Topic: What do you want or like to see in any of the BSD's ?

Ive often wondered what the BSD's need in order to be ahead of linux or other operating systems.
There are of course areas that BSD's are ahead like stability, consistently of the kernel api , the system design is clean and not fussy.less hacks to get things to work.

The areas i think need more concentration  include a real time kernel/low latency kernel , in order for the BSD's to get into embedded  devices and audio workstations.

Scalability is also a issue, i ve read it often in the mailing list that linux scales better.

More packages, keep the ports for flexibility, but maintaining a larger maintained package repositories is important, its quite annoying waiting hrs,days to update software.

A newer file system, dragonfly project has started work on a new one, weather its a good idea for all BSDs , ive no idea, but it seems a good move.

That's my first thoughts, i add some more when they come to mind.

Re: What do you want or like to see in any of the BSD's ?

After Maxlor showed me the possibilities offered by ZFS, I'm saying   ZFS

"UBER" means I don't drink the coffee... I chew the beans instead
             -- Copyright BSDnexus

Re: What do you want or like to see in any of the BSD's ?

A better scheduler, maybe some kind of multimedia framework, no giant lock at all (FreeBSD) and so on.

>Scalability is also a issue, i ve read it often in the mailing list that linux scales better.

Rather nonsense, because of FUD.

http://jeffr-tech.livejournal.com/6831.html

F!XMBR

Freedom is the sure possession of those alone who have the courage to defend it. --Pericles

Re: What do you want or like to see in any of the BSD's ?

Software applications, at least for the desktop.  Native Flash, a VMware host (or at least one decent software virtualization program), Acrobat (the full one, not only the Reader), Photoshop (GIMP has many severe limitations), a good 2D and 3D CAD program, a decent accounting package, and so on.

Most people sniff that the problem lies with the commercial vendors, but I don't buy it any longer.  The people responsible for the the OS have to ensure that people can use the OS for their needs, and that means software.  They have to take a more active role in getting software to run on the BSDs.

For me the OSS server applications have been adequate, but then I don't need anything like SAP.

Re: What do you want or like to see in any of the BSD's ?

DrJ wrote:

Software applications, at least for the desktop.  Native Flash, a VMware host (or at least one decent software virtualization program), Acrobat (the full one, not only the Reader), Photoshop (GIMP has many severe limitations), a good 2D and 3D CAD program, a decent accounting package, and so on.

Most people sniff that the problem lies with the commercial vendors, but I don't buy it any longer.  The people responsible for the the OS have to ensure that people can use the OS for their needs, and that means software.  They have to take a more active role in getting software to run on the BSDs.

Why is it not the commercial vendors problem? Where is the flash,vmware,acrobat,photoshop src tarball  ?that is publicly available ?

If you can get the vendor to give it to us then sure those apps will eventually get into the BSD's , until then there is null we can do about it.

Re: What do you want or like to see in any of the BSD's ?

>They have to take a more active role in getting software to run on the BSDs.

Why? It's no company, it's community work - if you're able to help - in some way - do it. Developers of *BSD are developing the operating system, nothing more, nothing less. It's up to others to fill the gap with applications.

F!XMBR

Freedom is the sure possession of those alone who have the courage to defend it. --Pericles

Re: What do you want or like to see in any of the BSD's ?

I've often seen this attitude, but individuals alone cannot do it.  FreeBSD as an organization has done it before with Java, and could choose to do it again with other projects.  Of the things I listed, VMware host support is probably the most important, at least for me, though Xen will help once it hits Free.

I have pretty much resigned myself to leaving the BSD community (unless OSX counts), since the software simply is inadequate for my needs.

Re: What do you want or like to see in any of the BSD's ?

DrJ wrote:

I've often seen this attitude, but individuals alone cannot do it.  FreeBSD as an organization has done it before with Java, and could choose to do it again with other projects.  Of the things I listed, VMware host support is probably the most important, at least for me, though Xen will help once it hits Free.

I have pretty much resigned myself to leaving the BSD community (unless OSX counts), since the software simply is inadequate for my needs.

