OSNews

I always liked CDE, and this is positive news. It's been over a decade since you could purchase it for Linux/FreeBSD. There has recently been an effort to create an open-source clone, which has been decently supported on FreeBSD.

Regarding the UI, CDE is a stable environment, and I always thought iconify to desktop was superior to a taskbar. If I could design a modern desktop, it would largely resemble the CDE layout, with live thumbnails of the minimized apps as opposed to fixed icons.

I don't know how much traction this will gain, except for nostalgic users. The beauty is, we have the option to keep using it.

Hi, I am the documentation lead for the CDE project.

The people behind OpenCDE are involved with CDE as well. A FreeBSD port is in progress.

There is an issue with dthello that is making it impossible for it to run on other distributions other than Debain Squeeze and Ubuntu. It's related to rpcbind and it's being worked on.

I'm curious, why it was released now? The petitioning worked? (from http://www.marutan.net/cde/ seems there was some decent amount of good will; hm, and I just stumbled on it recently from http://www.marutan.net/rpcemu/ ) The Open Group largely lost interest, and threw it out there for those still wanting to use it and/or maintaining older apps?

My first experience with it was on an AIX workstation donated to the IT lab when I was in college. I was one of the few students with prior *nix experience so the lab manager asked me to help him with setting it up. It was a blast! Until then I had only worked with FVWM, blackbox and twm and I loved CDE's approach to desktop management.

I suppose that's why I took so well to Xfce when I switched to it from Gnome so many years later.

Question: As someone who never used CDE back in the day I have to ask, other than nostalgia of course...why would anyone care now?

After all it isn't like there isn't tons and tons of DEs to choose from now in FOSS, your KDEs, GNOMEs, XFCE, E17, LXDE,JWM, etc, you have so many choices and all of them frankly are so much better supported and so far ahead i just have to ask what the point is of wasting time and resources on something so...well old.

So while i'm all for there being more stuff to play with I just don't get the point of trying to bring truly ancient history back from the dead. To me it always made about as much sense as trying to run a new machine with Win95, we've just come so far from those days other than strictly nostalgia I just don't get it.

You might as well ask why there are so many people using FVWM? Again very ancient desktop.

I personally fr example have turned all my Windows 7 installations tio basically Windows 2000 desktops. All themes and other bling off.

"so many" doesn't mean what you think it means...

Edited 2012-08-14 00:03 UTC

Pretty sure you'd still be able to run GTK+ applications, just like you can run KDE applications from Gnome or XFce

Perhaps, but what would I gain then by using CDE, in exchange for the visual and configuration inconsistency?

Years ago, I had some customers insisting on having their well-known CDE environment "somehow" on their new Linux and BSD workstations. I ended up installing XFCE 3 for them, and with some tweaking (especially colors and some menu editing), I was told that that was exactly what they've looked for. The system was later on changed by replacing lots of workstations with thin clients (using X11 networking sessions). That solution was "exactly like" what they knew from their former Solaris environment, even though they did use quite different programs (leaving XFCE to serve as window manager and program launcher).

Sadly, this solution seems to be impossible now as XFCE 3 isn't supported anymure (due to the death of Gtk 1), and Xfce 4 (using Gtk 2) is much more "too different". I still have a 300 MHz system running BSD with XFCE 3, office programs and multimedia stuff. Unbreakable. :-)

There is a certain elegance to having rock solid system that still gets the job done. Whizbang features just get in the way.

This is something from the *BSD/commercial Unix world that I miss in the Linux ecosystem.

W8, how did an application using non-free toolkit end up under GNU umbrella?!

For being alpha quality software, it's pretty stable. I use it as my desktop DE now. There are are some quirks and not everything works (dtcreate, dtappbuilder and probably a few other things).

Of course the colors are a bit garish but I created a palette how to so you can change that.

Squeeze has been the official Stable release for years now. It froze in 2010. Wheezy, the next release, is already frozen, so we won't see Debian Stable with CDE in Debian-main until 2014 or so.

Whoops, I get Debian's weird release names mixed up all the time. That's what I actually meant.

I swear, I should just refer to them by version number... it's a lot harder to mix up. Doesn't help that with Debian's long release cycles, it's not uncommon to run across news items on two or three different versions of the distro (stable, oldstable, testing), and people usually use the codenames instead of version numbers.

They are easy. Debian Squeeze is the current stable version. Wheezy is the testing version, the next stable version. Sid is the unstable version.

Yes, easy once you memorize them. But easy to mix up once a new version is around the corner. And several years later, it's a lot harder to remember exactly what codename a specific release was, but again, the version number is a different story.

Debian's Toy Story code names are a lot easier than Ubuntu's even wackier ones though, I'll give 'em that.

Edited 2012-08-07 02:11 UTC

If the names are unclear, you can always use the title; stable, testing or unstable. Debian Testing just froze a week or so back which means it's basically in Beta working towards becoming the new Stable version.

