Ideas and Opportunities: TikTok SDK
Posted: Sun Aug 13, 2023 2:46 pm
I have new ideas for apps and features every day and I'm generous in sharing them. Ideas are cheap, execution takes time and opportunity is time sensitive and concrete--self-employment/freelance for a client or a market inefficiency that can be capitalized on directly with one's own resources. I'm tight lipped about some classes of opportunities, but I talk about them and ideas if they're something I want to see in the world and I don't care who builds it. This journal is an attempt to illuminate my thought process, show activity and share these things alike in a raw/lightly/un-edited format.
I've never used TikTok and don't even have a mobile phone in my possession. But I'm aware of the popularity of the platform and the controversy around data collection. I'm indifferent about the ethical issues as I know it's possible to opt-out of mobile phone use completely, but of course I believe everything should be as open source and self-hostable as possible. The latter creates a more voluntary and informed computing landscape. But in doing research, I just want to see what's there for developers. So whenever I can get a glimpse into a platform through an SDK I usually look into it, and sometimes implement parts of it. I found TikTok's SDK today.
When I see that a platform has an SDK and deide to look into it, I mentally map out the operations and objects available in terms of social networks that I know. Hogumathi is my concrete realization of this mapping. I've added several providers with login and feeds, and through a fairly complete grammar or taxonomy of object and collection types that can be used to map the functionality. Often I'll just stub out functions and link documentation that would help realize the different methods. I have enough providers that I could make a matrix as part of the documentations. Development has been paused for a while and even I can't remember the details of each provider without going back to the code. I had a Trax project setup on my CarPuter but I lost possession of it when I fled to the homeless shelter.
Time is not free unfortunately and I'm a one man show, so I can't jump on every shiny thing I see. I have to objectively weigh the amount of opportunity that a pursuit could attract. Doing things like writing README files and maintaining an issue tracker both help myself in the future when I come back to a project after a while, and can also be public visibility if I want to share a project. I tend to not talk about most of my projects publicly anymore, and also I am not comfortable with using my work to train an LLM, so the platforms that I can host my private work on are limited. But Hogumathi is one that I've done some work to commercialize, with a free base distribution and premium extensions with all of the experimental providers.
Marketing isn't something I've ever put much thought into during my career. I organically marketed my self-employment services pretty well in the past by participating in user groups and publishing my blog. Something I've obviously seeking to reproduce lately. I got plenty of freelance and employment opportunities through word of mouth alone. But since I had the falling out with OpenTable in 2016 people have been pretty apprehensive about associating with me, which is totally rational. I'm a damaged stock and not a damaged company, so I will make a recovery. But it does limit my opportunities quite a bit. Some people have stayed loyal which I'm thankful for, but still when given opportunity I can't perform at the level I once did due to simple lack of resources.
TikTok is pretty interesting because people can get exposure with so little. I'm interested to know how engineered the algorithm is, what the actual production process is behind things that get traction, and how much return can actually be built on that traction. Since I've not personally used the app myself it's hard to know. I can look closely at screenshots and find some top brand case studies. That's a dimension I'm not accustomed to. My natural level is looking at developer documentation and working with a 'business' partner who takes care of all the non-technical details.
I've never used TikTok and don't even have a mobile phone in my possession. But I'm aware of the popularity of the platform and the controversy around data collection. I'm indifferent about the ethical issues as I know it's possible to opt-out of mobile phone use completely, but of course I believe everything should be as open source and self-hostable as possible. The latter creates a more voluntary and informed computing landscape. But in doing research, I just want to see what's there for developers. So whenever I can get a glimpse into a platform through an SDK I usually look into it, and sometimes implement parts of it. I found TikTok's SDK today.
When I see that a platform has an SDK and deide to look into it, I mentally map out the operations and objects available in terms of social networks that I know. Hogumathi is my concrete realization of this mapping. I've added several providers with login and feeds, and through a fairly complete grammar or taxonomy of object and collection types that can be used to map the functionality. Often I'll just stub out functions and link documentation that would help realize the different methods. I have enough providers that I could make a matrix as part of the documentations. Development has been paused for a while and even I can't remember the details of each provider without going back to the code. I had a Trax project setup on my CarPuter but I lost possession of it when I fled to the homeless shelter.
Time is not free unfortunately and I'm a one man show, so I can't jump on every shiny thing I see. I have to objectively weigh the amount of opportunity that a pursuit could attract. Doing things like writing README files and maintaining an issue tracker both help myself in the future when I come back to a project after a while, and can also be public visibility if I want to share a project. I tend to not talk about most of my projects publicly anymore, and also I am not comfortable with using my work to train an LLM, so the platforms that I can host my private work on are limited. But Hogumathi is one that I've done some work to commercialize, with a free base distribution and premium extensions with all of the experimental providers.
Marketing isn't something I've ever put much thought into during my career. I organically marketed my self-employment services pretty well in the past by participating in user groups and publishing my blog. Something I've obviously seeking to reproduce lately. I got plenty of freelance and employment opportunities through word of mouth alone. But since I had the falling out with OpenTable in 2016 people have been pretty apprehensive about associating with me, which is totally rational. I'm a damaged stock and not a damaged company, so I will make a recovery. But it does limit my opportunities quite a bit. Some people have stayed loyal which I'm thankful for, but still when given opportunity I can't perform at the level I once did due to simple lack of resources.
TikTok is pretty interesting because people can get exposure with so little. I'm interested to know how engineered the algorithm is, what the actual production process is behind things that get traction, and how much return can actually be built on that traction. Since I've not personally used the app myself it's hard to know. I can look closely at screenshots and find some top brand case studies. That's a dimension I'm not accustomed to. My natural level is looking at developer documentation and working with a 'business' partner who takes care of all the non-technical details.