tde-devels

devels@trinitydesktop.org

3 participants
3808 discussions

libltdl missing from arts
by L0ner sh4dou 12 Nov '11

12 Nov '11

While compiling trinity manually on arch I discovered that libltdl was removed from arts source tree. I got errors both on CMake configuration stage (http://paste.pocoo.org/raw/506509/) as on compilation stage (http://paste.pocoo.org/raw/504813/). Luckily i found on my HDD trinity code I pulled just before version froze and which I thought I deleted. After putting the missing libltdl dir into current arts dir both CMake configuration and compilation went without errors. Apparently libltdl from arts is not quite the same thing that is available in package form in most distributions main repos. As it is said in arts libltdl readme it has been slightly modified to make it work with KDE. As for now it should be put back in the arts, since miss of it prevents arts compilation. Whole missing thing compressed: http://dl.dropbox.com/u/26010362/libltdl.tar.gz

2 6

Re: [trinity-users] Re: OSS support ?
by Francois Andriot 12 Nov '11

12 Nov '11

Le 12/11/2011 12:21, Nick Leverton a écrit : > In article<4EBDC181.8080905(a)free.fr>, > Francois Andriot<trinity-users(a)lists.pearsoncomputing.net> wrote: >> Le 10/11/2011 17:02, Philippe Bourdeu d'Aguerre a écrit : >>> Le jeudi 10 novembre 2011, François ANDRIOT a écrit : >>> Yes. The choice is only "Automatic", "Nothing" and "Alsa". With KDE 3.5.10, I >>> get also "OSS" and a few others. >> Hello, >> I've just compiled an ARTS version for RHEL5 which should support OSS >> and Threaded-OSS. >> As I thought, it was just forgotten in the CMAKE port, but all the code >> is still in place, it was just not compiled. > Could you post the change, please ? It would help me too. > > Thanks, > > Nick Hello, first, I send a copy of the mail to 'trinity-devel' since it's not really an user question. You will find the patch I made attached with this mail. There is one drawback (I think), it's that I've hardcoded the following line in 'config.h.cmake': #define HAVE_IOCTL_INT_ULONGINT_DOTS 3 This value was guessed automatically when using autotools, but I did not find how to replace it with a CMAKE equivalent. The rest of the patch should be sane. Also, I did not test OSS at all, so I don't even know if it works. Francois Andriot

1 0

kdebindings 3.5.13: embedviewimp.h build errors
by Darrell Anderson 11 Nov '11

11 Nov '11

Slackware 13.1 Previously built kdebindings without this error. I don't know what is different this time around. Log: http://humanreadable.nfshost.com/trinity/build_logs/kdebindings-3.5.13-i486… The log also says the following: Detected TQt3... ... checking if TQt is available... no configure: WARNING: You are attempting to compile Trinity without the Trinity Qt Interface installed. Please install libtqtinterface-dev and try again! tqtinterface is installed. I don't know what this error messages keys on, but the phrase "Please install libtqtinterface-dev and try again!" seems Debian specific. Need help. :) Darrell

3 5

Changing list subscription
by Darrell Anderson 11 Nov '11

11 Nov '11

Is there a way to change my email address in this list? Or do I have to subscribe with a different address and unsubscribe the current one? Darrell

