Crap Nullpointer Exceptions

This is the blog of John Dulaney, a hacker of Fedora, SCAdian, player of Music, blacksmith, sailor, and consumer of Bacon.

Category Archives: Fedora

Fedora Internationalization and Localization testing

Hey, all.

Please support Fedora’s Internationalization/Localization teams by helping to test translations.

Live images may be found here:

There are three days to choose from:  30 and 31 August, and 1st September.

Instructions are included on the test day Wiki pages:

I personally will be testing French (the question is whether or not it will be a test of my French).

Thanks much for your help!



So, I have not been sleeping well for the past couple of months.  In fact, two to three hours of sleep is the norm, and only one hour is not uncommon.  This is not a new phenomenon; I go through basically the same thing every summer.  As it happens, I am sitting here at 2 AM, writing this and listening to Neil Young.

So, I wind up being rather productive.

There is something substantially different about this year, though.  I am no longer living where I grew up, but rather in what was my Grandfather’s house prior to his passing.  So, this means no more sitting on a lifeguard stand at the beach and pondering Life, the Universe, and Everything.

Anyway, my big project of late is testing AutoQA.

What is AutoQA, you ask?  It is the software suite in development by Fedora’s Quality Assurance team that is designed to test Fedora.  You can read about it here:

Any way, in order to test the test software, several things are needed.  Firstly is a mock-up of the Fedora build system (Koji and Bodhi), although the entire thing is not needed; just the back-ends that actually do the work.  I am using the actual back-ends and interfacing with them as necessary.  Secondly, one needs an instance of AutoQA running to be tested.  This is pointed to our mock-up of the build system as if it was the build system.  No new code needs to be integrated into AutoQA itself to handle this.  And third, test cases with predictable output are needed.

So, how does this mock-up work precisely?

1. Read test case from file (plain text) that describes location of packages to be run through the AutoQA process and the predicted output of AutoQA if those packages were put through the actual build system.  It should be noted that for some tests, almost the entirety of (default) Fedora may be required.  From here, it is not difficult to simulate the hooks that AutoQA uses, although different ones are required for different tests.  Some of them need to be updated after AutoQA is running, which triggers certain tests.

2. Start up an AutoQA instance and pass as options the test we want to test and point it to our packages in our mock build system.

3. Wait for output to be generated and then parse it.  Compare it to expected results.

So, none of this is really difficult.  The hard part is in creating test cases that thoroughly check AutoQA.  We need tests that will pass AutoQA tests, tests that fail them, tests that fail because of one dependency error, tests that test to see what happens if there is a failure in the build system.

Want to help?  We could use it with test creation, and Quality Assurance in general.  It’s a fairly easy way to contribute to a Good Thing (Okay, maybe not so easy with test creation, but testing software in general is not that difficult).  Go to to get started.


P.S.  Since I wrote this this morning, I’ve checked my email.  Tim Flink has also been working on this project and has got some of the back end stuff I hadn’t done yet up on Git Hub:

Why I use AutoCAD instead of QCAD

A couple of reasons, really.  Number one on the list, though, is that QCAD is useless with my drafting style.  Let me preach on it.

In AutoCAD, one can simply hit ‘l’ (or ‘L’, but the extra keystroke is not necessary) on the keyboard to draw a line.  No clicking in the text box at the bottom is necessary.  This makes drawing lines very fast.  In QCAD, one has to click on the text box at the bottom, which is very frustrating.  The alternative is, like in AutoCAD, to click on the little icon in the toolbar.  But wait, rather than then being able to immediately draw a line, one is presented with choices for a bunch of different lines.  When it comes down to it, AutoCAD has the same choices, but they all get their own separate icons.  It’s just one click, and boom, lines can be drawn.

But wait, it doesn’t end there!  When one is done drawing the line(s), one cannot simply hit <Enter> a couple of times and get out of line drawing.  It has to be manually done by the little icons on the left hand side.

