Archive for August, 2017

August 23, 2017

Baby Gear, Round 3!

It’s time to start thinking about baby gear! Z3 will be here before we know it.

A New Baby Monitor System. I loved our old monitor, really I did, but it’s showing it’s age and barely functional at this point. The parent console often loses signal and sometimes can’t process the image feed. We’re left with a ghostly white image that lingers on the screen. I’ve gotten used to enough that I can guess what’s going on in the room. Usually. If that’s not enough of a draw back, the night time vision of one of the cameras is about half power, and whenever the parent console is connected to that monitor it picks up an intermittent buzzing that keeps me awake. Since the girls are in the shared big girl room, we’ve just been using the working camera.

A new high chair. I hated our old one. We were going to pitch it and replace it before Alexis was born, but my frugal tendencies took over and saved it from the dumpster. The padding is now ripped, so I think I can finally give myself permission to part with it. I have no idea what kind of high chair I’ll look for next aside from cheaper and simpler.

A crib. The old one is ok, but with the non standard screws I can make a case why it should be replaced. It’s also been chewed on quiet heavily, and shows the wear of having been used for two kids already. Normally I’m fine with cosmetic defects, but I know I’m going to want to take photos of a sleeping baby in her crib.

A bottle warmer. These things always seem to get so gross. Looks like each kid will be getting her own.

Another rocking chair. One can never have too many rocking chairs. The first one we purchased is downstairs in our living room, and still comfortable but showing it’s age. We purchased a second one that is currently in the nursery, but that’s in the middle of the house. And also not very comfortable, if I’m being honest. I’m thinking about getting a third one for the master bedroom since Z3 will probably sleep in there for a few months (I know, it’s a total splurge.) I figure I’ll use the new one to replace the one downstairs at some point.

I’m kind of amazed that the jumperoo is still in good shape, and Alexis’ activity gym given how Nicole destroyed hers. It sometimes feels like those are the only two things that survived. The rock n’ play is doing okay, but the fabric just isn’t as nice as it once was. Alexis’ changing pad is also in ok shape. We recently purchased a new crib mattress before moving Alexis to a twin size, so at least we’re good there.

I kick myself every time I mess up the settings on my camera. On Nicole’s first day of school I mistakenly left my camera in full manual mode, with the shutter speed set for indoor photography of stationary objects. The photo on the left is my favorite pose wise. Nicole is so happy. But the photo is so over exposed with a slight motion blur given the slow shutter speed. Fortunately I realized my mistake, adjusted my settings and was able to also capture the photo on the right.

Current state of the art photo editing techniques cannot save the photo on the left. The skin on the left side of her face is so over exposed the sRGB value is pure white. The computer simply has no way to figure out what color was supposed to be there, so there’s no way to automatically fix the white patch. Nicole could have cyan colored skin for all the computer knows. Or checker pattern skin, for that matter.

As an machine learning person, I find this limitation incredibly irritating. Chances are, when you, a human, first looked at the image on the left Nicole’s skin did not appear pure white to you. Your brain filled in a likely color based on the right side of the photo where her skin isn’t as badly over exposed and what you know of human skin tones. With the two photos side by side you can easily imagine a combined photo with the pose from the left and the coloring on the right.

Current state of the art image recognition can recognize faces. It should be able to map point for point the location of the eyes, mouth, nose, hands, etc between the two photos. Given that, it seems plausible for a AI enabled photo editing software to merge the two photos and create the ideal image.

Even without the image on the right, a sufficiently advanced AI based system should be able to recognize the subject matter much the way a human would. The system should then be able to generate a hyper realistic plausible image. It might not be the “correct” image, the image that would have been created if my settings were correct in the first place. I doubt for most people that would matter. A realistic enough resulting image that was plausibly correct would likely be sufficient for most momtographers like me.

August 15, 2017

Girl!

I am over the moon. Three girls was exactly the family I’ve been hoping for and envisioning.

I know I’m not supposed to have a gender preference. Or maybe I’m supposed to have a preference for the other gender I don’t have? Everyone around me seems to think I’ve been hoping for a boy, and were rooting for a boy on my behalf. Truth is, I’ve had my heart set on a third daughter since before she was conceived. With Alexis (Ziggy) I only developed a strong preference for a girl a week before the anatomy scan. With Nicole (Zippy) I was sure she was going to be a boy and I never really had a chance to hope for one gender or the other.

Nicole is super excited she’s having another sister. I told Alexis and she corrected me saying “it’s a boy!” I guess we’ll have to wait and see what she thinks when Z3 gets here!

Our anatomy scan went well. I’m oddly not feeling much movement this time around. The technician was surprised when I said that since Z3 isn’t in a bad position for movement. Maybe it’s just that I’m so distracted charging after my first two that I’m just not noticing the movement as much? She was quite the little active bean for the technician, quick to prove she was alive and healthy.

The technician wasn’t able to get one of the photos she was after. The hospital is trying something new where the perinatologist is on hand, so there’s no longer a need for second appointment when these situations arise. (With Nicole I had to go back in at a later date when the same thing happened.) Our technician explained the new approach up front to alleviate any concerns should they need to summon the perinatologist. I’d like to think as a third time mom I’m a little less prone to worrying, but we all know that isn’t true. Neither the perinatologist nor the technician were able to capture the structure they wanted in a single ultrasound image, but both were confident that they saw said structure and everything was developing normally so they sent us on our way.

