« Back to Slack? | Home | The fundamental attribution er... »

今始まる物語

Wed 2 Jan 2013 by mskala Tags used: , , ,

The title is a song lyric; it means "the story that starts now," and that's more or less where I feel I'm at. A lot has happened between mid-November and now, and I'm hoping that this will mark a boundary or change in the conditions around me.

First I went West to visit my family for a week, on the occasion of my grandmother's 90th birthday; then I came back to Winnipeg, and then a week later (start of December) I went to Toronto for a week. My main excuse for the Toronto trip was to take the Japanese Language Proficiency Test (N4 level); but I also spent some time visiting friends there and in Hamilton.

My current job with the University of Manitoba had been planned to last exactly two years, expiring the end of December, and then I'd leave Winnipeg to go home. Home is the place where people knock on my door and I don't run around knocking on doors. I don't know where that is, and my hope of finding a new job by the end of 2012 - at at least the status I earned five years ago, if not quite the level I've earned today - didn't pan out. So here I am in a weird transitional state. My employers at U of M have offered to extend my contract as long as I want to stay (I understand this as months, not years), which is a Hell of a compliment especially because they've already hired my replacement and will have to pay us both for as long as I'm hanging around. I'm very grateful to be appreciated, but I still need to plan for the longer-term future which I think won't be in Winnipeg.

The most immediate problem created by my funny status was that I couldn't stay in my apartment, on which the lease expired at the end of 2012. Anywhere else, the standard deal with an apartment lease is that you have a fixed-term lease for the first year and then after that it automatically becomes month-to-month (i.e. the tenant can leave on one or two months' notice). If the landlord and tenant want to agree to another fixed-term lease, they can, but neither party is allowed to require that and so landlords usually offer tenants inducements like not raising the rent if they agree to extend the lease.

That's not how it works in Manitoba. Here, landlords are allowed to just keep demanding year leases. Coming up on the end of 2012, mine told me I could either agree to stay for all of 2013, or leave, and given that I hope to have a job and move away from Winnipeg a lot sooner than December 2013, I left. (This is a simplified description of a very complicated situation.) I found a new apartment that I could rent month-to-month and leave on a month's notice, and I spent much of December arranging to have my possessions moved, carpet cleaning, other cleaning, filing change of address paperwork, dealing with telephone and Internet hookups, sorting out the electricity situation, getting my mail forwarded, and so on.

My CIBC VISA card was due to expire and rather than accepting the replacement, I called them up and cancelled it. I'm still angry at them about the DVDsoon affair (since when I think I've used that card exactly twice). I'd been feeling pretty unhappy about things that day and I felt a lot better after I cancelled and shredded the card. Unfortunately, I don't have enough credit cards that I can cancel one every day.

I finally walked out of Seven Evergreen for the last time on the day before the Solstice. Since then I've been holed up in my new apartment, slowly unpacking boxes, poking around on the computer, seeing noone, and thinking about graph theory. Today the temperature warmed up a little; it is now 9 below zero. I wanted to get outside, so right now I'm sitting in a coffee shop typing this on a laptop. Tomorrow is my first day back at work.

When I booted up my main home computer after the move, it reported that two of the four partitions in my RAID5 array were failed. A RAID5 array, of course, is designed to be able to recover seamlessly from the failure and replacement of any one of its component hard drives. Lose two, and you lose everything - if that is really the situation. But there were a lot of reasons to disbelieve the report, not least being that I also have a RAID6 spanned across other partitions on the same drives and it didn't report any problems. I followed the emergency protocol for such situations, which consists of guessing which of the two supposedly-failed partitions is in better shape, taking the worse one out of the array, telling the driver to forget it ever thought there was anything wrong with the better one, and then wiping the worse one and putting it back into the array as a replacement for itself. If you guess right about which is which (and you had just better guess right) and if the "better" drive really is okay, then the array will rebuild itself and either detect that the "worse" drive really is dead (at which point you can replace it normally) or detect that the worse drive is actually okay too and restore normal operation. I did all that and everything seemed to come back to normal.

But then yesterday (about ten days later) I got another report of a double failure, this time from the RAID6 array. Note that the RAID6 array uses more redundancy; it can recover from two simultaneously-failed drives without losing data. This time the RAID5 still claimed to be fine. In fairness, I'd been pushing the system very heavily at the time, streaming audio and video off the array, rebuilding fonts, and running combinatorics experiments on all four CPU cores all at once. And again the problem seemed to sort itself out when I took the "failed" partitions out of the array and put then back in as replacements for themselves.