Saves you worrying about version numbers and nick-names.

Curiously, though I never used it, I still have this "that's a GUI for Real Work(tm)" perception ingrained in my mind ...maybe from ~films (non-horrible ones, WRT to depicting how computers looks like), or photos of scientific labs & experiments.

If you like the colours, there is a GTK3 theme now out for integration with HP-UX
http://gnome-look.org/content/show.php/HP-UXMotif+%28Gtk3+and+G...

Edit: It also includes gtk2 theme.

Edited 2012-08-06 21:20 UTC

Have used CDE on both AIX and Solöaris and I really didn't like the default color theme on Solaris but AIX was perfect. Specially the terminal colors. Have used above mentioned screenshot many times to pick the colors and make my terminals look the same.

Motif is the widget toolkit and currently it is still not yet open sourced, but there is OpenMotif and actually that is used to build CDE currently.

Here seems to be a list of apps using Motif toolkit: http://www.openmotif.org/wiki/random

Edited 2012-08-06 07:46 UTC

Looks like OpenMotif (which really just seems to be one of Motif distribution methods) is also not yet open sourced...

Motif will be on sourced in a few months.

Yes Motif, which has functions that rival Win32 in the amount of parameters.

It makes somewhat sense, Motif was supposed to mimic Windows' look/feel, I guess they wanted the behavior to be similar down to the API level...

What?!

Motif is older than Windows.

Nope, Presentation Manager predates Motif by a few years. Motif was developed as a partnership between several vendors (HP and Digital mainly) and Microsoft, the goal was to have a consistent look and feel between Windows/OS2 and Unix machines. At least that was the "theory," the end result was a dreadful mess of a toolkit.

I think it is ironic to see people praising Motif/CDE, given how much it kinda stunk during its heyday.

Edited 2012-08-06 21:07 UTC

From the Motif FAQ: Motif is a widely-accepted set of user interface guidelines developed by the Open Software Foundation (OSF) around 1989 which specifies how an X Window System application should "look and feel".

From Wikipedia: The Open Software Foundation (OSF) was a not-for-profit organization founded in 1988 under the U.S. National Cooperative Research Act of 1984 to create an open standard for an implementation of the UNIX operating system. The foundation's original sponsoring members were Apollo Computer, Groupe Bull, Digital Equipment Corporation, Hewlett-Packard, IBM, Nixdorf Computer, and Siemens AG.

While Presentation Manager (1988) predates Motif with a year, the first usable Windows version (3.0) only came in 1990.

Sigh...

From the wikipedia article:

"Motif's operation was designed to correspond closely with the then-familiar Microsoft Windows and OS/2's Presentation Manager interfaces, and Microsoft played a key role in designing the original style guide."

http://en.wikipedia.org/wiki/Motif_(widget_toolkit)


Windows 1.0 was released in 1985, BTW. "Usable" is a very subjective metric BTW, and irrelevant to the timeline. The point still stands the look and feel of Windows/Presentation manager predates motif by a few years.

Edited 2012-08-08 23:49 UTC

Yes, if you can get OpenMotif to run on Weyland first.

And it will probably do so in a near future, since this toolkit is opensource, it is relatively simple (against GTK or QT) and has a quite active community.

The license status is in the source. The Open Group isn't the fastest with announcements. It should be up today. It'll be here:
http://blog.opengroup.org/

It takes about 5 hours to build on the Pi.

Just because it's old doesn't mean it doesn't have good ideas. You reference Mac OS 8/9 - for me, that interface, Platinum, is miles and miles ahead of whatever UI disaster Apple is using now. Platinum was designed with usability in mind, it was consistent and logical. Mac OS X, on the other hand, is Microsoft BOB with garish skeuomorphic crap and incredibly inconsistent.

Old != bad. Platinum > Aqua.

Well, of course old doesn't mean bad. I myself am a huge fan of many old projects, like BeOS, OpenVMS. But that's just a different thing and different history. BeOS died and Haiku OS was born as an open source [from the start], and OpenVMS ... was never *open*, so there's no real problem anyway.
I would be more than pleased to know why exactly did they open their sources. This would give us valuable information and ... their motives.

Might I add I don't try to take your joy away, CDE users! I can imagine you really love it and I have nothing against it. I'm just being suspicious ... or critical if you will.

This begs the question: What then are the good ideas in the specific case of CDE?

No, it doesn't: http://en.wikipedia.org/wiki/Begging_the_question

They pretty much all suck these days - for me Mac OS X is the best of the worst and that isn't saying much for any of them to be quite frank. Having used IRIX owns desktop along with CDE on Solaris I would sooner the effort go into reviving and adding functionality to CDE than what seems to be the wasted effort by the two major desktops hauling along the monstrous memory hogging crap they call 'features' or stripping out preferences for the sake of 'ease of use' aka GNOME.

