More on xpra

My ticket (#79) on the xpra trac was closed as INVALID (a status which annoys me, because it fails to reflect sufficient granularity in why a ticket was closed). Boo.

The comment from Antoine was that xpra should be installed system-wide, in which case no funky

sys.path.insert(..)

is required. That’s true, and a great thing indeed. However, if you do not have root on the box(es) you want to run xpra on, you do need to futz around with sys.path, or you need to write a wrapper script which sets up PYTHONPATH for you. This is exactly the situation I found myself in earlier today, when I made xpra available on our network via an installation in my homedir.

Now that I’ve got that off my chest, it’s time to turn to the problem of the two shared objects which are part of the distribution. I’d like to make xpra available as a blended (in IPS terms) package. That means we need to have some way of putting

xpra/wait_for_x_server.so

and

wimpiggy/lowlevel/bindings.so

under $MACH. Guess I’ll be learning something about Cython and distutils.







A big day

Today’s the day- after J’s taken the kids to swimming lessons, there’ll be time for a cuppa and then it’s off to the Wesley Medical Centre for her Stereotactic Radiosurgery. She’s due in there 40 minutes before the start of the procedure, the procedure itself takes an hour … and then she’s done.

There’ll be annual followup MRIs for at least a few years, and it could take 2 years before we know whether it’s worked or not. At least she won’t have facial nerve paralysis, or loss of hearing in one ear as a result of an invasive procedure.




All done…. and now we wait

Um. Well, that was easy. Got J to the clinic a few minutes early at 11:20, went off to Indro with the kids to wander around Dymocks and get some lunch. While we were in Dymocks J rang to ask us to come back earlier so I could be there when the specialist gave his post-op talk. Had a quick lunch then went back to the clinic. After talking with the specialist (followup appointment in 2 weeks) we went back to the car and by 1:10pm we were on the way home.

Boggle.

Here are some pics of the mask that they made last week to keep her still in The Machine:

View of the front halfFront and back together

Photos processed with Darktable using a build from git rather than the 0.9.3 packages I put up last week.




pssst! wanna darktable package set for Solaris 11?

A few months ago I got Darktable built on my Solaris 11 Express system. Between then and now I was distracted by $DAYJOB (getting Solaris 11 ready for release seemed to take a few cycles ;->), so it wasn’t until last Friday night that I got around to trying the new version, 0.9.3.

Having learnt something from the previous effort, I used gcc 4.6 from the start, for all the prereqs and darktable itself. I want to make this available to others, too, so I’ve spent a few hours building IPS packages and getting things just right. They’re not _quite_ 100% yet, a few rough edges remain. However, I think they’re generally ready to offer up for feedback.

I’ve built them against Solaris 11 FCS on the x64 platform; they might work with OpenIndiana build 151 too.

You WILL need to add the SFE repo so you can get the GCC 4.6 runtime. To do this, utter

# pkg set-publisher -g http://pkg.openindiana.org/sfe sfe

Then download the gzipped p5p archive for Darktable 0.9.3, and utter

UPDATE: Shawn’s comment alerted me to a need to change the syntax here:

# gunzip /path/to/darktable-0.9.3.p5p.gz
# pkg install -g /path/to/darktable-0.9.3.p5p 'darktable*'

This will pull in the prereq packages that I’ve identified and should add the darktable elements to your system’s gconf schema.

Here is the content of the p5p archive:

$ pkg list -g /scratch/web/htdocs/Packages/darktable-0.9.3.p5p

NAME (PUBLISHER)

VERSION

IFO

darktable/darktable (JMCP)

0.9.3

i--

darktable/library/OpenEXR (JMCP)

1.7.0

i--

darktable/library/exiv2 (JMCP)

0.22

i--

darktable/library/flickcurl (JMCP)

1.22

i--

darktable/library/ilmbase (JMCP)

1.0.2

i--

darktable/library/lcms2 (JMCP)

2.3

i--

darktable/library/lensfun (JMCP)

0.2.5

i--

darktable/library/tiff4 (JMCP)

4.0.0

i--




Need to work on the heart rate

Did a shortish ride around the district with Tim this morning. Rather than going over Arrabri Ave and back via Mt Ommaney Dr, we headed left up Mt Ommaney Dr so we could freewheel down the hill to the Jindalee boat ramp. I really noticed my lack of fitness – had to stand all the way up to the summit, and my hr didn’t recover to my normal spinning level for about 90 seconds. I’m going to have to do lots of this before I start trying to train up Mt Coottha.

MtOmmaney_Dr



Happy to knock a minute off

Last weekened I did a 40k ride into town and back. Rather than stopping for coffee at Cup in West End as I frequently do when riding with friends, I thought I’d see how I went without any more stops than required for traffic lights and intersections.

I did almost the same ride again today (less 80m), and was interested to see that I was faster over that distance by about 100 seconds. Also, while my pace was better (22.4 vs 22.1 km/h), I burned fewer calories.

For a giggle, I looked at the last 4 rides I did along that course, where I’d ridden more than 40km, and then the last 4 rides I’d done over that course with at least 39 km. That covers 7 weeks or so, from 1st December. I’m intrigued to see that while my average pace has improved a little (19.5 to 22.4), my calorie count has decreased. I don’t feel like I’m more fit, and my weight is still fluctuating between 87.5 and 88.5kg on a daily basis. I still need to move more (note to self, GET ON THE TRAINER DURING THE WEEK), and I am trying to eat less.

last4_40krideslast4_40ishk






Stereotactic radiosurgery

We went to see an oncologist today, to get a second opinion about what could be done re J’s meningioma. After the pre-Christmas teariness seeing the neurosurgeon, it was really nice to get some potentially good news.

J’s going to get stereotactic radiosurgery. We’re not sure whether that’ll be all in one session, or fractionated (multiple sessions). We’ll know that after they’ve done the planning stage and performed their simulations. She won’t need to go into hospital for it, just rock up to a clinic on as many days as required.

Compared with surgery, this form of treatment has major advantages. With surgery the skull is opened, you need to spend 6-9 hours on the table, up to a week in ICU post-op, then around 3 months of rehab. With stereotactic radiosurgery, they don’t crack your skull open (the beams are highly and specifically targeted to the correct places within the brain cavity), if you can get it done in one hit it’s ~2 hours of being in a mask; if you need fractionated doses then it’s X times 15-20 minutes. The likelihood of hearing loss and facial nerve paralysis (let alone any other cranial nerve impact) is significant with surgery, but the chances of that happening with stereotactic radiosurgery are significantly less.

We’re a bit apprehensive about it all, but feeling generally positive.