- Oct 16, 2010
- 7,169
- 9
- 8,718
Greetings everyone,
I am reaching out today to provide an update on the transition to ASF v2. To quickly recap, we made the switch to ASF v2 last week. While the move has proven beneficial for most, some members have been a bit frustrated and taken aback by the abrupt transition from ASF v1's traditional look to the modern, material aesthetic of ASF v2. This dramatic generational jump has even prompted a few members to leave our forum, a decision I respect entirely. However, it's important to clarify why these changes were both crucial and necessary.
Given these circumstances, we were left with two choices: shut down ASF or build a new version, now known as ASF v2. Choosing the first option would have been easy but unconstructive, whereas the latter presented challenges, particularly due to ASF's limited budget.
Despite these challenges, after six years of dedicated effort, ASF v2 has been realized. I share this with you so you understand the context behind the decisions made.
I am reaching out today to provide an update on the transition to ASF v2. To quickly recap, we made the switch to ASF v2 last week. While the move has proven beneficial for most, some members have been a bit frustrated and taken aback by the abrupt transition from ASF v1's traditional look to the modern, material aesthetic of ASF v2. This dramatic generational jump has even prompted a few members to leave our forum, a decision I respect entirely. However, it's important to clarify why these changes were both crucial and necessary.
- ASF v1 was reaching end-of-life: The software dependencies it relied upon were either discontinued or greatly outdated, leading to substantial security vulnerabilities that were increasingly difficult to rectify. Additionally, features such as "Thumbnails," Likes, and Thanks, ceased functioning due to these outdated dependencies, resulting in blocks from numerous modern browsers.
- ASF v1 was riddled with spam: Security gaps allowed spammers to exploit non-critical breaches, culminating in a flood of unwated content. The fight against this wave was turning progressively more difficult.
- ASF v1 was not scalable: Its outdated platform made scaling up services a daunting task. Purchasing additional CPUs to match the forum's demands was costly. Additionally, the compatibility issues of older backend software like PHP 5.6 with 2022's operating systems, alongside the nearing end-of-life of supporting platforms like CentOS 7, made maintaining the old backend and frontend near impossible.
Given these circumstances, we were left with two choices: shut down ASF or build a new version, now known as ASF v2. Choosing the first option would have been easy but unconstructive, whereas the latter presented challenges, particularly due to ASF's limited budget.
Despite these challenges, after six years of dedicated effort, ASF v2 has been realized. I share this with you so you understand the context behind the decisions made.
Following the switch, we've identified several issues with ASF v2:
- 'Bad Gateway' Error: This is the topmost priority to fix. Essentially, on AlmaLinux 8 (a CentOS 8 fork, as IBM terminated CentOS 8) which utilizes Linux Kernel 4.x, the NFS client caused a deadlock, leading to the computing node's inability to read PHP files from the master server, and subsequently failing all requests.
- Inaccurate Japanese Search: Certain members have noticed that searches in Japanese do not always yield the desired results. This anomaly can be attributed to a lack of full-text searching for Japanese in the index. We are planning to remedy this issue soon.
- Interface: We have received and made some adjustments based feedback concerning the new interface — too much wasted space, excessively bright light theme. We assure you that all feedback is taken into serious consideration as we continue to make further adjustments. Your observations and suggestions are valuable to us in this ongoing refinement process.
Last edited:
