GW Guru is back !

Quote from Ecto Sacrifice

gw guru archive at least had pictures before and was searchable via ctrl +f.. now it's even more unusable and no pictures are left.

i don't understand the thought process behind "remastering" it when it's objectively worse than before. hopefully pictures and search is added otherwise it was all for nothing, leaving us with even less of a historical archive than before

I stand corrected, take a look here. I'm glad to see that there is ongoing updates to the archive to make it more usable and fixing images.

Article

Major update for Guru Archives coming soon (spoiler: search)!

Guild Wars Reforged is coming up soon, so it only made sense to brush up on our archives, too.

Iaerah

November 27, 2025 at 11:26 PM

The current version was the first step forwards to improving the archive. It's been quite a lot of work to get it to that state and it wasn't perfect and I was aware of that. But 2024 was generally a bad year for me, especially the last half of it and 2025 has been the worst year of my life, without a doubt. I lost my job in the beginning of the year, my mother got diagnosed with a serious disease, my relationship imploded, my nephew died, financial issues due to the loss of my job and so many bureaucratic challenges... my life as it was changed completely this year.

So, my focus this year has been to try and get through it all, trying to get my life back on the rails. While I spent several months at home, working on things was the last thing on my mind. It has only been recently that my desire to work on things again, returned.
I'm glad to have the beginnings of stability back with a new (challenging) job, a better understanding of myself and working on my self-confidence, trying to find my way in the new IT-world (which has drastically changed in the last year) and a new relationship as well. It's been a though year, but in the end, years like this are extremely unpleasant but they make you stronger and better as a person. At least, I hope so

But, back to on-topic, I wanted to introduce search into the archives as well, but I ran into a hard deadline when I built the previous version of the archive. The costs for the server were going to increase significantly so I wanted to migrate off it as soon as possible. But once the new archive was set up, I didn't touch the new server at all. In fact, I forgot a lot about how I set it up and after a while I didn't want to touch it anymore to avoid breaking something, which is never a great situation to be in.

The GuruArchiver was one of my first projects in Golang and I was never really a programmer, so it was quite an undertaking for me to build it. Over time, I started doing a lot more programming and I became a lot better in Golang, so each time when I touched the code, it improved a lot. The new version of the archive that will be going live soon, is a testament to how much I have grown. I have stripped out MariaDB for SQLite, but almost no database code itself needed to change, but other parts of the application required a lot more attention and time. Building search isn't rocket science, but with the size of the archive, you quickly run into practical limits: the database itself is 2.7 GB big. Searching through that without optimizations requires a lot of resources and a single search took about 15 seconds to complete, which isn't really feasible (especially since going to the next search result page is another search which took another 15 seconds). I have optimized this and it has made a huge difference, with searches now taking about 200ms, but it made the database nearly 7 GB in total. I've seen companies switching to specialized solutions like ElasticSearch for less, which is a whole other service to update/maintain and another tool that I would have had to learn. So that's why I didn't implement search before, because I wanted it to be a good implementation and not a half baked one.
Another reason why I didn't implement search back then is because of the impact it could have on the server. Imagine having to run several queries that take 15 seconds to complete on a server, it would quickly lead to issues with the database and impact every other website on the server that also used that database. It would very likely have impacted Guild Wars Legacy and I didn't want that.

Fixing the images was a hard thing to do as well, because the archive hasn't got a unified way to deal with images. Some are expected to be in specific folders, some in others. That's due to how it was archived, it was a static HTML dump. There were so many issues with the scraped data as well (broken HTML), which is something that needed to be fixed as well. Every little thing you touch required further changes elsewhere and it quickly became a spiral out of control.

So... it took a while before I felt mentally fit enough again to tackle a project of this size. It was a lot of work, I added a lot of things that I originally wanted to have in the archive. I wanted to build a definitive version of the archive that would include everything that you all would want. And that just took a while and it meant that I had to learn new technologies as well. It's my second project using SQLite, my first project using Templ, first timing building search from scratch. It's the first time that I got help from an AI agent to help me build the new lay-out files. I'm really glad of the result and I'm really proud of it, but it didn't come from nowhere. This has been days of late-night coding and something that I do for my love for Guild Wars and this community, and because I have the skills to do so.

As long as Guild Wars is around, and as long as I am around, I'll always make sure to do my best for this community. And yes, further improvements will be made to the archive. Now, the next steps I need to do for the archive is to set-up the new server, set-up and test the new deployment system for the archives, write the Ansible code to manage it all and finally deploy it to the new server. I hope to find the time somewhere this week

Kev