[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
GNUstep's Direction... Hopefully: GNUSTEP
GNUstep's Direction... Hopefully: GNUSTEP
Fri, 30 May 2003 15:56:21 -0700
Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.3) Gecko/20030312
I've been interested in GNUstep since I first stumbled upon it about a
year ago. I learned how to use the web on a NeXTcube, and thought
OPENSTEP was a lofty and worthwhile goal for the software industry. Very
happy to see the concept and goals of OPENSTEP continued in a libre form
of the OpenStep Specification as implemented by GNUste_The_API. I'd
really like to see the development take off, or at least grow, perhaps
with the successful development of LinuxSTEP or similar into
I am curious. Is there a GNUstep Steering Committe or similar? (Please
don't flame about how committees destroy software, I was just searching
for the most easily understood term). From my lurking in the mailing
lists, I am unable to discern if there is a group planning the direction
of the GNUstep Project, though that is implied by its being a GNU
Project, and Adam Fedor is listed as the Project Administrator.
I know there are a few parts to any OpenStep-Compliant systems equation:
the API, the frameworks, the libraries, and GUI. GNUstep appears to be
moving toward a fully integrated environment hosted on any OS that can
support it, as was OPENSTEP.
From looking at the two other main libre software integrated
environments, GNOME and KDE, they seem to have a validity due to the
appearance of a standardization of the environment through accountable
code maintainers and steering to keep a focused future direction.
Someone who doesn't dig into the GNUstep world wouldn't get this
impression of the project or environment.
Mainly... this is due to inaccessablity. The libre software that people
consider concrete have individuals that interface the public, and
control the direction of the project. This is what lends a sense of
accountability, confidence, dependability, future, justification,
quality, reliability, supportability, usability, and viability to open
sourced libre software. The perception is felt by contributors,
developers, users, and those that would support or make use of the
software for reasons _other_ than the virture, elegance, challenge, or
beauty of it. You could say the general public, including third-party
and commercial software developers, operating systems integrators, and
commercial operating systems venors.
If I happen to have insulted any GNUstep supporters/users/developers
with these comments, I apologize, for my intent is to point out that the
project seems nebulous and insubstantial in comparison to Apache,
Mozilla, PERL, GNOME, KDE, Linux, FreeBSD, Mac OS X (not a libre
project, but the closest currently developed and descendent
OpenStep-Compliant implementation, and the most visible topic of direct
comparison by many), for example. Properly collaboration between a
steering group and the designers both inside of the group and outside,
can be very benificial if they avoid stifling the growth of a project.
The best example is in the Hollywood Film Industry. Films have what are
called Continuity Experts. Too often they are not expert, and too often
they miss the things that directors and editors toss in that destroy the
illusion of "truth" or "reality" toward a film's possibility of being
taken seriously as something that could take place off the screen and in
the "real" world. NeXTSTEP and OPENSTEP had wonderful continuity. This
seems to be kept up in Mac OS X, and everybody knows that the
pre-OpenStep-Compliant Mac OSes were pinnacles of design continuity.
A structured software project lends toward the perception that the code
will remain viable and maintainable past the
interest/lifetime/hard-drive-crash/brilliance/vision of the original
architechts. So we have ASCII. Unicode. C. Smalltalk. Lisp. Objective-C.
UNIX. X-Windows. Linux. GNU/Linux. The BSDs. Mach. NeXTSTEP. OPENSTEP.
Mac OS X. What about GNUstep???
Not only would I _LOVE_ to replace Microsoft Windows (Windows 2000
Professional, in the case of the workstation I write this to you on via
Mozilla Mail) with a GNUstep implementation running on a _Reliable_ OS,
I'd love to see the win32 backend and implementation of the environment
flourish. I'd like to see a GNUSTEP environment whereby I could do
anything critical, like eMail, web browsing, document creation,
multimedia, etc. utilizing the beautiful(well, I am fascinated with the
NeXTSTEP GUI, it is like beautiful warm clay that reminds me of the
sheets of clay Cuneiform was developed and used with, but this is a
tangent =8-D ), user-friendly, cross-platform GNUstep applications.
This would prevent marketshare and (user-brainshare) from any
justification that development of GNUstep oriented software is a waste
or will not have a broad enough marketshare.
Now, JAVA is supposed to provide this software universality. It had the
roots to do it (Read: derivative of OPENSTEP), and it does, for what it
is used for, which is internet-oriented cross-platform software.
However, it primarily is used for internet-oriented software. The
desktop-oriented software (i.e. something _other_ than a JAVA applet in
a website transaction/interaction) for example, Limewire (if you like
it, great. I didn't) was horrible to my perception, it seemed very
kludgy, and with JAVA, there is no adherence to ANY generalized UI
Guidelines, so I never know WHAT to expect from an interface for a JAVA
app. OPENSTEP didn't have this problem. Neither does GNUstep. The
applications I have seen display that the designers have done an
excellent job of sticking to the OpenStep Spec UI Guidelines. *Big Pat
On The Back To All*
For GNUstep to continue OpenSteps vision of Universality, it must be
made to be universally accessable.
