Why Are You NOT Using FreeBSD ?

[Adam Strohl]

On 6/3/2012 10:09, Mark Linimon wrote:
> On Sun, Jun 03, 2012 at 01:43:43AM +0200, Fritz Wuehler wrote:
>> So there could be lots of overlap and just looking at the two numbers
>> you posted doesn't really tell the whole story.
> No, I agree that it doesn't.  I was just trying to add an aside, and
> point out that the task would not be trivial.
>
> Since I'm heavily invested in FreeBSD ports I think I need to step back
> and let other folks comment in this thread.

I manage and support a little over 50 FreeBSD servers (VMWare, Xen and 
native) and feel that the port system, on the whole, is excellent.  Its 
easily one of the best features about FreeBSD.   Portaudit reports 
issues and I can plan and upgrade them as needed.  Portupgrade works 
great 99% of the time and when it doesn't it has the good sense to roll 
back what its done.  If there is any question as to what it should do it 
errors and tells me, which is exactly what I want it to do.

I've been a FreeBSD user for about 18 years and supported it 
professionally for about 10.  In this thread I've read a few posts that 
contain blanket statements like "ports are broken" and "never work", I'm 
at a loss as to how to respond to this as it is completely counter to my 
experience.   I wish I could see what they were talking about and figure 
out what happened so I could understand what caused them to make such a 
statement.  It's like they're talking about a different OS than the one 
I know.

I've written a simple script to run portaudit and pop up a dialog with 
check boxes that then kicks off portupgrade for the selected ports which 
have issues.   99% of the time its that simple.  This is what I want in 
a server environment.  I do not want things auto-updating (a.k.a. auto 
breaking) or making decisions about supporting libraries behind my back. 
   PHP is a good and common example why: an upgrade can and does break 
web sites that ran fine before.   Updates need to be managed in a 
process which is outside the scope of the OS (because its a server not a 
desktop).  FreeBSD has all these great tools for managing the mechanical 
action of updating and imposes minimal process which is perfect because 
I have my own process.  And if things get mucked up (which mostly isn't 
the ports system fault when it does happen), its easy to back out and 
re-do if needed.

After reading this thread I am wondering if I should clean the update 
dialog script up and submit to the ports tree.  It seems like people 
think the port update process is harder than it is because it lacks a 
Windows Update like dialog which is essentially what this is akin to 
(and there might be a port which does this already, too .. anyone?).  
All the hard stuff has been done by the FreeBSD team, all I did was put 
a bash/dialog script on it.

I very rarely run into ports that don't build on supported versions of 
FreeBSD (ie; ones that haven't reached EoL).  I have a number of 
customers with a few 6.2 boxes [which I can't wait to upgrade] and still 
almost everything builds without tinkering.

All of this is in the scope of servers though (web, DB, application, 
etc) and not on the desktop.  I haven't used a FreeBSD desktop since 
probably 4.x, and while I don't begrudge the work people are doing for 
the desktop experience it just doesn't apply to me nor is it why I love 
FreeBSD.   I won't say something like "you're running a server OS on 
your desktop and expecting it to be like a Mac".  What will say is: I'm 
getting from this thread that a lot of the complaints people have seem 
to be based around the desktop.  My guess is that this is a super 
minority of actual use (by server count).

BUT: I feel like people are judging how fit an FreeBSD is for server 
work by how easy/Mac/Windows/whatever like (as many Linux distros try to 
emulate) it is to update.  Not good ... but it makes sense from a 
social/human perspective, and is probably another thing we should 
consider in terms of advocacy.

I'm interested in what people think about this, and yeah this should 
probably be in the advocacy list but its not so thhblt :P