Gnome3 and Unity desktops compared BIG PICTURE.

[bsniadajewski]

Re: Gnome3 and Unity desktops compared BIG PICTURE.

I also got the soon-to-be latest Ubuntu going just to try out Unity. It still feels somewhat awkward compared to KDE, Gnome 2.*, or even Windows, but is very responsive and smooth. I like having the Ubuntu button to the Launcher (maybe soon allowing the Launcher to be moved in later iterations) and, to me, a little more configurable than the GNOME Shell (though I haven't tried the 3.2 version of GS). I still prefer KDE, but Unity is starting to get up there in usablility and good looks, IMHO.

[woodsmoke]

Re: Gnome3 and Unity desktops compared BIG PICTURE.

great comments folks!

It would be really great if others would offer their opinions also!

woodsmoke

[bra|10n]

Re: Gnome3 and Unity desktops compared BIG PICTURE.

I think the current mood and state of play in Linux gifts the likes of e17 an enormous opportunity to win over a huge number of adopters.
I recently tried a live version of Macpup and I was reasonably impressed with it

I was a little lost on the desktop being unfamiliar with the handful of apps it included by default, but I thought to myself at the time if anybody was to take a leaf from corenominal (#!) and include a setup script or even a text document based 'how-to' for new users it would go a long way to winning them over.

I find that while I can install and modify the e17 desktop to a certain extent, I remain concerned about finding suitable apps to suit my needs, questions regarding available repos, hardware etc.
It seemed to me to be a lot of work for destination not exactly known.

Yet for those who accept their Gnome2 days are numbered and are otherwise unwilling to embrace either GS or Unity, the catch phrase of such distros wanting to catch their attention could be something like,

"Want the features of traditional Gnome, and more? Click here."
Just some thoughts...

[woodsmoke]

Re: Gnome3 and Unity desktops compared BIG PICTURE.

Hi bra|10n

I think you last line up there WILL be used by a lot of distros!

woodsmoke

[bra|10n]

Re: Gnome3 and Unity desktops compared BIG PICTURE.

Perhaps, but from what I see it maybe only enlightenment that could really provide it.

[GreyGeek]

Re: Gnome3 and Unity desktops compared BIG PICTURE.

This is a general comment about Unity, Gnome, KDE and Mono. A "BIG PICTURE", if you will.

Canonical (Shuttlesworth) has said from the beginning that Unity was designed to be a "one size fits all" desktop that would run on a desktop or (eventually) a smartphone, and all devices in between. The target user is not you or me, it is the vast majority of "device" (mainly the smaller devices) users, whom we affectionately refer to as "Joe and Sally Sixpack".

In the FOSS community the mantra is "release early, release often". The reason is because the average GPL project is a labor of love, not cash or a proprietary EULA, and thus cannot afford to pay for beta testers. The purpose, of course, has always been to give feedback (bugs and suggestions) to the developers in order to improve the application. This mantra has worked wonderfully for 20 years and continues to do so. What hasn't worked as well in recent times because of the MASS INFLUX of Windows users into the FOSS community is the education of the new Linux users into the quid quo pro relationship between the availability of high quality free software and their willingness to post bug reports at bugzilla sites as apposed to posting whiny rants around the web.

The KDE desktop is a classic example. The versions between 4.0 and 4.2 were in massive flux and required a lot of testing, bug fixing, and user suggestions. The flood of rants from 3.x users could not change the fact that the new Qt4 API was so different from the old Qt3 API that starting with and building on top of the old 3.x was not possible, not only because of the difference in code base but because Trolltech announced that it would not support or improve the old Qt3 code base, and only support bug and security fixes for a limited time. The KDE dev crew had no choice but to switch to the new Qt4 API. IMO, that was the only choice they could make, and it gave them a great opportunity to move the KDE desktop to a modular base which could support the rapid evolution that was witnessed in the initial releases of KDE4. No one can deny that the result of the KDE development process is a desktop which is unsurpassed in its power, beauty and flexibility.

I say that to say this: the Unity interface (I hesitate to use the term desktop because a smartphone or tablet display doesn't, IMO, qualify as a "desktop" ... perhaps an "icontop", or a "fingertop", or a "dumbtop") can only get better, and it will as development proceeds. Its improvement may be slower than a regular FOSS project because the target audience isn't as technically savvy as Gnome's or KDE's so there won't be as much or as good a feedback to make rapid development possible. Most development will be done by paid developers. While I couldn't care less, Unity will be a considerable draw for Joe and Sally, who couldn't care less about Gnome or KDE.

