Still no actual sewing to be seen. Tonight I washed 8m of bemsilk lining, which will be used to line this first dress and at least one other. Once it was washed, it crinkled up like a mofo and I had to iron ALL of it while it was still damp.

Seriously, my wrist hurts now. It took me over an hour.

Rather than cutting the lining, I started prep work on another dress. This one is a shirtdress from a traditional, well-known pattern company. (I’m not naming them for reasons you’ll soon see, but it’s obvious in the photos.) I purchased a “downloadable” pattern, and it’s the first and last pattern I’ll ever buy from them. It came as a DRMed PDF, and I had to install some really dodgy proprietary software to even view it. And here’s the thing – this dress has a couple different options, so there are many, many possible pieces. The whole pattern is well over 100 A4 pages. And the software will only let you print the entire thing. You can’t specify pages, and even if you could, there’s no overview that shows you just the ones you want. Oh, and the DRM also prevented you from printing-to-PDF to try to work around these limitations. I was livid.

Of course, a livid geek is also a motivated one. In less than an hour I was able to, uh, figure out how to create an unencumbered PDF. That allowed me to create an overview which showed all the pages together, thus allowing me to see which page ranges I needed to print for my selected dress. Turns out I needed less than 60, so I saved 40+ wasted pages. YOU’RE WELCOME, TREES. AND SUCK IT, TRADITIONAL PATTERN COMPANY THAT I’LL NEVER BUY FROM AGAIN.

You can see my overview on the left there. The green pages were the ones I needed for my dress. On the right is the stack of printouts.

Print-at-home patterns are convenient but it means you have to go to the effort of assembling them. I have a good system down involving a mat cutter, glue, and tape. I managed to get about half the pages stuck together and assembled tonight.

Tomorrow I’ll finish assembly and then cut out the pieces properly. Then on the weekend I should be able to start sewing the outer part of the Hawaiian bias dress and cutting the lining!

After yet more crawling around on my poor knees I managed to finish cutting all the main pattern pieces tonight, which means all the masking tape and fabric are finally off the floor. For now. 😂

The final preparation step is to hang all of the pieces for at least 24 hours. This is because they were cut on the bias – that means the fabric (which is already really drapey due to the rayon in it) is very stretchy and can grow on you. So by letting it hang, you’re allowing the pieces to stretch before you start sewing them together.

I was a little worried about how the skirt pieces would come out. It was hard to tell as I was cutting it whether the flower placement would be nicely random. The three pieces across the front are more similar than I’d like, but it’s still offset a bit and the Snook said he didn’t even notice. I’m not sure whether I have a big enough piece left to cut another one, so I think I’m stuck with it regardless.

I reckon I’ll leave these to hang tomorrow while I get onto the next task: cutting out most of the pieces again, this time in the lining fabric. (My back is killing me already…)

My goal was to have the Oolong dress cut out tonight… and I didn’t quite get there. First I had to iron my fabric though. It’s a Hawaiian print that I bought in Honolulu more than 5 years ago, and of course I no longer remembered what the hell it actually was. Between my Internet sleuthing and the Snook’s chemistry knowledge (ie he literally burnt a chunk of it), we’re fairly certain it’s a cotton/rayon blend. It’s very drapey and should work nicely with this pattern.

Now – this dress is cut “on the bias,” which means you have to place the pieces at 45-degrees to the fabric’s grain. This means that I actually needed to square up my fabric. I ended up laying it out on the living room floor, using the wooden floorboards and a load of masking tape to get it stuck down straight. Then I started pinning and cutting.

Ugh. Crawling around on your hands and knees on a wooden floor really sucks. I managed to get about halfway done, so I’ll finish up tomorrow. I’m wrecked.

As you would’ve seen from the photo in the previous post, Frocktober is here again! I am once again taking part, and you can donate towards my efforts here. I have set a stretch goal of $2000 raised for the Ovarian Cancer Research Fund this year, but I reckon we can do it!

In addition to wearing a different dress every day, I’m also going to be documenting a couple frocktastic sewing projects throughout the month. I haven’t sewn anything for several months, so it was nice to tidy up the sewing room today and to hit Spotlight for supplies.

My big challenge with sewing (and it’s the same challenge I have with cooking) is that I lack the patience to set up my mise en place properly. Most of the work isn’t the sewing part; it’s in the painstaking preparation: tracing, ironing, and cutting. I need to learn to appreciate this part just as much as I do the actual assembly.

My tasks today were to wash my fabric, trace out my pattern – Colette Oolong – onto some interfacing, and then pin it together into a toile for sizing. I tried it on and the Snook assisted with some adjustments. (We added some width to the underarm pieces, deepened the darts at the upper back, and adjusted the shoulder slope slightly.) Then I had to transfer those adjustments back to my pattern pieces.

That’s all for today! The next step is to iron my fabric and cut out the pieces.