Java is a different child, what seemed to stop it being in FreeBSD was licensing issues, had sun released it on the BSD , none of that would of happened, its now on gpl so things are a little better there.

Vmware obviously feel FreeBSD is not big enough for there market, this might be resolved when Linux compatibility is upgraded .

Its interesting that OSX has everything you want because its market share is smaller. If it was not for the ipod , it could well be a dead company, although there have a market in the graphic design industry.

Re: What do you want or like to see in any of the BSD's ?

lobster wrote:

Vmware obviously feel FreeBSD is not big enough for there market, this might be resolved when Linux compatibility is upgraded .

Doubtful.  Orlando has had major issues with the port, and VMware has not been helpful in responding to his (fairly easy) requests.  Certainly the linuxator is a moving target, and maybe things will improve enough, but so far there is no joy.

Its interesting that OSX has everything you want because its market share is smaller.

That is my point.  Companies want to deal with other organizations, and not individuals.  That is why this has to come from FreeBSD (or Net or Open).  It just won't happen otherwise.  It might not anyway, since companies do what companies do, but without an organizational voice the effort is dead out-of-the-gate.

FWIW, together with Scottro and WIntellect here I lead an effort to get Wine ported to BSD -- Free to start, and then others.  The response was rather underwhelming.  And it took a huge amount of time that I no longer have to invest (on the order of 1000 hours).

Re: What do you want or like to see in any of the BSD's ?

DrJ wrote:

That is my point.  Companies want to deal with other organizations, and not individuals.  That is why this has to come from FreeBSD (or Net or Open).  It just won't happen otherwise.  It might not anyway, since companies do what companies do, but without an organizational voice the effort is dead out-of-the-gate.

In that case it must come down to what the Free,net,Open etc organsations who deal with companies want in the distros.
If they do not want vmware , then its not going to be top on there agenda.

11

Re: What do you want or like to see in any of the BSD's ?

lobster wrote:

In that case it must come down to what the Free,net,Open etc organsations who deal with companies want in the distros.  If they do not want vmware , then its not going to be top on there agenda.

Exactly right.  I fear that none of what I mention is on the agenda of any of the BSDs.  I'm not privy to insider discussion, so this is only speculation, but I have seen no evidence of any such desire to date.

Free has by far the largest installed base, at least for desktop use according to my work on the CrossOver Office petition.  It pretty much has to come from them.  The others don't have the user base.

The only other organizations that could do it are PC-BSD or Desktop BSD.  Now that PC-BSD is quasi-commercial, they may also be well-positioned to lead this sort of effort.

Re: What do you want or like to see in any of the BSD's ?

DrJ wrote:

I've often seen this attitude, but individuals alone cannot do it.  FreeBSD as an organization has done it before with Java, and could choose to do it again with other projects.  Of the things I listed, VMware host support is probably the most important, at least for me, though Xen will help once it hits Free.

I have pretty much resigned myself to leaving the BSD community (unless OSX counts), since the software simply is inadequate for my needs.

Java isn't comparable to other applications, because Sun was on the way to open it anyway. And last not least they had to pay for it. Do you have the money to pay companies like Vmware to port their application? I doubt it and most of the open-source users aren't found of spending money for a free operating system. Money or manpower, it doesn't matter but you do need at least one of them.

>The response was rather underwhelming.

There you go, maybe it's a special *BSD attitude. I do see it of course. If you're speaking in terms of community, I would say Linux has got the better community than *BSD. Of course there are some people like Scrotto and so on, but there are only some. FreeBSD e.g. has got one big problem it doesn't know its future direction, where to go - that's true. They don't want it to be a mere server OS, but they don't know what do you need for a full-blown desktop environment.

It's for example a desktop etc. for myself, but my wife isn't found of *BSD and uses ArchLinux.

F!XMBR

Freedom is the sure possession of those alone who have the courage to defend it. --Pericles

13

Re: What do you want or like to see in any of the BSD's ?

Oliver wrote:

FreeBSD e.g. has got one big problem it doesn't know its future direction, where to go - that's true. They don't want it to be a mere server OS, but they don't know what do you need for a full-blown desktop environment.

