|
| Greg's diary |
| November 2001 |
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Thursday, 1 November 2001 | Echunga | |
| Top of page | ||
| next day | ||
| last day |
More stuff getting in the way of my tutorial. Ah well, I think I have enough to keep me going, but I still need to polish the slides a bit.
Spoke to Greg Rodgers again in the afternoon, and decided to try to get a complete Linux system on AIX volumes, since we have access to them. Spent a bit of time learning how to set up AIX volumes in the first place, and then booted into Linux. The good news: we could see the volumes (named linuxroot and linuxwap). The not so good news: EVMS reported their size as zero. swapon refused to have anything to do with the volume as a result, but mkfs was more flexible, and started building a file system on linuxroot, but unfortunately hung the system before completing. Ah well, it's progress of a sort.
| Friday, 2 November 2001 | Echunga | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
Carried on trying to investigate the EVMS problems I discovered yesterday, and since I still don't have a kernel debugger, I decided to attack it from the userland perspective. That was an unbelievable amount of work. Not only does Linux not normally come with userland sources, just getting hold of them and building them is more pain than I could have imagined. There seems to be no way to download the entire base system, and depending on where you start building, the Makefile may or may not build. Spent a couple of hours on IRC talking to Debian hackers. My favourite comment was:
> hacker: Who's in the office who can help me build Debian userland? <hacker> groggy: why would you want to build it? > So I can control it. * groggy . o O ( Why would anybody ever want sources? ) ... * groggy wishes somebody would just package up all the base system sources so that he could untar it and compile it. <hacker> deb-src http://http.us.debian.org/debian stable/non-US main contrib non-free <hacker> that would be tooo easy :-) <hacker> n.b. the /non-US after stable > Works for BSD. > Yes, I note. I just don't understand. <hacker> yeah ... but on Debian not everyone wants (or is able to) build stuff from the sources <hacker> There is another directory level in the repository ... > I can certainly understand why. <hacker> I don't know why > It's such a pain.
Finally I got it done, and the compile failed because of header file problems. With a bit of tweaking, I got swapon compiled, and was able to run it under gdb. Probably not what I wanted, but I didn't know that when I started. Tried to patch the EVMS sources, but they failed to build for reasons which appear unrelated to the patches I made. sigh Not much to show for a day's work.
| Saturday, 3 November 2001 | Echunga | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
The BSDCon Europe is dominating my activity at the moment. On the one hand, I don't have any particular concern about holding a total of 6 hours of talk, along with the demonstrations, but it's still a concern to get enough material together. Spent most of the day working on that, in the process connecting sydney, my Dell laptop, to my existing 5 display setup with x2x, getting a total of 12,573,248 pixels in the process.
We were planning to go riding today, but somehow we didn't get round to it. Instead did some riding in the arena, which was probably necessary, but not to Darah's taste. We should do more of this.
| Sunday, 4 November 2001 | Echunga | Images for 4 November 2001 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
A lazy day today. Slept in late, and did some work on the tutorial. Also a bit of hardware rebuilding: monorchid still didn't have any tape units, and my backups were all a bit erratic. Also put together the Duron box, still without a name (though I suppose I should call it daemon), and tried to find a Vinum bug which a couple of people had reported: after shutting down to single user mode and returning to multi-user mode, the system panics. Not for me, unfortunately. It appears that it's happening to Federal Express, though, so I should give it some attention.
In the evening, Richard Sharpe and family came along; first time I've met his family. They're planning to augment the exodus of technical people from Australia and move to Fremont CA; I can't imagine why.
| Monday, 5 November 2001 | Echunga | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
Final day before leaving for England, and despite the work at hand, I didn't get very much done. It looks like I'm going to have to finish the tutorial slides on Wednesday and Thursday.
| Tuesday, 6 November 2001 | Adelaide –> Melbourne –> | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
Up relatively early to tidy up the last things, and it was a good thing too. After removing the 802.11b wireless card from air-gw (leaving the Aviator, 802.11 FHSS for Yana), air-gw refused to talk to Yana's machine, firefly. Turned out to be a problem in the DHCP configuration, which saw firefly on one network (the wrong one) and the card on another network (the right one). I wonder how that happened, and why it had to show up now.
To the airport, and everything went relatively smoothly. I never fail to be amazed by the security people: they let me go through security in Adelaide with the same pouch that they sent separately last time, but in Melbourne they impounded it again. In Adelaide there was metal cutlery, in Melbourne plastic. On the flight from Melbourne to Singapore we got metal cutlery, from Singapore to London plastic. sigh
| Wednesday, 7 November 2001 | –> London –> Brighton | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
I wish that Qantas would find better schedules to get people into England at a reasonable time. I was on QF 9, which arrived at 5:20 am. That meant that we were woken somewhere like 3 am for “breakfast”.
Finally arrived in London, and it took forever to get my priority baggage. Finally a Qantas man showed his face and apologized, explaining that the freight people hadn't shown up on time. Gave up waiting for my pouch, which people said would arrive last. Instead went out to sort out the rental car, which also took a surprising amount of time. After that, decided that my pouch might be available by then, so found out how to get back inside, somewhat hampered by the lack of personnel. Finally the same Qantas bloke answered the British Airways phone, explaining that nobody from British Airways was available. Since I was a Qantas passenger, I was able to convince him to let me in.
A point I forgot to mention at the time: the enquiry desk had a thermometer on the counter, showing the desired room temperature: 16°. I would dearly have loved to take a photo, but it wasn't the time or place to try something like that.
Picking up the car worked OK, but by the time I left it was 7 am, and I decided that going round the M25 would take too long. Checked the distance to Southampton: 57 miles. I keep forgetting how tiny this country is. Decided to go down to Winchester and then back East on the A 272, a road I knew like the back of my hand in the late 60s. It's obviously not the late 60s any more. Yes, I recognized some of it, but it's really fascinating what I have forgotten. For example, there's a stately home in Petworth. I only recalled the bad traffic.
On past Haywards' Heath to Chailey, where 33 years ago my girlfriend Lesley Cannings lived. The place has changed beyond recognition (or my memory has), but I found my way past where she lived and then on to Fletching, so confused by the layout that I went back to Piltdown to check my directions. It seems that I have forgotten a fair amount about the places, but the framework is still good. I suppose this should tell me more about the nature of my memory than I can digest at the moment.
On through Nutley, where Guy Belsham's family used to live, and then to Tunbridge Wells, which I still recall relatively well. Finally back to Brighton, and rather to my satisfaction found the hotel without needing a map.
Less to my satisfaction was the hotel itself. I confirmed that they had given me a “special” rate of £133, whereas my father had had the “normal” rate of £121. After a bit of discussion, the receptionist—who comes from Adelaide—decided that I didn't need special treatment and gave me the £121 rate. He also told me that there was an Internet connection in the room, which turned out to be an RJ45 phone jack, which to boot wasn't even connected.
Parking the car was more fun. When I finally found a parking place, there were no signs out. I finally found a lift, which left me outside the hotel, and when I found the (rear) entrance I couldn't get in because the doors were locked.
The room is also no revelation, especially for that price. Not only light bulbs burnt out, but apparently non-functional electrics. I get the distinct impression here that nobody cares. Certainly there was no guest questionnaire in the room.
In the afternoon I succumbed and slept until the evening, when I thought I should go out and get something to eat. There are a number of French bistros here, apparently for the benefit of the visiting French, and they look relatively authentic. Somehow the prices worry me: in Australia I would pay about $25 for a good meal; here it's more like £20, which is more than double. Instead I had a “meal” at another place, for £11, basically fried sardines and bacon and eggs; the latter would go for $6 in Australia.
| Thursday, 8 November 2001 | Brighton | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
Woken at 5 am by a wakeup call. Grrr. Called reception and asked them to turn it off, but disconnected the phone anyway, just in case.
Woke again at a more reasonable time and off in the direction of Tunbridge Wells and Pembury, this time driving through the town and noticing that I remembered almost nothing of it. Found my way to Pembury (well, Colt's Hill) without a map and looked at Pamela's musical instruments. There wasn't much there apart from an oboe after Triébert's System 3, for which she wanted far too much money. There was one on Ebay a couple of days ago, but I forgot to bid. Decided to think about it—I can return to London via Pembury on Monday if I decide to take it—and back to Brighton to return the car.
Wasn't feeling too well in the afternoon, and got an early night.
| Friday, 9 November 2001 | Brighton | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
Woke up early enough this morning that a phone wakeup call at 5 am wouldn't have been a problem, and tried to finish my slides, in which I was reasonably successful. Then to breakfast and the tutorial, which went about as well as I could hope; in particular, the time worked out pretty exactly.