IMO, Gnome shot itself in the foot when it set its course in the same direction as the Unity desktop by dumbing down version 3.0 without increasing its power. If anything, KDE has gotten more powerful and flexible, not to short its stunning eye appeal. By going in the opposite direction that Gnome went, KDE has, by virtue of its increased power, actually become easier to use.

I don't follow Gnome's development very closely, but the demise of Novell caused De Icaza to lose his well paying VP job, along with about 100 other Gnome & Mono developers. Canonical is switching from Gnome to Unity. Gnome has losts its favored position with a financially powerful Linux player because that player has died. To make matters worse, with Win8, Microsoft is switching from .NET to HTML5 & Javascript as its primary dev tool. .NET will be "along side" of HTML5, but I have experienced that relationship when MS cut VFP out from under its users, so I see nothing but a lingering death for .NET. And, after .NET's colossal BILLION dollar failure with the London Stock Exchange .NET trading app, its appeal has been in steady decline. As goes .NET so goes Mono. That means that Mono will also experience a decline. What has been totally devastated is the grand plan expressed by De Icaza nearly a decade ago to create a "Gnome" desktop built entirely out of Mono and interfacing directly with the kernel, bypassing the clib libraries. That would have been a fulfillment of James Plamondon and Microsoft's desire, as expressed in the combs-3096.pdf, to have Microsoft's code control the Linux desktop. Not many are aware of that, but it was one reason why, IMO, De Icaza kept parts of .NET not covered by the ECMA 334 & 335 standards (or the "Microsoft Promise") in Mono. Without them Mono had to create a GTK# hybrid interface in order to throw up gui dialogs user windows. While they weren't used by the Mono team, including those WinForm and ADO components in Mono was such a blatant violation of Microsoft IP that it put at risk those users of Mono who did implement such protected features. IMO, it was a classic IP trap, which most of the FOSS community didn't fall for because despite nearly a decade of Mono development the number of Mono based applications are remarkably small.

As a desktop, KDE's future is bright because contrary to a lot of "device" users and prognosticators, even after Microsoft attempts to lock down the PC OEM boot paradigm for "security" reasons there will always be vendors supplying open PCs, even among major PC OEMs.

An aside: Sorry, Steve, I don't believe it is a matter of security because it doesn't address the major part of Microsoft's security problems, the promiscuity of the ActiveX controls. In fact, if nothing else, resorting to a locked boot sector is, IMO, an admission that Microsoft is unwilling to leave the ActiveX paradigm behind because it is behind their "easy to use" factor. Their Security Essentials AV software is good at stopping known infective agents but cannot block the countless zero-day exploits, which is how the MBR gets infected by a Trojan. Once the agent has infected the MBR and the phantom recovery partition only a clean wipe of the HD will remove it. MSSE can't do that. To keep the ActiveX controls demands the boot lock down. IMO, it's a "win-win" for Microsoft. They lock out BOTH the MBR Trojan and Linux. Of course, it's not Microsoft's doing, and they were the first to tell us so. The PC OEMs will carry the blame for any failures to boot a Linux OS on Win8 certified hardware. Only time will tell if getting the PC OEMs to honor Linux certificates will be successful. I hope you find a way to boot Oneiric on that tablet without the need for a certificate, but I doubt that will happen unless you can break a 1024 byte RSA key.

[SteveRiley]

Re: Gnome3 and Unity desktops compared BIG PICTURE.

An aside: Sorry, Steve, I don't believe it is a matter of security because it doesn't address the major part of Microsoft's security problems, the promiscuity of the ActiveX controls. In fact, if nothing else, resorting to a locked boot sector is, IMO, an admission that Microsoft is unwilling to leave the ActiveX paradigm behind because it is behind their "easy to use" factor. Their Security Essentials AV software is good at stopping known infective agents but cannot block the countless zero-day exploits, which is how the MBR gets infected by a Trojan. Once the agent has infected the MBR and the phantom recovery partition only a clean wipe of the HD will remove it. MSSE can't do that. To keep the ActiveX controls demands the boot lock down. IMO, it's a "win-win" for Microsoft. They lock out BOTH the MBR Trojan and Linux. Of course, it's not Microsoft's doing, and they were the first to tell us so. The PC OEMs will carry the blame for any failures to boot a Linux OS on Win8 certified hardware. Only time will tell if getting the PC OEMs to honor Linux certificates will be successful. I hope you find a way to boot Oneiric on that tablet without the need for a certificate, but I doubt that will happen unless you can break a 1024 byte RSA key.

