The Grumpy Editor's GNOME 3 experience [LWN.net]

The Grumpy Editor's GNOME 3 experience

Posted Mar 15, 2011 23:54 UTC (Tue) by sramkrishna (subscriber, #72628)
In reply to: The Grumpy Editor's GNOME 3 experience by sfeam
Parent article: The Grumpy Editor's GNOME 3 experience

It's dead.. because nobody wants to work on a perfected tool. If it is in fact perfected, there is nothing else to do is there? Just sit around and move chess pieces around. We won't be keeping many volunteers around with that attitude.

to post comments

The Grumpy Editor's GNOME 3 experience

Posted Mar 16, 2011 0:13 UTC (Wed) by Tara_Li (guest, #26706) [Link] (3 responses)

Do you *NEED* as many volunteers? If it's down to just minor bug and security fixes, is that so out of line? Just churning to keep people busy is make-work, not real innovation.

The Grumpy Editor's GNOME 3 experience

Posted Mar 16, 2011 1:37 UTC (Wed) by sramkrishna (subscriber, #72628) [Link] (2 responses)

Your question doesn't make any sense. For some people, the desktop experience reached perfection when bash3 was released. Without the drive of "just works" you would not have had the kind of ease of use you get today.

I suspect distros would also be unhappy (you know all those companies that hire free software people?) if we just stopped doing things.

The Grumpy Editor's GNOME 3 experience

Posted Mar 16, 2011 11:26 UTC (Wed) by sorpigal (subscriber, #36106) [Link]

The key difference is that when bash3 was released the bash team didn't decide that bash4 needed to have incompatible syntax and require 3D acceleration in order to remain relevant. Instead bash4 builds on bash3 in a backwards compatible way, adding and not subtracting.

> I suspect distros would also be unhappy ... if we just stopped doing things.

How about solving new problems and not re-hashing old problems? A lack of feature density is a sign of immaturity, not a sign of usability.

The Grumpy Editor's GNOME 3 experience

Posted Mar 16, 2011 13:50 UTC (Wed) by dskoll (subscriber, #1630) [Link]

Without the drive of "just works" you would not have had the kind of ease of use you get today.

Umm? I do not believe computers today are significantly easier to use than they were 10 or 15 years ago. In fact, I find some of the modern desktop trends in GNOME and KDE making computers harder to use.

The Grumpy Editor's GNOME 3 experience

Posted Mar 16, 2011 0:24 UTC (Wed) by dgm (subscriber, #49227) [Link] (5 responses)

Volunteers for what? Having people reinvent the wheel just to keep them busy is possibly the worst idea ever. Think about it.

The Grumpy Editor's GNOME 3 experience

Posted Mar 16, 2011 5:35 UTC (Wed) by bronson (subscriber, #4806) [Link] (4 responses)

Volunteers to invent spokes, vulcanized rubber, and all sorts of other improvements. You're welcome to continue using your stone wheels as long as you'd like.

The Grumpy Editor's GNOME 3 experience

Posted Mar 16, 2011 9:56 UTC (Wed) by dgm (subscriber, #49227) [Link] (3 responses)

Great, now instead of inventing paper or penicillin, let's just reinvent our wheels every five years, just because. Next iteration we will do them squared, on the basis that:

* they are more compact and easy to store.
* people are great a handling square things, they do not roll away by mistake.
* they are more beautiful.
* they *are* more *beautiful*.
* Did I mention square is more beautiful?

Don't mention that old roads will not adapt very well to those new wheels, but hey, it's all in the name of progress.

The Grumpy Editor's GNOME 3 experience

Posted Mar 16, 2011 22:12 UTC (Wed) by bronson (subscriber, #4806) [Link] (2 responses)

How lame. I'm sorry I paid any attention to you.

The Grumpy Editor's GNOME 3 experience

Posted Mar 17, 2011 1:17 UTC (Thu) by dgm (subscriber, #49227) [Link] (1 responses)

No, lame is using the interests of volunteers to excuse a bad technical decision, that is, throwing out the window all experience gained with a currently working environment, all in the pursuit of some questionable usability nirvana.

Don't get me wrong. I'm all for improvement, but I seriously object to two things:
1. the mentality that once something is done, it's dead. Not. It's done. Most software projects should aim at being done some day, at least the parts others have to rely upon. That's what originated my initial post.
2. the way Gnome has chosen to release Shell. Make a new point zero release and rush a half baked idea with a half baked implementation. And make it mandatory! What's that familiar smell? Oh, yes! smells like KDE4 all over again! Why would it be so difficult to maintain the current, working, environment for the (millions!) of current users, and give the radically new stuff as an option for the adventurous?

But I guess they will make the same mistake again, who is going to test it if it's not mandatory, anyway? Excuse my lameness, but I have been through all this before, and it was not pretty.

All this would be moot, though, if Gnome tried to finish something. But can Gnome 2 be considered "done"? Nope, it was never written with that in mind. There are piles of bugs that will *never* get a fix, because in the minds of those developers Gnome 2 is not mature, but deprecated.

