Re^4: Writing Popular Perl Software

I've got CI down for sure, between this, which uses external resources, and Test::BrewBuild which does things like this, all internal (note my RPi unit test platform has expanded a tiny bit). I can throw at my internal CI platform anything, and even test reverse dependencies that require my (or someone else's) code to ensure changes don't break things.

Started on AWS, but got sidetracked (I actually had a reasonable API started). I've also dabbled in trading a bit numerous years ago, but I abandoned that quickly, as nobody allowed me to use anything as a real-life test platform, and nobody would permit me to have historical raw data. Medical interests me. What do you have in mind?

Comment on Re^4: Writing Popular Perl Software

Replies are listed 'Best First'.
Re^5: Writing Popular Perl Software by Your Mother (Archbishop) on Oct 10, 2018 at 08:06 UTC
I had a pretty good Perl interface written for the TD Ameritrade API quite a few years ago, it was right when they launched it. The test suite used their historical data as fixtures. I got sidetracked with a new job and home life. Let it slide. Peeked at it about 2 months ago. They changed the interface. Haven’t dug into whether my code is retrievable; or garbage because that was when I was just becoming a halfway decent hacker. Took a couple simple passes at https://developer.tradier.com/ a few days ago. Again, time is the problem. It’s sad (for me) too because I had a scalping algorithm, based on some Perl statistics modules, that was working extremely well on historical data. Medical. I would love to see the DCMTK get a module family like the libxml2 stuff. I have taken some weak runs at it. I have zero XS chops and had no trouble wrapping up the command line tools I needed with Capture::Tiny and such so… A MeSH installer, license manager, and tool suite to go with UMLS::Similarity but make it all easy; if it could be tied in with ICD-10 and if there were a solid HL-7 distribution, Perl would probably become the backbone of choice for a few different kinds of medical applications. I started what I think is a nice piece of code for Coinbase which I played with during Coinbase API. Trouble there was their docs suck and I had no energy to pursue it. I think I may be taking the Lazy dictum wrongly these days. :P	[reply]

Replies are listed 'Best First'.

Re^5: Writing Popular Perl Software
by Your Mother (Archbishop) on Oct 10, 2018 at 08:06 UTC

I had a pretty good Perl interface written for the TD Ameritrade API quite a few years ago, it was right when they launched it. The test suite used their historical data as fixtures. I got sidetracked with a new job and home life. Let it slide. Peeked at it about 2 months ago. They changed the interface. Haven’t dug into whether my code is retrievable; or garbage because that was when I was just becoming a halfway decent hacker. Took a couple simple passes at https://developer.tradier.com/ a few days ago. Again, time is the problem. It’s sad (for me) too because I had a scalping algorithm, based on some Perl statistics modules, that was working extremely well on historical data.

Medical. I would love to see the DCMTK get a module family like the libxml2 stuff. I have taken some weak runs at it. I have zero XS chops and had no trouble wrapping up the command line tools I needed with Capture::Tiny and such so… A MeSH installer, license manager, and tool suite to go with UMLS::Similarity but make it all easy; if it could be tied in with ICD-10 and if there were a solid HL-7 distribution, Perl would probably become the backbone of choice for a few different kinds of medical applications.

I started what I think is a nice piece of code for Coinbase which I played with during Coinbase API. Trouble there was their docs suck and I had no energy to pursue it. I think I may be taking the Lazy dictum wrongly these days. :P

[reply]