What was the benefit to the UNIX user when NextStep was introduced, then
extended as OpenStep? The benefit to John and Jane Doe Computer Users?
They didn't _need_ to know how to use anything but a mouse and a
keyboard because the GUI didn't impose any of the
software/driver/firmware/BIOS/architechture/hardware mechanics or
internals upon them. Basically, aside from the power of Rapid
Application Development and arguably the best Development Environment
yet devised, it made NeXTstations and NeXTcubes very powerful appliance
tools. It made powerful computational engines, and stable, reliable,
well-designed Operating Systems accessable to Everybody Who Didn't
Design It. Understanding HTML, HTML parsers, the JPEG
spec/implementation/file-format, how GIFs were static or animated image
formats, how the HTML engine actually rendered web pages didn't apply to
me when I was learning how to use the web on that 'Cube. The 'Cube was a
tool which provided an interface in the form of a web browser with which
I could access universally available information from anywhere on the
planet. Sorry, had to go with the Universality theme to prove a point.
For me, GNUstep remains inaccessable. :-(
The Windows Installer is a good step in the right direction. Just as a
Mac OS X Installer, or an installer for any other operating system that
can host GNUstep.
I understand that GNUstep is still under development, and is an
"incomplete Work-In-Progress." So is Mozilla. They release milestones
that anyone on the supported platforms can install and use, _without
knowing or even understanding_ how the parts fit together to work.
Installing GNUstep follows a certain order and pattern, irregardless of
the individual OS that will be hosting it, correct?
It would be extremely useful to have installers for all supported OS
hosts for GNUstep milestones. Why? I'm not approaching this from a
standpoint of laziness. I know a good deal more about computers and
software than my family and extended family do. I have no clue, even
from the manuals, how to install, from source, or package, non-Microsoft
Windows software. Haven't been able to figure it out, and haven't found
anything that has been able to spell it out in a way I am capable of
absorbing. If I can't figure it out, how's my
Mom/Dad/Siblings/Grandparent supposed to?
Please don't send me a "RTFM" flame. Or that "If personABC is incapable
of understanding softwareXYZ, then they shouldn't be using it". That
flies in the face of OpenStep's intent, and would make me wonder why you
are wasting your time developing something this good ONLY for the (not
my own words, just heard it from John and Jane Doe user MILLIONS of
times) "Geek Elite(TM)". OpenSteps intent was Universality, and they did
intend for Grandma to use the system. In fact, the OpenStep derivative
that is Mac OS X is used by Many a Grandma who isn't even afraid of it.
What's to become of GNUstep?
I'll be damned disappointed if it becomes a failed, by-the-wayside,
half-baked, Computer Geek/Hacker attempt/exercise/clone to model
"TrueRealSoftware(TM)" just for the sake of flexing their brains and
giving their wrists carpal tunnel syndrome.
Myself, my Mom, my Dad, my Sister, my Aunts (including Aunt Genea, who
has Down's Syndrome, and isn't a Computer Science Major), Uncles,
Cousins, Friends, Coworkers, Bosses, Strangers, Microsoft Windows users,
"OperatingSystemFillInTheBlank" users are also in the scope of GNUstep's
target audience, aren't we?
Now, I've done my part. I've been convinced by my experience with
NeXTSTEP & OPENSTEP to buy into GNUstep's promise and I've gotten
interested. I'd love for it to encapsulate my Windows 2000 Pro machine,
I'd love to be able to install any and all of the following:
GNUSTEP/GNU/Linux, GNUSTEP/(GNU?)/FreeBSD, or GNUSTEP/(GNU?)/Darwin, or
GNUSTEP/(((GNU?)CygWin?)MinGW?)/M$ Windows, or GNUSTEP/(GNU?)/Mac OS
Milestone releases with installers and facilities such as "Mozilla with
Talkback (to send errors to the developers/bugfixers)" for GNUstep would
only help grow the user base (by me, a coun't of 3 computers, in the
LEAST) and provide GNUstep Core and Application developers with quality
control data and a means to improve and expand GNUstep's
message/vision/implementation for software out into "TheRealWorld(TM)."
The current installation files (such as the Installation Guide) don't
even stick fingers up my nose to drag me, step by tedious step, how to
set-up facilities for obtaining, and setting up, _all_ of the necessary
items for a fully fleshed out GNUstep implementation capable of
supporting _all_ of the GNUstep applications currently available.
It sucks to be this interested in something, "SeeTheLight", and be
unable to "touch" it, except by way of the SimplyGNUstep demo CD (which
by the way is That project's concern technically) which is being
redeveloped from scratch, for one, and for another didn't give me the
option of saving work to my hdd (a la KNOPPIX), and be incapable of
evangalising it to others (especially those less computer-literate than
I) without being a hypocrite.
I'm a strong proponent of imagination, but I, for one, prefer GNUstep to
become a dream fleshed out as a miracle, rather than a fantasy which
evaporates as a mirage.
Questions, Comments, Rotten Produce?
|[Prev in Thread]
||[Next in Thread]|
- GNUstep's Direction... Hopefully: GNUSTEP,