When building your city you should not view the city-design part as a way to solve your plumbing problems. Not if you want someone to live there, anyway.
Nor should you build a city for which plumbing cannot be designed.
But rather, in your dream, in your vision of what a city should be, consider that you will still need plumbing and it shouldn't cost a million dollars to fix a leaky faucet.
Beyond that, the plumbing shouldn't be much of a consideration, unless you are building a city for plumbers.
This is what I am calling an MVP, which stands for Minimum Viable Poems. I have put together what I feel is good content (poems), and I have basic site navigation and responsive design working.
I realize there isn't huge overlap for those who might be interested in this blog for the technical or work related stuff and those who might be interested in poems and literature.
None the less, it's an area of interest for me: poetry and technology. I am curious about these two things, how they overlap, and how they can help us in our everyday lives.
This took me longer than I thought it would to get right. In the end the solution is simple.
The main idea is that each line of a poem needs to wrap in such a way that you can tell it is a line. So, to accomplish this each of the lines of the poem are contained in a <p> tag.
This works for both mobile and non-mobile. The crux is that text-indent: -20 will outdent the first line by 20 px. The margin setting ensures that the text doesn't accidentally overlap something else on the page.
If the screen is small enough and any lines wrap, text-indent will not impact any subsequent lines causing them to appear indented, which is the poetry convention for a wrapped line.
This is web-only, and not designed for e-readers. I started with an article on that here, but ultimately the first-line stuff led me astray for a while. For my purposes, text-indent alone was enough.
I apologize if you came to this site and it was slow.
I just upgraded the VM that runs this site to address the problem. I am hopeful there will be fewer problems now.
It appears that the old VM was memory-bound or near-memory-bound a lot of the time. This was likely leading to the CPU problems because of paging.
Now the VM has room to breathe on the memory side and I believe that will resolve the problem. I have no proof that memory was the actual problem, only a correlation between slow times, higher memory usage, and CPU spikes. Time will tell if this resolves the issue.
In the meantime, paying a few more bucks to see if things get better seemed like the right thing to do.
Jonathan Fries on .