Greetings, all!
I'm at the Wikimedia Hackathon in Milan and, thanks to the help of Bernhard Krabina (who introduced us) and Daniel Dobriy (who is doing the actual work), we'll be moving WikiApiary and its bots to a kubernetes cluster.
You can see the work in progress at https://wikiapiary.dobriy.ai/.
The hope is to use these resources to start WikiApiary crawling and discovering wikis again.
Of course, none of this is to diminish Bawolff's much needed work last month to get the site actually serving pages again.
Thanks,
Mark.
It is always delighting to hear about WikiApiary!
By the way, while the old site will be open to humans and bots soon, what would the destiny of the new site be? Will it receive continuous development to become fully functional? Or will it be dropped due to the lack of manpower? The development of the new site has been stalled for a long time, and we are all eager to hear something from the developers!
TripleCamera.
在 2026/5/2 21:00, Mark A. Hershberger via Wikiapiary 写道:
Greetings, all!
I'm at the Wikimedia Hackathon in Milan and, thanks to the help of Bernhard Krabina (who introduced us) and Daniel Dobriy (who is doing the actual work), we'll be moving WikiApiary and its bots to a kubernetes cluster.
You can see the work in progress at https://wikiapiary.dobriy.ai/.
The hope is to use these resources to start WikiApiary crawling and discovering wikis again.
Of course, none of this is to diminish Bawolff's much needed work last month to get the site actually serving pages again.
Thanks,
Mark.
TripleCamera via Wikiapiary wikiapiary@lists.wikimedia.org writes:
By the way, while the old site will be open to humans and bots soon, what would the destiny of the new site be? Will it receive continuous development to become fully functional?
This is something I honestly have not thought about.
But, since you asked, I did look at Phabricator[1] and GitHub[2] (something I haven't done in a long time) and did see many outstanding issues and tasks.
Some of them are even recent!
I’ve realized that the "destiny" of the new site depends entirely on the hands currently on deck. While the MediaWiki Stakeholders' Group and Dobriy.ai (CC'd) provides a home for the project, the actual momentum comes from volunteers.
Right now, the gap between the old site’s revival and the new site’s completion is simply a matter of manpower. If we want to see the new site become fully functional rather than remaining stalled, we need to move from interest to active contribution.
Since you are eager to see movement, I’d love to have you join us in tackling some of these tasks. Whether it’s triaging those Phabricator tasks, submitting PRs on GitHub, or even just helping to document the current state of the new site, any bit helps.
If you (or anyone else on this list) have the time to pick up one of the outstanding issues, let’s coordinate. The "destiny" of the site isn't set in stone--it's waiting for more people to help write it.
[1] https://phabricator.wikimedia.org/project/board/2341/
[2] https://github.com/WikiApiary/WikiApiary/issues
Mark,
Thanks for your reply! I have scanned though the emails from the past three years, as well as some Phabricator tasks and GitHub issues. Here are the three most important tasks I can think of:
What is the plan? =================
The new site ------------
Before taking actions, I'm really interested in the developers' envision about the new WikiApiary. What functions will the new site have? What would its infrastructure be? Which features are already realized, and which are still under development? It would be great if there were a detailed roadmap about the development of the new site, and it would be better if there were tasks that volunteers could participate.
The old site ------------
Besides, it is surprising that Dobriy AI will be the host for the old site. Can its staff make a short introduction for the company?
According to previous mails, the biggest issue blocking the revitalization of the old site is a fatal Semantic MediaWiki issue: Semantic MediaWiki can't handle the volume of the old site[1]. Is it solved now?
Continue developing the WikiApiary extension ============================================
Arount the end of 2023 and the beginning of 2024, when the new WikiApiary was being set up[2][3], Cindy developed the WikiApiary extension[4] to replace the old infrastructure. Unfortunately, development had quickly stalled, probably due to Cindy's lack of time. The repository resides at gerrit.wikimedia.org now, only receiving commits occasionally from bots. Does Cindy have time to continue developing this extension? If not, I believe there are extension developers who are willing to help, but they don't know what to do with the codebase. Could Cindy provide some guidance?
