Monday, 8 March 2010

Music hath charms to soothe a savage breast, to soften rocks, or bend a knotted oak.

Since I last mentioned music back in January I have accumulated another ten albums and unlike last time where there were only a couple of stand outs, this time I have he opposite problem.

The unordered list:

Justin Sandercoe - "Small town eyes"

I am learning to play the guitar, I have been using Justins' course, it is very very good, this album? Also very good. If you like melodic guitar lead music with varied influences this is for you. A couple of tracks made me immediately think of some Crowded House riffs (which is not a bad thing). Only minor niggle is the uneven levels on some of the louder pieces, but it really is a minor observation on an otherwise fine first album.

Molly Lewis - "I made you a CD, but I eated it."

Although this is only a short selection of original material from Molly, it is a very promising first album. I really like her voice and although a ukulele is not generally the most well respected of instruments, in her hands, it has an odd charm. This album is available from DFTBA records.

Rhett and Link - "Up to this point"

A pair of talented comedians who use music very effectively to highlight their humour. I originally stumbled across them on youtube and decided to take a punt. The album is 27 short pieces which fit together surprisingly well. Difficult to categorise but think a cross between Flight of the Choncords and Jonathan Coulton with a dash of youtube immediacy.

They Might Be Giants - "Flood", "Apollo 18" and "John Henry"

Strictly a replacement of the old tapes which have completely disintegrated in the intervening couple of decades since first purchased. Flood is still one of my favourite albums ever, certainly in my top 10. If you do not know them TMBG are just ace, please try their music!

Seasick Steve - "Started out with nothin and i still got most of it left"

Well its a kinda fun album primarily based on blues electric "guitar" (some of the instruments are little more than a stick with a nail in and a guitar pickup.) Nothing bad, easy to get along with, definitely worth a listen.

La Roux - "La Roux"

This synth pop album was on remainder in ASDA and I took a gamble. Its OK I guess and for 3quid I cannot really complain.

Red Hot Chili Peppers - "By the way"

Not their best, but competent enough.

Aqua - "Aquarium" , "Aquarius"

Um...yes, I have a soft spot for 90's cheese OK? Nothing more than a gross self indulgence of my silly side. But they are fun ;-)

So that is my new music since January all 166 tracks of it . Most of it pretty good, certainly no lemons (well aside from the Aqua but that is supposed to be silly!)

Oh and The XX has really grown on me from last time and I am looking forward to their next release.


Well my last post elicited a response from Mr Lougher the squashfs author. Just not one I was expecting. Apparently he did receive one of my emails (I sent five in total) to which he has not replied as I have accidentally come across as critical.

This is absolutely not my intent and I wish to publicly say that, It would have perhaps been more constructive to actually tell me this by email and this misunderstanding could have been avoided.

For reference The final email in the series is reproduced below, if I have been overly unhelpful please let me know in the comments so I can avoid this mistake in future.

Hi, we are using Squashfs and have come across several issues. We
initialy tried to use the Debian source package of the 4.0 release but
then moved to using the the CVS edition which fixed some bugs but had
issues of its own.

Couple of things to start with:

- The commit you made recently titled "Change get_basename() to use
getcwd rather than getting the PWD env var." does not seem to be
what you intended?

- I would like to assist in improving these tools so they work better
for our use cases. To aid in this have you considered updating the
revision control system the project is kept in? an SVN or GIT
repository is much easier to work with than CVS.

- I will probably assist with maintainership of the Debian and Ubuntu
packages (I am a Debian Developer ) and would
like to bring a couple of patches to your attention, one to avoid a
division by zero error and a second to enable building with
alternative libc. These are both attached to this mail.

- The tools currently make a number of assumptions about structure
alignment which are incorrect on some architectures. I am crafting
a patch to address this which should make the tools work correctly
on ARM (they currently simply segfault).

I hope this is seen as constructive and we can work together to
improve this software. If you do not feel you want to interact with me
and choose to take a differnt route, please let me know at your earliest

I do hope that Mr. Lougher will accept that I intended to be helpful and constructive and not cause offence. I have chosen to use the blog format for this as that is the form I made the previous complaint and also email between myself and Mr. Lougher appears somewhat erratic.

It should be noted that some of there points have already been addressed, however there are several more issues which I hope to be able to contribute towards.

Wednesday, 3 March 2010

Limited Success

I spent the day working with Daniel. He has been trying to get the correct runes for a GCC build in a strange environment. Turns out all he needed was me to throw stupid ideas at him until one caused him to examine some of the generated intermediate symbol maps...

Turns out that busybox awk fails in an interesting way which (eventually) causes half the symbols to be missing in the stage 1 object. Switched to using gawk and managed a successful build.

Seems to be a relatively good time for solutions, last week I managed to make squashfs-tools work on ARM. Unfortunately the author seems to be completely unresponsive to my (repeated) emails (if anyone wants the patch series feel free to mail me). The main issue with the tools is their complete blind assumption that a c structure can simply be cast to an arbitrary character array and get sensible results.

The structure casting method simply does not work on ARM (or indeed any platform which has alignment restrictions) without at least adhering to some minimal rules, primary among these is that the structure *must* be aligned to a word (32bit) boundary and that if you have 16bit or 8bit quantities within the structure they may cause gaps between members i.e. the structures are not packed by default.

No this behaviour is not a bug, it is perfectly acceptable by the c standard. Just because on x86 the practice of arbitrary casting works does not mean it is safe or sensible. Nonetheless many programmers simply do not want to believe its wrong.

Anyhow, rant over, back to work!