I haven't had time to make additional progress on my project yet. I have to relegate such activities to weekends, because I'm a member of three symphonic wind bands, which take up most of my weekday evenings.

Like you, I remain unconvinced that secure boot, as a general security feature, is necessary. What threat vectors does it promise to eliminate for most users? None that I can see so far. Mobility appears to gain some benefits from secure boot -- a laptop/tablet thief will have a much more difficult time extracting information from his latest purloined toy, so corporate information assets will be better protected. However, what benefit does secure boot bring to "Joe and Sally Sixpack," as affectionately name them? Honestly, I can't see any.

[GreyGeek]

Re: Gnome3 and Unity desktops compared BIG PICTURE.

...
Mobility appears to gain some benefits from secure boot -- a laptop/tablet thief will have a much more difficult time extracting information from his latest purloined toy, so corporate information assets will be better protected.
...

Good point! IF the PC OEMs honor Linux certificates then Linux would benefit from that kind of protection.

Linux is extremely secure from remote logins, port attacks and email payloads. BUT, if a knowledgeable user gets physical control of a PC running Linux they can hold the left shift key during power up to bring up the grub menu, edit the kernel line and and add "single" to the end of it, and then continue the boot up. That will drop them into a console as root without the need for any password, even if root has an obfuscated password. As root they have complete control of the system and can change any user's password or account information. The only thing still hidden from them would be the individually encrypted files which have a password different from that of the user's account. This is a good argument for encrypting your password list file or other sensitive documents residing on your system using GPG encryption with a 2048 byte RSA key.

To avoid this "single" boot break-in one can put a boot up password in the BIOS, if the BIOS supports such. Then, the grub menu would never show and the computer could not be booted. The BIOS boot up password is a good defense against plundering of a stolen laptop (unless the thief knows how to expose the mobo and either remove the mobo battery or short its terminals for about 5 seconds), but it is not a protection against an MBR Trojan on a Windows box. Alas, for every defense there is another offense.

[SteveRiley]

Re: Gnome3 and Unity desktops compared BIG PICTURE.

I'm imagining a laptop/tablet configuration in which the following are in effect:

• UEFI secure boot
• BitLocker (to encrypt the hard drive, including the OS)
• Domain membership
• Group policy objects appropriately configured

Such a beast would be pretty much useless to a thief. However, even among the largest corporations, the bring-your-own-device trend is has grown to the point where it's unstoppable now. Indeed, some organizations simply give $1,500 to new hires and say, "Go buy whatever you want." Given that the consumerization of IT is in full swing and shows no signs of abatement, I'm coming of mind that entrenching into locked down configurations means you've chosen the wrong side of that battle.

[bsniadajewski]

Re: Gnome3 and Unity desktops compared BIG PICTURE.

IMO, Gnome shot itself in the foot when it set its course in the same direction as the Unity desktop by dumbing down version 3.0 without increasing its power. If anything, KDE has gotten more powerful and flexible, not to short its stunning eye appeal. By going in the opposite direction that Gnome went, KDE has, by virtue of its increased power, actually become easier to use.

KDE also has kept the default desktop look-and-feel, which will definitely help anyone coming in from Windows. It sure worked for me.

As for the UEFI secure boot deal, there's another reason to stay away from Win8, as well as a Metro interface that even makes Gnome 3 look good. That, and I'll be upgrading my computer from now on, making sure I can disable the "secure boot" setting if need be.

[oshunluvr]

Re: Gnome3 and Unity desktops compared BIG PICTURE.

For those of you wishing to try E17 I can recommend (I as have several times )Bodhi Linux. Ubuntu LTS based 32bit only (for now) Enlightenment desktop. It's very well done.

[GreyGeek]

Re: Gnome3 and Unity desktops compared BIG PICTURE.

...
KDE also has kept the default desktop look-and-feel, which will definitely help anyone coming in from Windows. It sure worked for me.
...