Providing server-side dump for the old site ===========================================
A server-side dump for the old WikiApiary would have significant historical value. There was a Phabricator task created in 2022 requesting it[5]. Besides, I can recall that Winston also requested it on the 2024-02 MWstake meeting.
Although we do have a client-side dump[6], it only contains the history and images dump. This is because client-side dumpers have limited functionalities. A server-side dump would be more complete.
What's more, the server log would be useful when troubleshooting the SMW issue mentioned above. It would be more than helpful if you could provide the server log as well.
That's all I can think of. I'm also curious about the opinions of other volunteers, so feel free to share here.
I'm looking forward to your early reply.
TripleCamera.
[1] https://lists.wikimedia.org/hyperkitty/list/wikiapiary@lists.wikimedia.org/t... (Status of WikiApiary) [2] https://lists.wikimedia.org/hyperkitty/list/wikiapiary@lists.wikimedia.org/t... (Wikiapiary maintenance) [3] https://lists.wikimedia.org/hyperkitty/list/wikiapiary@lists.wikimedia.org/t... (WikiApiary status update) [4] https://www.mediawiki.org/wiki/Extension:WikiApiary [5] https://phabricator.wikimedia.org/T315935 [6] https://archive.org/details/wiki-wikiapiary.com_w-20231222
在 2026/5/3 15:58, Mark A. Hershberger 写道:
TripleCamera via Wikiapiary wikiapiary@lists.wikimedia.org writes:
By the way, while the old site will be open to humans and bots soon, what would the destiny of the new site be? Will it receive continuous development to become fully functional?
This is something I honestly have not thought about.
But, since you asked, I did look at Phabricator[1] and GitHub[2] (something I haven't done in a long time) and did see many outstanding issues and tasks.
Some of them are even recent!
I’ve realized that the "destiny" of the new site depends entirely on the hands currently on deck. While the MediaWiki Stakeholders' Group and Dobriy.ai (CC'd) provides a home for the project, the actual momentum comes from volunteers.
Right now, the gap between the old site’s revival and the new site’s completion is simply a matter of manpower. If we want to see the new site become fully functional rather than remaining stalled, we need to move from interest to active contribution.
Since you are eager to see movement, I’d love to have you join us in tackling some of these tasks. Whether it’s triaging those Phabricator tasks, submitting PRs on GitHub, or even just helping to document the current state of the new site, any bit helps.
If you (or anyone else on this list) have the time to pick up one of the outstanding issues, let’s coordinate. The "destiny" of the site isn't set in stone--it's waiting for more people to help write it.
I still have access to the wmcloud version of the site, i guess i could run the dump script if people want that. Probably would take a few days.
I do not think the SMW log would be useful in debugging any issues mentioned here. One of the problems with wikiapiary at one point was mysql not being assigned enough ram (both innodb page cache size but also max temp table size before spilling to disk, were too small giving suboptimal performance). Of course that isn't the only issue - smw scales rather poorly. I wrote some potential ideas for smw performance improvements at https://github.com/SemanticMediaWiki/SemanticMediaWiki/issues/6559 - but they are just ideas that need to be validated . One thing to maybe look into is using fulltext indexes in mysql for doing ask queries, but that would be a major rewrite of the smw query backend.
-- Brian
On Saturday, 9 May 2026, TripleCamera via Wikiapiary < wikiapiary@lists.wikimedia.org> wrote:
Mark,
Thanks for your reply! I have scanned though the emails from the past three years, as well as some Phabricator tasks and GitHub issues. Here are the three most important tasks I can think of:
What is the plan?
The new site
Before taking actions, I'm really interested in the developers' envision about the new WikiApiary. What functions will the new site have? What would its infrastructure be? Which features are already realized, and which are still under development? It would be great if there were a detailed roadmap about the development of the new site, and it would be better if there were tasks that volunteers could participate.
The old site
Besides, it is surprising that Dobriy AI will be the host for the old site. Can its staff make a short introduction for the company?
According to previous mails, the biggest issue blocking the revitalization of the old site is a fatal Semantic MediaWiki issue: Semantic MediaWiki can't handle the volume of the old site[1]. Is it solved now?