Dad showed up in the middle of it, listened for a few minutes, and disappeared. In the evening, was rather surprised by the arrival of Harti Brandt, whom I didn't recognize until I saw his badge. Off with him, Dad and Paul Schenkeveld to dinner, then to a quiz evening in which we formed teams. When it was clear that we were losing, we decided to lose spectacularly and won the booby prize, a whole lot of Brighton Rock (solid foam sugar in cylindrical form) with “FreeBSD” written down the axis, and spent some time getting rid of that. Far too late to bed.
| Saturday, 10 November 2001 | Brighton | Images for 10 November 2001 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
The perfect complement to a late night is an early morning call, and that's what I got, at 5 am, the same one I had two days ago. Threw the phone against the other wall and it stopped ringing. Grrrr.
Up a little too late and somehow missed Dad. Wasn't too impressed by the keynote, but then Poul-Henning Kamp discussed devfs, something I didn't want to miss. It was mainly a waste of time, of course: it was directed at people who didn't know devfs, and what I really wanted was internals.
Mark Murray on after that about PAM, rather more use, since I really didn't know anything about PAM. I suppose I need to investigate more, and at least get it to work correctly with rlogin.
Had lunch by myself, chicken biriyani in a passable Indian restaurant. I think I've had just too much of people, or maybe I'm still not feeling my best. To support the latter theory, flaked out in the afternoon and didn't come back to life until 5 pm.
Conference dinner in the evening, not the way I would have done it. I wasn't the only one who objected to the music, not because of its nature, but because it was there and stopped people from talking. Sat with the German contingent: Christoph Herrmann, Thomas-Henning von Kamptz, Harti Brandt, Bodo (whose name I continue to forget), and Paul Schenkeveld as token German. Turned out we had a lot in common: Thomas-Henning plays the cello, Bodo the piano, I knew that Harti played the guitar, and Christoph rides horses (Brandenburger Warmblut). Early to bed.
| Sunday, 11 November 2001 | Brighton | Images for 11 November 2001 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
Up relatively early this morning and had breakfast with Poul-Henning and his girlfriend. Poul-Henning amused me by suggesting that they should have had a string quartet last night; that's just what we had at the AUUG conference in September. Obviously I agreed.
Spent a lot of time outside the conference rooms talking about debugging. My tutorial on Friday seems to have met with a lot of interest, and we've had at least one commit as a direct result of the tutorial.
Lunch at a Chinese place round the corner. As Bodo said, it should be rated as three skulls and crossbones. The afternoon began to drag a little, and I spent time catching up on my mail, which ended up not being too bad. Also collected the photos I took during the conference.
The final session was “BSD Questions and Answers”, dominated by “how do we sell BSD?” and, more to my interest, “how do we get the number of PRs down”. I recalled the horrible story of my efforts at Tandem, and how little they got in the way of results, but in the end Jordan and I suggested a new kind of commit bit, a PR commit. That could have a number of positive effects.
In the evening to dinner at the same Indian restaurant I went to yesterday lunchtime. I was less happy this time, and again went to bed early.
| Monday, 12 November 2001 | Brighton –> Tunbridge Wells –> Pembury –> London –> | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
Up relatively late today, and had a thoroughly unsatisfactory breakfast. After that gradually checked out, and despite all attempts was not able to meet with Paul and Brian.
Gave Pamela a call and came to an agreement to pay her £225 for her Système 3 oboe, rather against my intentions. Set off towards Tunbridge Wells, but took the wrong exit on the first roundabout and found myself driving down the promenade. Since I had plenty of time, continued along there, discovering in the process where Roedean school is, and discovered a dearth of signs pointing to London. Leaving Seaford, a memory suddenly clicked, and I stopped and checked my address book: yes, I was directly at the end of the road in which Pat Chapman lives, the bloke who sold me a total of five clarinets in June 1990, including my bass clarinet and my Eb clarinet. On the spur of the moment, drove down and found that not only did he still live there, but he remembered me in quite some detail. Quite impressive. He had a few instruments for sale, including a set of three single-piece Buffet Crampon rosewood clarinets (C, Bb and A), and an Italian high Ab clarinet, which was so tiny that neither of us could get our fingers round them. Also a couple of other instruments, including a couple of band oboes and a six-keyed flute with covered keys. I was sorely tempted, but in the end left without anything.
