pin to htcondor v1 bindings

[kpedro88]

The latest htcondor package on pip only has the new v2 bindings. That means it only allows import htcondor2; import htcondor will not work.

In lieu of trying to update everything, let's just pin to the old v1 bindings for now.

Eventually, updating to v2 will be needed because v1 relies on Boost.Python and therefore requires separate wheels for each Python 3.X, which (to my understanding) will not be provided for newer values of X.

Details on v2 API and transition:
https://indico.cern.ch/event/1386170/contributions/6127739/attachments/2934325/5155703/PythonBindingsVersion2.pdf
https://htcondor.readthedocs.io/en/lts/apis/python-bindings/api/version2/migration-guide.html
https://github.com/htcondor/htcondor/blob/main/python-bindings.md

Possibly related to #38?

[ikrommyd]

@kpedro88 what do you think of #46?
Pinning to that old version is not really an option. My understanding is that htcondor 24 ships both versions of the bindings and htcondor25 only the v2 bindings. The htcondor version that ends up in the images is usually dictated by the af-images https://github.com/CoffeaTeam/af-images. If we pin there to <24 or <25 we may have environment resolution problems with latest python versions and we need people to be on latest python. <25 may be doable but <24 definitely not.