Feedback from Sep 25 from chat:
TP: Oh and one more thing that just came to my mind:
The reproducibility lecture is the first that actually uses the conda environment.
Do we need a actual intro on how to use it in there? Or should this be covered in week 1?
RB: i think it's better introduced as part of this lesson instead of as part of week 1. it is not really needed in week 1 and is more localized in time and context and memory to discuss it as part of reproducibility.
RD: (I started writing this earlier in the day but didn't send)
yeah, it's the first time it's used. Actually "how to use conda well" would be an important and interesting thing to focus on for reproducible research.
RB: and i wouldn't mind offering and/or recording a separate session where we share how to use this tool well. it's not easy to do in 10-15 mins but would be valuable to have.
RD: thinking more, could the whole lesson sort of be made around "how and why to use command line": first organization, then we see you can make scripts, then conda (from command line but if possible we should show how to use it from IDSs/other tools), then snakemake (as final demo)
IP: I like the proposed flow - though there's also containers to cover. Maybe it's just me, but If the intro+motivation+writing excursion parts would remain, this might easily get a bit too much for a 2-2.5 hour session?
RD: but it sounds like if there should be something that gets a big update, this might be it? and yeah, it can't be everything. there will be some people for whom command line and project organization is too easy, some it's too hard
IP: I would think that "how and why to use command line" would be a great intro for beginners, but looking at the feedback, it's snakemake and containers more 'advanced' users appear to find useful. Would it be crazy to also make that a longer day and split it in two parts - part (1) is the more basic intro and part (2) could take a bit more time to discuss and demo snakemake and containers?
Of course, that also means that advanced users probably won't be around to listen to the motivation...
Feedback from Sep 25 from chat:
TP: Oh and one more thing that just came to my mind:
The reproducibility lecture is the first that actually uses the conda environment.
Do we need a actual intro on how to use it in there? Or should this be covered in week 1?
RB: i think it's better introduced as part of this lesson instead of as part of week 1. it is not really needed in week 1 and is more localized in time and context and memory to discuss it as part of reproducibility.
RD: (I started writing this earlier in the day but didn't send)
yeah, it's the first time it's used. Actually "how to use conda well" would be an important and interesting thing to focus on for reproducible research.
RB: and i wouldn't mind offering and/or recording a separate session where we share how to use this tool well. it's not easy to do in 10-15 mins but would be valuable to have.
RD: thinking more, could the whole lesson sort of be made around "how and why to use command line": first organization, then we see you can make scripts, then conda (from command line but if possible we should show how to use it from IDSs/other tools), then snakemake (as final demo)
IP: I like the proposed flow - though there's also containers to cover. Maybe it's just me, but If the intro+motivation+writing excursion parts would remain, this might easily get a bit too much for a 2-2.5 hour session?
RD: but it sounds like if there should be something that gets a big update, this might be it? and yeah, it can't be everything. there will be some people for whom command line and project organization is too easy, some it's too hard
IP: I would think that "how and why to use command line" would be a great intro for beginners, but looking at the feedback, it's snakemake and containers more 'advanced' users appear to find useful. Would it be crazy to also make that a longer day and split it in two parts - part (1) is the more basic intro and part (2) could take a bit more time to discuss and demo snakemake and containers?
Of course, that also means that advanced users probably won't be around to listen to the motivation...