.
Contrast this with OpenBSD, whose primary purpose is to be a stable, secure, and correct operating system--I'm sure the developers would continue working on it even if they were the only ones using it. While more users is a sign of health and moral support for the project, it's not the metric by which they gauge their success. So if they were given the choice to (say) improve SMP support or IPSec performance, vs updating a port of big, bloated, buggy, and now subject-to-malware application, which do you think the devs will do?
Now, if by "Desktop", you mean "operating system that can be run by a single user for everyday tasks", then yes, OpenBSD is a fantastic desktop operating system with regular updates, freely available source, and a wide-variety of third-party packages for quick and painless installation. However, if by "Desktop" you mean "a clone that behaves exactly like Windows to such an extent that it IS windows, except that technically it isn't", then OpenBSD makes a terrible desktop. But for all the time that Linux developers spend moaning about how bad everything is about windows, they certainly put a lot of programming time into making sure that everything behaves so much like Windows as to be indistinguishable from Windows.
At least the OpenBSD developers know what system they're developing for.
Comments
By couderc (212.234.204.97) on
OpenBSD is my desktop OS since more than 7 years now and of course it fits my needs.
Now this is a known fact that many people keep windows for games. This is the same reason why Linux users are also keeping those binary BLOBs for their graphics cards.
It will become worst with the new hype of 3D accelerated desktop features (see useless cpu-sucker eye-candy app).
By Anonymous Coward (64.81.82.25) on
> Very true. Even sadly true, because if anyone wants/expects
> people switch to OpenBSD because it's better, they must
> know what people expect from that mythical desktop-thing.
> And currently it's not what system is providing.
>
Why do even seemingly smart people always assume that free, open source projects wants to cater to the masses and attract more people?
By Anonymous Coward (62.252.32.11) on
This may come as a huge surprise to you, but presentation is as important as content. The old "it's not what you say, it's how you say it" applies to text just as much as it does to speech. Sure, things like tex are powerful tools, not disputing that, I even like tex. But when I want to send off my CV and a covering letter, I'll fire up OO Writer, type and print.
Plenty of people need OOo, don't confuse your deluded version of reality with reality, please.
Comments
By couderc (212.234.204.97) on
>
> Plenty of people need OOo, don't confuse your deluded version of reality with reality, please.
Interesting, when i want to write a CV i just need an editor. I was used to save it in RTF format (with Ted editor) as it was the only standard format available.
But one thing is sure, few people need a full bloated editor (read OO) to just write a CV.
I hope that a clean and light editor will be available for ODF files soon.
By Anonymous Coward (64.254.225.66) on
The frequent occasion to export your CV, thus justifying the existance of patently decrepit software, dictates your entire contribution [SIC] to this thread.
By Leen Besselink (194.213.15.120) on
> >
> > http://undeadly.org/cgi?action=article&sid=20060508055853&pid=14
>
> This is really sad... Mark was Debian person, I wouldn't expect
> this from him.
>
Yes, very sad.
> On the other hand Ubuntu/SUSE is targeting in very differeng group
> than OpenBSD (and less Debian). While I love OpenBSD's consistency,
> using it on desktop isn't nice experience. (I mean OpenOffice.org,
> why I need that bloody emulation? Yes, different target audience.)
>
> If you (or we, if you count me :) want to make OpenBSD
> valiable alternative to desktop-linux there is really lot to do.
>
> DesktopBSD/PC-BSD are nice projects but as you know for sure
> users' freedom isn't their priority. (ath_hal comes to mind.)
They even supply certification for hardware with blobs, if the vender pays for the extra support load:
http://www.ubuntu.com/partners/certification/hardware
I guess in that case it's not as bad ? As the vendor is paying extra for the problem they create.
By Matthias Kilian (84.134.34.59) on
> > and inclusion of BLOBs.
>
> Please don't spread FUD. See: http://lwn.net/Articles/184118/
> While it might be used to simplify keeping propriteary drivers
> *up to date* it might be used for free (linux out-of-tree)
> drivers too.
Sorry, it wasn't meant as FUD. I read about it on heise.de (german) some days ago, and they sometimes "shorten" the facts. However, they also wrote that Novell encourages the vendors to provide open source drivers. Only I doubt that the vendors will care.
> > So if you think open source is a good thing, better don't
> > count on Linux. Even if the Linux kernel developers start
> > to fight BLOBs in a way similar to OpenBSD, the big
> > distributions won't be stopped from doing the wrong thing.
>
> Because it's about *freedom*. Isn't that the same freedom that
> you, *BSD-only people like to bring when GPL vs. BSD battle begin?
You mean, take it and do whatever you want to do with it? Sure, this is the never-ending license war, but I'm really not interested in this one.
IMHO, the problem isn't licensing here, but it's vendors vs. distributors. If a vendor doesn't provide documentation or open source drivers, the distributor has several options: a) don't support the vendor; b) try to kick the vendor in the ass; c) reverse engineer; d) use vendor's BLOB.
Unfortunately, many users (and some distributors) happily accept d) for convenience, and they aren't aware that this leads vendors to think BLOBs are fine.
