[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Open Letter @ the developers of oBSD

Hash: SHA1

I've some critism, some questions and some requests.
At first my english is not very well. So ignore some misstakes I give my

After writing a mail in the smp_(_at_)_openbsd-mailinglist I became some
critism. I asked if there any useable code for x86 and if their
intrested in hardware I could LEND (Theo knows wich hardware, but more

I became some... not realy friendly answers.
Yes I'm not a coder, yes I'm just a user and yes I can't change the
But: Please stop shit like this on ALL mailinlists here.
I recognized such answers in all mailingists.
Just things like: If you're using x86 ist't (bla). But "real" mashines
(then the list: sparc, sgi and co)dosn't have (bla).

I'm sorry but... bullshit. Answers like this one dosn't help.
That's just a waste of time for everybody. Even for me could I read
answers like this and hope to learn anything which could helpfull later.

So that's a request to ALL people around this world wich read this mail:
Stop it, please...

Another thing is the port- and packagesystem.
I know oBSD hasn't as much developers as debian.
And I think there less developers working on oBSD as on
I using oBSD since 2.8. And in my point of view: Great work... one of
the best OSs I ever use.
But I here the things wich are improveable (I think so).

1. Things i miss (SMP is the most thing I miss... but I'm not alone)

I miss SMP realy, I miss HT, I miss PowerStepping (Pentium4-Mobile).
I miss actual ports, packages.

I know there not so many people and I know it's hard to make all
packages up2date (or ports).
So please (just MY point of view) reduce the packages to set manpower
I dosn't know who manage the ports at all. I dosn't know who's the big
boss wich judge each package. Yeah I dosn't know...
But if the developers (coders) have to hold the packages/ports up2date
please give other guys this job.

Another thing I miss are drivers at all.
Somewhere I could read that Intel dosn't support the project with docs.
Yes...well... but: A server without an intel-nic? Ok.. 3com, but RealTek
or VIA?
So I think it could be helpfull if the responsible guys ask other
opensource-projects like fBSD. If fBSD got these docs ask these guys.

I miss much... even you (developers) give us (users) much.
Yes SMP is not easy to set up.. but even it's there it should be useable
for the future, or?
I dosn't know.. maybe SMP on SPARCs differ from SMP on INTELs.
Maybe it's too hard? I don't know.. :o/
But I know I miss it. Couse I've an INTEL SHG2 Motherboard.

After a walk in the computerroom our university I asked the FW-Admin
which OSs are in use there. He told me "Solaris, LINUX and fBSD".
He's a solaris freak but he told me if he has a choice he would use oBSD
for the firewall. So fBSD is in use becouse SMP.
I dosn't lie... I know much of these people. People who would use oBSD
but who can't.

So some guys told me (many months ago) oBSD isn't an workstation OS.
I use it on my notebook (so I miss acpi and feautures of the P4-M) and
it works very well.
I would use it on my workstation too... but I've to remove it from my
server couse SMP.
SMP is nothing which is not useable for all...
It's a missed feauture..

An oBSD firewall will never work on large networks. I think so...
Yes oBSD can handle large traffic, but not 2 CPUs. Some companies need
this power on Servers. Sure..the firewall on my university is not a real
example (dual XEON, just for the mainbuilding (1TB connection)) ..sure.
But it shows that oBSd isn't useable for very large networks.

So... please put some developers... and "start" the SMP-part of oBSD.
You've great cryptocard-drivers.. well.
But my university, the companie of my mother and all the other companies
I saw dosn't use them. But they've SMP-mashines in use.

I talked with over 100 people in this year..
I take my laptop, sit down.. and then the people start to ask.
"Hey which OS you're using? is it linux?" (and so on).
So I became contact to peoples like the FW-Admin of my university and
other companies (newspapers an so)...

And why didn't you copy code? SMP-Code from fBSD or so?
Hey guye... oBSD, fBSD...nBSD?
Just reduce it.... BSD.
One licence, or? So where is the problem? :o/
If you can't code it yourself at this time, take other code and replace
it later... so you've a feauture which is useable and you dosn't break
any law if you use BSD-code.
Ok..there's NO honor... like "hey I developed the SMP-bla in oBSD".
Yes... but what's your goals?
A secure OS...well.
But a secure OS used by 5k people on the world?
Which is not useable in large networks, on big servers?
Yes security is very importent... but other things importent too. So
please never forget this fact.
Yes... some guys told me: "We dosn't code for users". Well... but the
world isn't a place full with coders. There less coders as users. The
less big-servers without dual, quad (or more) CPUs as servers with 1
cpu. Sometimes I wish some guys open their eyes... but just sometimes.

