What I want from 10.04

Only two things really: More bug fixing and more hardware regression testing.

I think the Ubuntu community at large would take very kindly to pushing the release back to 10.06 so that there would be a couple of solid months of doing nothing else besides testing and fixing.

I really want Ubuntu to succeed. 9.10 was way to buggy to be a proper release, and I'm afraid that Canonicals commitment to a regular schedule will trump making it as solid and bug-free as possible.

Consider that Windows 7 had what seemed like almost a full year of beta testing, a couple of extra months to make Lucid the best Linux distro release to date would be a refreshing and appreciated change.

Consider it, please.

Demagogue

From Merriam-Webster:
demagogue: a leader who makes use of popular prejudices and false claims and promises in order to gain power

There is probably no better description for what it is that Rush Limbaugh does.

The Afghanistan Decision

I don't know what Obama has decided, but I'm very very worried. Its a terrible decision to have to make, and it will haunt his presidency either way. War is a horrific thing, and I can never really truly support a war, even if I think it is necessary. But I don't know if what is about to happen is necessary. I don't believe its largely about combat missions (though I'm sure its implied), but more about patrolling, providing security, and building a strong army/police force to combat the Taliban. Nonetheless if he increases troop levels, it will become a defining moment of his presidency.

Not only he is going to have a very hard time with the base of his party, but it might taint all the good that he is trying to do (and prevent much of it if - and when - we get mired in a long, violent struggle). The Republicans will continue to lie about him - relying on the stupidly inane memes that Democrats don't support our troops, that he's cutting the military budget, that he hasn't done enough to secure 'victory' (for whatever the fuck that is). They lie about him constantly, will continue to do so, and there is nothing he can do to change that.

I know Obama made the war in Afghanistan a part of his candidacy, I was slightly troubled by it then, but I'm especially troubled by it now. I even admire his resolve to keep to his commitments, but I'm sure this alone is not enough to go to war. No doubt this would be a promise that would be joyful to break if the conditions no longer called for it. I know the mission has changed - and being very explicit about what the mission is is of utmost importance. If its no longer about tracking down the Al-Qaeda terrorists that attacked us - and how can it be since Al-Qaeda has largely moved from the region? - then is it mainly about preventing the Taliban from regaining control (and thereby being a safehaven for terrorists)? Or is it about Pakistan somehow? And if so, wouldn't it just be better to support them directly? Sure its all of the above, but fuck, is expanding the war effort the best way? There are lots of nasty and corrupt regimes out there that don't like us very much - but that doesn't justify sending troops in.

I was very impressed when Obama came out not that long ago and rejected all of the proposals that were presented to him regarding Afghanistan. It was a sign of real leadership to demand something better, and not simply choose from a list of bad options (like Bush the 'decider'). But was a better option really presented? If not, then despite all Cheney's blathering stupidity about 'dithering', he should continue to work to find the best possible option even if it takes longer than his detractors say it should. I will listen closely what he has to say, but I'm already scared by the rhetoric being used - 'finish the job'. I can only interpret this as meaning that we need to secure a truly democratic government (as opposed to a merely corrupt sham) that has both broad appeal to the populace to engage with and also has enough power to sustain itself from incursions of the Taliban. But even when I put it in words it sound idyllic, and have a hard time believing that we can 'finish' this - even if we are there for a lot lot longer. And they know we won't be there for a lot longer. And we shouldn't be. Fuck.

Brief Review of Karmic and Hopes of Lucid

Karmic Koala is out! I've been using it though the beta release, and overall I'm really impressed. There are a LOT of changes under the hood, which has given rise to some unfortunate instability, but the direction the distribution is going in seems to be good. There are the beginnings of a visual refresh going on as well - I hope for more in the future! The main annoyance I have is the large number of serious bugs that came in Karmic. For example, I couldn't burn a CD from Rhythmbox (it has subsequently been fixed, but for a couple of weeks was an annoying regression). There are also reports of data loss do to the switch to the new ext4 filesystem (bug report here) - this makes me very worried about switching to ext4.

For the rest of this post, though, I want to discuss what I hope for in the next version of Ubuntu, codenamed Lucid Lynx. Given that this is an LTS release, I think that, along with the normal package/application updates, the main foci of this release should be: (i) being as stable and as bug-free as possible, and (ii) provide a more thorough revamp of the look and feel of the desktop.