August 8, 2017

Mommy & Me Days

With Nicole’s Kindergarten right around the corner, and the baby’s due date in the not to distant future, I’ve been thinking a lot about how best to ensure each child gets the individual attention he or she needs.

This spring Nicole wasn’t eating her school lunches. I told her if she eat at least two bites of everything on her plate every day for a week, I’d pick her up for lunch the next Monday and we’d go to a restaurant. Just the two of us. That was all the motivation she needed. She was super excited and eat like a champ, even the vegetable, often eating more than the stated minimum.

A few weeks later the elementary schools were all out on break for the spring holiday. Daycare attendance was surprisingly light as a lot of parents decided to keep both kids home, rather than just the school age child. The more I think about it, the more this creates a nice opportunity for a Mommy & Me day.

The kids are at different developmental stages. Nicole loves to play with beads, and make necklaces and bracelets. That’s hard to do on home days with a little sister around who sometimes puts things in her mouth. Alexis’ attention span for coloring is roughly ten minutes, where Nicole can go on for hours. Nicole is old enough to sit quietly through a movie at the theater. Alexis, not so much. The usual result of these differences is the older child is relegated to the activities the younger child.

If Nicole’s school is out, but Alexis’ is still in session, that doesn’t have to be the case. The opposite is also true, and Alexis will have the opportunity to have the big doll house and kitchen all to herself. We can spend time reading just the books she enjoys, and she won’t have to wait for Nicole’s longer chapter books to be finished.

To put this philosophy to the test, Domingo and I decided to enroll Nicole part time in her final week of preschool. She’ll get two special home days with Mommy before starting kindergarten. We’ll go to the movies, pick out her new school bag and lunch box, and generally do whatever she wants to do. Next month Alexis will get the same treatment.

August 5, 2017

1st Day of School Sign

With Nicole starting school soon I started thinking about all those First Day of School photos we could take. I really liked the idea of doing a chalk board photo, but my calligraphy is about what you’d expect of someone who spends most of her time typing. And is dyslexic. There are places where one can buy a faux sign (like etsy), but then I’d need to keep buying a new one every year, and if the seller I was buying from decided to close up shop I might be not be able to find someone who could duplicate their work. A slightly different font choice, layout or color would drive me absolutely batty. I need things to match, it’s a compulsion. Besides, I’m pretty handy with the graphics program. I should make my own!

I started with this tutorial which included the chalk background.

In my experience, a true white (#FFFFFF) color text was a little too white, and made it obvious the chalk board was fake. Another tutorial recommended #CCCCCC for the color of the text, but that ended up being too gray. A good compromise was #E3E3E3.

Once I had the image designed I was on the hunt for a wide wood (or faux wood) picture frame where I could remove the glass. I settled on this one from target after holding up a few next to an actual chalk board to see how well the wood grain would good with the chalk board texture. I bought three since they were on sale, and I will one day have three kids in school simultaneously.

I printed the image on heavy card stalk. The card stalk had a nice grain texture which helped with the realism of the chalk board. The heavier the cardstalk, the stiffer the paper and the less likely it is to bow without the glass to keep it pressed flat.

Finally, I removed the stand from the back, and taped over the hooks so they wouldn’t catch on Nicole’s clothes when she held up the sign.

I am very happy with how the sign turned out. Best of all, since I used my own graphics program it will be a snap to go in and change the grade level, or the career aspirations every year. I like it when I make things easy for my future self.

July was a more encouraging month, at least in terms of user growth. In July I had a 19.3% growth in users over June, however I made $3 less! Oh well, I keep saying I care more about user growth than income growth. Users tends to be more stable, which means I expect income to catch up at some point.

In terms of progress I made a major change under the hood. I had previously been using a two stage deployment. I had two different work environments, a development environment were I could work on changes and a production environment that hosted the live website. To reduce user bandwidth, I’d compress files while moving them from development to production. This worked fine up until now as I could always fine a time of low to no user activity.
As long as I followed my own invariants, the compression worked seamlessly. When I broke them, I could quickly undo the compression before anyone noticed and avoid much disruption. Downtime was no more than minutes at most, and there were always some minutes a day (usually in the middle of the night) where it could be risked.

Datayze has grown to the point where that is no longer the case. Some apps don’t really have down time anymore. Less trafficked times of the day are no longer no traffic times. In some cases I’d wait days before having an opportunity to push a change. It’s good news for datayze, but means I need to split up and test the compress step independently of the push to production.

Now I have three independent silos, one for development, one to test the compression, and one for deployment. Technically I have a fourth “sandbox” silo for crazy, half formed ideas, but there’s no automatic method for pushing from there to development. I keep the sandbox completely separate from development so I don’t accidentally deploy something very not ready for prime time. I don’t want to get a reputation for broken code and wonky, misbehaving apps!

The change should help increase productivity. Now I have much more flexibility when I deploy.