The sad thing about all this is that, according to all those mails from Theo on misc@ (and according to my own experience several years ago, when I tried to get documentation for a digital camera), in many cases there isn't just a single one good reason for a vendor not to publish documentation. No IP or patent issues. Just completely retarded "policy" dictated from the legals departments.
I've tried to get documentation once, and it was just a long, frustrating process ending with me just giving up on it. To be honest, if I had to do what Theo and the other people @openbsd.org are doing, I'd going mad within a few months. [You could argue that this already happened to *@openbsd.org ;-)]
But back to topic: people just want their hardware to "work". Fine. They accept BLOBs. Bad, but probably they just don't understand the consequences. AFAIK, RMS was one of the first who *did* understand them, long long ago. And people like Marc Shuttleworth should understand them, too, especially if he has a debian background.
'nough said.
ps: sorry for not filling in my name in my previous post.
pps: no, I don't bash on linux regurlarly. Only if I'm legally drunk.
Comments
By Anonymous Coward (83.17.211.222) on
> distributors. If a vendor doesn't provide documentation or
> open source drivers, the distributor has several options:
> a) don't support the vendor; b) try to kick the vendor in
> the ass; c) reverse engineer; d) use vendor's BLOB.
There is big difference between linux and bsd land hidden in
this statement -- BSD distributors are creators, Linux
_distributors_ aren't.
Linux distributors don't create content (oh, they add own
set of icons sometimes ;) It's not their task to write drivers,
I think.
Distributors may choose a), b) or d).
Depending on distribution's target audience they choose a) or d)
(But see below comment on Marks decision)
> Unfortunately, many users (and some distributors) happily
> accept d) for convenience, and they aren't aware that this
> leads vendors to think BLOBs are fine.
In my opinion not accepting blobs results in just no working
driver. I'm not sure that vendors care what users think at all,
Nvidia shown that they don't. See madwifi -- it wasn't accepted
be kernel maintainters -- nothing happened since then.
I really begin to think that vendors just don't care.
> But back to topic: people just want their hardware to "work".
> Fine. They accept BLOBs. Bad, but probably they just don't
> understand the consequences. AFAIK, RMS was one of the first
> who *did* understand them, long long ago. And people like Marc
> Shuttleworth should understand them, too, especially if he has
> a debian background.
Maybe Mark think that ignoring these drivers has no effect?
While I don't like it, it is what I really see happening.
...and sadly, he might be right. :/
> 'nough said.
>
> ps: sorry for not filling in my name in my previous post.
> pps: no, I don't bash on linux regurlarly. Only if I'm legally drunk.
I'm sorry for not filling my name in any post.
I'll probably stop commenting now, because my opinions seem
disliked here. Just like my opinions criticizing linux on
linux forums, oh, well...
By Wim (194.78.167.231) wim@kd85.com on https://kd85.com/notforsale.html
> > BLOBs that will be included in Ubuntu.
>
> Sad, if it's true. Could you provide the link?
I was sitting in the audience next to the guy who for some reason was compeled to ask the question.
I'm not sure if it's a written down rule for Ubuntu in their policy, but the explanation we got was "we'll bend over if needed to give our users the best experience. We don't care if they are binary blobs are not".
Basicly, it's not their fight. Their fight is market share. Not Opensource.
http://undeadly.org/cgi?action=article&sid=20060508055853&pid=14
By Anonymous Coward (24.34.57.27) on
By cnst (217.12.147.5) on
Comments
By Anonymous Coward (217.12.147.5) on
I've posted it on slashdot, but they've rejected it. :(
Comments
By Anonymous Coward (156.34.219.26) on
>
> I've posted it on slashdot, but they've rejected it. :(
Well, it is on Slashdot now. The story intro is (here is a shocker) inaccurate. You would almost think the story submitter didn't read the article, and the editors didn't bother to check it out. Not to worry ... the article was up for much of the day before any Slashdot readers seem to have noticed.
Comments
By Anonymous Coward (217.12.147.5) on
> >
> > I've posted it on slashdot, but they've rejected it. :(
>
> Well, it is on Slashdot now. The story intro is (here is a shocker) inaccurate. You would almost think the story submitter didn't read the article, and the editors didn't bother to check it out. Not to worry ... the article was up for much of the day before any Slashdot readers seem to have noticed.
Here is the link, btw:
http://hardware.slashdot.org/article.pl?sid=06/06/12/1837238
By Marc Balmer (213.189.137.178) on
Donate now
By Dave (203.45.13.17) on
It might even improve the quality of the kernal :-)
Comments
By Charles (216.229.170.65) on
>
> It might even improve the quality of the kernal :-)
>
>
Believe it or not, there are lots of people who have read and understand both licenses and *PREFER* to use the GPL.
The entire GPL concept can be summarized as: "I'll share if you will. Don't want to share? Write your own damn code, then."
I only can see three viewpoints that advocate using BSD:
1. Corporations -- "Please code stuff for us so we don't have to pay you." To whomever keeps posting the rants "these people aren't your employees and don't owe you anything" many big companies DO see you as unpaid employees that produce product and get no benefits in return. The perfect corporate slave.