What would I like to see?

1) CDE based configuration tools - networking configuration tools, video card settings etc.
2) Getting the major software titles like Firefox, Thunderbird, LibreOffice and so on to use OpenMotif - get some consistency.
3) Porting it to Wayland but ideally I'd like to see something better than Wayland so that other operating systems aren't left out in the cold - yes, I am a FreeBSD fan.
4) Niceties of modern desktops such as anti-aliased fonts etc.

So basically you just want to keep reinventing the wheel....

No, we already have a wheel - it's called CDE but the children are hell bent on going their separate ways re-inventing wheel with all the faults of the original design.

CDE is not a wheel you want to use as a foundation for anything, it was a mess of a desktop environment. It basically was the pinnacle of "designed by committee." IMO CDE was one of the reasons why commercial Unix workstations went the way of the DoDo.

I think a lot of people's views in this thread about CDE are tragically warped by the lens of nostalgia...

UNIX workstations died not because of the lack of commonality but the fact that SGI, SUN and IBM couldn't get it through their thick skull that no one is going to pay $15,000 for a workstation when a 'good enough' workstation running Windows on an x86 chip could do the job quite nicely. Sorry to break the bad news but the world runs on 'good enough technology' and not the best, most sexy, most awesome pinnacle of engineering in a particular field. Sorry but had the UNIX workstation pricing dropped at the same rate as the PC and reached parity people would have been able to look over the imperfections just as people were able to look over the imperfections of Windows NT and Linux in the early days.

Thing is, any parity in price-drops (or, more generally, in price/performance ratios) wasn't even really achievable for traditional UNIX workstations. Meanwhile, the economies of scale (of that "good enough" approach) afforded the PC working itself upwards, eventually surpassing the "pro" hardware/software.

Joining the PC world did seem to help the only remaining prominent *nix workstation, Mac Pro, for some time ...but it's still visibly slipping away as of late, uprooted also by more traditional PC OEMs.

Platinum was familiar to YOU, that does not meant that it is superior to Aqua.


One could make the case that given the significant increases in user base Apple experienced after the introduction of Aqua, whereas Platinum was stagnant at best, that it would seem that people prefer Aqua (OSX) to Platinum (Sytem 8/9).

Hahahahahahaha!! Such a joker!! Did you ever use Microsoft Bob? If you think OS X is like Microsoft Bob, I doubt it.

Please don't mix graphical representations the general of real life in isolated apps with Bob. They are far removed. I find OS X far less offensive than the Crayola XP interface. The default control panel was fcuking insulting. The window dressing actually messed with the physical size of Windows and broke apps that made assumptions about the width of the title bar. For me, Windows 7 is every much as Crayola, and Windows 8 and the UI formerly known as Metro is more of the same. Simply depressing. I'd take the Windows 95/2000 classic theme over the god awful XP/7/8 themes any day of the week.

Well with Windows you can easily turn them off... (and control panel stuff is one click away, button visible in the same window).
Good luck changing what's messed up like that in OSX.

I think the main point is that there's so much code out there designed specifically for certain horribly outdated systems that is near impossible to port to something modern. The more older stuff that gets open sourced, the easier it is to manage a transition to running on newer operating environments.

The need for compatibility with one program often keeps entire networks back

Assuming the developers are able to make a leaving out of it.

Just because it's old doesn't mean you throw it away. There are many fans of CDE. The code is open, who knows what will happen?

I'm not sure what the problem is, it's not "closed", as it's now Open Source. I'm not sure how it's "outdated" as I use it all the time, and I don't know of anything it does not do, that it really needs to. Sure there are features I'd like, but that's the case for all desktops. I don't find it buggy, I use it every day, I see more glitches in Windows 7 and Mac OS X, and I sure as hell see more in Android.

Any real benefit? Maybe not to you, maybe not to me either, even as a user of CDE every day.

With respect, I think you make a lot of assumptions just because it's old. CDE is very solid, and runs quick. It's not a big "look at me" desktop with 3D effects or "dashboards" nobody wants. It does it's minimal job very well and stays out of the way.

In my experience, older software seems to be somewhat *less* buggy. It's probably due to a smaller code base (less code, fewer bugs), longer product cycle for stamping out issues, and pretty much no features added in a hurry because it's in fashion at the time.

By all means, try out CDE and decide it's crap, but again, with respect, you're stating a lot of problems with CDE which do not exist.

Nah, we mostly remember the relatively few positive examples (and those which survived, which still can be used; long-maintained hence, duh, decently debugged) - while forgetting tons of negative ones.

Similar effects with the popular myths about old films and music, or the general "old times were better" ...we just don't remember so well all the crap that was pushed, how the nice stuff wasn't so available and discoverable (a'la imdb or last.fm)

Didn't think it ugly at all, not saying it couldn't have been improved but at the time I wished it worked on windows as well. Not sure I wouldn't now.

