I don't see why that would be "abuse", it is perfectly capable of storing flexible documents

there are CMS stuff made with XTDB, I've worked on one as well https://github.com/solita/lipo

Do you need bitemporality, though? That sounds like logseq, athens and roam do, which are also made with clojure and datascript

Common term for those are PKMS, personal knowledge management system. With that short description is a bit difficult to see how a personal CMS would differ

there are ready made mind map and zettelkasten software, so there's no need to make your own... unless you want to

@U8ZQ1J1RR fwiw, I've been using xtdb for nearly 2 years and I've never touched it's bitemporal features other than to try them out for fun. They're totally optional and (imo) not that important. If you need them, they exist. If you don't, you're not exposed to them.

@U06FM9QN9 I've toyed with this a bit, using xtdb as a post-anki language learning tool and PKM/note-taking doodad. In general, I've come back to other zettelkasten with a proper UI. At the moment, I'm using GitJournal and Zettlr over markdown because my data just isn't that graph-shaped, but the other Roam clones work just as well. I think the only reason to jump straight to xtdb would be if you need (or anticipate) custom functionality, beyond what the off-the-shelf tools give you. Or if you prefer a REPL to a GUI, I suppose? But that might get tiresome after a while. 😉

@U06FM9QN9 A sensible middle-ground between hosted services you don't control and using a raw database as a CMS might be to start with Athens (https://github.com/athensresearch/athens) and customize it if/when you need to. As @U8ZQ1J1RR mentioned, Athens is made with DataScript, so it should be pretty familiar.

Not immediately but possibly in the future. I'd like to have the flexibility of having it.

I actually cry because I had ideas for an ‘intelligent notebook’ as far back as 2008 but ideas are useless until put into motion. But to get back on topic I'm just looking to build my own thing that will function as a personal system and parts of it will be public facing. Toy project for learning and entertaining myself and a few friends. So if I do give xtdb a chance am I fine just using the rocksdb backing? At the moment I'm using postgresql but if I can get away from it and be safe with just the xtdb/datalog defaults that would be fine.

@U11SJ6Q0K definitely a personal project and it's not really a mind map. Not in the way you're thinking…

One more question…would I be ok storing basic images? Literally book covers and more than likely no more than a few hundred. Or should I just continue to put them on the file system and store the string/path?

for a small scale project I guess storing images would be ok in the db

but for a "real" site, I would say it is better to put images in S3 or similar for hosting

as datalog doesn't really give you anything for binary blobs, you are just bloating the indexes

Yeah I originally was going with an s3 bucket then I thought of the amount of images I'd have after a years time and I figure I may as well use some of the space my server will allot. Plus I hear a lot of bad experiences with aws/billing if an account is compromised etc and it's just not worth the risk to me. I'm a hobbyist programmer so getting hit with a significant bill would potentially really hurt me

Though if traffic scales then of course I'd adjust. I'm keeping that in mind as I build…

if you don't have clustering needs (only 1 server), you could put them in the filesystem as well

Yeah. Im thinking that's best…the file system

This lipo project is very clean by the way. Thank you.

Hey again, we haven't considered backporting that particular change so far. Is there a concern about consuming the beta? The proper release should happen in the next week or two

Hey @U028ART884X thanks for mentioning this - it looks like it could be a regression specific to beta2 due to the recent tx-time import changes. Could you share your node config and a small snippet with the submit-tx call you are using please? Feel free to DM me if you're more comfortable with that. I'm surprised if there was a regression here for beta1 though, as I can't see/think what might have changed :thinking_face: Presumably your same code worked fine with 1.20.0?

What's funny is that it doesn't happen anymore. I was using an empty config and it was crashing every time. And then I couldn't import a namespace after that (or something like that). I hope that you redeployed because if you didn't and it fixed itself then it may be a problem 😄 But I think it should crash every time because it's a problem with how things are imported and executed (the method declaration, if I'm not wrong). If it'll come back I'll paste it into this thread.

Huh, weird. I know protocols and REPL reloading get's confusing & broken pretty quickly. I'm guessing you weren't attempting any strange monkey patching or whatever? Or using some hot-swapping dependency magic?

I have this in my user profile: [com.clojure-goes-fast/clj-memory-meter "0.1.3"] and this one does all sorts of interesting things. But I didn't import it into my current project. Also I restarted the REPL and then tried to start only the node. So it couldn't be that dependency. Also it crashes in various ways when I try to import it or measure XTDB on other JVMs than java 8. There are many issues with it.

Aha, well, please do shout if it crops up again(!)

Hm. The exception came back when I started the REPL today. I didn't change anything in particular. And it persisted when I restarted the REPL. But then I stopped my REPL and ran lein clean on the project. And then the next time I started the REPL it was gone.

Hmm, is XT definitely only appearing once in the classpath?

feel free to dump the output of lein classpath and lein deps :tree

I think I fixed it by running lein clean. I'll try it next time it appears. Currently classpath has only one of each: :/home/_/.m2/repository/com/xtdb/xtdb-core/1.21.0-beta2/xtdb-core-1.21.0-beta2.jar: :/home/_/.m2/repository/com/xtdb/xtdb-kafka/1.21.0-beta2/xtdb-kafka-1.21.0-beta2.jar:

[com.xtdb/xtdb-core "1.21.0-beta2"]
   [org.agrona/agrona "1.11.0"]
   [org.clojure/data.json "2.3.1"]
   [org.clojure/tools.cli "1.0.206"]
   [org.clojure/tools.logging "1.1.0"]
   [org.slf4j/slf4j-api "1.7.30"]
   [pro.juxt.clojars-mirrors.com.stuartsierra/dependency "1.0.0"]
   [pro.juxt.clojars-mirrors.com.taoensso/nippy "3.1.1"]
     [org.iq80.snappy/snappy "0.4"]
     [org.lz4/lz4-java "1.7.1"]
     [org.tukaani/xz "1.8"]
   [pro.juxt.clojars-mirrors.edn-query-language/eql "2021.02.28"]
 [com.xtdb/xtdb-kafka "1.21.0-beta2"]
   [com.cognitect/transit-clj "1.0.324" :exclusions [[org.msgpack/msgpack]]]
     [com.cognitect/transit-java "1.0.343"]
       [javax.xml.bind/jaxb-api "2.3.0"]
   [org.apache.kafka/kafka-clients "2.6.1" :exclusions [[org.lz4/lz4-java]]]
     [com.github.luben/zstd-jni "1.4.4-7"]
     [org.xerial.snappy/snappy-java "1.1.7.3"]
   [pro.juxt.clojars-mirrors.cheshire/cheshire "5.10.0"]
     [com.fasterxml.jackson.dataformat/jackson-dataformat-cbor "2.10.2" :exclusions [[org.clojure/clojure]]]
     [com.fasterxml.jackson.dataformat/jackson-dataformat-smile "2.10.2" :exclusions [[org.clojure/clojure]]] 

Cool, thanks, and I don't fully understand how classes get compiled and cached across REPL sessions, but running lein clean definitely is a good idea when trying deps upgrades

I don't know either. But the thing is that it worked on Friday when I was using the same configuration. And today I started everything again and got the exception. That's what was strange.

Edit: I didn't try to run this via leiningen. It's only the REPL that was giving me this error. I think I didn't get it via CLI. So probably it's more of a REPL's problem (I use CIDER in spacemacs).