I snipped out your comment about the community, but I have found the BSD community to be very good.  But given that the topic is "what do you want to see in the BSDs", this is what I would like to see.  Getting the support of commercial software vendors.  FreeBSD could do it, but they have to decide they want it, and go for it.

It's for example a desktop etc. for myself, but my wife isn't found of *BSD and uses ArchLinux.

I am starting to evaluate Arch to replace the BSD desktops my company uses.  It would solve some of the problems I mention in the interim.  That would lead inevitably to BSD leaving my company entirely.

Maybe I'm just in a grumpy mood.  But I have had this discussion with scottro for a long time.

Re: What do you want or like to see in any of the BSD's ?

Oliver wrote:

FreeBSD e.g. has got one big problem it doesn't know its future direction, where to go - that's true. They don't want it to be a mere server OS, but they don't know what do you need for a full-blown desktop environment.

That is not true, FreeBSD has an agenda for 7 ,http://www.freebsd.org/releng/index.html , all the information on where FreeBSD is going is there.

You are certainly right that FreeBSD is not a full-blown desktop, that is not one of there priorities.

pc and desktop have that at interest but at heart there are just FreeBSD cosmetic changes.

I would much prefer the effort put into FreeBSD than what there are doing because those projects rely on Freebsd for the underlying system.

15

Re: What do you want or like to see in any of the BSD's ?

lobster wrote:

That is not true, FreeBSD has an agenda for 7 ,http://www.freebsd.org/releng/index.html , all the information on where FreeBSD is going is there.

I find *nothing* on where FreeBSD is going at that link.  I find organizational information, and no more.  That is not an agenda.

You are certainly right that FreeBSD is not a full-blown desktop, that is not one of there priorities.

pc and desktop have that at interest but at heart there are just FreeBSD cosmetic changes.

I would much prefer the effort put into FreeBSD than what there are doing because those projects rely on Freebsd for the underlying system.

I see them as integrating a desktop environment for newbies much like Ubuntu is doing for Debian.  If there is to be commercial software for the desktop, you need a good desktop presence.  As you say, it is not Free's priority to be a full-blown desktop (and I hope they never lose their server orientation) but then someone has to step into the void.  I personally have no problems in setting up as good a desktop as can be had, but to get the numbers necessary you need it to be easy.  So I see these efforts as good and necessary things.

Re: What do you want or like to see in any of the BSD's ?

lobster wrote:
Oliver wrote:

FreeBSD e.g. has got one big problem it doesn't know its future direction, where to go - that's true. They don't want it to be a mere server OS, but they don't know what do you need for a full-blown desktop environment.

That is not true, FreeBSD has an agenda for 7 ,http://www.freebsd.org/releng/index.html , all the information on where FreeBSD is going is there.

You are certainly right that FreeBSD is not a full-blown desktop, that is not one of there priorities.

pc and desktop have that at interest but at heart there are just FreeBSD cosmetic changes.

I would much prefer the effort put into FreeBSD than what there are doing because those projects rely on Freebsd for the underlying system.

I do know the agenda, I read the mailinglists, I'm in the DesktopBSD team and I do some work in terms of howtos and advocating FreeBSD. I'm using plain FreeBSD with Fluxbox and mostly console applications since two year, before that I was an avid Linux user for about 10-11 years - you see my needs are very small wink

http://ivoras.sharanet.org/freebsd/freebsd7.html

-> tell me, what's usable for average joe?


http://ezine.daemonnews.org/200402/dadvocate.html

Greg Lehey 2003,

What should the BSDs do? As I've said several times in the last five years, that depends on what the projects want. But if we want to be seen as a viable alternative for use by non-developers, my big issue is that we need to understand the end user perspective. We can't just say of any feature

Last edited by Oliver (2007-03-21 17:57:42)

F!XMBR

Freedom is the sure possession of those alone who have the courage to defend it. --Pericles

Re: What do you want or like to see in any of the BSD's ?

- (Near) Real-time file system replication
- Scalable network file system, with multiple master servers, robust enough to run through flaky Internet connections
- ZFS-like file system

Another thing I'd like to have, but not in the base, is a good HTTP spam filter to protect the users sites. Everyone has dynamic pages with editable content, and getting tons of spam. Manually handling rules with mod_security sucks.

Re: What do you want or like to see in any of the BSD's ?

Thanks for that links, i could not find it.

Oliver wrote:

-> tell me, what's usable for average joe?


http://ezine.daemonnews.org/200402/dadvocate.html

Greg Lehey 2003,

What should the BSDs do? As I've said several times in the last five years, that depends on what the projects want. But if we want to be seen as a viable alternative for use by non-developers, my big issue is that we need to understand the end user perspective. We can't just say of any feature

Re: What do you want or like to see in any of the BSD's ?

jage wrote:

- (Near) Real-time file system replication
- Scalable network file system, with multiple master servers, robust enough to run through flaky Internet connections
- ZFS-like file system

Another thing I'd like to have, but not in the base, is a good HTTP spam filter to protect the users sites. Everyone has dynamic pages with editable content, and getting tons of spam. Manually handling rules with mod_security sucks.

That's cool suggestions

Re: What do you want or like to see in any of the BSD's ?

>but if there are not the people to support it or want to support , what can you do?

But that's actually the problem, they want it (sort of) and the lack of perception - they are doing almost nothing for it. Look at the mentioned Wine, it's compared to Linux almost unusable in FreeBSD (don't know the status in Net or Open) - this would be for example a so-called killer-application.
The gap between Linux and *BSD will be huge in Future in terms of technology. There is quality - okay, I don't see it in any Linux distro - otherwise there is the massive lack of technologies in FreeBSD.

Linux for example made a jump in terms of performance from 2.4 to 2.6. Most of long-term FreeBSD users are lamenting about the decrease of performance from 4.x to 6.x or after 5.2.x and now they are waiting for the new and shiny FBSD7 but many of them are already gone.

F!XMBR

Freedom is the sure possession of those alone who have the courage to defend it. --Pericles

Re: What do you want or like to see in any of the BSD's ?

Oliver wrote:

Look at the mentioned Wine, it's compared to Linux almost unusable in FreeBSD (don't know the status in Net or Open) - this would be for example a so-called killer-application.

Wine is hack to get windows programs to work, native applications should always be favoured. My experience with wine on FreeBSD is that if the application is single threaded its seems to be ok.

Oliver wrote:

The gap between Linux and *BSD will be huge in Future in terms of technology. There is quality - okay, I don't see it in any Linux distro - otherwise there is the massive lack of technologies in FreeBSD.

There is most likely 10 linux developers to every one FreeBSD developer , more hands on the deck, more work gets done.
There is little point in talking  about the lack of X when developers numbers are not there.
A  building cannot be built in a reasonable amount of time without enough builders, FreeBSD is no different.



Oliver wrote:

Linux for example made a jump in terms of performance from 2.4 to 2.6. Most of long-term FreeBSD users are lamenting about the decrease of performance from 4.x to 6.x or after 5.2.x and now they are waiting for the new and shiny FBSD7 but many of them are already gone.

I have not been around that long, so i cannot reasonable comment on that, what i can say 2.6 was  known to be unrealisable in its early days, slackware for example still ships with a 2.4.x kernel.
I am too looking forward to the shiny and sexy 7 to come out, lets hope performance improves.

Last edited by lobster (2007-03-21 20:06:10)

22

Re: What do you want or like to see in any of the BSD's ?

lobster wrote:

There is most likely 10 linux developers to every one FreeBSD developer , more hands on the deck, more work gets done.
There is little point in talking  about the lack of X when developers numbers are not there.
A  building cannot be built in a reasonable amount of time without enough builders, FreeBSD is no different.

This also is exactly the point.  It is a classic network effect (in a business sense, and not a computer sense).  I'm in biotech, and the three big biotech communities in the US are the San Francisco Bay area, San Diego, and Boston.  Nearly everyone who starts a company in that space goes to one of those three places.

Why?  Because everyone else is there.  That gives you all sorts of benefits, from access to labor, capital, savvy attorneys and accountants, supply networks, the whole nine yards.

It is also why "everyone" uses Windows.