Digging through the logs, I see some errors that look like they could possibly be associated with media problems on one (just one) of the two drives where there've been complaints. I think, even though these are not the very similar error messages I would expect for a loose SATA cable, that I may possibly have a loose SATA cable. It could have gotten jostled during the move. So it's on my to-do list to take apart the box and reseat all the SATA cables. I also upgraded my Linux kernel, because I suspect (based on reading others' reports about this same version, and a kernel panic I saw) that the 3.6.6 version I was using may have some RAID-related bugs. So far there've been no further problems since the kernel upgrade, but it's only been a few hours since I finished doing that and based on the previous experience I think I'm going to have to see at least a couple weeks without errors before I can guess that the problem is solved.

In other news, I'm continuing to make progress on the Tsukurimashou Project. Moving, travelling, and looking for a job have all taken their toll on my productivity. I had for a while been hoping to push out version 0.7 by the end of 2012 and that obviously didn't happen, but I've very close to the desired kanji count for that, and I've made a lot of progress on infrastructure stuff.

Stay tuned for updates on "ecchi" - another software project I've had under my hat for a couple of years. Just recently I've started to get enough interest from the people at work that I may really be able to go forward with ecchi in a serious way.

I'm still at the query stage for my novel Shining Path. In the latest round of queries I finally got some non-rejection responses. At present, three agents are looking at requested full copies of the manuscript. There are no reliable statistics on what the industry average acceptance rate is for manuscripts at the "requested full" stage, and I suspect it's not high enough that I can solidly expect an offer of representation now, but at least there's a significant chance. So the wheels are turning. It does seem reasonable to expect that Shining Path will eventually get represented, and later published; but it may be years yet before that happens.

8 comments

Google Employee
Just a thought: if you gave up on the ungrateful academia and accepted a job from some respectable company (I'm sure any one of them would love to have you) you would never have to knock on any door ever again. It may not be as glamorous as being a real researcher, but let me tell you, if feels really good. And when you feel good, you look good. And when you look good... well... good things happen.
Google Employee - 2013-01-02 20:59
Matt
I wish I could believe that. But "knocking on doors" doesn't mean "applying for jobs" - if I meant that I would have said - and having a permanent position wouldn't magically put an end to running around knocking on doors. It more refers to what kind of job I have, and it's absolutely not true that industry jobs are free of running around knocking on doors. I've even had friends in industry describe their experience of running around knocking on doors in their jobs as a positive thing they think I would be pleased to be forced to do.

Much more importantly, it also refers to what kind of life I have outside of work. Am I going to a place - a geographic location and culture outside my workplace - where I'm socially popular, or to one where I'm pressured to do more seeking of personal contact on my own initiative, running around knocking on doors, than I'm able to sustain because "oh, well, it's what you put into it"? A vague hope for good things to happen is laughably inadequate, because I can hope for that and have nobody other than myself be responsible for it, anywhere.
Matt - 2013-01-03 05:48
Matt
Hm. And maybe the previous comment will finally give me enough incentive to fix this Web logging software's overquoting bug. I am understandably paranoid about input sanitization, and that sometimes has effects.
Matt - 2013-01-03 05:54
Google Employee
What I meant is that in academia you are merely an average researcher, while in industry you will be unparalleled genius that basically doesn't have to do anything at all. You can sit there like an oracle and people will come to you with their trivial issues. Something as stupid as writing down a formula for the inverse of 3x3 matrix without writing down any intermediate steps impresses even the most highly paid architects. If you also manage not to be an asshole about it, your confidence will skyrocket. And chicks dig confidence, no matter what package it comes in.
Google Employee - 2013-01-03 06:04
Celes
Hope everything turns for the best. It may sound very bland and cliched, but I guess it comes from a good place. Nothing is planned in life even if we humans foolishly think and delude ourselves but reality has a knack for smacking us good for burying for heads in the sand. Remember we are here but in a blinking of an eye, and in borrowed time, due to be back to where we came from. Make the best of it.
Celes - 2013-01-05 01:35
Raymond
Would you work in Montreal?

I know two bright people, with the same Erdős number as yours 8-)
working there (one in academia and his startup and the other at Idilia)

In the computational linguistic field, IIRC... I don't even know what _that_ is 8-)

They both got their Ph.D. with Gonzalo E. Reyes at UdM (now retired) http://reyes-reyes.com/2011/04/03/the-history-of-categorical-logic-1963-1977 for some lecture...

Sooooo let me know if the city is OK and I'll email my contacts.
Raymond - 2013-01-08 18:40
Raymond
Oh, by the way, let me introduce myself: Raymond Lutz, I translated the http://www.digital-copyright.ca/petition/user in French. And I'm reading your blog since then (march 2002)
Raymond - 2013-01-08 19:01
Matt
Raymond: at the moment I'm only really considering academic jobs, but it does happen that one of the ones for which I'm currently preparing an application is at McGill.
Matt - 2013-01-09 09:02


(optional field)
(optional field)
Answer "bonobo" here to fight spam. ここに「bonobo」を答えてください。SPAMを退治しましょう!
I reserve the right to delete or edit comments in any way and for any reason.