2. Answers...

Yes I miss answers.. not from the mailinglist.
I miss answers like people as Theo.
I wrote Theo a mail..
I asked if it's possible to LEND hardware... 
I asked him if the project need webspace...a server or other things.
But.. i never became an answer. Why?
Hey guys.. it's like in the SMP_(_at_)_OPENBSD-Mailinglist_(_dot_)__(_dot_)__(_dot_)_
If you DOSN'T need anything... tell me. Tell us all... write it down on
the homepage "We dosn't need anything even manpower".
I was and I'm very disapoint to get ANY answer.
That's not the way....
Or is this the way to find people wich would help... wich could give you
something. So that you're able to work better... :o(

3. Future and old drivers.

I think we've to think about the future...
There will be SMP... there will be HT. There will be a lot of new
devices. PCI-X and co are comming to be standard.
EISA and ISA will be outdated..

I don't know but how much people dev. on the feautures of the future?
Like PCI-X,G-LAN, SMP (yes..again) and so on?
So when it's time to kick off old stuff?!
I dosn't have an 386er as FW, I dosn't have an 486er...
The oldest PC in the companie of my mum is an P2 400 mhz.
The oldest in the university I ever found is an K6-2 (I asked the man
which is responsable for all hardware on the mainbuilding). The oldest
PC of my friend is an P2~333 Mhz.
My oldest PC will be an P3-S 1400Mhz.
I will replace my server (AMD) with an INTEL-XEON.

I've no need for EISA/ISA. I dosn't know which notebooks work (PMCIA and
co) but I know NO person wich use an ISA-NIC (even it will be enough
for standard DSL connections).
I don't know persons wich buy ISA-devices today... nobody use it realy
today. Yes.. I hear some guye here wich told me "I use it" or "My router
works with an 486er...". Yes.. ok.. but I dosn't mean you.
I think about the mass of users or.
Who build networks with 10Mbit? Nobody.. not today.
I'm not poor and I'm not richt. I've to save money for the intel
motherboard...yeah. But I need the calculation-power.
But even I'm not rich I've 3 EURO for an PCI-100Mbit RL-Device.

I replace the RLs 2 or 3 months ago. Couse there's a NIC (vr)
 on the AMD-motherboard of my Server/Router. So I took the only
NetGear-NIC I have into the server.
The VR is enough for DSL and the DC it for the lan.

But at all: Remove old or senseless code if it need coders which take

I think if the project will reorganize itself it could be more
effective. It will able to produce an OS wich is more then an FW-OS for
middle companies or homeusers.
The securest OS at all dosn't make sense if nobody can use it.

I read more then 1 time that you've not enough coders.
So.. why dosn't you let us (world) know?
Write it down.. under the pic of oBSD on the mainpage "WE NEED CODERS".
Yes users could spend...hardware (but I will lend it... I've not the
money to spend it), money... yes we could buy it. Ok....
But there's one thing normal users CAN'T do... code. Like me.. I'm not a
kernelhacker. I could spend money, lend hardware.. talk with peoples in
the university.. yes. But I can't code.
So if you need coders talk to the world...
Write it down on the mainpage...
We search coders/crypto-experts/xyz...

I don't know how much people visit the donations-webpage. But all visit
the mainpage of the homepage...

That are only ideas... and my point of view. I hope I dosn't become
answers from idiots with text wich nobody need...
But I hope I became answers from people which think about these points.
It's not all critic, oBSd works well.. even without ACPI or
But.. I hate to remove it from my server comming soon.
And it's not useable on my workstation (sound: audigy2) too. :o(
I would use it.. couse in my opinion it's a good OS.
Yes.. it's not so fast like BeOS, well.
But for MY work it's perfect. Secure, smal and not too lame.

I hope these things will change...
I hope I could use it someday on my workstation (but I think then I've
new hardware *g*) or my server...
I waited 1 version (3.3 to 3.4) to use it on my notebook.
Yes... fBSD works well..but I waited and now I could use it.
I can't do other things.. so I've to wait.
But how long should I wait? How long could I wait?
Will I've to say my son he should use my old servermotherboard with
openBSD 20.5 to use SMP? Or sound? :o(



I forgot something: I saw into the generic--sound-drivers-section many
"sections". Like OSS and so on. It is possible to take all these
FREE-drivers and put it into ONE sound-deamon?!
I don't know if oBSD use OSSd (the last free) but if it use it pleace
remove it with an own one. I think this could be helpfull to hold the
code (when it's done) more useable for the future.
Version: GnuPG v1.2.2 (OpenBSD)