Layout re-design April 27, 2006
Posted by a1291762 in : site , 4commentsIf you’re a regular reader (do I have any of those?) and it feels like something subtle has changed with the site, rest assured it has.
I have changed my layout code. The primary reason was so that I could have a real footer, one that was the width of the page, not the content div. The previous layout code was not able to handle this due to its inability to detect the length of the side bar. If all my stories were long it would have been ok but some are short and it’s unacceptable to me for those pages to look like crap.
As far as I can tell, most browsers shouldn’t have a problem with the new styling. It may even be better than the old code WRT local CSS hacks, font overrides and zooming.
Progress Updates in the comments…
Update 28 April 2006
I’ve settled on my final solution. It looks fine in Konqueror, Safari, Camino, Firefox. There is a minor glitch in IE 5/Mac (but everything still looks fine). I couldn’t figure out how to work around a bug in IE/PC so the columns are swapped. Other than that, it is also fine. If you know how I can get the menu to stay on the right hand side of the screen in IE (without breaking the footer), let me know and I can put the columns back how they’re meant to go.
We all sleep in April 24, 2006
Posted by a1291762 in : Uncategorized , 1 comment so far7:50 am. That’s got to be some kind of record for us. Sure, each of us has probably slept in longer than that (except for Isaac) but we’ve never all slept in that late. Bree said Isaac made noise earlier but she just threw a blanket on him and he went back to sleep.
At least I had an excuse though. I was up late last night trying to build a new release of iPodIcons. It’s now a Universal Binary and works with the iPod Nano and the 5G iPod.
How hard could it be to build a simple cross compiler? April 20, 2006
Posted by a1291762 in : Uncategorized , add a commentOh My Freaking Head! How bloody painful do the GCC folks think they can be?
It seems that without a certain amount of prebuilt libraries and headers, you can’t just build a gcc cross compiler for a foreign system. Of course, I don’t have a linux PC handy to build my cross compiler on (and I don’t have the bandwidth/time to find a suitable combination of linux/glibc to build against) so that means I’m screwed. Or does it?
It seems I may just be in luck. I did manage to build a cross compiler but it didn’t target linux specifically. Instead, I targeted i386-elf. Unfortunately, this isn’t enough to work with GNU-EFI because it has no knowledge of EFI (I guess that’s tied to the linux target). I’m now in the process of bootstrapping linux/glibc with my i386-elf compiler to see if I can get something that the real cross compiler can use to build against.
Ugh.
Update: 24 April 2006
It seems my strategy is the correct one, as evidenced by this article. I’m having trouble getting glibc or newlib to compile but once that’s worked out, I should be almost there.
Sorting out Isaac's eating problems? April 19, 2006
Posted by a1291762 in : Uncategorized , add a commentWe fed Isaac breakfast today. He is definitely opening his mouth for food now but you’ve got to be quick.
Apparently there’s a condition in some members of my direct and indirect family where the oesophagus can’t hold in food very well. Considering Isaac’s tendency to vomit food, sometimes hours after eating it, I’m inclined to think he’s got something like this. We have noticed that since feeding him dinner, he’s not vomiting at night so we tried feeding him breakfast to see if it will work at other times of the day. My theory is that the thicker consistency of the solids makes it easier for his oesophagus to keep the food down.
If this works it’ll mean that he’s keeping more food down and eating more at the same time (he tends to eat solids and his regular bottle).
D’oh! His first sleep after breakfast and he still vomited. We have already slept him in a chair (when we has sick) so we might need to do this to get him to stop.
3-day parenting April 15, 2006
Posted by a1291762 in : Uncategorized , 1 comment so farI’m starting to notice a strange pattern with our parenting. Every time we change something for our kids, it takes them 3 days (or less) to accept the change. All sorts of things have fit this pattern. Starting solids, dropping night feeds, changing routines. The most notable example was when we moved Emily to her bed, which took a day. She’s only gotten out of it twice since she’s been in it (Chrismas time).
Since I went on holidays, Emily no longer gets a dummy and is using a potty. By the third day, Emily didn’t even ask for the dummy (though she did cry at bedtime for a week). The potty thing is technically longer but we started it and Emily got a cold so we postponed it until after she got better (because she was miserable, was having trouble understanding what we wanted her to do and dehydrated, making it really hard to get her to actually use the potty). Today was the third day since starting potty use again and she didn’t do a single wee or poo in her nappy. She even told us that she needed to go to the toilet (I was thinking we’d have to keep asking her or she’d forget and go in her nappy).
