Older blog entries for Burgundavia (starting at number 77)

Plugins considered harmful

Dear Application Developers,

I have noticed a worrying trend recently. Too many application developers and packagers seem to think that putting core functionality into a plugin or extension is a good idea. It is not. The next time some user asks me how to open ODT in Abiword (you need the abiword-plugins package, which is NOT installed by default when you install Abiword) or how to do session saving in Firefox (you need an extension), I think I am going to scream.

So if you develop an application, please remember that plugins and extensions are for whizbang features not many use, not for core pieces, like opening a file format or recovering from a crash.

Yours,
Corey Burger
Department of Good Defaults

The community at work

With Flight 5, XChat-Gnome was removed, leaving only gaim able to do IRC. However, this made getting onto #ubuntu now very difficult, due to needing to setup gaim with multiple arcane dialogs.

Into this breach has stepped JoelBryan. He has written a tool called Ubuntu Live Chat Support, designed to make the process of getting onto #ubuntu a mere wizard dialog away. Rock!

Update: Apparently, the man is not done. How about an easy way to handle irc links?

A world of Mono hype

Why this sudden facination with Mono apps? Mono is a great technology, but it is not the only game in town.

• Think Diva is cool? Pitivi is a more mature women.
• Beagle-ing your way to the your docs? What about Tracker?
• Your heart screaming for Banshee? Rhythmbox might beat to a better tune for you

Then there are cool Python apps like Jokosher, Serpentine, gramps and the Gourmet Recipe Manager.

Don't get me wrong, Mono is great. But lets focus on what really matters, great apps, not what they are built with.

History sliders everywhere

When I first used f-spot, I thought the history slider was crack. I thought, this is a totally non-standard and useless. Nuke it. That was then...

This is now. Now I think the history slider is one of greatest "new ideas" in computing (It is probably an old idea, just resurrected). As such, one of the greatest things someone could write would be a giant history slider program, for everything on your computer that you use: programs, documents, coversations, people, webpages, everything. To see what I had done, I would just click on a button labelled "History" and slide back to the day, then zoom into the hour or even the minute.

I have created a mockup and put it on the Ubuntu wiki at Computer History Browser. Please excuse my 10 minute Inkscape job, but I think it gets the point across. It also uses the same colours as Gimmie, to reinforce which is which.

It appears we have a release

... of GNOME that is. 2.14 was just kicked out the door today. As always, the ever-uploading seb has already got it into Ubuntu, but you will not see it in a stable version of Ubuntu until June 1st, as Ubuntu 6.04, more commonly known as Dapper, officially became Ubuntu 6.06 today.

Yours truly had a minor role in the press release of GNOME 2.14 and all typos can be attributed to my terrible editing.

Launchpad is hurting Ubuntu

Launchpad is like Dr. Frankenstein's monster. Like Dr. Frankenstein, the launchpad developers are pretty much uniformly geniuses (In fact, some of them are even quite cuddly, a quality I don't think Dr. Frankenstein enjoyed). And as with the monster, Launchpad is a great idea. But sadly, like the monster, Launchpad is a beast. It eats kittens and causes good people to tear out their hair. In fact, I would rather have my fingers gnawed by rabid chimpmunks than use Launchpad some days. Today was one of those days.

But first, I should provide some history. I have been involved with Ubuntu since almost the beginning. I have watched Launchpad under development all this time. Schedules kept slipping until finally, Malone was turned on for all Ubuntu bug tracking a few months back. Soyuz was turned on shortly thereafter, and Rosetta has been running for awhile.

In the beginning, I liked Launchpad and used it a fair amount. But as time progressed, I watched with frustration and then sadness as persistant issues were not fixed, such as the way Malone lists bugs. This eventually led me to give up doing any bug triage in Ubuntu. I know I am not alone in this.

So how does Launchpad hurt Ubuntu? Well, its UI makes it much much harder to do simple tasks. For instance, today I was looking at my synaptics touchpad issue. After several minutes of fruitless search (another longterm annoyance with Launchpad), I finally found the bug report. I also found another bug report, identical. So I preceded to tell Launchpad that the new bug report was a duplicate of the old one. Except not, because the new one already had another duplicate bug. So I had to change that third bug and then the new report. This kind of crap costs me and the developers time. And I don't have to "experience" it as part of my work. (I get to "enjoy" Netsuite, but that is another issue)

So where do we go from here? Honestly, I have no idea. Launchpad is too entrenced to realistically consider another option. The damage it has done has been done. We need to find some way out of this self-inflicted morass of good ideas and broken implementations. I just hope we do, for the good of the entire Ubuntu project.

PS. This really has nothing to do with the license (or lack there of) that Launchpad is under. No, really. The UI would suck just as hard if it was open source.

Your desktop is a mess, please bug the developers to do something useful with it

Your desktop is a ghetto. It is a wasted and unloved space. It looks like the your average inner city, run down with garbage everywhere.

Why am I making this bold prediction? Because I have watched Average People (TM) use the GNOME desktop. Their desktops are either in one of two states: So clean you could eat off of or a random collection of files and folders. Either way, not very useful.

So how do we redevelop this slum of your computer? I think SymphonyOS has some interesting ideas. Rather than using the desktop to store stuff, why not use it display information? It is doing the former very badly anyway.

Anyway, you can go back to making your desktop a mess. I know I will.

Gervase, I will agree with you completely. As far as I can tell from the Ubuntu side, we really only gained one contributor, the excellent Manu Cornet. We also had similar issues with people not being part of the community and thus not really joining it during or after. However, I would add the following suggestions:

1. Code should be developed in a public source repository
2. Projects should post a clear page with who is working on what, and what the status is
3. SoCer's should post status reports regularly (ie, more than just once or twice)

I should note that I was not involved with the planning or execution of any of the Google SoC stuff here at Ubuntu, this is merely an outsiders view. I should also note that Google is not really at fault here. One of the problems that I think Ubuntu ran into was simply that all the development team was too busy to effectively tutor the SoC people. Again, that is just an outsiders view. I could be completely wrong.

However, I don't think there is anything systemic wrong with Google SoC that is not fixable this time around.

Saddened

We in the Free and Open Source software community spend a great deal of time talking about free codecs and supporting those codecs, at least in words. So why on Planet Gnome do I see three different examples of people demoing cool new things with non-free codecs/tools?

Update: Make that four

Apparently my blog post from July, about the consoles and bugs is called "Ubuntu 's Haiku" on a French forum. I feel so wise.