[lsb-discuss] Thoughts on LSB ISO standard

Theodore Ts'o tytso at mit.edu
Fri Jul 27 22:15:52 UTC 2012


On Fri, Jul 27, 2012 at 08:15:52PM +0200, Keld Simonsen wrote:
> On Thu, Jul 26, 2012 at 06:45:04PM -0400, Theodore Ts'o wrote:
> > So I can't speak for the Linux Foundation, but I was CTO (on loan from
> > IBM) during the time period when we made the decision to let the PAS
> > status lapse.
> 
> Excuse me, I am not fully familiar with the lingo here. So you were Chief
> Technical Officer from the Linux Foundation on the ISO LSB project?
> And then what would that mean?

I wasn't just working on LSB; I had to keep on top of all technical
activities at the Linux Foundation (i.e., at the time, Moblin, Maemo,
the various conferences run by the LF).  But yes, at the time I was
driving the LSB, and one of the things that meant is that I had to
work with the various corporate sponsors of the LF to see if they were
willing to pay for all of the costs associated with the renewing the
PAS process.

Previous to that, I was one of the representatives on the US National
Body, so I was one of the people who went to the INCITS meeting (which
meant that we had to have multiple corporate representatives paying
$$$ to the INCITS body) to formulate the US National Position which
was then used when we voted at the ISO/IEC JTC1 Linux Rapporteur
Group.

The various corporations involved also made sure they had engineers
representented on the various other countries National Bodies, since
voting happened by country.  So essentially when the LSB got the ISO
stamp of approval, it required a lot of behind the scenes work by
multiple countries, which was not unlike what Microsoft did when it
ramrodded OOXML through ISO, although admittedly on a much smaller
scale and there were multiple multinational corporations involved, not
just a single one (Microsoft).

A big focus at the time of all of this back room maneuvering was to
make sure that if anyone else tried to start up Linux standardization
efforts, it would get redirected through to the LRG, and then to the
LSB.  This way we wouldn't have to worry about someone else deciding
they wanted to start a competing standards effort, or at least, could
have a chance of redirecting to the LSB work group should someone try.

One of the worries I have about waking up ISO is that it may cause
some other people to decide that they want to spend time creating
other Linux-related standards, and ISO is such a heavyweight, hard to
manage process, where things are done by individual countries voting
(fortunately we were a relatively small ISO group that moved under the
radar; I have friends who were involved with the M, aka MUMPS,
standards effort, and they had representatives from the US State
Department involved while formulating the US National Position.....
which might cause you to suspect that sometimes it's a lot more about
politics at the international level than merely about technical
matters....)

> > On the "costs" side of the equation, just simply getting the document
> > into the right shape so that it met all of the ISO requirements would
> > have required a lot of editorial work.  The people who have the skills
> > to do that editorial work were at the time, fairly few and far
> > between, and they all wanted to paid significant sums of money per
> > hour to do that work.  (Hey, they were consultants with a relatively
> > rare skill set; I don't begrudge that, but it was still a cost.)
> 
> that is understood.

So who is going to do all of this work?  Or who is going to pay to do
all of this reformatting and adding references as required by ISO?
And would that volunteer effort be better spent making the LSB better,
as opposed to making things acceptible for ISO's policies and procedures?

> > In addition, merely engaging with ISO would have taken a lot of time,
> > going to various ANSI/INCITS meetings beforehand, and needing to play
> > politics with (for example) the Sun representative, who had no love of
> > Linux, and who was distressed that Linux wasn't fully compliant with
> > POSIX.  There were other all sorts of other silly things; such as
> > strategizing at the ISO level to have the meetings in the Far East, to
> > prevent certain known troublemakers from being able to attend (because
> > they might derail the proceedings) because the travel expenses would
> > be too expensive for them --- but that also meant that those of us who
> > had to actually *attend* the ISO meeting had to pay $$$ to travel long
> > distnaces to far off lands (and if you were employed in by company
> > with niggardly travel policies, you either suffered in economy class
> > or upgraded to business class on your own nickle).
> 
> So you would favour a scheme that would not introduce this overhead on
> travelling and meetings?

Ultimately the fundamental Flaw about ISO is that voting is done at
national level.  Which means you need to organize representatives at
multiple national bodies, and to become a member of an accredited
national body also costs money.  (At the time, IBM paid for me, and
they also paid for Akio who was on the Japan national body; the Linux
Foundation paid for Nick who represented the UK national body; and
there were others.)

This is the sort of thing which is much more easily handled by
multinational corporations.  If we wake up ISO, and someone
mischievous (like, say, Microsoft, or even Oracle) has the resources
to pack various national bodies, we may not be able to control the
voting, and God only knows what might come out of such a process.

(BTW, I think this is part of the whole business model of ISO
consultants; they try to hit up corporations and point out the huge
risks if the standards goes the wrong way, and for a mere $200/hours,
they could help make sure the right thing happens.... and then they
get to travel to exotic locations on corporate expense accounts.)

> 
> Yes, we can risk that. But then, if we have a POSIX Austin Group
> like organisation, having all experts in one group would counteract
> such a tendency. It has not happened in the Austin Group yet, and
> they have been functioning for more than 10 years. And the Austin
> group was formed because we have had similar troubles as you report
> for the LSB at ISO work.  So there already were people and politics and
> other sharks in the water for the POSIX ship. We got away with those
> troubles by more adequate organisation.

Even so, that only worked because companies were willing to pay for
the Austin Group's representatives.  I was on the Usenix board when it
decided that they could no longer afford to pay for a certain Austin
Group representative, and then when that person showed up to try to
hit up the Linux Foundation, Jim Zemlin, who is the executive
directory of the Linux Foundation asked me whether or not the LF
should fund this person (and at the time the LF was undergoing
significant financial pressures, and we ultimately said no).

The bottom line is that even an Austin Group style effort takes money
and resources, and it's hard to fund that on a pure volunteer basis.

> I would like an ISO standard that is exactly the same as the LSB standard,
> and I think I see ways to do this. The ways I foresee have been played
> out a handful of times before with good results (IMHO).

But what does an ISO standard get us that wouldn't be achieved if we
instead channelled that resources to produce a better, more
comrprehensive LSB standard?  Certainly after the OOXML fiasco, the
ISO stamp means a lot less than it ever did.  Do you really think it
would influence ISV's into adopting the LSB who wouldn't otherwise?
Personally, I find that hard to believe.

Regards,

					- Ted

P.S.  A significant issue is that there is a strong argument to be
made that since ISO has already standardized POSIX, it shouldn't be
trying to promulgate a second standard which is not in harmony with
POSIX.  We had to swat down an effort by the Sun engineer (in part by
making sure we had enough corporate votes by having engineers to
attend the INCITS meeting, which costs time and travel and INCITS
membership fees) to force the LSB standard as promulgated by ISO to be
changed so that it didn't conflict with an already standardized Posix
specification.

Granted, Sun and Solaris is a dead letter at this point, but who knows
from what other national body might arise someone who decides that it
would be proper for ISO to "add value"?  And do we really want to
spend all sorts of energy making sure we have representatives on all
of the right national bodies, and doing all of the vote counting so
things go the way we want?  Remember, in the ISO world, Trinidad has
as much voting power as all of the United States!


More information about the lsb-discuss mailing list