DownFromTheTrees

Don't Jump Blindly to Upgrade Complex Software

The MATE desktop crew released their new 1.6.0 upgrade a few days ago. I've been running it on Ubuntu 12.04.02 LTS. So far, I'm pretty impressed: Smooth, fast, and reliable. A definite improvement over the 1.4 release.

However, the update from 1.4 to 1.6 is causing real problems for many people. Apparently, the 1.4 -> 1.6 migration wasn't actually tested before the release of 1.6. In addition, it seems some 1.4 users may have been upgraded to 1.6 when they thought they were only doing routine updates to 1.4.

So, yes, the 1.4 -> 1.6 migration should have been thoroughly tested. And, no, users ought not to be able to trigger major version upgrades without their express, and obvious, consent.

That said, jumping quickly and blindly into updating any large and complex software system is something to be avoided.

Even with the best intentions and the most thorough testing, no one has any idea what's really running on all the user machines that will be upgraded. Developers cannot know what's been installed on your machine that might blow the upgrade. This is especially true with open source systems like Linux.

Here are three strategies for coping with big upgrades like MATE 1.6.0:

1. Wait. The Dot One version, of whatever it is, will be along presently, with bug fixes and tweaks. Plus, you will be able to benefit from the experiences of other people who updated early, had problems, and found fixes.

2. Don't upgrade in place. Instead, do a clean install.

3. Maintain a separate /home partition and stash configurations files you've edited/created somewhere safe.

My /home partition is on a separate partition (on a separate drive, as well). When I edit or create a config file that resides outside /home, I stash a copy on Dropbox.

So, to update to 1.6, I did a new install of Ubuntu, chose manual partitioning, mounted but did not format /home, installed MATE 1.6, installed Dropbox, copied a few files to the right places, and was good to go.

I've used this approach for a long time on many distributions. While I do upgrade to point releases (e.g., 9.2 to 9.3) in place, I just avoid in-place major point upgrades of complex software like MATE. There are invariably problems, and the best problems are the ones I don't make for myself.

Waiter! There's Pork In My Elk Lasagna!

Europe's been having a problem with horsemeat showing up in places where it ought not to show up. Like hamburgers and sausages.

That's not really a health threat. Most people, though, don't think it's very appetizing and, if done knowingly, it is fraud.

Now, Ikea has a problem with its elk meat lasagna. (Lasagna made from elks sold by a furniture store might be a problem all by itself.)

The Swedish firm is withdrawing thousands of elk meat lasagnes that are contaminated with pork. Again, no known health issues, but a serious issue for those with beliefs that forbid pork consumption.One test showed one batch of lasagna contained 1.4 percent pork. Perhaps the grinders were not properly cleaned? Perhaps we have better measurement techniques that can detect smaller amounts of contaminants? Perhaps we've all been eating pork in our elk lasagna and not knowing it?

Maybe I'll just stick to buying DIY bookshelves at Ikea.

MATE: Open New Apps Centered On Screen

Got the new and very nice MATE 1.6 release installed and running? Want applications to open in the center of the screen, instead of over in the left corner? Open a Terminal and execute this:

gsettings set org.mate.Marco.general center-new-windows true

That's it. You're done.

Is Internet Radio Linked to Ignorance of Local Issues?

Survey says: Fewer people are listening to music from AM/FM radio and more people are listening to music from internet radio.

No surprise there. But, I wonder what that means for the future of local news production, and for a citizenry sufficiently informed about local affairs?

Many (most?) over-the-air radio stations are still going after local audiences. Even if their local listeners hear only a few brief seconds of headline news about local issues, that's certain to be orders of magnitude more than they will hear on internet radio. Internet radio is, by its nature, global in scope. No local news there. No news at all, in my experience.

Don't Care About Google Reader, But I Care a Lot About RSS

Google's announced plan to close Google Reader means little to me. Reader is a tool that delivers something I rarely need: synchronization.

Comment elsewhere on Reader's eventual demise breaks down into two broad categories.

One camp is outraged that Google is going to take away a free service they depend on.

That is outrage spawned by naïveté. We live in a market economy. What businesses provide, they stop providing. (BTW, the "free" part of Reader is of little relevance here. Google could just as easily kill it if they had been charging for it.)

The second camp of reaction sees RSS as an outmoded technology that ought to, and will be, shoved into the background as we all come to depend on commercial vendors rather than discover and manage our own RSS feeds in our own RSS applications.

I don't agree, at all.