So how did we get this 3-day things happening? I think it’s the expectation we have for our kids. We mean what we say. We don’t give in. Bad behaviour is not tolerated. Basically, we expect our kids to do certain things and they know this and will try to meet our expectations. This sharply contrasts with what I think is a more common attitude which seems to involve letting the kid do what they want while trying to convince them it would be a good idea to do what you want.
Of course, it also helps that we waited until they were ready for the things we were asking of them. For example, most people spend months doing potty training but if the kid can’t tell when they have to go, there’s little point. We’ve been encouraging Emily to tell us when she does a wee or poo and we were pretty sure she could tell that she had to go so we just told her that wee and poo don’t go in the nappy anymore and that she should tell us so we can put her on the potty. The first two days (three if you count the day before she was sick) were adjustment days where she didn’t want to go to the potty. She’d say “no” when we asked and then immediately go in her nappy. We would re-inforce the new rules when she did this and try to sit her on the potty at opportune times (eg. just after she wakes up she usually needs to wee). By today, she understood the new rules and wants to do the right thing. She’s even a bit over-enthusiastic. She’ll get us to put her on the potty so she can fart!
Yet another 3-day thing, Isaac has started on solids. He doesn’t just push the food out of his mouth now and even opened it for me today. We fed Emily solids in the morning, moving from the rice cereal to kid porridge and finally weet-bix. At some point we started feeding her dinner too. However, we’re doing Isaac at night to try and help him sleep better. He’s always eaten less than Emily and this makes him get hungry more easily. When he wakes up in the early mornings (which is common, due to cold or noise from outside) he’ll get hungry and require settling and a dummy (breakfast is at 7, not before). Emily would wake up in the early mornings too but would go back to sleep just fine. On dark and quiet mornings, Isaac doesn’t get up until breakfast time and when he does wake, it’s not at a regular time (which would indicate a feeding problem). This morning, Isaac sleep in until nearly 8 after being re-settled so the food might be working.
I think another part of the problem is that by 4 months, Emily could put the dummy in her mouth. Isaac still can’t do this so he’ll whinge until someone else does it. We encouraged use of the dummy instead of his fist because dummies are easier to get rid of. Of course in the last few weeks, Isaac has been getting ready to crawl so we need to tie him up or he’ll end up squashed into the cot bars. He’ll have to learn to stay still at night before we can stop tying him up and let his hands be free to put a dummy in. We’ve got a road near us getting re-surfaced and all the traffic goes down our street. Put together it’s not so surprising that he’s not handling the early morning properly yet. I think the road works are due to end in a month which should be good for him since he should be able to lie still and put in his dummy. The later sunrise for winter will keep the birds quiet till later so all should be good. I’m crossing my fingers.
MacBook Pro fan and heat observations April 14, 2006
Posted by a1291762 in : Uncategorized , add a commentI noticed an interesting thing about Eve while compiling MinGW. The CPUs get taxed pretty hard when you do a big make and the fans came on. However, unlike Bree’s iBook, the fans seemed to get faster in small increments, not the all or nothing approach the iBook has. It was strange to hear the fan slowly spinning up as the compile continued.
I was initially concerned with the heat that Eve generates. However, I’ve since come to the conclusion that it’s not a big deal. I use it on my lap all the time without problems. The trick is to place the edges of the machine on your legs (ie. legs apart). This even worked when I used the machine in bed.
So far, the hottest the machine has been was when I re-installed Mac OS X. I think the optical drive creates quite a bit of heat.
Mac-hosted EFI development April 14, 2006
Posted by a1291762 in : Uncategorized , add a commentIt seems very strange to me that after all the fuss over EFI, nobody bothered to get Mac-hosted development going for it. Perhaps people have tried but I can’t find them with Google.
There seem to be two possible alternatives. One is cross compiling MinGW and using that. I tried that but something wasn’t right because I couldn’t run the binary. The other alternative is cross compiling GCC for Linux/glibc (to get both ELF shared libs and an objdump with EFI support) and using the GNU-EFI linker script. I’m trying that method now.
As far as I can tell, EFI binaries are not like regular binaries because there’s no VM system in the EFI environment. This means that the PE binaries need to be relocatable but I think that both MinGW and GCC only allow the creation of non-relocatable binaries. The GNU-EFI linker script takes your app, compiled as a shared lib (which is relocatable) and moves the object code over to a PE binary with a relocator stub attached. It sounds like a giant hack but if it works, that’s good enough for me.
