UPDATE: The novel, Pleasing María, is blogged at website PleasingMaria.com. To view the Table of Contents, click here. To go directly to the first chapter, click here. To read the latest novel post, click here.
If you thought the last post was boring, this post is much worse – hit the back button now and escape, otherwise…
Quotes, and the processing thereof:
I’ve continued to scrape quotes from various websites, and have several thousand quotes, of varying quality accumulated. These quotes fall into many different categories, which presents several management difficulties. Because the novel I’ll be soon publishing is romantic erotica, I want to tweet quotes primarily in these categories:
- Travel/erotica – for subsequent novels in the series.
The last category is chosen on the advice of Michael Rogan in his eBook, “Twitter Marketing That Doesn’t Suck”. With thousands of quotes to manage, I’ve had to find and build some programs to help manage them. Working in the Linux and Windows environments, building my Author’s Marketing Platform has been interesting and fun, especially coding the C programs on Linux, an area where I worked for many years. Some specific tools:
- sort -u – to eliminate exact duplicates in the quotes database. A standard Linux program.
- shuf and sort -R – programs to shuffle and randomize quotes. Standard Linux programs.
- vim – a text editor with Regular Expression capabilities. This makes cleaning-up web scrapes pretty easy. A standard Linux program now ported to Windows. This is a text editor intended for programming, not a word processor. Its command set is archaic and obtuse, but if you need to work with large quantities of messy text, its implementation of Regular Expressions is unbeatable.
- similar – a program to find duplicate quotes with small variations (they’re not exact duplicates) using a string distance algorithm. String distances between 25 and 30 work best. This is a small C program that I wrote.
- strip140 – to separate out quotes that are too long for tweets. A trivial C program.
- SeparateQuotes – a program, to review (by eyeball) quotes and assign them to my categories. A small C program I created.
- VirtualBox – allows me to run Linux on my powerful Windows computer (and my Netbook). Available for free from Oracle at https://www.virtualbox.org. I installed a minimal Linux, Tiny Core. This is not for the faint-hearted.
- ManageTweets – my personal program to combine quotes, twitter user addresses, URLs, and hashtags into Hootsuite format. This is a larger C program that I’ve continued to refine as my understanding of Twitter has evolved.
Regarding tweets and Hootsuite, there are several issues:
- Tweet real-estate is very limited – 140 characters max.
- Hootsuite reduces tweet length to 125 characters if you include a URL.
- Hootsuite allows a maximum of 350 tweets to be bulk scheduled. This is ridiculously low, as socialoomph.com allows 5,000 tweets to be scheduled.
- Hootsuite doesn’t allow duplicate quotes.
- Hootsuite auto-shortens URLs (good), but only puts their auto-shortened URLs at the end of the tweet (bad).
- Hootsuite’s strongest advantage over other bulk-scheduling programs is the ability to schedule tweets at specific times. The negatives are: 1) only one tweet can be scheduled at a specific time, and 2) times are at 5 minute boundaries.
- I’ve discovered that many of my followers are located in the UK and farther east; Hootsuite’s ability to schedule tweets at specific times is useful for reaching those folks during their time of peak usage.
My tweet management program, ManageTweet, works around issues 2-5 by implementing substitution macros within the tweets that are substituted for twitter addresses, URLs, or hashtags as the tweets are created. For example, the macro %DA*% is changed into a twitter address; the macro %GO36% becomes “http://GuyOrdinary.com/?p=36”; the macro %HER% becomes the hashtag “#EroticaAuthor”. I can redefine these macros dynamically to change their meaning and can put the macros anywhere within a tweet, so a URL can be put at the front or in the middle of a tweet, for example. Regarding issue 6, I can set schedule interval times and increase tweets at peak times. (I warned you that this would be boring!) My personal Twitter management tool is now basically complete; I still need some hashtag improvements.
Immediate projects I need to work on:
- Understand the full power and etiquette of hashtags, and incorporate them better into my tweet management program. At the moment, I mainly use #EroticaAuthor to inform readers of my non-kinky quotes of my principal focus.
- Categorize my quotes, find more sex related quotes, and improve their quality.
- Launch-time promotions for my novel. I’ve got some interesting and novel ideas for these promotions, but will keep them under wrap until they are better developed.
If you’d like to receive announcements of future blog posts, join the mailing list on the right side of this page. To contact me directly, reply to this post, or click on the CONTACT GUY ORDINARY link at the top of this page. Visit http://PleasingMaria.com to read the novel as blog posts starting in November, 2015. If you know of good solutions to the problems I’m trying to solve, please contact me. Good luck!
All the contents of this web site are Copyright © 2015 by Guy Ordinary, all rights reserved. The contents have been registered as a published work with the U.S. Copyright Office.