Then, we get to splines.  I use splines probably more than lines when drawing ships.  I’ll trace out the stations of a ship with splines, arrange them three dimensionally along the profile, draw in the water lines (which are also splines) and fair everything out to get a nice, smooth hull.  Well, since in QCAD all the points in the splines must snap to existing entities, it makes it difficult to trace things out.

Looking around in QCAD, it isn’t bad feature-wise for a 2D drafting program, and has come a long way since the last time I mucked about with it.  However, the user interface does not present itself very well for a professional draftsman (which definition I fit, even if in the current economy that’s not how I make my living).  For non-professional home/hobby use, it can be made to work, but I wouldn’t even think about doing some of the the things that I do with AutoCAD with it, even 2D (and considering I do the majority of my drafting three dimensionally…).

You know, it’s odd that a Free Software supporter such as myself would choose the proprietary software over the open source.  However, in this instance, the open source program just doesn’t cut it.  QCAD is probably about twelve years or so behind AutoCAD in terms of development.  Sounds like I’ve just given myself a new project…

John Dulaney

I am returned

So, I got back into town in time for poetry night, and, for once, I am not impressed. Glen was, as usual, good, and after him went this dude that was going up for the fist time, thusly deserving some slack.  Now, however, there is a ‘guest’ that is ‘rapping’, and I must admit that I am not at all enjoying it.  Normally, I’m not a huge rap fan, anyway (with several very notable exceptions).  This time, it’s some girl that isn’t even in time, much less putting out anything really original or even marginally inspired.

<Rant>If you don’t like/want systemd (and I think most distros will wind up with it eventually) then don’t use Fedora 15.  Don’t go shouting on the Devel list about how it shouldn’t be installed by default.</rant>

Now that my anoyances are out of the way, on to SELF.  I reckon I’ll break it down to what I liked:

  1. All the cloud stuff.  That will be useful for QAing F16, and that alone is enough for me to have gone.
  2. Networking.  I met up with a couple of KDE devs (one of whom also does embedded and alternative architectures, in which I also have interest) that has gotten me to install KDE.  There’s some cool stuff there since the last time I messed with it.  I’ll admit, it’s been a while.  That said, I still prefer Fluxbox in terms of workflow.
  3. The Genetics talk was supremely interesting.  I would very much like to learn more on the subject, and may at some point dabble in the code some.
  4. I missed Spot’s talk because I was networking, but that was cool (see 2 above; I also talked about real estate (random)).
  5. Helping folks understand how Fedora works was decidedly win.
  6. Speevan back was fracking awesome, except for the hangover.

There really wasn’t too much I disliked except the aforementioned hangover, but since that was self-inflicted, I can’t complain, can I?  I also managed to bomb my kernel on Friday morning, but it wasn’t difficult to fix (those that were there know what I did).

I tried to introduce Ryan Rix to grits, but he was supremely uninterested.  I consider that the fail of the weekend.


Southeast Linux Fest So Far

Sitting here at the Fedora booth, explaining Harker’s Island English, installing KDE (/me is interested in what it looks like these days) and generally having an awesome time.  Earlier today and some of yesterday I attended some cloud sessions in anticipation of all the cloud stuff that will be coming out in Fedora 16.  I figure QA will be easier if I know a little bit about what all is going on.  Later on I intend to go to a talk on Advanced Regular Expression.

Hanging out with Sparks, Spot, Southern_Gentleman, jsmith, etc. has been tres interesting.  Last night’s burn party was pretty cool, too, especially the goings-on whilst DVD’s were burning.  Hanging with these folks and the exchange of knowledge that has resulted has been very, very good.

I plan to attend Spot’s talk on how we fail later on, and then afterwards party.

More later.

New kernel, New instrument

So, what do flutes and the Linux kernel have in common?  Me.  Specifically, once I start a big compile (such as the kernel), I’ll practice playing whilst my computer is bogged down.  Since I am now learning the flute, and since I’ve just recompiled the new kernel (3.0-rc1) due to leaving out a driver during my first round, I was playing whilst my computer was compiling.

