packer | creating identical machine images for multiple platforms

You can use the vsphere-clone builder for this. After your base build produces a VSphere template, you can then use that output artifact as the source for the child Packer template build:

The issue is probably with tensorflow==2.2.0. This is a very heavy library and you can't install it on t2.micro. You need at least t2.medium (not in free tier) which has more RAM to successfully install tensorflow==2.2.0 on EB.

The error says MemoryError. You must upgrade your ec2 instance to something with more memory. tensorflow is very memory hungry application.

Change the approach of collecting your data a bit and select the elements in order they appear under the date seperators.

Step #1

Select all the date seperators:

Try changing x.lstrip('at') to x.lstrip('at ')

You have not provided any evidence that the parameter is being ignored. ansible_python_intepreter applies to the execution of modules on the targets, not to the execution of Ansible itself.

The Ansible control process always uses the Python interpreter it was installed under; the best way to change this is to change your install process.

• StreamReader.ReadLine() is blocking by design.

• There is an asynchronous alternative, .ReadLineAsync(), which returns a Task instance that you can poll for completion, via its .IsCompleted property, without blocking your foreground thread (polling is your only option in PowerShell, given that it has no language feature analogous to C#'s await).

Here's a simplified example that focuses on asynchronous reading from a StreamReader instance that happens to be a file, to which new lines are added only periodically; use Ctrl-C to abort.

I would expect the code to work the same if you adapt it to your stdout-reading System.Diagnostics.Process code.

Currently according to your code above when you click on a ListView item you are trying to open a url using the below Intent which is failing to construct a url from an ArrayList:

[proposed round-robin solution removed as it was not applicable]

The problem can be converted to a graph where each distinct game is a vertex and the edges are links to other games that don't conflict (i.e. other games that don't involve the same teams). Each week's schedule will be a maximum clique in that graph and there will be 16 disjoint cliques of 16 games in the 256 node graph.

Finding maximum cliques is a well known NP-Complete problem but given the constraints, there may be a dynamic programming solution. Having 16 disjoint cliques of 16 games means that the full graph is covered by these cliques so every game will go in exactly one of these cliques. We can also leverage the fact that we know of 16 games that each belong to a different clique and can serve as anchors to expand the 16 distinct cliques (e.g. the 16 matches with the 49ers for example).

My initial attempts at solving this with DP were unsuccessful and would take forever to converge to an answer. They worked on a small schedule generated with the round-robin algorithm but failed with the 32 teams dictionary.

[EDIT] Found a solution that may work

After realizing that there were repeated teams in the lists I figured out why my initial graph solutions were failing. This gave me an idea for one final attempt that actually worked (at least for the data sample you provided in a comment and for shuffled round-robin schedules I tested).

The solution is an optimized variant of the graph algorithm I was playing with. The graph's vertices are now unique games (team pairs) that allow for repetitions. The algorithm uses a list of game sets (one per week) and attempts to place each game (vertex) in one of the weeks while respecting constraints. It backtracks the game placements when it hits a dead end. Combined with a dynamic prioritization of games to place, the program reverse engineers the schedule within seconds (it was still taking an inordinate amount of time without the prioritization).

Note that, although it did in all my tests, I cannot be sure that the program will always respond as rapidly with different data. That's why I'm only saying this MAY work.

Input