On to Tunbridge Wells, which took far longer than intended. I had intended to pick up the oboe from Pamela round 1:30, but I didn't get to Tunbridge Wells until 3 pm, where I needed to pick up some cash. On leaving, I got a flat tyre:
and discovered that Hertz had not put enough tools in the car to change it: one of the wheel nuts was a lock nut, and I didn't have the key:
|
|||||||||||||||||||||||
As a result I lost nearly an hour before the RAC came and changed it for me. Grrr
Picked up the oboe, and then round the M25 to the West of London. Considering I left Pembury at 4 pm, I made it in record time: I was passing the M4 at 5:15, too early to go to the airport and too late to go to Oxford. After some vacillation went to Rose Hill Instruments in Beaconsfield, also an old friend, and bought some flute music.
Back to the airport, and discovered that Hertz treated the flat tyre as an accident which, to add insult to injury, they expected me to pay for. Spent some time venting my displeasure, but ended up having to pay £17.60 for it, a substantial percentage of the price of a new tyre. I should send them a bill for my time.
In none too happy a frame of mind up to terminal 4, which must rank as one of the ugliest terminals I've been to. Like terminal 2 in Frankfurt, it is woefully signposted, probably one of the reasons for the long queues in front of the information desks:
|
|||||||||||||||||||||||
High time to get home. Even in the frequent flyer lounge, things were strange. They had smaller white wine glasses for white wine, and larger white wine glasses for red wine, coincidentally exactly the kinds we have at home. In addition, though they had a nice selection of cheeses, they only had oatmeal biscuits to eat with it:
|
|||||||||||||||||||||||
It could only happen in England.
Even when I got on the plane, the irritations continued: I found I had been shoe-horned into a bulkhead seat at the very front of the upstairs cabin. When will people realise that this kind of seat is not comfortable? Fortunately I was able to get another seat.
| Tuesday, 13 November 2001 | –> Singapore –> | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
Woke up in mid-afternoon Singapore time and noted the interesting route we had had to take to avoid the current conflicts in Afghanistan and Pakistan:
|
|||||||||||||||||||||||
Had breakfast at 6 pm, then on to Singapore, where I did some shopping for once in a while, and bought a couple of books, an annotated Q'uran and a book about the second world war cracking of the Enigma code.
| Wednesday, 14 November 2001 | –> Melbourne –> Adelaide | Images for 14 November 2001 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
Arrived in Melbourne a little late, and had to wait nearly 30 minutes for my priority baggage. After customs out to look for where to put it for the onward flight; no signs, so it's a good thing I know the airport. It turned out I had to wait in the normal checkin queue just to hand over the pre-checked baggage. Despite the size of the airport, they still only had two people for business class checkin, the same as in Adelaide, and I had to wait 15 minutes until I could hand in my baggage. People didn't seem to think that this was a problem.
Back home and lots of work had piled up. Why do all the important mail threads spring up while I'm travelling?
| Thursday, 15 November 2001 | Echunga | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
I bounce spam to the abuse@ address of the ISP on a regular basis. If the abuse@ address doesn't exist, I just block the domain. Today, though, I got a couple of interesting replies. The ultimate response must be this one:
Delivered-To: abuse@stargate.net Received: (qmail 20641 invoked from network); 15 Nov 2001 01:18:01 -0000 Received: from unknown (HELO monorchid.lemis.com) (192.109.197.75) by smtp2.mx.pitdc1.stargate.net with SMTP; 15 Nov 2001 01:18:01 -0000 Hi. This is the qmail-send program at dissolved.noc.sgi.net. I'm afraid I wasn't able to deliver your message to the following addresses. This is a permanent error; I've given up. Sorry it didn't work out. <spamadmin@noc.stargate.net>: AUTORESPOND: Junk mail received.
Well, yes. Goodbye, stargate.net.
Here's another:
<abuse@hostcentric.net>: host mail.hostcentric.net[209.25.142.68] said: 550
<abuse@hostcentric.net>... Relaying denied
I wonder where they think their abuse@ address is.
Spent most of the day reading mail. Lots of things have happened while I was away.
| Friday, 16 November 2001 | Echunga | Images for 16 November 2001 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
Another lazy day. I suspect I've come down with some kind of virus which, though it doesn't make me feel sick, is robbing me of energy. Or maybe it's just the fact that summer is gradually arriving and pulling me away from work. Spent a fair amount of time reading about Enigma and Bletchley Park. I don't know whether it's the description or something else, but I have difficulty visualizing the encryption process and how to crack it.
| Saturday, 17 November 2001 | Echunga | Images for 17 November 2001 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
Yet another quiet day. I have to accept that I'm somehow under the weather. Still, it probably does me good to slow down occasionally.
| Sunday, 18 November 2001 | Echunga | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
Gradually I'm coming back to normal. Today I got a mail message from somebody in England who had a Triébert système 3 oboe for sale. Grrr, she really could have told me that a couple of weeks ago before I bought the other one, and while I could have gone to pick it up.
