Table Of Contents
|1.||16�Dec�2001||(16 posts)||Exporting 'Syntax Enlightened Source'|
|2.||17�Dec�2001||(1 post)||Out of kdenonbeta: KBugBuster|
|3.||23�Dec�2001�-�29�Dec�2001||(3 posts)||KOffice Filter Framework Upgraded|
|4.||20�Dec�2001�-�31�Dec�2001||(6 posts)||What's a "PIM"?|
|5.||20�Dec�2001�-�28�Dec�2001||(6 posts)||Icon Server A No Go|
|6.||28�Dec�2001||(1 post)||Emacs Style Keybindings|
Welcome to KC KDE and Happy New Year to one and all! Split between holiday family obligations and the flu, KC KDE took a Christmas holiday. With the arrival of 2002 KC KDE returns and we look forward to a great year ahead.
KDE3 Beta1 is out and CVS HEAD continues to gain solidity and speed. The KDE developers have moved into the "only commit bugfixes and minor features" phase of the release cycle. Binary compatibility and strings that need to be translated are still open for changes but there is an effort to start stabilizing these fronts.
Of course KDE2 isn't dead, as many people are still using it and developing for it. In particular, there has been a surge of eye candy for KDE2 in the last few months including the alternative icon set iKons 0.5 (http://users.skynet.be/bk369046/icon.htm) , Mosfet's Pixie Plus (http://mosfet.org/pixie/) image viewer and the fledgling KDE Passion project (http://ben.digitalsimplicity.com/passions.php) not to mention updates and forks of several native widget themes. There was quite a wait for these flourishes to appear after the release of KDE2, which was understandable given the large changes involved and the initially smaller user base. With KDE3 being more evolutionary than revolutionary (though the style API has changed considerably for native widget themes) there hopefully will be a much shorter wait for third party eye candy this time around.
Enjoy this week's summaries and Happy Hacking!
1. Exporting 'Syntax Enlightened Source'
16�Dec�2001 (16 posts) Archive Link: "[Kwrite-devel] [PATCH] nice feature for kate.."
Summary By Timothy R. Butler
People: Emmanuel Touzery,�Daniel Naber
In an effort to add a feature he discovered in VI to Kate, Emmanuel Touzery (mailto:email@example.com) posted a note saying:
Recently i discovered a nice feature from vim: export a syntax enlightened source in HTML.. great isn't it? i wrote a method for KateDocument that does that (it requires information that you can't have in a plugin).
IMHO, and if you decide to take it in, the export to HTML should go in the submenu where you can choose the enlightenment mode (document->syntax enlightening). Also, a potential improvement would be to export the selection or the whole document, depending on the user's will, but I don't have the time for this at the moment... maybe soon :o)
Daniel Naber (mailto:firstname.lastname@example.org) responded, "Very cool! However, your version currently produces invalid XHTML, i.e. it's not well-fomed. There's e.g. "<font ...></i>" in your patch.html." After a quick exchange of messages, Emmanuel replied, "here we are. the attached html file is XHTML-1.0 compliant according to validator.w3.org (http://validator.w3.org) (tell me if there's still anything wrong, though), and you also have the relevant sources." Daniel finished up the discussion saying "Ok, I assume Jospeh is going to commit it to HEAD."
2. Out of kdenonbeta: KBugBuster
17�Dec�2001 (1 post) Archive Link: "KBugBuster"
Summary By Rob Kaper
Topics: Bug Tracking, New Application
People: Martijn Klingens
KC KDE primarily covers development mailinglists, but sometims the latest developments in KDE development can better be seen in the CVS tree, where the actual code is stored.
One of the big changes last week was the move of KBugBuster from kdenonbeta (breeding grounds for new KDE projects) to kdesdk, the module combining several handy tools as a KDE Software Development Kit. Already present were KBabel (assists translators), several useful scripts, kompare (kdiff), kapptemplate (KDE application template) and other goodies.
KBugBuster was originally created by Martijn Klingens but now maintained by several other KDE developers. The application is only useful for KDE developers: it is a convenient interface towards the KDE Bug Report System (http://bugs.kde.org/) and allows developers to manipulate the contents of the system. Bugs can be closed, reassigned or simply viewed. Pending changes ar queued and when the developer wants to commit them KBugBuster sends the appropiate e-mails to the control addresses of the bug report system.
While some wishlist items remain open, KBugBuster makes bug busting or squashing a lot more convenient and can save a developer quite some time, which can be used for new features or even more bug busting, improving KDE applications everywhere.
3. KOffice Filter Framework Upgraded
23�Dec�2001�-�29�Dec�2001 (3 posts) Archive Link: "KOffice filter Changes"
Summary By Aaron J. Seigo
Topics: KOffice, Filters
People: Werner Trobin
Probably the number one quoted weakness of KOffice is its import and export filters (or lack thereof). Fortunately the filters in KOffice are getting attention with new ones being added and the existing ones improved. Also of import are the filter framework changes that Werner Trobin (mailto:email@example.com) has been working on. Werner updated the KOffice developers with his progress saying, " As I'm already quite happy with the progress I made on that new filter stuff and basic importing and exporting works, I thought it's a good idea to break KOffice today. I'll commit the stuff I have here right now and disable all the filters from compilation. Then I'll start porting the "simple" file to file filters, one after the other." Werner went on to list what remained to be done on the filter framework. Point number four in that list was quite interesting: " Implement support for embedded usage of filters. I didn't do that up to now as it would have been a hack like the "old" solution. IMVHO we need to take a closer look at KoStore and friends to find a sensible solution. We should also take into account that this is supposed to change in the near future, when we switch to the unified storage model (OpenOffice stuff). [Did we reach any consensus, yet? Thomas?]"