Of course, I may equally well be wrong. It happens to me all the time.

The Grumpy Editor's GNOME 3 experience

Posted Mar 17, 2011 3:23 UTC (Thu) by jcm (subscriber, #18262) [Link]

As usual, Jamie Zawinski had something to say about this years ago...

http://www.jwz.org/doc/cadt.html

His choice of language and attack vector aside, many people have written about this kind of problem, for a long time now. And I don't mean GNOME, I mean the general problem of never really getting to a finish state.

Jon.

The Grumpy Editor's GNOME 3 experience

Posted Mar 17, 2011 9:43 UTC (Thu) by nim-nim (subscriber, #34454) [Link] (8 responses)

The point is not to work on the shell indefinitely. The shell itself it utterly uninteresting by itself.

The point is to have more apps that perform useful tasks for users.

That won't happen if the platform is changed upside down every five years. That's why people are still writing TCL/TK and motif apps, and no sane isv wants to touch GTK stuff if it can avoid it.

The Grumpy Editor's GNOME 3 experience

Posted Mar 17, 2011 16:11 UTC (Thu) by jcm (subscriber, #18262) [Link]

Exactly right.

The Grumpy Editor's GNOME 3 experience

Posted Mar 17, 2011 17:34 UTC (Thu) by walters (subscriber, #7396) [Link]

We obviously don't have a lot of ISVs (in the sense you mean) - but what applications are you talking about?

The Grumpy Editor's GNOME 3 experience

Posted Mar 17, 2011 19:02 UTC (Thu) by jmorris42 (guest, #2203) [Link] (5 responses)

> That's why people are still writing TCL/TK

That is one reason I still use it. That and documentation.

Challenge: Point me to a source for hardcopy documentation to develop a GNOME application. Can't, can you. Now in a scripting language.

Closest I came was an online GTK+ 2.x in Python document that looked pretty good but GTK isn't GNOME. All those wonderful GNOME technologies that churn almost annually and the only way one can learn how to use them correctly is to look at the source of an existing app and HOPE it is using it correctly since odds are that author learned the same way.

So last time I needed to knock out a graphical app for Linux I grabbed my worn copy of Practical Programming in TCL and Tk and got r done.

The Grumpy Editor's GNOME 3 experience

Posted Mar 17, 2011 20:36 UTC (Thu) by Darkmere (subscriber, #53695) [Link] (2 responses)

http://www.amazon.com/Official-GNOME-Developers-Guide/dp/...

Is that good enough?

The Grumpy Editor's GNOME 3 experience

Posted Mar 17, 2011 22:48 UTC (Thu) by anselm (subscriber, #2796) [Link]

It came out in 2004. The latest review on Amazon is from 2007. Whether what is in that book is of any use with current GNOME is anybody's guess.

The Grumpy Editor's GNOME 3 experience

Posted Mar 19, 2011 22:44 UTC (Sat) by jmorris42 (guest, #2203) [Link]

Nope. As someone else has already noted it is old. It is also out of print. Even the web docs on www.gnome.org are bad. Describing a libraries under a banner noting it's deprecation. And look at the language bindings.

C? Duh.

C++ also looks well maintained.

Java? Tutorials "Coming Soon" for about five years now. "has been used to develop non-trivial applications" and "coverage level is reaching maturity" doesn't inspire a lot of confidence to jump in and find out what works and what doesn't while developing code intended for production.

Perl? "Our documentation isn't what we'd like it to be..." is truthful. A look around leads me to think most of the GNOME APIs are supported at some level.

Python? A lot of GNOME apps are written in Python so one would think there would be documentation out there.... one would be mistaken. All pointers in the FAQ are to information dated between 2001 and 2003 so considering how many APIs have been deprecated since then...

The Grumpy Editor's GNOME 3 experience

Posted Mar 19, 2011 23:47 UTC (Sat) by dbnichol (subscriber, #39622) [Link] (1 responses)

Sorry, but that's not a fair assessment. TCL/TK, like GTK+, is a _toolkit_ not a desktop environment. I can't think of any desktop environment based on TK that you could still write a program for today.

Tcl/Tk

Posted Mar 21, 2011 15:36 UTC (Mon) by dskoll (subscriber, #1630) [Link]

I can't think of any desktop environment based on TK that you could still write a program for today.

I don't know of any desktop environments based on Tk, period. However, I love Tcl/Tk and I think the GNOME developers could learn a lot by studying it. Here's why:

Tcl/Tk has extensive and well-written documentation in good old UNIX troff format. The API is completely documented both at the Tcl level and at the C integration level.
The C code is very clean and well-documented.
Tcl/Tk has continued to evolve over the decades, but retains the essence of what Tcl and Tk are. New features are carefully considered and added, but only when they fit in with the existing design philosophy. There's never been a "the entire Universe has changed" release of Tcl/Tk. GNOME could do well to study this last point.