2 1

Re: [trinity-devel] Need official clarification
by Timothy Pearson 11 Nov '11

11 Nov '11

> Does Trinity team want a role of an upstream for KDE3? Yes we do, but at the same time we do not want to be saddled with the KDE name. If we were to remain "KDE3", instead of rebranding to TDE, we would forever remain in the shadow of KDE4 (and KDE5 when it comes out, etc.). As stated on one of our pages, we aim to keep the spirit of the old KDE3 interface alive, but will gladly accept innovation within the established interface style. For example, while we have no desire to copy Plasma or other similar features from KDE4, a new program that makes it easy to configure and control monitors was included in the TDE source; user visible changes were made to bring the desktop lock system up to modern standards, etc. We are not against change in general, we simply ask that the user is given as much information as possible about (and tight control over) the task at hand. I have definite theories on interface design involving how to maximize I/O to and from the slowest link in the system: the human operator. So far these theories line up exactly with the TDE interface (with some exceptions of course, but these are being rectified). As long as the new feature being added does not collide with these theories, chances are it will be welcomed with open arms. Also, if it does collide I am more than willing to explain why I would consider it to be suboptimal. > Is seems that as you claim Trinity is a totally new product you do not want such role. We do actually. Trinity is an exciting new development of KDE3, it has KDE3 roots, and it is the most likely candidate for upgrade from an existing KDE3 system. We simply do not want to continually live in KDE's shadow, or be hampered in our efforts to improve this excellent desktop environment. > In that case I will request removal of KDE3 from the next openSUSE release. > I do not want to be attacked for including software with no upstream. From what I can tell your software does NOT have an upstream. If you were to build TDE 3.5.13 on OpenSUSE then it would. Tim

1 0

Fwd: Re: [opensuse-factory] The release notes/product highlights for 12.1
by Ilya Chernykh 11 Nov '11

11 Nov '11

Hi! This is an extensive objection by a KDE4 developer against marketing and promoting KDE3 and/or Trinity. I would like to see what can you say in response as it seems he expresses quite common sentiments of distributions developers towards KDE3. He also links an extensive article in German in a respectable Linux journal named "Trinity - desktop with no future ". The machine translation is here: http://translate.google.ru/translate?hl=ru&sl=de&tl=en&u=http%3A%2F%2Fwww.f…

4 6

Re: [trinity-devel] Fwd: Re: [opensuse-factory] The release notes/product highlights for 12.1
by Baron 11 Nov '11

11 Nov '11

On Friday 11 November 2011 17:19:01 Ilya Chernykh wrote: > Hi! > > This is an extensive objection by a KDE4 developer against marketing > and promoting KDE3 and/or Trinity. > > I would like to see what can you say in response as it seems he > expresses quite common sentiments of distributions developers towards > KDE3. > > He also links an extensive article in German in a respectable Linux > journal named "Trinity - desktop with no future ". The machine > translation is here: > > http://translate.google.ru/translate?hl=ru&sl=de&tl=en&u=http%3A%2F%2 >Fwww.freiesmagazin.de%2Fmobil%2FfreiesMagazin-2011-09-bilder.html I could be wrong, but, I get the impression that there is the realisation by the KDE people they have made a mistake ! In doing so alienating lots of its users ! -- Best Regards: Baron. -- Best Regards: Baron. -- Best Regards: Baron.

1 0

Need official clarification
by Ilya Chernykh 11 Nov '11

11 Nov '11

Does Trinity team want a role of an upstream for KDE3? Is seems that as you claim Trinity is a totally new product you do not want such role. In that case I will request removal of KDE3 from the next openSUSE release. I do not want to be attacked for including software with no upstream.

1 0

kdelibs potential security vulnerability
by /dev/ammo42 11 Nov '11

11 Nov '11

Hello, The kdelibs/qt3 combination I built (from the stable sources) has a security hole: when the dynamic loader loads libkdecore.so, the libmng.so.1 is first searched in a directory in /tmp, which could potentially result in a local root privilege escalation if there are suid binaries which depend on kdelibs on the target system. Even without suid binaries, on a multi-user system one could gain another user's privileges by letting them load a kdelibs-using binary while a forged libmng is present in the /tmp directory (and usually can't be deleted by a regular user because of the sticky bit in /tmp). The build script I use does $ export LD_LIBRARY_PATH=$TMP/build-qt3/qt3/lib:$LD_LIBRARY_PATH (actually Calvin Morrison's PKGBUILD does the same thing). Perhaps this hack has a relation with the issue, but anyway Patrick Volderking (the creator of Slackware) mentions in his old build script for Qt3 (when KDE3 was in Slackware) that building Qt3 in /tmp gives ****, and his script builds qt3 directly on $PREFIX/lib{,64}/qt (or qt3, depending on the exact script). But I'm not sure this is compatible with all distribution policies… as another means of avoiding /tmp, the gcc Slackware build scripts which use /$(mcookie) as a temporary directory. Are official binaries affected by the issue ? I don't know Debian/Red Hat building procedures at all and then I have no idea of the answer.

