Nicholas Petreley Slams Gnome

FreeLinux writes "Mainstream computer rag ComputerWorld, has posted a review of Gnome 2.6 by Nicholas Petreley. This opinion piece review, titled Living Down to a Low Standard, positively lambastes Gnome 2.6 over the new spatial Nautilus and Gnome's design choices. The review is quite the opposite to a previously reported review from PCWorld, last month. While this latest review is bound to be a polarizing and heavily debated issue (read flamebait), it is important in that this review will be seen by so many mainstream readers and corporate types who may have been considering Gnome."

Comments

Vicious

[two_stripe]

Why doesnt he pick on someone his own size?
Those poor gnomes. :(

Re:Vicious

[ThrasherTT]

Shouldn't your sig be:

Dont drink and derive. Alcohol and calculus dont integrate!

Re:Vicious

[gujo-odori]

I made that migration, too.

My path was FVWM95->Afterstep->Window Maker->Gnome->KDE3.0.

I liked Gnome better than the other things that were around at the time, and steered away from KDE partly because of the licensing issue with Qt that existed at the time, and partly because KDE prior to 3.0 was just so generally ugly and amateurish-looking that I couldn't stand to look at it. Who designed those icons? Blech!

Then KDE 3.0 came out. I tried it out of curiousity and found that it was *worlds* ahead of Gnome. Gnome up to and including 2.4 was nowhere near catching up. Whatever chance they might have had was buried by KDE 3.2.

I will take a look at Gnome 2.6, just to see how they've done, but I have my doubts. I read some of the ideas that were going into the design for Gnome 2.6 and all I could think was "That sounds really stupid."

So, while his review of Gnome 2.6 (or more accurately, of Nautilus in Gnome 2.6) may be written in rather inflamatory language, it should not be dismissed outright as being crap. Even if it's not as bad as he says, the idea of having every double-click open a new window and be so difficult to override is criminally stupid.

People tend to dis KDE by claiming it works too much like the Windows UI, but you can customize that any way you want, and so I do. It's something I like a lot about KDE.

If it has behaviors that are much like Windows by default, so what? That can help new users make the transition. Is that a bad thing? We also need to keep in mind that Microsoft does know a thing or two about UI design. Unlike most open source projects, MS does have UI specialists. Lots of them.

I have a laundry list of things I hate about Windows, but only two things on that list are UI-related:

1) You can't customize the UI much. It just works the way it works;

2) This is the bigger one: you get only one desktop. On my notebook, I have 8 virtual desktops. On my desktop machine, I have 10. This allows me to organize my work by assigning different types of tasks to different desktops, and I have a set pattern of where I put different types of things. Ctrl-F[1-10] takes me to the desktop I need. I cannot do this with Windows and it's a major PITA.

That's it. Those are the only two points of MS UI design that bother me, and the first one is pretry minor, really. If KDE copies some of their ideas from Windows (but don't forget that KDE has a lot of capabilities that the Windows GUI does not have), it could be that KDE developers just know how to recognize a good decision when they see it.

Don't panic... it's not that bad

[Allen Zadr]

Sadly, the article brings up some very good points, albeit in a very inflammatory way.

The most damaging part of the "review" is that it says nothing aboout Gnome as a whole. It's just a rant about this user's opinion about how Nautilus was designed ( changed) to work in 2.6.

This sort of rant, if done constructively could certainly help the developers make better choices, but to put it directly to mass media as a review just sucks.

Well, as a Pointy Haired type myself, I can assure you, these mags hit the coffee table in the lobby - and very few people actually read the articles... However, if this review makes the front page, Gnome is toast.

Re:Don't panic... it's not that bad

[andyross]

Obviously its just my opinion, but the way Nautilus browses the file system is backwards compared to 2.4, and the removal of the tree in a left hand pane was a very bad decision.

The tree has not been removed! Right click on something and select "browse" and you have your old Nautilus back. Not enough? Make a launcher on your desktop and have it run "nautilus --browser". Still not enough? Put it in a menu (FC2 already does). Still not enough? Delete all the spatial icons from your desktop and you will never see it again.

Good grief. You would think that "expert users" who can handle the complexities of a browser-based file interface might be able to, y'know, configure their desktops before whining about it in public.

Re:Don't panic... it's not that bad

[spectre_240sx]

"Seems like people are always complaining that GNOME and KDE don't innovate, and when they do everybody gets upset."

Did you miss the part that OS/2, Win95 and early Mac OS versions worked this way too? How is it inovation if it's already been done before?

The whole idea of the spacial file management system is to bring the metaphor of files and folders closer to what it is in the real world. However, that comes at a loss in usability, and there's no reason to try to do this if people are already comfortable with the way that file managers work at the moment.

Re:Don't panic... it's not that bad

[cweditor]

Glad to see so much interest in one of our columns :-)

For the record, this column appeared on page 30 of Computerworld's print edition this week, on the last page of our Technology section. Nicholas Petreley is one of several different columnists who rotate writing for that Tech opinion page. (There is an opinion column on the last page of the print publication's Management section as well.)