Kid Update April 14, 2006
Posted by a1291762 in : Uncategorized , add a commentEmily does not sleep with a dummy anymore. We took it away at the start of the holidays and it only took a few days for her to stop asking for it and crying. She went to kindy and didn’t even ask them for it.
We think that Emily is ready for potty training. We tried pull-ups but they’re too much like a nappy so Emily can’t feel when she’s done the wrong thing. We got a bunch of training pants from Best and Less. They’re padded underpants so they have a bit of leakage protection but they don’t feel very nice when she has an accident. She’s used the potty but seems a little reluctant to do so. We started just before Emily got sick and had to give up on it since she wasn’t in a state to handle change very well. Hopefully we’ll be able to get a bit further now.
Isaac is very restless. We have to tie him up or he won’t go to sleep. Emily did this too. Isaac rolls over all the time and can move around (not really crawling yet but it’s close). If we don’t tie his legs, he’ll end up squashed at the top of the cot, crying because his head can’t move.
Holiday Update April 13, 2006
Posted by a1291762 in : eve , add a commentI’m almost a week into my holiday and the kids are only just getting better. Worse, Bree and I are now feeling less than wonderful. I think I am catching up on my sleep a bit and I’ve gone through the caffeine withdrawls so hopefully the next week will be better.
After spending some time playing with Eve, I thought about partitioning, mostly because of Boot Camp and the new non-destructive resize facility in the diskutil program. I’m not sure what’s installled but I’m already using 40G. That seemed a bit big so I figured I’d play around with various “cleaning” programs to remove things that aren’t needed (eg languages that aren’t English). Unfortunately, one of the things I tried was a utility that removed non-Intel code from the binaries. That’s the point at which my system became unstable. Some things even refuse to run.
No problem, I thought. Just chuck the home folder onto Lukusdyret, re-install from scratch (without installing all the extra crap) and away I go. Of course, at that point, Luksusdyret decided to die. It was dead. Really dead. The machine seemed to POST ok but the screen never got a picture and the boot process hung before the OF keys were checked. I tried swapping components in and out until I was sure it was either the CPU or the motherboard, both of which I’m not willing to replace (after all, this is now a spare machine). Bree was panicky about the hard drives so I popped the boot volume into Max (with the system’s RAM, to help verify it wasn’t the cause of death) and the iMac sprang to life. I haven’t used it for ages but it totally blew me away. It didn’t seem all that slow and it doesn’t support any of the “acceleration” features found in 10.2 and 10.3. I checked out some of the extra RAM I had lying around (fished out of a dumpster when my work threw out some “bad” machines) and found a combo of 256M+64M that worked nicely together. I had long thought to make Max a media station/kids machine and it looks like it’ll do the job nicely. Since my monitor went bad, it also happens to be the only desktop machine with a working screen. I could hook Luksusdyret up to the TV but that’s got crappy resolution.
Anyway, I turned off Max and powered up Luksusdyret one last time and low and behold, it booted. I slowly added in alll the original components, powering on each time to be sure, until the machine was fully loaded again and running fine. I’m still not sure exactly what the problem was but I think I know why it went away. The battery on the motherboard is dead. It holds a charge for maybe a minute but that’s it. When you unplug the machine for longer than that, the clock is reset. I can only think that something in the volatile memory was making the machine refuse to boot. When I booted Max, I just used the power cord from Luksusdyret so it sat there for a few minutes with no power.
I’m now in the process of copying all 8.3GB of crap from my home directory on Eve to Luksusdyret. I briefly entertained the notion of doing the copy over the wireless network but the estimated time was 10 hours. 100Mb ethernet reduced that to 16 minutes. I can’t wait until we get a backup disk (ie. massive IDE disk in a FireWire enclosure) so that backups can be made wherever I am at 4x that speed.
Update Friday 14 April 2006
Don’t use Monolingual to remove non-Intel binary code from an Intel Mac! As noted above, the system became unstable. I couldn’t even connect to an AFP share! Luckily SMB still worked so I could backup my files. After doing the install I copied everything back via Wireless. It really did take around 10 hours to complete (it went while I was sleeping).
2.5 weeks of holidays April 7, 2006
Posted by a1291762 in : Uncategorized , add a commentI’ve just started 2.5 weeks of holidays. It was originally planned as a “rest + MacBook Pro” time but the timing fit with Bree’s holidays too so I’ll be doing some child minding so she can have a nice time too.
Of course, murphy’s law and everything, both of our kids got colds in the last few days so the holidays haven’t started very well.