This is amazing. In 1996, a high school senior in California shot a documentary about a group of punk-feminist 8th grade girls. It’s a time capsule. I love this…

My impulse when watching was to figure out where I would’ve been in that particular social pecking order. (I graduated in ’95, so a few years older than most of the students in the film.) I would’ve admired those girls, I think, but I didn’t have the self-confidence to fully identify. I was trying so hard to not stand out, while simultaneously really wanting to. I was probably one of the milquetoast hangers-on, to be honest.

I was really happy to read a follow-up to see that the women are still kicking ass now, 20+ years later, and they’re fine with the reaction to the video.

Note to self… look into hosting WP on AWS. 🤔

Last week I was delighted to receive an honest-to-goodness feature request for MUGicalNode from my new colleague Paul. He said that he really liked the calendar, but he had issues when viewing on his iPhone. Namely, on an day with lots of events, you end up with something like this:

That’s not very nice, is it? He wondered if there was a way to create a single “event” per day and then have that be a list of all the meetups. Sounds like a good idea, I said!  This afternoon I set to work making that a reality.

Introducing digest calendars

I modified my existing “makecalendar” Lambda to also create a separate digest calendar and upload it to the same bucket. For each day, it creates an all-day event with a description that lists the events and their URLs. You don’t get as much info (like the individual meetup description or venue details), but you can click on the link to go to the relevant Meetup.com page. I think it’s a really nice solution!

Note: depending on how many events there are on the day, you may need to click “Show All Notes” to see the full list on your phone.

If you’d like to subscribe, the URLs are:

• https://krishoward.org/sydneymugs-digest.ics
• https://krishoward.org/melbournemugs-digest.ics
• https://krishoward.org/brisbanemugs-digest.ics
• https://krishoward.org/perthmugs-digest.ics
• https://krishoward.org/hobartmugs-digest.ics

Please let me know if you encounter any issues. Thanks to Paul for the suggestion, and to the Snook for helping me untangle a mess of timezone nonsense.

Back in January 2017, I wrote a PHP script I called MUGicalPHP that used the Meetup API to create iCal feeds (ie calendars you could subscribe to) for tech events in various Australian cities. The idea was that it would update the feeds every day with new events. It was kind of flaky though, and it often failed silently without me knowing about it. (Sometimes the Meetup API just decides to return ZERO RESULTS for some reason. 🤷‍♀️) Still, it was useful and mostly worked, so I shared the links with lots of folks in the community and made plenty of use of them in my job at YOW!.

Fast-forward to 2018, and now I’m working at AWS. As I was going through the hiring process, it occurred to me (not for the first time) that I could probably use “that Lambda thing” to rewrite my meetup script to be “serverless.” So… I did it!

MUGicalNode is now working and it’s up on Github here. If you were using the old links, they should point to the new version and hopefully you shouldn’t notice any change at all (beyond them being more reliable and frequently updated!). If you just want to subscribe, add these in your calendar application of choice:

• Sydney – https://krishoward.org/sydneymugs.ics
• Melbourne – https://krishoward.org/melbournemugs.ics
• Brisbane – https://krishoward.org/brisbanemugs.ics
• Perth – https://krishoward.org/perthmugs.ics
• Hobart – https://krishoward.org/hobartmugs.ics

How it works: I started the project with a straight monolithic reimplementation of the old PHP script, and I was super proud when I got it working locally. Unfortunately, the Meetup API did not like me firing off lots of requests in parallel and promptly banned my access key for a bit. As a result, I ended up breaking the script into three parts: “getgroups,” “getevents,” and “makecalendar.”

• “getgroups” runs once a day for each city. It requests all the groups in a location matching a topic list and then stores those groups in Amazon Simple Queue Service.
• “getevents” runs every minute of every day. It checks the SQS queue to see if there are any groups. If there are, it gets the top one and requests the next 10 events for it from Meetup.com. If events are returned, they’re saved in an Amazon DynamoDB database.
• “makecalendar” runs once a day for each city. It retrieves all the events for the city from the DynamoDB and then constructs an iCal feed from them. It then uploads the file to an Amazon S3 bucket.

This seems to spread the requests out enough that Meetup doesn’t complain. That said, as of three days ago SQS is now a supported event source for Lambda! That means instead of triggering “getevents” every minute, I could simply use the queue as the event source and have it trigger automatically. My only hesitation is the throttling on the Meetup side, so I’m going to have to do some testing to see if this is feasible.

Costs: I don’t have an operating cost estimate for this project, but I suspect I will be well within the free tier for all these services. I’ll let you know at the end of July.

Additionally: As part of this project, I also rebuilt my old personal site https://krishoward.org and migrated it to AWS for hosting. That code’s on Github here. I was indebted to this great post from Vicky Lai that describes how to get your custom domain and SSL certificate working with S3 buckets.

This is very, very cool.