Unless the BSDs can turn around their market position, they will continue to fall behind Linux, and probably Solaris in the long run.  How best to do that is a real question, and it may well be that its position on the desktop is already lost.  I don't think so, though, and for me it would only take a good paravirtualization product.  Those we have really are not very good other than the very old version of VMware, and that currently does not work on 7.0.

However, I think it will take a *lot* more than simple technical excellence.  Marketing, partnerships and that sort of thing simply have to be explored.  Otherwise I don't hold out a lot of hope for the future as a general-purpose OS that is broadly useful.

Re: What do you want or like to see in any of the BSD's ?

>slackware for example still ships with a 2.4.x kernel

Yes this is because he doesn't know better. I do know lot of production servers with kernel 2.6.x - have a look at the Isps most of them using Linux with kernel 2.6.x. No problem whatsoever. Maybe killing a Linux server is easier under heavy load, but a proficient administrator will probably handle such difficulties. We're using e.g. Solaris and Debian Servers at university - guess why? They're doing their work day by day without any hassle.

>Wine is hack to get windows programs to work

It's an emulation layer and almost as essential as Linuxolator. Without the latter most people wouldn't even use FreeBSD. I'm using Mathematica, without binary compability Linux would be the only possiblity.

>There is little point in talking  about the lack of X when developers numbers are not there.

I mentioned it above, the lack of manpower - but if you do tell people we are ready for everyone, people lock at it this way. And if you have a greater userbase you can do much more things, apart from far more "influence" on companies. Last time (2006) the maintainer of Gnome port made the silly advertisment "FreeBSD will be comparable in 1 year". This for examples is nonsense, just hype to lure people into thinking something. But the reality ...

If this is too much work, FreeBSD core team has acknowledge it ... the power to serve and nothing else.

>A  building cannot be built in a reasonable amount of time without enough builders, FreeBSD is no different.

How old is FreeBSD, yes and I do know the court matter in the 90s? You have to make it more attractive even for possible developers, you have to advertise it, you have to push it - look at some postings above, the fact that most of the people don't even know it or mistake it with a Linux distro is a pity. *BSD is a several decades old UNIX derivative, so it has a proper base - sometimes you have to have courage to speed up development. No fun without risk.

I could certainly use OpenBSD or NetBSD, apart from Mathematica, so I'm not the typical user. Vim, LaTex and so on doesn't count nowadays as "office software" wink

But you throw in FBSD7 and your saying

I am too looking forward to the shiny and sexy 7 to come out, lets hope performance improves.

http://ivoras.sharanet.org/freebsd/freebsd7.html

What do you think this will be? I'm testing FBSD7 too, the scheduler is maybe nice for Mysql or scaling at all with 8 cores etc. But it's almost of no use for my dualcore. ZFS e.g. is somewhat of a hype, maybe useful for server farms, but not for the average user. Journaling is hype to, if you understand what UFS2 + softupdates really is, you do know that you already have quality.
And the rest? Nothing to bother with.

With really bad luck it could be possible that ULE2 (the new scheduler in current) will be deactivated like ULE1 in FBSD5 - if he doesn't prove himself in daily work.

Linuxolator will be certainly without kernel 2.6.x because of problems and lack of time,

http://lists.freebsd.org/pipermail/free … 03206.html

So know think of it, what is really new and shiny? And remember you can test ZFS maybe at home, but nobody with a sane mind would test it in FBSD7 at production servers for the next two years. If you keep all this in mind (yes I'm always pessimistic wink) there isn't much for the the average user left.

F!XMBR

Freedom is the sure possession of those alone who have the courage to defend it. --Pericles

Re: What do you want or like to see in any of the BSD's ?

As a sysadmin, I'd like to see true, perfectly convenient binary upgrades, something like ArchLinux's combination of abs and pacman--pacman is Deb's apt-get install and abs are like Gentoo's ebuilds--over simplification, but basically, the entire system can be upgraded with binaries, or, if you really need to customize a program, you can build it from source.

