vTual v.0.1 is Ready For Public Launch - For Real!
As we step into 2025, it's remarkable to reflect on the swift progress that I made in the past year: The development of vTual, a project that has captured all of my attention, was completed in less than a year!
This January marks a significant milestone, as version 0.1 of vTual has been finalized and is now generaly available for public launch.
Write More Insight!
Aight. As noted in the vTual changelog, the project has been in development since early 2024, specifically around March. This period coincided with my decision to resign from my previous job for various reasons.
This transition marked the beginning of an exciting once-in-a-lifetime journey for me, which allowing me to dedicate my time and energy to bringing this passion project called vTual come into life. However, the journey was not without its challenges.
You know, a significant number of shortcomings arose. Particularly due to a lack of thorough planning, which led to several complete overhauls of the codebase.
Additionally, I discovered that my experience was not quite sufficient to manage a project of this scale effectively. While the concept may appear straightforward at first glance, translating it into a functional, optimized and user-friendly platform has proven to be a multifaceted challenge.
As a result, the development process took nearly a year to complete, far exceeding the initial estimate of just 3 to 6 months.
But the storm has passed, as the development of vTual version 0.1 has been successfully completed and is now available for general public use. At least that's the plan.
What Does it Mean?
Okay don't be angry. On the surface, the vTual project is already fully operational and can be accessed and utilized through its official website without any obstacles. However, the information regarding this launch has not yet been publicly announced.
So only those who are already aware of the project can take advantage of its features at this time. YTTA.
Why Do That?
During the brainstorming process for the initial ideas and concepts, it was concluded that vTual will adopt two unique types of approaches regarding user management.
The first approach allows users to register and manage their own accounts independently, giving them control over their experience.
And the second approach enables administrators to manage accounts on behalf of the creators, acting as temporary representative for them. When the said content creator wants to manage their own profile, they just have to submit a claim.
But due to our very limited marketing budget, our current focus will be on adding profiles of various content creators to the platform, since a barren battlefield is not beautiful.
And once we have a substantial number of profiles, I will consider fully sharing the message to the world that vTual is not just a mere dream, but a tangible reality.
Also for the convenience of its users, I need to finalize the documentation related to this project. Since comprehensive documentation is crucial for helping users understand all aspects of vTual, including its features and functionalities.
By ensuring that this information is clear and accessible, I aim to enhance the user experience and empower individuals to make the most of the platform.
"So please try us when we are ready."
Is Everything Good?
Depending on the context, I can explain it as follows:
About the Codebase
Currently, vTual is at version 0.1, which means that the codebase is still in a stage where it functions adequately but is not yet optimized to meet all needs.
It's akin to trying to sleep comfortably in a room cluttered with trash; While its possible to get by, the environment is far from ideal.
However, I have previously worked with a codebase that was far worse than a decaying radioactive septic tank from Chernobyl. Afaik this particular "codebase" was even rented out for almost whooping $40,000 a year to numerous unfortunate clients, many of whom ended up suffering significant losses as a result of using it.
So in light of that comparison, I believe the vTual codebase is in relatively good shape.
About the OpEx and CapEx
Recognizing that a project like vTual requires a robust and complex server infrastructure and configuration, I have allocated OpEx amounting to at least few hundreds of dollars just for the first quarter of operational.
While this amount may not seem substantial to some, it is quite significant for me personally, especially considering that this is just a small hobby project. At that time my justification was that "This investment is crucial to ensure that the platform can handle user demands and function smoothly from the outset."
However, I later conducted a reality check to reassess whether such an investment was truly necessary at this stage. Like you know, can someone guarantee the success of this project?
The answer is, of course, was a hard big no. Then it became clear that while the initial allocation was well-intentioned, it might not be essential just yet.
This realization prompted me to reevaluate my priorities and focus on more immediate needs for the project. So I decided to cancel all plans regarding various plans and configurations about the whole server infrastructure design.
Such plans like setting up a high availability environment for a centralized storage servers, series of distributed database server or any other functions that theoretically could operate in a serverless environment; It is all gone now, so no more overcomplexity and overspending.
Goodbye to you annoying but loved portainer, haproxy, patroni, citrus and the rest of another little devil. You all somehow forced me to be an overnight DevSecOps.
Before revealing my own server benchmark, for comparison, I conducted a benchmark using YABS against servers from Hetzner Cloud with significantly higher specifications:
- CPX41 series, which features 8 cores of Epyc 7002, 16 GB of RAM, and 240 GB of SSD storage, priced at €24.70 per month, per server.
- CPX51 series, which features 16 core of Epyc 7002, 32 GB of RAM and 360 GB of SSD storage, priced at €54.40 per month, per server.
And that's the final benchmark of the server available on Hetzner Cloud.
So, what about the server I am using? Its specifications are likely something along these lines:
- 4 cores of Epyc 9554, 16 GB of RAM, and 200 GB of NVMe storage.
- 4 cores of Epyc 7443, 16 GB of RAM, and 200 GB of NVMe storage.
And that's the final benchmark among the many servers that I could, should, and truly regretted having to cancel. These two bad beautiful bitches is just born wild.
I won't mention the price or the provider, but I certainly had an overshadowing goal in mind when including this comparison.
And just as a fun fact, the server mentioned above was purchased during Black Friday, which means the price was significantly "more" reasonable.
In addition to the main server mentioned above, there are several other microservers that I cannot cancel because they have already been purchased for the upcoming year.
