[GH-ISSUE #54] Onboarding docs mention local Ollama but doesn't hint at how to handle or detect snap-installed Ollama models #26

New Issue

2026-02-15T17:04:19-05:00

yindo commented

2026-02-15 17:04:19 -05:00

Originally created by @arcaven on GitHub (Aug 3, 2025).
Original GitHub issue: https://github.com/cloudstack-llc/msty-docs/issues/54

For a typical Ubuntu user a local install of Ollama is as likely to be snap rather than apt or "curl url | bash" remote shell injection install method.

The docs at https://docs.msty.app/getting-started/onboarding provide a tab-based view for those new Msty users who likely already have Ollama installed. The options are: get started using Ollama (presumably set it up again) or use Ollama remotely or ... if Ollama was installed in user space, then they'll have and msty will detect a ~/.ollama/models directory. However, "ollama serve" served models will be in /var/snap/ollama/common/models or something like that.

Recommend updating the docs for the Ubuntu-snap-ollama crowd, as it's a major part of your demographic. I haven't even used your software yet, so I'm not sure I know what is best here, tell them to use snap installed Ollama as remote via http://127.0.0.1:11434, create a ~/.ollama/models and copy or ln (and fight with permissions, snap-installed Ollama is root:root) or copy the models aside from snap, uninstall snap Ollama and install it a different way.

If I had an opinion, I'd write it up, but let me get some experience with Msty before I start having opinions.

Originally created by @arcaven on GitHub (Aug 3, 2025). Original GitHub issue: https://github.com/cloudstack-llc/msty-docs/issues/54 For a typical Ubuntu user a local install of Ollama is as likely to be snap rather than apt or "curl url | bash" remote shell injection install method. The docs at https://docs.msty.app/getting-started/onboarding provide a tab-based view for those new Msty users who likely already have Ollama installed. The options are: get started using Ollama (presumably set it up again) or use Ollama remotely or ... if Ollama was installed in user space, then they'll have and msty will detect a ~/.ollama/models directory. However, "ollama serve" served models will be in /var/snap/ollama/common/models or something like that. Recommend updating the docs for the Ubuntu-snap-ollama crowd, as it's a major part of your demographic. I haven't even used your software yet, so I'm not sure I know what is best here, tell them to use snap installed Ollama as remote via http://127.0.0.1:11434, create a ~/.ollama/models and copy or ln (and fight with permissions, snap-installed Ollama is root:root) or copy the models aside from snap, uninstall snap Ollama and install it a different way. If I had an opinion, I'd write it up, but let me get some experience with Msty before I start having opinions.

yindo referenced this issue

2026-06-05 16:27:00 -04:00

[GH-ISSUE #26] illegal hardware instruction (core dumped) #16

yindo changed title from ~~Onboarding docs mention local Ollama but doesn't hint at how to handle or detect snap-installed Ollama models~~ to [GH-ISSUE #54] Onboarding docs mention local Ollama but doesn't hint at how to handle or detect snap-installed Ollama models

2026-06-05 16:27:03 -04:00

Sign in to join this conversation.

1 Participants

Notifications

Due Date

No due date set.

Dependencies

No dependencies set.

Reference: cloudstack-llc/msty-docs#26