This report will be a little bit longer than normal cause it will comprehend all the week changes too. So let’s start.

  1. Status : There was three big features that I’ve been doing during the last two weeks. First one, I’ve changed a little how the new information is inserted at make_graph function as we already had an .update method for LaizyJSON files. During the last week we could test how the new working environment for the new_update_upstream_version is doing, also if the information regarding the JSON files stored at versions folder is being correctly loaded and inserted at your graph. The second one, I also found a test script for the update functions called test_update_upstream_versions at a test folder on cf-scripts repo, this served as a good way to test if the new alterations with the .update method were working well. The third big feature, occurred during the Scipy 2020 wich I had the great opportunity to participate, with excellent presentations and very interesting discussions as also a session of discussion with two of my mentors about conda-forge and the packaging systems. My overview and experience with Scipy 2020 will be in an upcoming blog post during the week.

  2. Abstract : The alteration was very simple, just changed the way the information was been inserted at the attrsusing an already implemented method for LazyJSON .update, then now we have (see the discussion here):

    for file in list_files:
   node = str(file).replace(".json", "")
   with open(f"./versions/{file}") as json_file:
       version_data = json.load(json_file)
   with gx.nodes[f"{node}"]["payload"] as attrs:
       attrs.update(version_data)

    As a simple test to check if the .update was inserting the information as we expected, I run locally the updates_nodes_with_new_update_version function using a versions folder with few nodes.json files from cf-graph-countyfair to check if it will be correctly assigned to its correspondent place at the graph. In the end it run perfectly and the new information was correctly assigned, I also run the correspondent test function for update_upstream_Version that can be find here where I realized that the current test was not aligned with the new altaration made i.n the new_update_upstream_version so I changed the test accordingly (see #1064), and after this is merged we can finally replace the old configuration with the new one and test it!

During the last meeting (14/07) I was concerned with some few future steps of the schedule so I asked about some ideas and task to solve this doubts, in response beside an amazing conversation with everyone (where I coulda discover some other aspects of the bot and the future projects to maintain and automate the bot) and the future projects to maintain and automate the bot) I asked a little about the current Migration process and its structure , the Manba and BOI solvers see(which I also need to write so question and ideas to talk whit Wolf one of the mainteines of conda-forge and the creator of the Mnaba Solver ), some comments about the Scipy 2020 and a few questions about the web services of conda-forge.

  1. Next Steps : For this week I will try to follow the merge process of the new_update_version as the principal stuff, besides that I need to think a little about the pining and migration process, as already said I want tocreate a big detailed guide of conda-forge and the bot main works/process. Then, when finally migrate, we can check for eventual bugs, if everything occurs smoothly we can pass for an optimization phase of the code as also visit some related functions, e.g the update_sources script.

As cited by Christopher I will try to follow the “PR” comments updates at Grayskull audits as also I am very interested at this particular issue, about the bot organization.