At first, the initial purpose of these microservers was to serve as part of a series of servers dedicated to database-related needs.
And lastly, as for the CapEx, I think that there is nothing important I can share since this project does not have fixed assets that require payment, such as buildings or other significant expenses.
About the Server That Will be Used
Sorry, I may have cut that point quite absurdly.
In a previous changelog, I shared an update regarding the plan to relocate the server infrastructure from Asia to Europe, which is apparently Germany, due to the relatively lower costs but with higher specification.
While it is true that I subsequently purchased a considerable number of servers in Europe, this does not mean that the servers located in Asia have become inactive. You know why? This is because the servers in Asia were also purchased on an annual basis. sobs.
Initially, these Asian servers were intended for specific system needs, such as handling email-related services and managing SSL certificates provision and renewal.
But since the servers in Asia were purchased on an annual basis, It is safe to say that, as a result, I do not need to spend a lot of money on operational costs at least for the next year.
Yep absolutely as you think. All operational needs of vTual, including the website, database, websocket and more, will rely on these two significantly inferior servers located in Asia, when compared to what used to be in Europe.
Several other functions and operations, such as the need for sending notification emails or load balancing for storage servers, will be managed by appropriate serverless services that specialize in these special tasks.
At least until vTual demonstrates clear signs that it requires higher server resources, then upgrading the server will be prioritized once again.
About the Remaining OpEx Budget
I once said that "I have allocated OpEx amounting to at least few hundreds of dollars just for the first quarter of operational" so, what will happen to the budget after concluding that we will utilize the previously available servers in Asia?
First and foremost, I plan to use the funds to subscribe to Twitter Blue as part of the promotional strategy for vTual, particularly targeting the vTuber community.
Promoting the platform while engaging with potential users seems to be an effective approach I intend to take.
By leveraging Twitter's features, I can enhance visibility, connect with potential users, and foster a sense of community around vTual.
Secondly, the budget may find its way back into my personal savings, which seems to be growing colder after being left with fewer occupants for quite some time.
About the Future Update
vTual will continue to be actively and regularly maintained, although major updates may no longer be the top priority.
For example, processes such as bug fixing and routine maintenance will remain ongoing to ensure the platform runs smoothly.
As a case in point, there was a significant overlap between two schedulers yesterday, which led to the server becoming overloaded by clogged tasks.
This unexpected situation also caused a dramatic spike in YouTube API usage, skyrocketing from the usual 1,500 requests per day to an astonishing 212,000 requests between January 14th - 15th.
But by midday on January 15th, the issue was resolved, allowing YouTube API usage to return to normal levels. As a result, from January 16th onward, the API requests seems stabilized.
This situation arose because testing in the staging environment was conducted using a relatively small dataset, around 2,000 to 5,000 data.
However, once the system transitioned to the production phase, the volume of data increased significantly up to 30,000 just in a matter of days, thus leading to unforeseen challenges.
Therefore, once again, vTual will continue to be actively and regularly maintained due to the presence of numerous potential risk scenarios that could threaten the system's operations. So don't worry because I'm still active maintaining, even though the addition of new major features may be slightly deprioritized.
But as a result, major ideas such as rewriting the better version of the codebase, localizing the language, adding new authentication methods, and implementing several other features may take quite a while to come to fruition.
However, one of the major update plans that may be implemented in the near future is the addition of a TikTok tracker, which I hope can be realized smoothly.
Update on TikTok implementation: After conducting research on the official documentation of the TikTok API, it can be concluded that TikTok API does not offer any public information for developers. ALL DATA related to an individual can only be accessed by the person themselves.
This means that ALL DATA accessible through the TikTok API, including general information such as names, bios, and video content, can only be accessed by the individuals themselves. Thus, vTual is unable to manage any information regarding TikTok profile of a user.
Fyi, services that have previously been integrated such as Twitch and YouTube also impose restrictions and limitations on certain information. But these restrictions and limitations only apply to data that is considered private, such as membership lists that CAN ONLY be accessed by the person themselves.
Tldr: Twitch and YouTube allow access only to public information, while all data on TikTok is categorized as private.
There are several options available for implementing TikTok API as "public information" however most of these options are unofficial, paid, and/or expensive.
We will not force implementation by taking this option, so that the TikTok integration was officially canceled.
Where Are You Going?
At this point, this posts may seem like a rather melancholic farewell letter. However, the reality is that I am not going anywhere. In fact, this marks a new beginning, albeit one that doesn't align with the original vision I have envisioned.
As I mentioned at the beginning, vTual is a project I have been working on during my job search after resigning. After all, radioactive nuclear is toxic, you know?
Unfortunately, job opportunities did not come my way in 2024, possibly because I lacked an active portfolio to showcase my skills and experience. You know what's funny? This simple project could even be completed while I'm looking for a new job since early 2024. Lmao.
However, now that vTual is active and has taken on its own form, I intend to use it as a portfolio for potential future employers who didn't employ me in Chernobyl-like environment. This is why I will continue to maintain vTual actively, since vTual is still far from mature as a product.
However, if vTual has the potential to become a primary choice in my career, I believe I could dedicate the best version of myself to making it an outstanding platform. Well, who knows?
Verdict
In conclusion, I hope that vTual can have a positive impact on all of us, aligning with the core essence of everything: Silverspoon.
My wish is for everyone connected to Silverspoon, you and me, to experience goodness and prosperity. May this initiative bring about beneficial outcomes for all involved, fostering a sense of community and shared success.