Anyway, on to the kernel.  It’s running nicely, I haven’t had any issues other than leaving out the FAT driver (as I found out when I tried to use my USB flash drive, oops).  Since I left out the vast majority of the drivers available, the kernel as I’m running it is very small and seems to be a bit faster than the generic F15 kernel.  Of course, I’m not sure how much I’ll wind up using it in practice since for testing purposes I have to use whatever kernel is in the testing repository.

In other news, looks like my friend Larry is going to have a shindig tomorrow, so I’m going to be playing (sadly, not mandolin since that’s in the shop).  I’ll also be recording, so look out for that.  Expect to hear a bunch of folk, blues, some country, etc.  I imagine it will be pretty heavy on the Old Crow Medicine Show.

On the flute, I can sort of bang out something that with some (quite a bit, actually) imagination can be recognized as Yankee Doodle.  I’ll probably bring it and see what I can do about jamming with it.  I’m not quite sure how much I’ll be able to play, since all the notes I know are in the key of G.  Ah, well, I’ll get there, eventually.

Mucking about with btrfs

Today I changed one of my partitions over to btrfs (1)(2).  At first I had trouble getting mkfs to work, but then it was pointed out to me that the tools to actually do btrfs are not installed by default unless the filesystem is selected at install time (thanks fenrus02 for pointing out what should have been obvious, DOH!).

So far, I’ve had no issues and I have dumped a good 11 GB on it.

For those that are interested, the (very easy) steps are as follows (/dev/sda3 was the partition I used):

  1. su –      (easier to just do this since there are several steps that require root permission)
  2. yum install btrfs-progs
  3. umount /dev/sda3
  4. mkfs -t btrfs /dev/sda3
  5. blkid to get the new uuid for the drive
  6. edit /etc/fstab to automatically mount where I wanted it (include the new uuid and, in my case, replace ext4 with btrfs)
  7. mount /dev/sda3 /home/jdulaney/Media
  8. You may need to edit the permissions for Media (or whatever mount point you go with)
  9. You’re done!

So far, I’ve had zero errors, and I’ll probably start converting the rest of my file system shortly.




Fedora 15

As usual, the odd number release is what I update to.  I don’t know what it is, it just seems like even when multiple new huge features that the odd releases go smoother.  Prior to my joining the QA team, this was the case, all the way back to FC1 (of course, I did skip FC2 because I didn’t realize it had been released until after FC3).

I’ve been running Fedora 15 for about a month now, and I think it’s great.  I love the new features, even Gnome 3.  The only thing that really doesn’t seem to provide me any real improvement is systemd.  systemd’s parallelization is only a real benefit on multi-core hardware, which does not apply to my graying machine.  That said, I’m not going to knock it like some folks seem to be doing.

Gnome 3 is great for what it is.  I think that this is the way Linux needs to move if it is going to capture any more market share from Windows (advertising would help here, too).  There is plenty of griping about the ‘dumbing down’ of Gnome.  I have two answer for you:  not everyone is interested in command line for everything, or even anything, and there are still plenty of other choices (my favourite being Fluxbox).  I really do believe that the simplification and GUIation of Linux in general and Fedora in particular-as long as other options are left available-is a Good Thing.

So, how does F15 look from my QA perspective?  I haven’t had any major bugs since about a week before release and only one minor one with Empathy.  This is with the testing repository permanently enabled on my box (as I write this, I am downloading more updates to test).  Of course, this newest update set might break something, but that’s what I’m here for, to find this stuff.

How does it compare to Ubuntu 11.04?  Well, besides having newer technology, Fedora lacks the little things in Ubuntu that mommick me for dear life.  There’s all sorts of writing about how Ubuntu doesn’t push upstream, but I’m ignoring that.  There is just a long list of little naggy things in 11.04 that just don’t go away.  For the most part, Fedora lacks these to begin with, and if they are there, they are very easy to kill.

Maybe this all comes out as blithering praise for Fedora, but it really isn’t.  I admit that I’ve not been entirely free of suffering, mostly dealing with the new Network Manager API (if you’re involved with KDE, Sugar, or QA, you know what I’m talking about).  But, that’s the price paid for staying bleeding-edge.