Continue developing the WikiApiary extension
Arount the end of 2023 and the beginning of 2024, when the new WikiApiary was being set up[2][3], Cindy developed the WikiApiary extension[4] to replace the old infrastructure. Unfortunately, development had quickly stalled, probably due to Cindy's lack of time. The repository resides at gerrit.wikimedia.org now, only receiving commits occasionally from bots. Does Cindy have time to continue developing this extension? If not, I believe there are extension developers who are willing to help, but they don't know what to do with the codebase. Could Cindy provide some guidance?
Providing server-side dump for the old site
A server-side dump for the old WikiApiary would have significant historical value. There was a Phabricator task created in 2022 requesting it[5]. Besides, I can recall that Winston also requested it on the 2024-02 MWstake meeting.
Although we do have a client-side dump[6], it only contains the history and images dump. This is because client-side dumpers have limited functionalities. A server-side dump would be more complete.
What's more, the server log would be useful when troubleshooting the SMW issue mentioned above. It would be more than helpful if you could provide the server log as well.
That's all I can think of. I'm also curious about the opinions of other volunteers, so feel free to share here.
I'm looking forward to your early reply.
TripleCamera.
[1] https://lists.wikimedia.org/hyperkitty/list/wikiapiary@lists .wikimedia.org/thread/AMVVOP72ONA2YI7XJ7KV6O3K42BBBES4/ (Status of WikiApiary) [2] https://lists.wikimedia.org/hyperkitty/list/wikiapiary@lists .wikimedia.org/thread/I7VI2JXZ4Q3FJGFDCX4YES4JOWY2Q4DW/ (Wikiapiary maintenance) [3] https://lists.wikimedia.org/hyperkitty/list/wikiapiary@lists .wikimedia.org/thread/SSWGW7Y3OYDOWYAFBRFR4WMFS3BEKV7L/ (WikiApiary status update) [4] https://www.mediawiki.org/wiki/Extension:WikiApiary [5] https://phabricator.wikimedia.org/T315935 [6] https://archive.org/details/wiki-wikiapiary.com_w-20231222
在 2026/5/3 15:58, Mark A. Hershberger 写道:
TripleCamera via Wikiapiary wikiapiary@lists.wikimedia.org writes:
By the way, while the old site will be open to humans and bots soon,
what would the destiny of the new site be? Will it receive continuous development to become fully functional?
This is something I honestly have not thought about.
But, since you asked, I did look at Phabricator[1] and GitHub[2] (something I haven't done in a long time) and did see many outstanding issues and tasks.
Some of them are even recent!
I’ve realized that the "destiny" of the new site depends entirely on the hands currently on deck. While the MediaWiki Stakeholders' Group and Dobriy.ai (CC'd) provides a home for the project, the actual momentum comes from volunteers.
Right now, the gap between the old site’s revival and the new site’s completion is simply a matter of manpower. If we want to see the new site become fully functional rather than remaining stalled, we need to move from interest to active contribution.
Since you are eager to see movement, I’d love to have you join us in tackling some of these tasks. Whether it’s triaging those Phabricator tasks, submitting PRs on GitHub, or even just helping to document the current state of the new site, any bit helps.
If you (or anyone else on this list) have the time to pick up one of the outstanding issues, let’s coordinate. The "destiny" of the site isn't set in stone--it's waiting for more people to help write it.
Wikiapiary mailing list -- wikiapiary@lists.wikimedia.org To unsubscribe send an email to wikiapiary-leave@lists.wikimedia.org
dumpBackup.php seems to be broken on this version of MW and is filesorting things (And i don't really want to spend the time to figure out why). So i guess i won't be making a dump :(
On Sat, May 9, 2026 at 10:04 PM Brian Wolff bawolff@gmail.com wrote:
I still have access to the wmcloud version of the site, i guess i could run the dump script if people want that. Probably would take a few days.
I do not think the SMW log would be useful in debugging any issues mentioned here. One of the problems with wikiapiary at one point was mysql not being assigned enough ram (both innodb page cache size but also max temp table size before spilling to disk, were too small giving suboptimal performance). Of course that isn't the only issue - smw scales rather poorly. I wrote some potential ideas for smw performance improvements at https://github.com/SemanticMediaWiki/SemanticMediaWiki/issues/6559 - but they are just ideas that need to be validated . One thing to maybe look into is using fulltext indexes in mysql for doing ask queries, but that would be a major rewrite of the smw query backend.