Our print readers have seen quite a bit of coverage of Linux and open source in addition to this column. Two weeks ago, for example, one of our cover features was A Sunny Forecast For Open Source, about how weather.com has cut IT costs by one-third after moving away from proprietary software and hardware. It was one of the most-read stories on our site that week.

Sharon Machlis
Online Managing Editor
Computerworld

Re:Don't panic... it's not that bad

[IntlHarvester]

Another example would be "Team OS/2" -- who had some whackos that would actually make death threats against computer journalists. And they wondered why nobody covered OS/2...

(Even today, you hear people talk about "Steve Bartko", some MS employee that posted a handful of times on a compuserve board 12 years ago. Sorta a legend in the OS/2 and Anti-Microsoft Zealot community.)

Back in the 1990s, there was a real effort among the Linux Community to encourage "good advocacy". Flamers were pointed to a Advocacy HOWTO document and there was a real effort to keep discourse polite in the Usenet tradition. It seemed like they might have learned the lesson of Team OS/2 & Macinsitas. However, now days with a proliferation of web forums, and Linux accumulating all of the OS/2, Amiga, and M$ Hater wackos, things have degenerated.

The thing to realize, in the real world of IT Dept Politics, zealous advocacy often hurts one's cause more than it helps. People tend to think "This guy is not objective, I don't like him, therefore I disagree with whatever he says." Yes, that's not logical like Mr Spock, but its how the real world works outside of internet boards. God forbid people have to work with a lot of you folks.

Don't RTFA!

[Anonymous Coward]

### Warning! ###
### CATCHPHRASE ALERT ###

Nicholas Petreley uses the tired term "paradigm shift" in his article!
[not that anyone will actually read the article...]

### CATCHPHRASE ALERT ###
### Warning! ###

Unbiased

[Embedded Geek]

Mainstream computer rag ComputerWorld...

I'm glad the author of the slashdot story managed to keep his biases concealed until the third word of the story. If the article had praised Gnome, however, why do I suspect we'd be hearing about "Esteemed technical journal ComputerWorld..."

Re:Unbiased

[Dielectric]

No, you wouldn't, because there really isn't an esteemed technical journal for mainstream computing. The IEEE puts out some good stuff, but no one outside of the engineering community reads it.

I go to the cockfights when I need to make a decision on this sort of thing. I label one chicken Choice A and the other chicken is Choice B, and that has pretty much worked for me. This explains why I'm using a C-64 right now. That was one tough chicken.

Reply from one of the Ars Technica crew

[unmadindu]

Jorge Castro, one of the Ars Technica writers has written a very nice article refutng Petreley's claims at his site.

I dislike Gnome

[Uma Thurman]

Because Windowmaker is all I want. But Free Software gives us a bountiful array of choices. I don't get why Nick P. needs to run down someone else's desktop.

He needs to mind his own business and write about something he DOES like rather than running down something that he doesn't like.

Nothing new here

[Mars Ultor]

Not too surprising really - here's an earlier article when GNOME 2.2 was still hot. From the article:

KDE is delivering a better version of what GNOME's goal has apparently morphed into: becoming a great component framework that you can write to in multiple languages. Nicholas Petreley rebuffs the common GNOME battle slogans and explains why the window-manager's name needs reworking.

Other than boosting ad views, I'm not sure what continuing a KDE/GNOME flamewar here on /. really contributes to open discussion (pardon the pun)

and it's right

[Anonymous Coward]

I am all for simplification, but there is no reason to go back to kinder and ABC wooden blocks.

The biggest argument against spatial navigation, as produced by gnome 2.6, is that it requires the user to learn TWO different styles of navigation: one for their browser and one for their files.

That is NOT simplification. And they didn't ask the community, and they are going against the gain of EVERY other OS.

If spatial is going to pay dividends when "database" filesystems arrive.... introduce spacial THEN. And even then, have it as an option. Besides won't a database file-system be based on searches? So won't we need "back" and "forward" buttons???????????

I am not going to swear here, but I am MAJORLY pissed at gnome. I am on 2.4 atm because of it. It is at worst elitist insanity, at best a poorly executed jump of the gun.

Re:Simple Solution.

[stratjakt]

How dare he criticize something as trivial as "how you use the fucking thing".

Most people find that clicking-opens-a-new-window behaviour annoying. It makes browsing around your directory as annoying as closing popup ads - its the same experience, pretty much. Your screen clogs with shit you dont wanna see.

He makes the point that no modern desktop OS does that, and for a reason.

Why is everyone so defensive? It's a perfectly valid criticism. It makes the desktop frustrating to the point of unusable for many folks.

The guy has a point...

[ave19]

I agree with this article. (I even read it.) I want to give it mod points. Can we do that?