As a user, I'd like to see something that Just Worked (TM).  Oliver and the rest of the DesktopBSD people, as well as the PCBSD  folks, are all working towards that, but--I would not want to see it become like Ubuntu, where I watched 600 plus packages being installed, but not my beloved scim-anthy.  (This has changed, it's now available.) So, I downloaded the source and got

make not found. 

I would love for some of these things to work--as it is, I use linux-opera for most things and firefox for java and non-flash plugins.  To get linux-opera to work with non-flash plugins or get firefox working with flash is just too much effort for my rather minimal needs.  It would be nice flash 9 worked with sound--in Arch Linux, it's a simple matter of, without doing extra configuration, typing pacman -S alsa-oss or whatever it is. 
For an end user, googling shouldn't be that necessary, if it were planning to seriously get attention as a Desktop O/S.  The user, and I include myself, would like to install it, have X work, be able to plug in their scanner, iPod, digital camera, web cam and the rest and have it work.  Of course, one can argue that this isn't the case with Windows either, but the manufacturers aren't going to provide drivers unless the market justifies it.

Emulation--as Drj said, vmware doesn't seem interested, and I suspect that for people who need more sophisticated use of WIndows, qemu isn't going to make it.  It works for me, but my needs are minimal--I need to use a vpn client to get me into my network, I need to be able to open apps to walk my users through it, that sort of thing.  Wine simply doesn't seem capable--the two apps that I really need don't work with Wine, or even with crossover office in Linux.

Yet to me, saying, Ok, let's concentrate on the Power to Serve, is more important than the rest.  (That's speaking for my own personal needs, not from an advocacy standpoint.) If getting everything else to Just Work (TM) meant a bloated system where I have remove 300 packages after installation, I'll suffer with the lack of a desktop. 

Again, paraphrasing Drj, it's relatively easy for all of us in this thread to make our favorite BSD an easy to use, in most cases, desktop.  Yet, I need those few MS apps, so I have qemu (as vmware has stopped working on CURRENT) with, not only a WIndows installation, but an ArchLInux one as well.  From what I understand, not just from these forums, the VirtualBox program is better and smoother than qemu, but at present, at least, they don't plan to adapt it to FreeBSD.  PCBSD has the best chance of changing that, as it has that large user base and now has some commercial support.   

So, it's a tough thing--most of the arguments here are quite valid and true.  For getting the manufacturers' support, it has to become a larger market.  Linux is doing it and getting there.  As has been said, there are far more Linux developers. 

I think multimedia and peripherals are two of the biggest problems for the average user.  Drj is a doctor, and his needs from PDFs and other graphics programs are far beyond that of the average person.  In general, the average person can get by with the pdf readers involved, and with a little bit of work--not much, just a little--could probably use the office suites as well.  However, they aren't going to want to google to get their camera to work, when with MS or Mac, they can just plug it in, pop in a CD and have it run.

If it were my choice, though, I would rather that they concentrate on smooth binary upgrades of the system and all ports--however, I'm speaking in part as hobbyist--in other words, I don't mind googling to get it working as a desktop--and partly as an admin who doesn't want to spend a few hours doing buildworld--because test machines are the ones that  no one else wants to use, so buildworlds can take a long time on a test box.  My workstations are far faster, but they aren't adequate testboxes for production, because I have all my Desktop type stuff on them.

<@andre> i would be so much more efficient if i wasn't so stupid

Re: What do you want or like to see in any of the BSD's ?

I think it's to soon to draw any conclusions about FreeBSD's future.
Windows still holds more than 90% of the desktop market (http://marketshare.hitslink.com/report.aspx?qprid=2 and http://www.onestat.com/html/aboutus_pre … hare.html)
Linux has somewhere inbetween .3% and .4%
Oh my, Linux is getting big (Yes, that was sarcasm)
Vista has managed to get bigger in just two months

Linux is like Opera, everyone is counting opera among the "major browser", while in fact it has only ~.7% marketshare ... perception and reality are two different things.

Trust me, I know what I'm doing.

Posts [ 1 to 25 of 125 ]

Pages 1 2 3 4 5 Next

You must login or register to post a reply



Currently installed 2 official extensions. Copyright © 2003–2009 PunBB.