I first started using KDE with the 1.0 beta in Sept, 1998 release of SuSE 5.3. I chose it because of its similarity to the Win95 desktop, which I used at work. I configured KDE to look and work exactly like my Win95 work desktop to avoid the mental dislocation that occurs when one switches desktop paradigms too often. I continued to do that through the various releases of KDE and of WinXX, until I retried in June of 2008. Since then, I have used KDE's classic menu and a desktop populated with a weather widget and a couple other system infomation wigets, plus a units conversion widget. What has amazed me is that Win7 appears to have followed KDE4 this time.

[bsniadajewski]

Re: Gnome3 and Unity desktops compared BIG PICTURE.

When I first started on Linux with Mandrake (9 IIRC in dual-boot with XP), I didn't know about DE's and such. SO when it come to choose which Desktop to use, I chose KDE (3.4?) just to try. Once the desktop came u pon fiest boot, I said to myself, "Wow, this looks and acts almost like Windows. I could get used to this." I have tried most every other DE/WM out there, especially after getting SUSE 9.3 from Best Buy back in '05, but I have stayed with KDE even through the 3 > 4 transition (thanks to Compiz as KWin 4.0 was way too slow on OpenSuse 11.0). I have a Ubuntu partition to keep up with Unity, which is running well in the 11.10 beta2, and an empty spot for anything else I may want to try (maybe Bodhi, Fuduntu, or even a BSD). Since K 10.10 I've had an empty desktop except for the System Monitor widget with CPU, RAM, and network monitors enabled.

As to the similarity between later KDE 4 releases and Win7, all I can is good ideas can come from anywhere.

[woodsmoke]

Re: Gnome3 and Unity desktops compared BIG PICTURE.

Kind of what bsn was posting

yes....Unity....RUNS GREAT!!! but....I just don't LIKE it.....

I know......waaayyyyy toooo many rum 'n cokes.....

but.....what about Sawtooth....

what about BSD?

I mean sawtooth has been around for a long time and there are people who really LIKE IT!!!

I mean.....................LIKE IT!!!!!!

why is that?

and BSD.....again.....it is wwwwwwweeeelllllll youuuuuuknoooooowww

its' THAT distro....but it was one of the ORIGINAL ones.....

and the people can put a LOT OF GUIS on it.,........

it just seems that a LOT of people talk about "this" and "that" but ......what about people with EXPERIENCE....

with "this and that".......

PEOPLE WITH EXPERIENCE WITH DIFFERENT GUIS? DIFFERENT ways of doing things?

ARE YOU THERE??

I mean all of those "other" ways of things MIGHT JUST BE FULL OF .....CR!@.....but lets HEAR FROM YOU.....

SHED SOME LIGHT!!!!

the old woodsmoker tends to shed his light with a bamboo lantern with fireflies in it walking on the beach...... WAAY too many rum 'n cokes.... 8) 8) 8) 8) 8)

dunnno....Linux USED TO BE ALL ABOUT......choice?

woodsmoke

[SteveRiley]

Re: Gnome3 and Unity desktops compared BIG PICTURE.

How many different ways can you manufacture a hammer?

OK, before you run to your neighborhood Home Depot and count their collection (192), yeah I know they have a whole wall bristling with aluminium-handled hammers and wood-handled hammers and dainty hammerlets and massive OMFGIMAGONNAHAMMERYOU hammers. I've stood paralyzed in front of that panorama of hammers (hammeranorama?) many times.

But they all basically do the same thing, so they pretty much exhibit the same or similar forms. So it is with desktops. There just ain't that many ways to hammer out (OK, lame) a UI. A keyboard and mouse on a horizontal surface with a perpendicular display kind of dictates how you're going to interact with those things. There's a reason the fundamentals haven't changed a whole lot since the Xeroc PARC user interface: they work pretty damn well.

Form factors that depart from this perpendicular pairing of input and output surfaces necessarily dictate alternate user interfaces. For example, Wii and Kinect really shine when the output surface (your TV) becomes large enough to fill most of your field of vision: now the input surface becomes the full three dimensional space you occupy. Pretty neat how that works.

Phones and tablets, with their input and output lying in the same plane, also require alternate user interfaces. What happens when you try to take one of these alternate interfaces and force it to behave in physical environments it didn't evolve for? Well, I hate to say it, but if all you have is a hammer... hahahaha.