[lsb-discuss] Thinking about future LSB features

Dallman, John john.dallman at siemens.com
Wed Feb 18 04:17:41 PST 2009


Jiri Dluhos wrote:

> Wouldn't the simplest solution be to simply link the Motif library
> statically into the application, or bundle it with the executable? I
> think it is not so big...

Not quite that easy, since there are multiple shared libraries in 
the app that use Motif. One of the reasons for this is optional-extra
features. There's also the matter of Motif not being LSB-compliant,
although that's fixable in theory, and additional libraries that Motif
uses.

Hum ... Back in the early days, there was a focus on keeping LSB 
monolithic, so that if a distro supported a version of LSB, there was
no further complexity. Given that new we have optional features in LSB,
is there any scope for Motif being another optional feature, so that 
the distros for whom it isn't free-enough software can simply not
support that feature? 

> I understand it's lots of annoying work, but man-decades are, IMO, way
> overstated. When considering that the graphical interface is only
> a part of the whole code, I have problems imagining a software project
> so huge to justify man-decades; IMHO that would suffice to rewrite the
> whole UNIX from scratch.

This is a really, really big app. It started development in the 
seventies. The 64-bit Linux version is just over 1GB of shared 
libraries. 

> A big expensive CAD programs with a Motif interface? Don't the
> customers consider the GUI, well, a bit dated? Although Motif might be
> excellent in its time, the current user interfaces are much better not
> only in look, but also in comfort and usability.

This is not a /basic/ Motif user interface. It's grown quite a lot
of elaboration over the years. The customers don't seem to consider 
it obsolete or difficult to use. There is plenty of movement away 
from RISC UNIX platforms to PCs at present, simply because PCs are
fast and cheap. Most of that movement is onto Windows, but we don't
want to become a Windows monoculture. Platforms come and go; VMS
was an important platform from the mid-eighties to the mid-nineties, 
for example. 

Robert Schweikert wrote:

> In this market changes are very slow and a vendor will not gain new 
> business by converting to a new toolkit. If given the choice between a

> conversion of the interface to a more modern look and feel or new 
> features customers would most likely want the new features.

Yup. 

-- 
John Dallman
Parasolid Porting Engineer

Siemens PLM Software
46 Regent Street, Cambridge, CB2 1DP
United Kingdom
Tel: +44-1223-371554
john.dallman at siemens.com
www.siemens.com/plm

> -----Original Message-----
> From: lsb-discuss-bounces at lists.linux-foundation.org [mailto:lsb-
> discuss-bounces at lists.linux-foundation.org] On Behalf Of Jiri Dluhos
> Sent: Friday, February 13, 2009 7:09 PM
> To: lsb-discuss at lists.linux-foundation.org
> Subject: Re: [lsb-discuss] Thinking about future LSB features
> 
> Hi everyone,
> 
> Just my 2 cents of thoughts... please don't take me too seriously...
> 
> On Wed, 11 Feb 2009 19:13:05 +0100
> "Dallman, John" <john.dallman at siemens.com> wrote:
> > > What would make LSB more useful to you?
> >
> > The lack of Motif support is what kills it for most of my employers'
> > apps. I understand the reasons, but while there is no Motif, many
> > people don't see any point in doing the work to be otherwise LSB-
> > compliant.
> 
> Wouldn't the simplest solution be to simply link the Motif library
> statically into the application, or bundle it with the executable? I
> think it is not so big...
> 
> > And they really don't like the idea of switching to a
> > Motif substitute, unless it claims to be a complete replacement, up
> > front, with someone who'll fix deficiencies, and they don't see an
> > adequate gain in re-writing huge amounts of complex GUI in Qt, or
> > something else. It would be man-decades of work, at a conservative
> > estimate.
> 
> I understand it's lots of annoying work, but man-decades are, IMO, way
> overstated. When considering that the graphical interface is only
> a part of the whole code, I have problems imagining a software project
> so huge to justify man-decades; IMHO that would suffice to rewrite the
> whole UNIX from scratch.
> 
> > No, these people don't have any ideological commitment to Linux.
> > To them it is just another UNIX platform. So far, customers only
> > seem to want the products (big expensive CAD and PDM software)
> 
> A big expensive CAD programs with a Motif interface? Don't the
> customers consider the GUI, well, a bit dated? Although Motif might be
> excellent in its time, the current user interfaces are much better not
> only in look, but also in comfort and usability.
> 
> Best regards,
> 
>     Jiri Dluhos
> _______________________________________________
> lsb-discuss mailing list
> lsb-discuss at lists.linux-foundation.org
> https://lists.linux-foundation.org/mailman/listinfo/lsb-discuss


More information about the lsb-discuss mailing list