2 1

thiel@wporg.com
by Darrell Anderson 10 Nov '11

10 Nov '11

Hi Brian, Interesting personal history you shared! As much as I enjoy tinkering with computers, I get very mad when anything gets in my way of being productive. Therefore I am very much in the group of people who see computers being a tool --- the means to an end. I don't know that I can help. For several years I have sheltered myself in Slackware, which is not known as a "mom and pop" friendly system. Although some time ago I started a series of essays about finding an operating system for older computers, I have not continued that journey. I would like to renew that journey this winter. A big challenge with what you ask is defining your audience. A basic bell curve more than likely would show most users wanting only the basics. There always will be computer-challenged people who, no matter what anyone does to help, will never understand anything they do or think they want to do. Then there are the power users, both smart and non-smart. The non-smart ones can take anything apart but can't fix their mess. They always call somebody to help. Mostly though you are focusing on the people in the middle of the bell curve. My struggle is not so much selecting a distro. Even Slackware, if preinstalled and configured by a subject matter expert, can serve the purposes of these people because all of them want basic point-and-click access to applications. The majority of computer users do not care about how everything runs underneath. The challenge is older hardware can't deal with the modern internet. Back in the mid 1990s a 486 machine running Netscape 3 or 4 had no problem surfing the primarily text based web. Today too many web sites use JavaScript and Flash. Even if those two features are disabled, most older hardware still can't surf the web. The limited RAM and video cards on these old machines can't render a typical web page fast enough to be suitable. Running flash on these old machines is hopeless. If a user does not need the internet, then the older hardware runs quite well with the traditional apps of yesteryear: word processing, spreadsheets, etc. Anything that requires serious video rendering will bring these old machines to their proverbial knees. I have Slackware 12.2 and KDE 3.5.10 installed on two such machines: a Pentium I and Pentium II class machine. Startup speeds of any app is slow but tolerable. Using an app is acceptable after the app is started. I can improve desktop speed a bit by using a window manager such as IceWM, but I still need to start and use apps, which tends to be slow. The moment I try to surf the web, even with images disabled in the web browser, the systems show their age immediately. In short, the problem is not the operating system but the hardware. I don't know an easy answer. What is the minimum specs that allow surfing the modern internet in an acceptable manner? Are such users willing to accept that they can't watch online videos? Are they willing to accept that a dial-up connection means they can't download videos or receive such videos as email attachments? Perhaps the answer is to use nothing less than a Pentium III class computer. If I find the time this winter to return to my "old hardware" project, I should find one or two of those types of computers to experiment. The PI and PII machines can't deal with the modern internet. That said, anybody still using dial-up probably never would notice because the connection speed is more of a bottleneck than the computer. I wonder whether I can rig up a way to simulate that kind of connection speed and then test my old hardware. If we get past the hardware questions, which distro to choose? Any Ubuntu based system, regardless of hardware or desktop environment, will kill any old hardware. That includes Mint. Forget about cloud-based distros such as Peppermint because the hardware and connection speeds used by such people can't deal with the overhead. Otherwise I probably would stick with a Debian based system because of the size of the repositories for additional software. Puppy is an interesting possibility and one I wish I had more time to investigate. Which desktop environment? None of the new desktops will work on older hardware, such as KDE 4 or GNOME 3. Those new desktops require 3D acceleration hardware. That leaves Trinity, Xfce, or LXDE. To me, LXDE remains too experimental and likely would frustrate many mom and pop users. A window manager approach would work, but only if highly customized by a subject matter expert. Regardless of which desktop is chosen, everything would have to be preinstalled. Therein lies the real problem which no Linux based distro developer has solved: creating a system that mom and pop users with old hardware can use. The Puppy people might be close. As far as I can tell, the only solution is select a distro and then customize everything. Guess what? At that point you have created your own distro and must face all of the related support headaches of such a project. Any such project requires serious usability testing. Somebody like you or me can tinker and find a solution. Most mom and pop users can't --- and won't. I wish I had a simple answer. :) Thanks for stopping by. Darrell

3 2

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

tde-devels