I used to configure the crap out of gnome, making it do all kinds of weird stuff I liked. Then, version by version, my toys were taken away. I don't get it. If the toys made it unstable, why not fix them? What ever happened to the idea of "advanced" vs. "novice" settings for a UI? Every version that comes out has LESS functionality than the one before, railroading me into a certain way of interacting with a desktop.

In Soviet Russia, the desktop clicks on YOU!

Make it easy by default, but don't take away our toys and call it progress.

-ave

Yeha, that's *real* usability

[kikta]

The fact that linuxquestions.org or a web search is needed to answer such a question should be your first clue that something is seriously amiss.

Re:I don't use Linux

[stratjakt]

I'm glad you noticed that it wasnt much of a review.

That's because it's NOT A FUCKING REVIEW

Look:

Living Down to a Low Standard
Opinion by Nicholas Petreley

So while the zealots line up to flame him for his "unprofessional review", keep in mind it's an OPINION, and he can have whatever opinion he wants.

Let's talk about slashdots "unprofessional article" that criticizes this guy for being an "unprofessional reviewer" for merely voicing an opinion, which happens to be that $YOUR_PET_PROJECT sucks.

GNOME is the example, the point much bigger

[miketo]

The article spends most of its time on Nautilus, and I'm not going to rehash the debates here. But he makes a valid point, one that I've wrestled with since Day One of Linux:

Engineers design programs that work for them, not for end users.

I've seen this time and again during my work as a software product manager. Everything from base functionality to key UI choices are made by the development team based on what they find useful, or what they think will be useful. It is a very, very rare team that actually conducts any workflow analysis or UI usability studies during the design phase. And, once it's coded, it will cling like a limpet to a rock, difficult if not impossible to change.

I know enough about my own predispositions and biases to know that my judgment about what's best for me isn't always what's best for everyone. While both Microsoft and Apple make poor function / UI choices, with Linux the problem is magnified because each piece is built by a different design team using a different methodology.

Server-side and admin people aren't bothered by this, but your average end user is easily frustrated by applications that don't behave in an expected way, or don't have settings that can be easily changed to adapt to the user. If you give your software to a reasonably knowledgable end user, watch the interaction with your product. Don't argue, or explain why the actions aren't correct. Take notes, and figure out a way to accommodate the user. Don't use the mantra of "Read the man pages, foo!" That only leads to reviews like Petreley's, and the ensuing does not / does too debates on /.

"There is nothing more permanent than a temporary solution."

--Mike

Re:No big surprise

[Minna Kirai]

Here is a decent rebuttal.

It's a poor rebuttal. Aside from the obligatory accusation of bias, it mainly focuses on attacking Petrely's understanding of "spatial browser".

Problem: he never used the word "spatial".

Also, when Petrley complains that you need to edit GConf to change the default behavior, instead of finding a prominent checkbox, Jorge (a) lists 3 ways to change the current behavior, and (b) attack's Petrely's technical understanding of GConf. He says that aside from GUI, GConf is nothing like the Windows registry. Well guess what? From the end-user's perspective, the GUI all that matters! If you need to use Registry or GConf to alter a setting, then it's impossible to call that setting easy-to-change.

The oped comes down to a very simple position: when a piece of software first gets a radically different, optional interaction mode, common-sense dictates that the new mode should be OFF by default. To do otherwise will scare users who were accustomed to the existing behavior. (Or at minimum, the checkbox to "Act like the older version" should be prominently placed, such as an option at install)

PS. An additional funny part is that both Nick and Jorge manage to mistate what the motiviation for Gnome was: Nick says "freedom from Windows", Jorge says "kickass desktop"... when in reality it was meant for "freedom from KDE" (as is well-documented historically)

Nobody gives a good explanation

[Anonymous Coward]

The thing with "spatial navigation" is that it is only effective with a small group of commonly used folders. Back 20 years ago when you were lucky to fit 800k on your Mac floppy, you could justify the spatial finder as making it much easier to navigate your few folders.

The problem is that this thing doesn't scale! As a pathological example, say I have a 800GB volume with 400,000 files (mostly photos I've taken as a professional photgrapher) spread out over 3,000 directories. I'm not going to memorize the screen location of each of those 1000s of photo shoots. Dragging my mouse back and forth across my 24" monitor half-a-dozen times to get to the photo shoot I'm looking for is almost the worst scheme I can imagine. The Windows Explorer 2-paned tree model (as opposed to the MacOS tree where there's only 1 pane) is about the most efficient I can imagine for this scenario.

Now that disks are 1,000,000 times bigger than they were 20 years ago, why is somebody trying to introduce the metaphor that was only appropriate for use back then? Granted, it works fine if a novice user has maybe a dozen commonly used folders, but beyond that it is unwieldy.

I think the best solution is perhaps to use the "spatial" metaphor only for folders created on the user's "desktop". That way your ad hoc folders work the way your real desktop does (spatially), while proper hierarchies are still navigable the way they were intended -- as a tree.

aQazaQa