On Dec 24, 2013, at 2:57 PM, Doc O'Leary <
droleary@7usenet2013.subsume.com> wrote:
In article <mailman.10195.1387832754.10748.discuss-gnustep@gnu.org>,
Gregory Casamento <greg.casamento@gmail.com> wrote:
I believe part of the problem might be that no one takes the opportunity to
summarize some of these discussions at the end (i.e. take action items from
them and them figure out a plan of action on them). I¹ve worked for many
organizations and where this was the practice and it worked quite well.
Perhaps something similar is needed on this project. For each discussion,
if they want it to be acted on, they should summarize all of the points and
send that to the leads so that they can be considered.
I see no reason to wait until an "end", especially when such discussions
can go on at length and branch out into many related issues. And a time
limit needs to be set on a response from leads. Whether the answer is
yes or no, it is worse for moral for things to be ignored with no
explanation.
Indeed. This is why this practice hasn’t been put in place and, even if it where, it would be difficult to enforce.
The problem is that it is
unclear who should do this. It¹s not possible for one person to do it since
they would, undoubtedly, need to digest many different discussions,
understand them and then come up with action items. It might work if, when
someone knows
Such is the burden of leadership; if you can't afford to hire a janitor,
you might have to clean the toilets yourself. But I think it's not as
bad as that. I think that anybody and everybody would be willing to
make the case to management, just so long as they know management is
going to listen. Every time you use the word "consideration", it grates
on me because I've worked in too many places where such sentiments meant
a way to ignore making decisions.
Not here. The problem is that, over the years, the leads of this project have gotten some pretty inane and idiotic suggestions from people with little or no experience and having no bearing in reality whatsoever. Myself and those directly responsible for keeping the libraries and applications on this project in working order need to consider the impact of any changes which come our way.
The greatest power of any maintainer of any open source project is the power to say “No” and, while it must be exercised judiciously and sparingly it must be exercised to prevent crappy changes from getting into the codebase. So when I use the word “consideration” it is not code for “We’re going to ignore you” it’s sincerely meant to say that the leads will look over your changes or suggestions and consider if it’s worthwhile or not.
The track record of this project at accepting patches speaks for itself. We have adopted over 90% of the patches sent our way because most of them were worthwhile. As a group we’re very accepting of changes if they help make GNUstep better.
Quite often when I have tried to set goals in the past and shown the
³resolve² you speak of I have been alone on my endeavors. This is not
something I¹m blaming anyone for, but this seems to be the nature of the
beast. People need to be interested in achieving the goals that are
determined from any given discussion.
If you don't want to go it alone, the burden is on you to communicate
your goals in such a way that other people agree with them. And if you
send the right message, *new* people can express that interest. How
different might the Kickstarter campaign have been if it were directed
outward instead of inward?
I’m unaware of how the kickstarter was directed inward. It was posted on a public crowd funding site and was publicly announced on twitter, Facebook, slashdot and a few other public portals. The contributors to the campaign were not GNUstep contributors, they were the general public. People out there who believe in GNUstep. In spite of it’s failure, it raised awareness.
I am aware that the burden of the failure of the kickstarter rests solely with myself. I made a few mistakes in setting it up and considering how its goals should have been structured. It was, nevertheless, a learning experience for me and an overall positive experience since it let the world know we are still here.
Actually the reason I believe we should work on a UIKit implementation is
because UIKit is the most widely used ObjC API at this point in time.
iPhones and iPads far outnumber Macs (per Apple¹s own filings with the SEC).
Building a UIKit implementation would bring a large number of developers to
GNUstep and could serve to spark further interest in GNUstep¹s Cocoa
implementation.
And that's the flat-out reality of it. All that needs to be done,
really, it adding "hot air" to the right places that either fit it into
GNUstep's current goals, or alter the goals so they fit the reality. As
I have said, I still think GNUstep has a *lot* of changes to make
organizationally if it wants to be attractive to people who mostly use
Macs to mostly develop iOS apps. No doubt such changes will be
unpopular, but in should be obvious that GNUstep as a whole benefits in
the long term.
