Show HN: Gis.chat – a Geospatial Community
gis.chatHi folks! I'm excited to show you gis.chat, a geospatial chat platform in both senses: a platform about geospatial topics and a geospatial platform itself, referencing the location of our communities.
The setup is fairly simple and reproducible: a plain Zulip instance and a homepage with geospatial search capabilities.
It seems almost trivial but it has some very nice features. I guess you should be familiar with Zulips stream/topic model to follow along (https://zulip.com/help/streams-and-topics).
The core idea is that there are city-specific streams (currently represented by a pin), but there could just as well be streams about points of interest, line geometries (e.g. a river) or polygons (e.g. national park).
- Every local stream can have the same topics, e.g. "general", "news", "meetups", "jobs" etc. - With Zulip's search you can either search for a particular topic, e.g. "news" in a local stream or instead in all streams and have some kind of news feed of the community with "topic:news" - Once more communities are added, specific filters could be added, e.g. country-wise or by drawing your own area of interest - Eventually, for the ones who like, users could associate themselves with a local community in their profile or add there main location so one could not only search for the local communities but instead also for individuals
There are many nice features in Zulip's pipeline that would foster gis.chat:
- Further nesting of streams/topics - Semantic search
If for example Zulip would allow for saving coordinates (or better an entire geometry) in the Postgres DB, with the help of PostGIS, Zulip's search could allow for bounding boxes (or custom geometries).
Let me know if you have any kind of other ideas or feedback!
Nice! I think it's a good initiative to organize more local types of meetups in the geospatial community. Geospatial experts are a bit too few and inbetween to bump into many of them on a regular day, so anything that gets these people together is a big win.
As a side note to HN readers, geospatial is a super interesting application/ specialization that you might consider getting into if you're looking to add more meaning and purpose to your programming or tech career. You're often working to improve quality of life for someone, there are lots of interesting companies, people are generally super friendly and accessible, and there's a wealth of interesting problems and challenges to work on. My 2 cents!
Due to the nature of my job, I interact with some of the more noteworthy geospatial experts on a weekly basis. These people will take any time of the day to explain to you all of the microscopic nuances about how any GNSS works.
I would listen to this podcast.
Not a podcast, but a Stanford MOOC (massive open online course): https://www.youtube.com/watch?v=o1Fyn_h6LKU&list=PLGvhNIiu1u...
The basic operation of a GNSS with ~30m accuracy is literally explained on one slide in lecture 1.2, and from then on the rest of the course is about implementation specifics, and enhancing performance.
The basic concept at its core is so elegant and beautiful, but no less interesting are all the ways this system is augmented with different techniques to improve the accuracy down to ~1cm.
This is one of those resources I've downloaded and made sure to have backed up, because it's so invaluable to me.
This makes my weekend, going to binge this. Email in profile if I can buy you a coffee or beer for sharing. Thanks for taking the time!
GIS seems interesting but it also doesnt seem like it pays well. Being niche isn't bad per-se but I would expect it to be more lucrative.
In some ways it seems like GIS is to software engineering as business analyst is to data science. Not in nicheness, because business analyst isnt so niche. But doesn't pay as well and less programming, more GUI/business tools.
Are these impressions wrong?
Not wrong necessarily, if "traditional GIS" is your chosen application of geospatial skills. (By "traditional GIS" I mean creating one-off analyses by interacting primarily with enterprise desktop/GUI tools like ArcGIS).
But when combined with software engineering expertise, geospatial knowledge can be as lucrative as any other software job. (I am a geospatial software engineer.)
Thanks for the insight. I gained a sliver of insight into GIS when trying to do some data analysis on lakes. I discovered that there is a lot to it. Its an interesting space, especially for those interested in geography or the outdoors.
It is niche, but it's almost always a cost center role and heaviest users of GIS are govt and utilities, which have constraints on remuneration.
It can be lucrative in the right industry with domain knowledge (eg mining or oil&gas).
I totally agree with your latter assessment. I wrote off software engineering as a career that I would deeply enjoy for a few years, went to become a scientist. I came back to software when I discovered geospatial; it's way fun.
Oh, finally. There's plenty of GNSS/GIS folks lurking around the OpenStreetMap wiki and I've heard their telegram too but never explored it. There's some GNSS-but-not-GIS folks on the Time-Nuts mailing list. Galmon has its own little handful. There doesn't seem to be a good RTKLIB forum, which is weird.
I hope this takes off and unites a whole bunch of little communities.
You may be interested in joining the Spatial Community slack if you haven't already. There's a self-invite link at https://thespatialcommunity.org. There's a ton of channels there spanning a range of GIS related interests.
https://mapstodon.space/ is the niche Mapstodon instance for GIS
This seems neat! Your tagline encourages that you're open to everyone, not just engineers and data scientists. I think that's great, you'll find a much better engagement if you keep your funnel wide.
My question though is how could I send this to any of my friends in GIS who are only semi-technical? You mention topics like "social sciences, climate change, urban planning" - the people in those areas have wildly different product expectations than you and I. I'm a FAANG engineer and have never heard of Zulip, and if I'm being honest, clicking around the interface was not especially inviting. It was more reminiscent of a bug tracker or internal docs tool than a social network :)
I love the idea. I think that if you're serious about creating a community for GIS folks though, you may want to start talking to non-technical users and get their thoughts on the user experience and discoverability of the product. My guess is they'll have some suggestions which could really help your early growth.
I wanted to set up a similar chat for an online community.
Do you host yourself or do you use the free version? What scares me is often the cost, especially since there is no way my users will pay the $6 a month and we are not making a profit on this project, so we can't pay for something too expensive.
The idea is often to reuse our existing infrastructure and install a self-hosted chat.
What discourages me every time I want to move forward with this project is that there is never a rough estimate of operating costs when it comes to self-hosted open source chat platforms.
If you could share some pointers, I couldn't thank you enough!
I didn't know about Zulip stream/topics but heres my naive tl;dr after reading about it: Streams are like slack channels, topics belong to streams and have no analog in slack. They are essentially sub channels. Or tags/categories within the stream scope for your messages. Although threads (if people use them) arguably serve a similar purpose on slack.
I wonder, is it possible to message on the base stream, or only topics? I feel like if you cant comment on the base stream you always need a general topic. Otherwise you're going to keep making new super-niche topics or make topics "off-topic." And what happens when discussion on one topic ventures towards some other topic? To bridge the gap it seems like I need to make a new message in the other topic and link in the original one.
Generally speaking, I do not like forcing these structures on the "write" side. I had the same problem with Arc browser (with respect to types of tabs, workspaces). I don't want to have to essentially define some metadata about the action I want to perform. I have the same attitude towards taking notes. Generally speaking I'd rather have powerful tools to discover, re-organize as I see fit. I guess maybe this is less true in a messaging app though because you already have to choose a channel either way.
If you write message without defining a topic it will be in (no topic) topic. You can move messages from topic to topic https://chat.zulip.org/help/move-content-to-another-topic