2. Honest developers who ONLY care about quality code -- They write it for themselves, not for everyone else. If people find it useful, great, but that isn't the purpose. Often they aren't interested in others sharing code back because your average GPL patch-submitter writes shit quality code and it'd take more time to clean up that it is worth. I believe I've seen several comments over the years from OBSD devs that put them in this category.
3. Honest developers who hope others will pick up their code just so the products coming out of those others (corporations) will suck less. For example, Microsoft's use of BSD's TCP/IP code. Can you imagine if those guys wrote their own networking stack from scratch? Ugh.
Which viewpoint am I missing?
I'm constantly amazed at the number of people who bitch when some company takes a bunch of BSD code, makes a successful product and all they see is a credit in 4-point font in a manual somewhere or maybe some free pizza. No contribution? The big corporation won't behave in a civilized manner if they don't absolutely have to or aren't publically shamed? And you're surprised because...?
Comments
By tedu (69.12.168.114) on
> Which viewpoint am I missing?
the "you paid for the modifications to my code, so you are entitled to sell your value-add for whatever you get people to pay for it".
>
> I'm constantly amazed at the number of people who bitch when some company takes a bunch of BSD code, makes a successful product and all they see is a credit in 4-point font in a manual somewhere or maybe some free pizza. No contribution? The big corporation won't behave in a civilized manner if they don't absolutely have to or aren't publically shamed? And you're surprised because...?
i don't know of anyone writing bsd code that's been surprised that soembody used it.
By Anonymous Coward (70.27.15.123) on
> >
> > It might even improve the quality of the kernal :-)
> >
> >
>
> Believe it or not, there are lots of people who have read and understand both licenses and *PREFER* to use the GPL.
How is that relevant? Those people can still use the BSD drivers, because BSD code is free. Free for everyone. Not just free for other GNU/communists. The linux kernel can include BSD drivers no problem.
> The entire GPL concept can be summarized as: "I'll share if you will.
Don't want to share? Write your own damn code, then."
No, it can be summarized as "I'll share if you agree to this contract, which stipulates that all your code based on my code must be licensed under this exact restrictive contract". People releasing their code under free licenses like BSD/MIT/ISC are sharing. But people releasing code under the GPL still don't share with them, their code is only shared to other GPL users.
> Which viewpoint am I missing?
4. This code has no monetary value. Who the hell cares if someone uses it in their closed source project, it will have no impact on me in any way.
That is my view, and that is why I find the idiots who release every trivial bit of code under the GPL so annoying. You are not helping people, you are not creating free code. You are creating obnoxiously licensed code that is no good to many people, all because you are a paranoid dumbass who thinks microsoft is going to create a closed source fork of your awesome text editor.
> I'm constantly amazed at the number of people who bitch when some company takes a bunch of BSD code, makes a successful product and all they see is a credit in 4-point font in a manual somewhere or maybe some free pizza. No contribution? The big corporation won't behave in a civilized manner if they don't absolutely have to or aren't publically shamed? And you're surprised because...?
I am amazed at the number of people who bitch that BSD people don't like the GPL. You can't make someone stupider by beating them over the head figuratively, so quit trying. We're not going to turn into retards, you can stop now.
By Anonymous Coward (68.148.253.101) on
The reality is what happens "you must entitle to all of my work derived from your code, when I publicize or sell it. (If I don't publicize it, my derivative work is unknown and, thus, unimportant to you.) Don't agree? Sign your EULA contract and we'll have a deal."
> Which viewpoint am I missing?
Avoiding the caricatures (unpaid employee, using your source against you, BSD source completely disappears in closed derivatives, etc...) because the big corporations, be it IBM or MPAA or RIAA or game companies, will exploit Free/Open Source anyway due to the nature of Free and Open; therefore, freedom and technical merits are more worthy goals.
Respecting others' derivative works as their own rather than hijacking their works under an inalienable pseudo freedom that all code must be open. People's freedom override the code being free.
> I'm constantly amazed at the number of people who bitch when some company takes a bunch of BSD code... No contribution?
What does monetary contribution/donation have anything to do with the BSD license or GPLegalese? When a Free project is running out of money, how does code contribution feed the developers? What's your point for switching topic from code contribution to financial contribution? Since when did OpenBSD developers ask big corporations for code donations?
By ciphernaut (58.6.115.82) on
By Anonymous Coward (12.216.179.46) on
What? Linux was ahead of OpenBSD somehow?
Last time I checked, it was still a pile of crap. The reasons are self-apparent.
>FreeBSD ? You mean the Linux like distribution with a BSD license ?
Amen to that.
Comments
By Anonymous Coward (70.28.107.35) on
By Daniel Cedilotte (209.5.113.34) daniel@cedilotte.com on
> What? Linux was ahead of OpenBSD somehow?
> Last time I checked, it was still a pile of crap. The reasons are self-apparent.
>
> >FreeBSD ? You mean the Linux like distribution with a BSD license ?
> Amen to that.
Funny how people working on similar OSes(the BSDs) could bitch about one another instead of realising that there is no reason to. Now go smoke some peace pipe with magic indian herbs and stop bitching like dumb bimbos.