[nSLUG] Why you are not seeing software ported to Linux

D G Teed donald.teed at gmail.com
Thu Jun 23 00:09:40 ADT 2011


On Wed, Jun 22, 2011 at 10:19 PM, Daniel Morrison <draker at gmail.com> wrote:

> On 22 June 2011 21:30, D G Teed <donald.teed at gmail.com> wrote:
> >> > When a CDROM won't boot Solaris or Windows, I know the
> >> > drive or media is bad.
> >>
> >> You have an 'or' in there, which means you don't really know anything.
> >
> > There is another possibility.  You don't understand.  But this says
> > a lot about your intentions.  You don't seek to find an understanding,
> > the approach is combative.
>
> I'm sorry I was unclear in my statement. I did not mean to say that
> you "don't know *anything*". What I meant was: given the ambiguity of
> the situation with the CD, you cannot know with any certainty what the
> problem is. You don't know anything *about the CD*.
>

OK, sorry, I misread that completely.  Started the day too early.


> I do sometimes come across as combative; this is a character flaw, I
> acknowledge. However you are correct: I don't understand. You have
> started a thread concerning ABI (binaries), API (programming), package
> management (system administration), and a variety of other loosely
> defined standards (or lack thereof) and you have made a number of
> generalizations which spurred some responses. You don't really seem
> interested in the responses however. It sounds more like you want a
> platform to vent from (which is fine), but dislike it when others
> assure you that it isn't as bad as you suggest.
>

It isn't as bad?  Really?  My workplace sometimes uses mothballed hardware
from your workplace.  We do unusual things when we are lacking funds.
I might have a little more experience with how backward compatible
new software and old hardware really are.


>
> > It is just a question of going beyond the QA standard of "it worked for
> me".
>
> No, not necessarily. You assert that it doesn't work because the
> developers are lazy. This is an unfair claim.
>

I said "lazy" literally?  Don't think so.  I said they didn't
put a high priority on the issues.  Maybe you'd like to
offer a guess as to why the standards can be lower at times.
In my observations, developers dislike tasks like writing end user
documentation and QA with unit testing and all the nitty gritty.
Personally I believe it is because there is more glory in developing code
than all of the unsung chores that go around it.  The chores
are more likely to happen if people are paid to do them.  On the converse,
the paid chore might be done poorly, but someone's
job is on the line and they have to at least make the claim they covered
everything needed.  There are reasons why volunteer efforts can
be well done, and why professional efforts can be well done.


>
> > What I am getting at is aiming for a standard, like those in engineering.
> > Engineers don't build a bridge, drive a car across and declare it is
> > safe based on the car getting across.  But software developers do,
>
> LOL, that is exactly how the first bridges were made. Over hundreds of
> years, engineering discipline has improved.
>

Of course, but you don't see them determining the bridge safety
by trial and error afterward, whereas in software, especially
open source, there is still a great deal of what a person
could call a culture of experimentation, with a kind of
"gee whiz I built a helicopter in my backyard with some friends"
environment.  At some point, it has to mature, and roll up
into Bell Helicopters or something.  But since we don't
need bricks and mortar and a money making company, it
can be an organization like CentOS or Debian.  But how do we
avoid stuff like the kernel and grub disruption?


> Software developers, in contrast, are a new breed, and so yes, you're
> right... but they use only virtual cars. Software developers do not
> actually hold your life in their hands. Some software is engineered
> for mission-critical applications. It costs a lot more than Microsoft
> Office.
>
> > and especially those in the best efforts category of open source,
> > reverse engineered drivers, wrapper drivers, reverse engineered
> protocols,
> > etc.  It is the nature of the situation in many cases,
>
> No, I simply do not agree. In general, paid software developers are
> motivated by the pay cheque, and do the minimum required to get their
> code to work, and to pass inspection by the very small number of
> individuals who will ever look at it. Open source software developers
> are more likely motivated by the code, rather than the money, and are
> conscious that a virtually unlimited number of people may look at
> their creation. "Many eyeballs make shallow bugs", or something like
> that. Are you sure you understand the philosophy of Free Software?
>

As mentioned above, there are a bunch of pros and cons for
the quality you can get out of open source versus commercial software.
I would never say paid people always do worse quality.

I understand the philosophy of open source but I don't drink
the cool aid.  Philosophy does not override real world
experiences, unless one is deceiving themselves.

The many eyes quote is good.  Also is: "free as in free puppies".


> Also, the days of reverse-engineered drivers and protocols, and
> wrapper drivers, is mostly over. I'd wager >99% of open source code is
> not in that category. And code that is in that category is considered
> unreliable, hackish, stop-gap measures. As with many of your other
> examples, it seems you are cherry-picking examples of bad code,
> obscure or unique examples, and special cases, and comparing them with
> broad generalizations (e.g. all 13-year-old windows code still works).
> I don't buy it.
>

So when Microsoft change their network drive mapping or active directory
code
they send down the specs to the folks at Samba so they
can correctly emulate it?  No, it is reverse engineered.

Two different IBM xSeries servers have required the parameter
rootdelay to boot properly with 2.6.32.  There is a bug report
at kernel.org on it.  It has not been hard for me to experience
that problem.

In some hardware devices (wireless, printers) there has been a trend to
replace
firmware with software drivers for Windows.  In that case if it
works at all it would be via wrappers.

There is no cherry picking when it comes to the Sun Solaris binaries.  It is
there by design.

I don't know that things have become better or worse in regard
for design standards, but I just think there is room for improvement
as open source is maturing.  There has to be some mechanism of change.


> > but in others
> > there could be higher standards, and it is particularly demonstrated
> > when 7 year old hardware can't do things it used to.  The changes
> > have broke things.
>
> There could be higher standards. Like I said, there's lots I could
> complain about (and I do, at times!). However I still would rather a
> world with free software in it, than... I'll join Ben here. What is it
> exactly that you're advocating? More care, attention and dedication to
> duty, I think. This would also reduce traffic accidents. I'm all for
> it.
>

Discussion.  That's what I'm advocating.  And maybe freedom to
think outside the walls we tend to draw around what we do,
outside of what banners we like to carry, etc.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://nslug.ns.ca/mailman/private/nslug/attachments/20110623/9d35d108/attachment-0002.html>


More information about the nSLUG mailing list