Be as stable and bug free as possible

This is essential. Application developers have come a long way in providing the necessary functionally for getting things done using free and open source software. By in large Ubuntu has done a good job of integrating these applications into the distribution in a stable, easy to access way. That said, there are plenty of bugs that plague the distribution, and it would be nice to see a concerted effort (more than usual) to eliminate as many of these as possible. I shouldn't have to wait for a subsequent update to burn CDs from Rhythmbox - it should work out of the box (and missing obvious stuff like this indicates a lack of quality control). A small bug that I find crazy annoying is that glipper dies most of the time I login. Really?? This bug was reported over a year ago, and yet never has been fixed. Is it really that hard to find a fix for an application that crashes every time I login? Annoyed, I am.

I think that the decision to sync with Debian testing rather than unstable is a good decision in that various communities can cooperate on eliminating as many of these as possible at the same time (and, as implied by the name, there will hopefully be considerably more testing as well). That said, given that this is the first time Ubuntu developers have decided to sync with testing I'm sure unanticipated issues will arise, and they should not be unwilling to change the release date from 10.04 to 10.06 to make it as stable as possible.

The main reason for this is that if Canonical is serious about partnering with the corporate world, they need releases that can be relied on for a long time to provide a stable computer platform. And Hardy was a failure in this respect - annoyingly buggy to an extreme. But to this day I have to use Hardy on my Mini 12 because this is the only version of Ubuntu that Dell supports on this netbook (I have come to hate GMA 500).

I'm also wearying of updating my computer every 6 months. I want a stable operating system that I can rely on for a year or two without problems or a need to update. This brings me to another problem I've run into. Some applications require regular updates to remain functional. Yahoo, for example, changes its IM protocol on occasion - breaking Pidgin. When this occurs, Pidgin should be automatically updated to restore the functionality before the change. I shouldn't be required to sync up with an experimental PPA to get it to work again (in fact, I don't even think I should be required to use the unsupported backports repo). And this should apply to any application that would have problems like this. Don't leave me less functional because I want to continue using a stable supported release over multiple release cycles.

Provide a more thorough revamping of the look and feel of the desktop.

If you nail down the bugs, I'll be happy. If you make the user experience more satisfying and the visuals more pleasing, then I'll be enthusiastic. I can already start seeing some of the work of the design team that Shuttleworth hired to work on the user interface (for example, i love the new icons). The colors are better, but haven't yet found a sweet spot, and the desktop feels a little to blocky rather than slick. I don't exactly have a lot of good ideas, but I'll know it when I see it. I for example really like the refreshed look of KDE 4, though I'm not really inclined to use it.

The 100 paper-cuts was an excellent idea. I hope this will be continued! It definitely improved the distribution in lots of little respects. Do it again! In fact I'd say expand the idea a little. Maybe target, say, 75 trivial fixes, 20 fairly simple (though not trivial) fixes (like this one in particular), and 5 larger fixes. If all of these were related to the user experience, then over the next couple of iterations of the release cycle, huge strides in usability will be made.

Once people see how slick, functional, and easy to use Linux is, they will realize that, with the exception of things like gaming, they have very little reason to stick with Windows and might even get hooked on Ubuntu

Fuck, lost a lot of pictures!!

I'm so pissed at myself. My computer totally crashed (because I was using a development version that wasn't very stable), and I forgot to backup my fucking pictures. I'm super pissed. For future reference, here is what I need to backup on a crash or before big updates:

Photos folder
Music folder
dev folder
.purple folder
.tomboy folder
.tmp2 folder
scribefire.sqlite file

Splitting mp3 files on Ubuntu (esp useful for audiobooks)

check out mp3splt. Its not in the Jaunty repos, but it is in Karmic!

Tools lacking on Ubuntu

Here are some tools that Ubuntu, or Linux generally, seems to lack:

A good WYSIWYG html editor.