The key, fundamental, advantage of RSS is that it is an open technology that can be used by anyone to make their content available to anyone else. RSS is in keeping with the openness and freedom of the web. The commercial and proprietary vendors some are touting as replacements for RSS are the antithesis of that openness and freedom. They are to the web what cable news is to TV.

For example, Justin Williams argues that the "future of RSS isn’t in the feeds itself", that his "hope is that people don’t use this second chance at a decade old technology just to build NetNewsWire with popovers, a Tweetie-like sidebar and Twitter and Facebook sharing."

I have close to 500 feeds in my RSS app. I like that. I want to discover and control what feeds I read on a regular basis. I do not want, as Williams suggests, to leave that up to commercial lowest-common-denominator stovepipe vendors.

Williams lists Twitter, Flipboard, and Prismatic as his primary sources of RSS content. I've tried all three. I cancelled my Twitter account last year. Gems there may be in that morass of muck, but I lack the patience to troll for them. Flipboard is cute with the way it presents an RSS feed as a web page in an old-fashioned flip book. Unfortunately, Flipboard finds little content of interest to me, and what it does find makes it read something like Parade magazine. Besides, flipping those pages quickly becomes tiresome. Prismatic assumes the user is signed up to Twitter and Facebook. I'm not, so it's useless to me.

Here's the thing: I have been a heavy RSS user, professionally and personally, for years, since well before Reader and synchronization buzz came on the scene. Because I scan through hundreds of feeds on a daily basis, I do not want to suffer with the inevitable slowness of a web application. I want my reader to show me only the text of an item unless I decide I want to see the piece as it appears on its site. I want to be able to move rapidly from feed to feed, and item to item, preferably with the keyboard. I want a dead simple way to add new feeds. (Those I can discover on my own, thank you.)

In the end, as long as sites generate RSS feeds, I'll be able to continue using RSS just as I am. The threat on the horizon is this: Sites abandoning RSS and replacing it with proprietary deliveries via commercial vendors. Commercial sites could readily switch to an exclusive tie-in with a proprietary stovepipe for the prospect of added revenue. Non-commercial sites and blogs might just as easily sign up with a bundling service for the prospect of even a few pennies of revenue.

That's a world of content distribution controlled by commercial vendors in the interests of commercial vendors. In the end, that's a world of content creation controlled by commercial vendors.

Dear Debian, I'm Not Morally Obligated to Report Bugs

Debian Project Leader Stefano Zacchiroli says bug reporting is a moral obligation for open source users.

I kinda take offense at that. I usually do when strangers evangelize me.

People would report more bugs if it wasn't so annoying and time consuming: Get an apport/abrt alert, click "Yes, I'll report it", wait and wait for data to be gathered, click "OK, send it", be confronted with a Bugzilla/Launchpad page asking you to log into your account.

What account would that be, eh?

Bug reporting is set up with developers in mind. Hence, the hassle. Hence, the assumption the bug report is coming from someone who has registered an account with a specific bug collecting site.

All part of the notion, always lurking just beneath the surface, that open source is more about belief than it is about making better software. Easy, then, to obligate the faithful with occasional small inconveniences.

If Debian, et al, wants to see more users creating more bug reports, they need to look at it from a user's perspective and make it happen quickly, easily, and unobtrusively.

Shuttleworth & Canonical Should Leverage Ubuntu One To Sell RSS Synching to Replace Google Reader's

A good number of people are in a tizzy because Google says it is going to pull the plug on Google Reader, the web app that pretty much every consumer of RSS used to synch their feeds on multiple devices.

Plenty of pretty good RSS applications exist to run on your desktop or your device. The problem arises for people who really need that synching ability. So, here's a pitch to Mark Shuttleworth and Canonical: Sell RSS synching at $5 per month. You know those servers hosting Ubuntu One? Set aside some of them for this. Don't bother about doing a web app like Reader. No one liked it, anyway. You're already synching customer data with Ubuntu One. Here's a chance to pick up some cash synching our RSS feeds.

Pointless Tech Stunt-ery? What's The Point?

Here's some guy who did all his work for a year on his phone. What's the point? Are we surprised he could manage it?

Do everything for a year without the phone. Now, that would be impressive.

Lockheed: Working Fusion Power Plant By 2017

Lockheed says it can have a prototype fusion power plant running by 2017. They seem to have developed a way to use radio waves to heat and control deuterium gas that allows better control than other approaches. I hope they're right, and I hope crazies who don't know, and don't want to know, the difference between fission and fusion, stay out of their way.

Without Gene Modification, We'd All Be Amoebas

The Guardian runs a good piece about Mark Lynas, an anti-GM radical who found science, changed his ways, and lost his friends.