-- Brian
On Saturday, 9 May 2026, TripleCamera via Wikiapiary < wikiapiary@lists.wikimedia.org> wrote:
Mark,
Thanks for your reply! I have scanned though the emails from the past three years, as well as some Phabricator tasks and GitHub issues. Here are the three most important tasks I can think of:
What is the plan?
The new site
Before taking actions, I'm really interested in the developers' envision about the new WikiApiary. What functions will the new site have? What would its infrastructure be? Which features are already realized, and which are still under development? It would be great if there were a detailed roadmap about the development of the new site, and it would be better if there were tasks that volunteers could participate.
The old site
Besides, it is surprising that Dobriy AI will be the host for the old site. Can its staff make a short introduction for the company?
According to previous mails, the biggest issue blocking the revitalization of the old site is a fatal Semantic MediaWiki issue: Semantic MediaWiki can't handle the volume of the old site[1]. Is it solved now?
Continue developing the WikiApiary extension
Arount the end of 2023 and the beginning of 2024, when the new WikiApiary was being set up[2][3], Cindy developed the WikiApiary extension[4] to replace the old infrastructure. Unfortunately, development had quickly stalled, probably due to Cindy's lack of time. The repository resides at gerrit.wikimedia.org now, only receiving commits occasionally from bots. Does Cindy have time to continue developing this extension? If not, I believe there are extension developers who are willing to help, but they don't know what to do with the codebase. Could Cindy provide some guidance?
Providing server-side dump for the old site
A server-side dump for the old WikiApiary would have significant historical value. There was a Phabricator task created in 2022 requesting it[5]. Besides, I can recall that Winston also requested it on the 2024-02 MWstake meeting.
Although we do have a client-side dump[6], it only contains the history and images dump. This is because client-side dumpers have limited functionalities. A server-side dump would be more complete.
What's more, the server log would be useful when troubleshooting the SMW issue mentioned above. It would be more than helpful if you could provide the server log as well.
That's all I can think of. I'm also curious about the opinions of other volunteers, so feel free to share here.
I'm looking forward to your early reply.
TripleCamera.
[1] https://lists.wikimedia.org/hyperkitty/list/wikiapiary@lists.wikimedia.org/t... (Status of WikiApiary) [2] https://lists.wikimedia.org/hyperkitty/list/wikiapiary@lists.wikimedia.org/t... (Wikiapiary maintenance) [3] https://lists.wikimedia.org/hyperkitty/list/wikiapiary@lists.wikimedia.org/t... (WikiApiary status update) [4] https://www.mediawiki.org/wiki/Extension:WikiApiary [5] https://phabricator.wikimedia.org/T315935 [6] https://archive.org/details/wiki-wikiapiary.com_w-20231222
在 2026/5/3 15:58, Mark A. Hershberger 写道:
TripleCamera via Wikiapiary wikiapiary@lists.wikimedia.org writes:
By the way, while the old site will be open to humans and bots soon,
what would the destiny of the new site be? Will it receive continuous development to become fully functional?
This is something I honestly have not thought about.
But, since you asked, I did look at Phabricator[1] and GitHub[2] (something I haven't done in a long time) and did see many outstanding issues and tasks.
Some of them are even recent!
I’ve realized that the "destiny" of the new site depends entirely on the hands currently on deck. While the MediaWiki Stakeholders' Group and Dobriy.ai (CC'd) provides a home for the project, the actual momentum comes from volunteers.
Right now, the gap between the old site’s revival and the new site’s completion is simply a matter of manpower. If we want to see the new site become fully functional rather than remaining stalled, we need to move from interest to active contribution.
Since you are eager to see movement, I’d love to have you join us in tackling some of these tasks. Whether it’s triaging those Phabricator tasks, submitting PRs on GitHub, or even just helping to document the current state of the new site, any bit helps.
If you (or anyone else on this list) have the time to pick up one of the outstanding issues, let’s coordinate. The "destiny" of the site isn't set in stone--it's waiting for more people to help write it.
Wikiapiary mailing list -- wikiapiary@lists.wikimedia.org To unsubscribe send an email to wikiapiary-leave@lists.wikimedia.org
wikiapiary@lists.wikimedia.org