• Up until about a year ago, I was pretty satisfied with quanta plus. Since then I have become increasingly disappointed with it, mainly because no one seems to be porting it to Qt4, or developing it substantially. What had such great potential seems to be languishing.
• KompoZer crashes a lot - I'm waiting for the new version to come out (still in alpha), but I'm less than certain it will be better (and NVU is dead).
• Bluefish - although its not a WYSIWYG editor, I still think it could be good if they'd ever release a new version that updates a lot of antiquated crap in there (like not knowing how to open a browser easily)

Decent blogging software

• Mainly I use scribefire (I'm using it now) - its ok - but it only offers the simplest options for blogging. If I want any type of formating that goes beyond a bullet point, I've got to know my html (which it visually formats in a very unreadable fashion).
• There is an extension for OpenOffice called Sun Weblog Publisher. With it, its possible to format your blog post in openoffice and then upload it to your site. Unfortunately, its closed-source, and only seems to be updated sporadically when someone at Sun feels like it. If I knew there were an active developer community, then it would make me somewhat happier, but it doesn't seem like there is. I expect it to become increasingly antiquated the longer its not updated (and I wish they'd freaking open-source it).
  

Ubuntu LTS Support

I'm beginning to get frustrated with certain aspects of the Ubuntu development cycle, and in this post I'm going to try my best to spell out my complaint and what I think should be done about it.

My Complaint

First, let me say that I'm a huge Ubuntu fan. Since Gutsy (when I first started using Ubuntu) each of the iterations of Ubuntu has made large, important strides to becoming a preeminent OS. Many of the things that annoyed me about the desktop experience have been have been fixed or minimized, and with the release of Jaunty I've had very few complaints about things not doing what I want/expect them to, very few crashes, and less annoyances than at any other time using a linux based OS. As I've written before, Jaunty is the first time a Linux vendor has released a distribution that I'd pretty much recommend to anyone. As you can see, I'm a very happy user.

So my complaint then isn't about the great work being done on Ubuntu, its rather an end-user complaint about what the distribution requires me to do to use the latest and greatest releases of software applications. In particular, my complaint is specifically about LTS releases. Take Hardy, for example. Hardy, being an LTS, has three years support, whereas the non-LTS releases (like Intrepid, Jaunty, and Karmic) only have 18 months support. In both cases the support is pretty much limited to security fixes and does not include application version updates except through the Backports repos. Even then, the Backports repos offer a limited number of application updates often not including the application updates that I most desire (like the newest versions of Openoffice), and they are accompanied with the proviso that they are not supported by the Ubuntu team. This the background, now let me get to my complaint.

There are two main types of Ubuntu end-users, which I'll refer to as enthusiasts and lay-users. The enthusiast does not mind reinstalling or upgrading to a new version of Ubuntu each release cycle. In fact, they often start early with one of the alphas to watch the progress of Ubuntu development (this is what I do). They are generally careful about backing up their data and they generally are more familiar with the risks of updating than lay-users. Lay-users, on the other hand, don't want to have to worry about reinstalling or upgrading an OS, they just want what they already have to work, and they want to be able to easily update certain applications when newer versions come out. Not only are less aware that updating can go wrong in any number of ways, and often do not properly back up their data, they don't want to spend hours updating, and they definitely don't want to trouble-shoot anything when things go wrong. But the only way they know how to get new versions of applications is if they upgrade or reinstall.

This is where LTS releases come in, or at least where I think they should come in. Given their extended support they would be perfect for lay-users - except that their support is way too limited. Security patches are needed, but they are not enough. LTS releases also need to be able to allow for the easy installation of new versions of programs. This way lay-users can install an LTS version of Ubuntu, and then not have to concern themselves with the six month release cycle of the Ubuntu developers. They can live content with it until a new LTS version is released.

Solutions?

In my view, LTS releases should work in the following way. As it does now, an LTS release should offer a stabilized API for developers to work with along with a continuous supply of security fixes for the supported software, but it should also widen what it considers supported software for an LTS release to include newer versions of important programs that are released through the LTS lifecycle. I'm thinking of certain programs in particular, ones like OpenOffice, Firefox, and Pidgin. There should be an option to upgrade to the newer version of the programs, but no compulsion to do so. In other words, Ubuntu should offer a way to update to the newest versions of certain programs (which programs can be a mtter determined by demand) without relying on more risky options (one's they don't often even know about) like installing PPA repos, using getdeb, or waiting for backports to offer it (which it may never). Various versions of these programs need to be officially supported in an LTS release so that users can gain the benefits of these updates without having to upgrade their entire OS or rely on less safe deb repos. In this way, lay-users can be happy with an LTS for years without becoming frustrated that they are not getting the latest applications developers have to offer. Let the developers worry about the changes in undelying OS, this is largely invisible to most lay-users anyway. Lay-users care mainly about the applications, and they want an easy way to keep up with new releases of them.