I’m not sure what organizational changes can be made. It’s not that I’m not open to making them, it’s just that I am not sure what they should be. I believe the options here should be discussed. I am open to whatever serves this project best as it has always been a labor of love for me.
There are significant differences in how UI and NS classed behave. UI based
classes expect touch events and gestures whereas NS based gui classes are
looking for mouse movements, clicks and keyboard events. There are method
differences, delegate differences and so on. There are similarities in some
ways, but the differences are significant enough to make it necessary to come
up with UIKit as a separate GUI implementation.
I'm still not convinced. Abstractly, event handling is event handling.
It certainly might be easier or cleaner to start with a fresh code base,
but I wonder if that actually serves the bigger picture. I mean,
fundamentally, if GNUstep is going to allow both UI- and NS- classes to
be used alongside each other, there *must* be some commonality that
merges the two. Whether that's done in a base class or tacked on as
categories or via proxies or whatever, I don't see the goal as just
duplicating what Apple has done.
Have you ever done any iOS programming? If you have then it should be very obvious to you what the differences are. If you simply look at the documentation for UIView vs. NSView you can see the differences there.
Until you do something productive. You're only hot air to me.
Nice way to express seeing value in discussions.
What I mean is.. while I see value in what you¹re saying, on an open source
project it is a meritocracy.
I maintain the problem remains how you assign merit to things.
Merit is assigned based on one’s personal achievement and contribution to the project. The old saying is code talks, bullshit walks. ;)
As you can see, the same criteria for merit is used with the Linux kernel which was used as a basis of comparison earlier in this thread.
Like it or not many people on this project who
have been committing for years will view what you say as nothing more than
hot air unless it is obvious that you are invested in the project in some
way.
And they are *wrong*. To the point of having an unscientific method.
If you're building a perpetual motion machine, you are just plain wrong
to ignore the views of people who aren't "invested" in your delusion.
I’m wondering why you keep emphasizing science. GNUstep is not a scientific endeavor, no software project ever is. They are *engineering* endeavors. What we are guilty of on this project is failing to follow some *engineering* best practices. Perhaps we’re talking about the same things, but calling them something different, but I’m not sure.
To date you are not invested in any way other than to come here and
tell us how wrong we are.
And you are *wrong* to demand that I invest anything more. Until you
fix that disfunction, you will *continue* to be wrong. And you will
continue to drive away people who would otherwise be interested in the
types of things you're doing.
You are correct in the sense that you don’t necessarily have to contribute code in order to be correct about pointing out faults in the structure of something in the same way that I don’t have to be an architect to know that there’s a problem when I see a crack in the foundation of a building. However, you would be in a better position to help if you were more invested.
While you may be right on some points, it is
difficult to listen to you or to take you seriously when you have nothing
invested in this project personally. Sorry, but that is how I see it and how
many others will as well.
Change your thinking.
The easiest act by far is to walk away. You¹re correct. The BEST act is to
put your money where your mouth is, find something that you believe is
interesting and champion it¹s improvement.
That's why I'm still here. The burden is on you to realize there *is*
money where my mouth is.
I took the initiative years ago to take over Gorm and make it functional. I
championed that cause, and continue to do so, and it made GNUstep a better
project. Gorm is now a reliable and functional GUI builder. In doing so I
also took the initiative to flesh out many of the classes in GUI as well as
the general mechanisms which are used to load gui model files (gorm, nib,
model and xib). Because I took responsibility for these things it improved
the project.
Sure, but that doesn't that all improvements come in the form of code.
It doesn't even mean that the *optimum* priority should be code. Anyone
experienced enough with development projects knows how much necessary
effort goes in both before and after the coding is done. Hell, one look
at the Kickstarter shine-jobs that get millions in funding will tell you
that the ability to get the job done hardly even registers as a blip.
There have only been a very few projects on Kickstarted that have garnered that much attention. GNUstep should have been one of them. The problem is that only games and things people are interested in buying do well on such sites. You have to have a message and a product that gels with people in a way that they can see the tangible benefit of it. GNUstep was a hard sell on KS because it is just a set of libraries. Even if we were a porting environment for iOS (ala Apportable) it would have been a hard sell because such a thing has a limited set of potential users and many of those users may not be on something like kickstarter and be willing to contribute.
So, I believe that the perception
you have that the project is ³walking in a random direction² is just thatŠ a
perception.
And that *is* my point. It's not just my perception, but the perception
of most people who are on the outside looking in. You have a great
project that hardly anybody uses, and it has been that way since at
least 1996. If you want to change that, things will have to change.
Things other than code.
Running any open source project is like herding cats because people are self-determining. All I can do is set goals and break them down as you said once before. It is up to people other than myself to join me in what I feel is important as lead. Many times, in spite of knowing I was right I have been very much on my own.
My point is, quite simply, help us by becoming
part of the team. It is not up to us to try to be good enough for you or to
live up to your standards.
It kinda is. I, and people in general, can volunteer to be part of all
kinds of teams; more than you can shake a stick at in open source these
days. If what you're projecting is needlessly hostile (and it is), it
should be no surprise that I'm not eager to jump in to be part of that
kind of team. I like the technology, and I like the goals, but you *do*
still have a ways to go to be "good enough". Once you can frankly admit
that to yourself and make meaningful changes (based on nothing but hot
air), then you'll be getting somewhere.
Not really. Everything else we’ve discussed previously notwithstanding I have always been of the opinion that it is easier to change something from the inside out than from the outside in.
By saying the above I am not discounting any of your statements regarding
what you believe should be done on GNUstep. I¹m saying, quite simply, put
your money where your mouth is and help.
That you don't see the contradiction is telling. Let me know if that
ever changes.
I have been trying to determine what the larger community wants from GNUstep
for a very long time.
Why? There is ample evidence to determine which directions should be
given the highest priorities, both in the long and short terms.
Some people want a full
desktop, some people want a UIKit clone, some people want just a plugin that
will allow them to compile for windows (al la Cocotron). Some people don¹t
like the license LGPL. Some people don¹t like it because it¹s associated
with GNU or because they hate RMS. There are a myriad of sensical and
non-sensical reasons why people, even developers, act the way that they do.
I have spent a long time trying to please everyone and ended up pleasing no
one.
Don't bother trying to please wrongheaded people. It's as simple as
that. Face the reality of things, and structure the project in a way
that satisfies that. I tried to do that in a top-down fashion by
kicking out a slogan. You can either break things down in that way, or
go with a bottom-up approach. Whatever you do, you need to have a
meaningful message that is easy to communicate. From that position, it
is easy to address all the people pulling in different directions.
Indeed.
Instead of continuing to criticize, please make this discussion productive
and give some constructive criticism where you see the problems.
I have. The problem remains that nobody wants to go anywhere with it.
It's all so much "show me some code" or "no negativity" nonsense. When
you start saying "yes" more (or, hell, even "no") to the things that I
propose, and give good reason for it, we'll be making progress. Right
now, you're still crapping out a lot of "I see no value in that”.
I don’t know what you mean by “nobody wants to go anywhere with it.” I think you may be surprised. It’s always a mistake to guess what other people are thinking.
You¹re not just stating things directly, you¹re being condescending about it
and that¹s what is sticking in many people¹s throat.
It should. Because they are wrong. I don't think it's condescending to
point out reality. Certainly not as condescending as it is to deny
reality.
There's no denial. Many if the topics you've talked about have been discussed before.
--
iPhone apps that matter: http://appstore.subsume.com/
My personal UDP list: 127.0.0.1, localhost, googlegroups.com, theremailer.net,
and probably your server, too.
_______________________________________________
Discuss-gnustep mailing list
Discuss-gnustep@gnu.org
https://lists.gnu.org/mailman/listinfo/discuss-gnustep
GC