Spent a bit of time contemplating multimedia. It looks as if the best solution for the electric gate we're planning at the end of the driveway (130 metres from the house) will be to put an old laptop down there with a webcam, audio and wireless networking. Then we can access it from any network connection in the house, effectively anywhere.
| Monday, 19 November 2001 | Echunga | Images for 19 November 2001 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
Finally I'm feeling a little more normal, and about time too. We're coming up to another AUUG executive committee meeting, and the FreeBSD project is up in arms again about its core team, of which I'm a member. sigh It would be nice to get back to doing some productive work.
In that connection, decided to see if I could finally read AIX JFS 1 file systems from Linux, and wrote some exploratory programs which showed some interesting divergences between theory and practice. In particular, the magic number in a JFS 1 super block is often wrong, which makes me wonder what good it is.
| Tuesday, 20 November 2001 | Echunga | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
More work on the JFS1 stuff today. After working on Linux, where everything seems different, it's a breath of fresh air to read the AIX header files. It seems relatively clear that JFS took a lot of ideas, and probably code, from UFS. Got as far as reading individual inodes; now I need to understand how to find subsequent allocation units.
| Wednesday, 21 November 2001 | Echunga | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
A substantial amount of my time is spent reading or otherwise disposing of the 1500 odd mail messages I get every day. Fortunately, I can delete most of them, but the remainder keep me quite busy.
Today, a number of fires significantly increased this quantity:
On 21 November 2001 you received 2147 mail messages.
The problem is that this increase mainly represented messages I needed to read and respond to. That, and a teleconference in the morning and a dentist's appointment in the afternoon, kept me busy all day. And that on a day when I really wanted to work on the JFS stuff.
The dentist was encouraging: I was convinced that the tooth would have to go, but he thought it was just a minor change in height, and after filing it down a bit, he said it would be OK. I'd like to know how I got so long in the tooth in the first place.
| Thursday, 22 November 2001 | Echunga | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
Another day full of interruptions, but at least I expected this one. Yvonne was off to Melbourne for a couple of days, and it was also Yana's last day at school (forever, assuming the exam results are good enough), so I spent a couple of hours driving them around in remarkably poor traffic.
Back home, and finally got the JFS 1 structures correct, so I can now think about writing a file system to read the files. It all looks remarkably like UFS, so maybe I should start with UFS and modify it.
In the evening, took advantage of the fact that Yvonne wasn't there and cooked some hot curries, of which I ate too much.
| Friday, 23 November 2001 | Echunga –> Melbourne | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
More admin stuff today, and didn't get much real work done. I'm still pretty tired.
In the afternoon to the airport and on to Melbourne. SkyNet Global now really do have correctly functioning DNS, even without using their SMTP server, which I think is safer.
In Melbourne, discovered that only Peter Gray was there. Yvonne came just after Peter went to bed, and wanted a bottle of champagne. Raided the mini-bar, which was relatively sensibly priced, and discovered one disadvantage of the scheme: the champagne must have been there for ever, and it was completely flat. Got another one from room service; the difference in the shapes of the corks was remarkable.
| Saturday, 24 November 2001 | Melbourne | Images for 24 November 2001 |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
The hotel proved to be pretty noisy, and I didn't get much sleep. To add insult to injury, just as I was nodding off about 30 minutes before I had to get up, the alarm clock went off, although I had looked at it and had thought I had turned it off. There should be a hall of shame for hotels which do this to their guests. The last one was the Thistle Hotel, Brighton, and this one is the Sebel Suites in St. Kilda Road. At least the rest of the hotel was OK.
Had breakfast in the hotel after all, because Yvonne had to head off to the airport. Then Peter and I went to the meeting, which was somewhat chaotic. Somehow I had left my laptop power supply behind again, which was not too bad (Sarah lent me one), but somehow something had gone wrong with copying data to the laptop, and I was left without some of my documents.
To the university during lunch, bringing back a number of old memories, and by complete coincidence stumbled across a Linux Installfest. In to say hello. Conrad knew one of the organizers, and we spoke for a while. Their numbers were down over last year as well, though the weather was good. They think it might be communication, which seems reasonable considering that neither Sarah, Conrad nor myself knew about it.
Lunch in Lygon St., then back to the meeting. There's something funny about the hotel we're looking at, the Rydges Carlton hotel. There seem to be a number of strange smells about it, a thing I had noted when I stayed there last year. We'll need to reconsider.