4. What's a "PIM"?
20�Dec�2001�-�31�Dec�2001 (6 posts) Archive Link: "[Kde-pim] glossary"
Summary By Aaron J. Seigo
Topics: KDE PIM
People: Klaus Staerk
Anyone who reads the KDE PIM developers list has probably discovered that the world of personal information management is full of strange words, concepts and acronyms. To help new comers to the PIM scene out, Klaus Staerk (mailto:firstname.lastname@example.org) announced the new KDE PIM online glossary saying, " okay, from the "technical" point of view, we have now our glossary on pim.kde.org. You can find it at http://pim.kde.org/development/glossary.php (http://pim.kde.org/development/glossary.php) This page is not integrated to the navigation bar of pim.kde.org as there's no real content yet ;-) Please feel free to contribute to our glossary by sending me some KDE-PIM-words that you'd like to find there. So we can make this soon an official part of our website." Klaus also noted that there is a form at the bottom of the page for submitting new terms.
5. Icon Server A No Go
20�Dec�2001�-�28�Dec�2001 (6 posts) Archive Link: "icon server"
Summary By Aaron J. Seigo
People: Waldo Bastian,�Rob Kaper,�Rik Hemsley,�Alex Neundorf,�Lubos Lunak
In private correspondence Lubos Lunak (mailto:email@example.com) informed Waldo Bastian (mailto:firstname.lastname@example.org) that an icon server, which was hoped could speed up icon loading times, was not going to provide any real world improvements and that the current icon loading scheme was pretty fast as it is. Waldo replied and CC'd the core development list saying, " Yes, I'm afraid we ain't gonna win much with this :-( I don't think icon loading is very slow btw, but it is something which costs considerable time, as such it is a candidate for improvement. On to plan B then.. that would be to try to delay loading icons as much as possible. E.g. icons used in menu's theoretically don't need to be loaded until the menu is actually shown. Konqueror loads about 150 icons on startup, (50 of which are for the cogwheel) but I only see about 30 icons on the screen. That seems to indicate that we can load about 70 icons later." Waldo also noted that making the cogwheel animation one file would speed things up as would making icon searching smarter for things like favicons.
Rob Kaper (mailto:email@example.com) suggested, " Another thing: Konqueror displays the original mimetype icon (image.png) even for files of which we have a thumbnail. Would it be faster to directly scan whether there is a thumbnail available and in that case not show the image.png first only to replace it later?" Rik Hemsley (mailto:firstname.lastname@example.org) offered, " Another good speedup would be to set a fixed grid size when you have thumbnails enabled. This way, you don't get all the re-layout happening."
Alex Neundorf (mailto:email@example.com) noted that icon loading was probably not as fast as it could or should be and addressed the issue of delayed icon loading saying, " this has to be done in the xmlgui stuff. Maybe also for the toolbars. Also often KActions are created with a call to one of the functions in kiconloader.h, which all load the icon immediatly and return a QPixmap or QIconSet. To change this the return type would have to be changed to something which loads the icon later when needed."
Some of the suggested improvements were made and committed to CVS as the ongoing push for better performance continues.
6. Emacs Style Keybindings
28�Dec�2001 (1 post) Archive Link: "Update on K*Accel* and friends"
Summary By Aaron J. Seigo
People: Ellis Whitehead
Depending on who you ask, Emacs key bindings are either heavenly or ludicrous. But most agree it is nice to at least have the choice to abuse your keyboard however you wish. To aid and abet those desiring Emacs-style keyboard shortcuts in KDE3, Ellis Whitehead (mailto:firstname.lastname@example.org) has been working hard on improving the relevant KDE classes. Ellis wrote regarding his progress saying:
I just committed about three weeks of local changes to the K*Accel* and KKey* classes. The changes to K*Accel* are mostly just restructuring and simplification. Documentation is still sparse. New KKey* classes have been written, but the X11-specific methods is still use old code.
I've gotten rid of the basePtr() method in KAccel & KGlobalAccel. There is still a KAccelBase class, but it's completely hidden from the outside world (i.e., kaccelbase.h as a noinst_HEADERS header).
Internally, emacs-style shortcuts are now functional, but still rough. There is no user-interface to set them yet, though, so you have to edit rc files by hand in order to actually use it. I've started work on an appropriate widget, so maybe it'll be working by Sunday.
As is to be expected, the new code had some bugs. This in turn caused a temporary rash of accelerator related breakages in CVS. Things have since gotten much smoother and Emacs fans can look forward to having fun with keybindings in KDE3.
Sharon And Joy
Kernel Traffic is grateful to be developed on a computer donated by Professor Greg Benson and Professor Allan Cruse in the Department of Computer Science at the University of San Francisco. This is the same department that invented FlashMob Computing. Kernel Traffic is hosted by the generous folks at kernel.org. All pages on this site are copyright their original authors, and distributed under the terms of the GNU General Public License version 2.0.