It could be improved with antialiased fonts and UTF-8 support, which are all available in OpenMotif.
I think that would be a plus in usability without introducing any bloat.

There is experiment support for that in CDE right now:
https://sourceforge.net/p/cdesktopenv/wiki/FontsWithXFT/

Well, true XFT support in OpenMotif requires more work than setting a couple of resource files ;-(
I have done this in my software, here are the screenshots for choosing XFT fonts for the application

http://www.arahne.si/images/stories/news/new_version_ArahWeave/inte...
http://www.arahne.si/images/stories/news/new_version_ArahWeave/thai...
If somebody needs code for this please contact me.

Also, the OpenMotif was converted for XFT and all widgets support renderTables. But third party widgets are not automagically converted just by using OpenMotif. So various dt widgets will probably need some work in order to support XFT and UTF-8.

It is "experimental support". It may or may not work. Obviously in your case, it didn't work. Its a requested feature but not a high priority right now.

There is experimental support for XFT:
https://sourceforge.net/p/cdesktopenv/wiki/FontsWithXFT/

Great! Thank you for the link.

I was in the same boat. I was very used to running olvwm and then sun switched over to CDE. It was a bear trying to configure CDE to act more like openlook. Even worse was all space wasted on the desktop FOR the desktop when I cared first and foremost about maximizing my code windows (which I did with openlook).

You weren't looking around much - Java, Star Office. Arguably also Qt, Blender, Duke3D and id engines, Netscape, Sim City.
And in the OS or DE categories, where CDE is - definitely CP/M, GEM, Symbian, Solaris.

Yeah, open sourcing mostly didn't really help the last four and/or it was done when they were long past their time. Still, kinda neat with CDE, and should be useful at least in ~legacy scenarios.

Motif is the key, IMHO. It was the most common tool kit that people used on UNIX gui apps. I'm sure some will benefit having it available for free. Lesstif almost always worked but there were a couple issues. If nothing else, it would make updating those older apps easier.

Its not great wording. What it means is that the code compiles and runs on Debian Squeeze and Ubuntu.

Nope, No code from OpenCDE is being used. OpenCDE is deprecated and it's developer is part of the CDE project now.

I'm pretty sure scripts can't be setuid, at least that's the case on OpenBSD.

Besides FreeBSD, right now Debian Sid is our most requested port.

I don't know that IBM could even if they wanted to do so atm. Last I checked OS/2 is still active in the form of eComstation and I'd imagine there's some agreements in place that would complicate the code release.

That would be nice, but OS/2 is still a commercial product under the guise of eCommstation. I think they would have a problem with IBM releasing those technologies. Then again, IBM might not even have the rights to the code anymore. I don't remember the terms of the sale, but I think releasing code would have to come from eCommstation.

In an ideal world, we would not need money or patents.

But since we don't live in an utopia, and OS/2 is pretty much a dead platform long surpassed technically. And even though it may be of great interest to a few dedicated enthusiasts.... the cost of the legal/patent review process for the OS/2 codebase is not trivial, and probably not worth it for IBM. Unless hobbyist could raise some capital and offer IBM (or whoever holds the copyrights now) some grease to sweeten the deal.

Nice you mentioned it. I've been using GeoWorks Ensemble 3.0 (not sure about the version number, but it was before NDO or BBE) on DOS many years ago, still being impressed on how they did get things done that today's "Windows" is (natively) still incapable of, like real drag & drop, or pinning menues.

For screenshts and comparison visit "The GUI Gallery":

http://toastytech.com/guis/geos12.html

http://toastytech.com/guis/sol.html

After seeing the lackluster response to opensolaris, I've come to the conclusion that there are more then enough free unix variants out there already.

linux/bsd(s)/minix/opensolaris/darwin ...

For devs who do this kind of work.. likely they already are apart of one of these projects... if thats the kind of thing their into.

Doubtless it would be cool to have irix running on new hardware but getting a new mips workstation might not be so easy, porting irix to X86 is possible but would need a big dedicated effort I'd expect.

On the otherhand - the userland may be able to be ported to other OS's but at this point Im not sure what the benifit would be really.

My guess is the people who own IRIX (rackspace?) know this and the cost of going over a huge codebase and getting permissions from various copyright owners is too prohibitive to be worth the effort.

He's talking about the Indigo Magic desktop environment, not IRIX itself.

That said, the reason OpenSolaris failed miserably has more to do with the braindead decisions with licensing and general lack of support from Sun/Oracle than the OS itself. I'm sure if they'd gone GPL or BSD style licensing there would've been a great deal more interest.

The petition is about to open IRIX's desktop environment and underlying things, if possible, like 4dwm window manager, not about opening IRIX OS itself.