Hey there, let's work through this problem step by step—this kind of dependency conflict and missing metadata is super common with third-party conda channels, so you're not alone!

Why isn't conda info showing dependency details? #

First off, conda info <package> isn't the right tool for checking dependencies—it only shows basic package metadata like version, channel, and size. To see actual dependency trees, you need to use:

• conda search --info <package>: This pulls full metadata including dependencies from the channel.
• conda repoquery depends <package>: A more powerful tool (requires conda-libmamba-solver or conda-repoquery installed) that visualizes dependency chains.

Another likely issue: The dsm channel's blpapi package probably has incomplete metadata. A lot of third-party channels don't properly package dependency info when uploading, which is why conda info comes up empty.

Why are blpapi and pypyodbc conflicting? #

The UnsatisfiableError usually boils down to one of these scenarios:

• Python version mismatch: Maybe the blpapi from dsm only supports older Python versions (like 3.8-3.10), while your installed pypyodbc is built for a newer version (3.11+). Conda can't find a Python version that works for both.
• System library conflicts: Both packages rely on ODBC-related system libraries, but they require different versions that can't coexist in the same environment.
• Third-party channel incompatibility: The dsm channel's blpapi is probably an older, unmaintained build that doesn't play nice with pypyodbc from conda's default or conda-forge channels.

Fixes you can try right now #

1. Switch to a different blpapi source
   Skip the dsm channel entirely. Try installing from conda-forge (if available) with:

   conda install -c conda-forge blpapi
   

   If that still fails, use pip instead—pip's dependency resolver works differently and often bypasses conda's strict conflicts:

   pip install blpapi
   

2. Create an isolated conda environment
   Sometimes isolating the two packages in a fresh environment fixes the conflict. Pick a Python version that's likely compatible with both (3.9 is a safe bet for most packages):

   conda create -n blp_env python=3.9
   conda activate blp_env
   conda install -c dsm blpapi
   conda install pypyodbc
   

   Try reversing the install order if this fails—sometimes installing the trickier package first helps.

3. Manually specify compatible versions
   Use conda search --info blpapi -c dsm to check what Python versions and dependencies it supports, then do the same for pypyodbc with conda search --info pypyodbc. Then install specific versions that line up:

   conda install -c dsm blpapi=3.19.0 pypyodbc=1.3.6 python=3.9
   

   Adjust the version numbers to match what you find in the search results.

内容的提问来源于stack exchange，提问作者Mace