In the evening to a good Chinese food place in Little Bourke St., and then back to the hotel with Conrad and Sarah. Despite my intentions quite late to bed.
| Sunday, 25 November 2001 | Melbourne –> Echunga | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
Up at the crack of dawn and out to the airport, where I yet again had problems (the last time was all of 11 days ago). Security discovered a pair of nail scissors in my carry-on baggage and behaved really stupidly, ending up refusing transport of the item at all. I now have a week to reclaim it from Melbourne airport.
Spent the flight to Adelaide writing a letter of complaint to Qantas. Got back home ridiculously early, about 9:45 am, and vegetated through the day. I wonder what it's like to take holidays.
| Monday, 26 November 2001 | Echunga | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
Work is hotting up again. There are various mail discussions in which I should be participating, like how FreeBSD is going to schedule kernel threads, and I just don't get enough time to look at them.
Spent a while looking at the JFS1 code and how to convert something like the Minix file system to grok JFS1, but it was a lot more complicated than I thought. In particular, it seems unlikely that I will get it to work without a kernel debugger, and we still don't have one of them for PowerPC, so decided I could probably do it better if I implemented it on Intel first (including endian issues), so installed another Linux system on my test box. Why is installing Linux is such a pain?
| Tuesday, 27 November 2001 | Echunga | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
More work on the Linux installation today. I'm really amazed at how easily I can screw things up. After installing Red Hat, I had to upgrade the kernel to 2.4.12 (but not userland; that still makes me feel uneasy). The first kernel didn't get past booting: it didn't have SCSI support, and the system disk was SCSI. The second one didn't find the Ethernet card. The third one only found one of the two host adaptors; after a long time on the phone with Anton and working my way through drivers/scsi/Config.in, I discovered that the driver I was looking for only appeared if another one, purportedly for the same purpose, was not selected:
if [ "$CONFIG_PCI" = "y" -a "$CONFIG_SCSI_NCR53C7xx" != "y" ]; then dep_tristate 'NCR53C8XX SCSI support' CONFIG_SCSI_NCR53C8XX $CONFIG_SCSI dep_tristate 'SYM53C8XX SCSI support' CONFIG_SCSI_SYM53C8XX $CONFIG_SCSI
Finally got things working. It really makes me appreciate FreeBSD.
In the afternoon finally did some work on the next edition of “The Complete FreeBSD”, to be published by O'Reilly. One of their absolute requirements was that I should convert it to DocBook, for which Michael Still wrote a perl script. Things are taking shape.
| Wednesday, 28 November 2001 | Echunga | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
There's still far too much going on. I really need to drop some of my workload. I'm certainly happier now that we have “The Complete FreeBSD” converted to DocBook, but there are still things like articles which need to be done.
Mail, a teleconference and the core wars kept me busy for the rest of the day.
| Thursday, 29 November 2001 | Echunga | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
Had a bit more success doing some real work today, and managed to install a Linux kernel with remote serial debugging with gdb. Tried running the Linux gdb on sydney, but there was some issue with the libraries. Just for the fun of it, tried the FreeBSD gdb, and ... it worked! That makes things a whole lot easier. It's also interesting that break across the serial debug line works with the Linux kernel, which also shows something about where the problem lies in FreeBSD.
In the afternoon, lost the satellite IP connection. After a lot of checking on my end, called up the help desk and discovered that a storm in the Napa valley (where the satellite uplink is located) was causing the problem. I hope this doesn't happen too often, but it did give me a chance to modify the monitoring software and check situations where the tuner wouldn't lock.
| Friday, 30 November 2001 | Echunga | |
| Top of page | ||
| previous day | ||
| next day | ||
| last day |
More sidetracking today. Richard Sharpe is doing some work comparing FreeBSD and Linux TCP performance, and there are obvious problems with FreeBSD. Today was also the ADUUG meeting, so met Richard there and discussed the matter. This new job of his seems interesting. It will certainly be interesting to find out why our TCP performance doesn't measure up to Linux.
Do you have a comment about something I have written? This is a diary, not a “blog”, and there is deliberately no provision for directly adding comments. It's also not a vehicle for third-party content. But I welcome feedback and try to reply to all messages I receive. See the diary overview for more details. If you do send me a message relating to something I have written, please indicate whether you'd prefer me not to mention your name. Otherwise I'll assume that it's OK to do so.
| Top of page | Previous month | Greg's home page | Today's diary entry | Next month | Greg's photos | Copyright information |