💾 Archived View for gemi.dev › gemini-mailing-list › 001073.gmi captured on 2024-03-21 at 18:17:01. Gemini links have been rewritten to link to archived content
⬅️ Previous capture (2023-12-28)
-=-=-=-=-=-=-
Hi all, I'm sure most of you know Stéphane Bortzmeyer's crawler and his geminispace statistics page: gemini://gemini.bortzmeyer.org/software/lupa/stats.gmi With his kind permission, I wrote a script to download and parse this page daily, store the data in a SQLite database and generate 38 different charts to visualize the data. The charts are regenerated every night by a script to show the result of the latest crawl. Historic data has been included from Stéphane's archive (monthly resolution). You can see the charts here: gemini://botond.online/en/stats/ You can also download the latest SQLite DB, updated daily, from here if you want to play with it / use it in your own projects: gemini://botond.online/en/stats/stats.db I hope you'll find the charts useful. Please report any bugs you encounter. All ideas to improve the page are welcome. I'm planning to open source the script but I have to do some cleanup first. Botond
Balázs Botond [2021-11-10 Wed 21:40] wrote: > I hope you'll find the charts useful. Please report any bugs you > encounter. All ideas to improve the page are welcome. I'm planning to > open source the script but I have to do some cleanup first. Very nice visualizations, great work! -- Daniel Fleischer
On Wed, Nov 10, 2021 at 09:40:28PM +0100, Balázs Botond <balazsbotond@gmail.com> wrote a message of 26 lines which said: > You can see the charts here: > gemini://botond.online/en/stats/ Excellent and very useful. At the present time, it seems things are improving, both in the number of capsules and URL but also for some technical things such as the TLS versions. I suggest that graphs which do *not* start from zero (such as the history of TLD) are a bad idea, they create a false image of growth.
On Thu, Nov 11, 2021 at 5:08 PM Stephane Bortzmeyer <stephane@sources.org> wrote: > > I suggest that graphs which do *not* start from zero (such as the > history of TLD) are a bad idea, they create a false image of growth. Good point, thank you. I've updated the code, now all y axes start at 0.
On Thu, Nov 11, 2021 at 05:03:34PM +0100, Stephane Bortzmeyer wrote: > > Excellent and very useful. I Agree. Nice and useful, Thanks! Bye! C.
---