search

OCNS / SoftwareWG

The primary housekeeping repository for the INCF/OCNS Software Working Group, and the sources for the web site.
https://ocns.github.io/SoftwareWG/
11 stars 3 forks source link

Draft of software tutorial survey form #57

Closed appukuttan-shailesh closed 2 years ago

appukuttan-shailesh commented 2 years ago

In the last meeting (Jan 10; minutes coming soon), we decided to carry out a survey (via the various mailing lists) to identify tutorials on which topics are sought after by the community. I have created a draft of a form that we could share around:

https://docs.google.com/forms/d/e/1FAIpQLSd1DYK0ordTZJU04yc20ou2pVeRUcgHKYtUnNtaE-gYbyVkLg/viewform?usp=sf_link

Edit access: https://docs.google.com/forms/d/1iu8XX4WyXquD3ESH0N9akC_soK1TzlugAE6BQ_Z8joM/edit?usp=sharing

The list of options currently are:

Please provide suggestions for adding/updating the above options; and also free to improve the form using the edit link provided above.

sanjayankur31 commented 2 years ago

CC: @OCNS/software-wg

sanjayankur31 commented 2 years ago

Are we planning specific tool tutorials as part of the software wg satellite tutorials too, or are we trying to keep our tutorials general (like we did last year)? I ask because if we do specific tool tutorials (or list them in the survey), I think we'll need to make sure we represent all the tools out there---even ones where the developers are not currently software wg members. Otherwise we risk having a biased set of offerings which could upset folks :)

sanjayankur31 commented 2 years ago

@kernfel : would you be open to doing your bash tutorial again this year? @joewgraham : ditto for python?

appukuttan-shailesh commented 2 years ago

Are we planning specific tool tutorials as part of the software wg satellite tutorials too, or are we trying to keep our tutorials general (like we did last year)? I ask because if we do specific tool tutorials (or list them in the survey), I think we'll need to make sure we represent all the tools out there---even ones where the developers are not currently software wg members. Otherwise we risk having a biased set of offerings which could upset folks :)

More general tutorials (like Python or bash) certainly are useful. But I get the impression that many who have already started working in this domain would be somewhat familiar with these general topics (obviously still a lot to learn), and maybe more specialized tutorials might seem appealing, as they are not as commonly accessible. E.g. a tutorial on NEURON vs a tutorial on Python, I would think that there are plenty of resources for the latter, and so not very motivated to attend our tutorial for Python.

We certainly are open to all tools and, as discussed, we might have to rope in "specialists" for topics that we (Software WG members) aren't entirely conversant with. I suppose the purpose of the survey is partly to identify what the community members are looking out for. So I feel the survey could guide us on what kind of tutorials we should plan for.

I was hoping that others could add to the list of tools that we have above (in the survey). I have added those that I myself am familar with and/or know the "specialists" who could help ;-)

neworderofjamie commented 2 years ago

More general tutorials (like Python or bash) certainly are useful. But I get the impression that many who have already started working in this domain would be a somewhat familiar with these general topics (obviously still a lot to learn), and maybe more specialized tutorials might seem appealing, as they are not as commonly accessible. E.g. a tutorial on NEURON vs a tutorial on Python, I would think that there are plenty of resources for the latter, and so not very motivated to attend our tutorial for Python.

I was thinking exactly this earlier actually - there are now so many resources available for learning Python etc that maybe we should concentrate more on the specialized knowledge we can bring.

appukuttan-shailesh commented 2 years ago

p.s. we can highlight that we would offer tutorials only for Free and Open-Source Softwares (which I believe is what we had discussed - and agreed upon? - previously)

mstimberg commented 2 years ago

I also think that general Python/git/bash courses might not be the most "attractive" offer (but of course figuring this out is the idea behind the survey). IMO, ideally we would offer something more custom-tailored, e.g. something that teaches, say, bash and git along the way, but in the context of making research reproducible, publishable, etc. Something more around these lines: https://carpentries-incubator.github.io/open-science-with-r/ (with Python instead of R, I guess :wink: ) https://carpentries-incubator.github.io/fair-bio-practice/ There's also https://www.repro4everyone.org/ , I know there were a few people working on a version for neuroscience (not computational neuroscience specifically, though). I think teaching individual tools (say, NEURON) would be rather a topic for the "normal" CNS tutorials?

sanjayankur31 commented 2 years ago

Yeh, traditionally, folks do tutorials for their own tools as part of the conf if they wish to. So, I guess my question is if we're looking to change this and bring it all together under the umbrella of the software wg?

I mean: what we're going to do with the information gathered from the survey: do we go about organising the tutorials (reaching out to devs etc.) to make these tutorials happen under the software wg umbrella in a unified way? That sort of becomes like us organising a workshop at the conf---inviting speakers, doing the scheduling etc. (only it won't be happening during the workshops) Or, do we do the survey and share the results with the community (or privately with the dev teams), and then let the various dev teams decide if/how they want to do a tutorial/satellite tutorial at CNS?

I think i tend to lean towards the latter because some folks may prefer satellite tutorials, others may want to do theirs as part of the main tutorials or workshops, others may not do tutorials at all? It'll be sort of weird if they're all under "software wg" tutorials but are done in different ways. Not everyone may want to do it under the umbrella of the software wg for various reasons, and I'm not currently sure what the advantage of doing it under our umbrella as opposed to doing it themselves is either.. :/

For the survey, off the top of my head, other tools to add:

More general ones:

+1 to FOSS only (but what about Matlab toolboxes which are FOSS?)

For reference, this is last year's list: https://www.cnsorg.org/cns-2021-tutorials

I'd like to do general beginner level tutorials too if folks are up for them---for example as Marcel noted---especially if we can get the satellite tutorials running independently (and cheaply) from the main conference---they act as gateway events to get more people into comp-neuro/science/research. For example, our bash/git/python tutorials last year were quite popular. They're more for Outreach and education I guess? (There's just something about a live class/tutorial that makes folks prefer them to going through videos/documentation)

appukuttan-shailesh commented 2 years ago

I think teaching individual tools (say, NEURON) would be rather a topic for the "normal" CNS tutorials?

I agree. NEURON (didn't use the best example earlier, but intentionally kept it out of the survey list for the above reason) has now been around long enough, receives a lot of attention from the community. (I probably should remove NeuroML from that list on similar grounds?!).

My intention, when proposing specific tools/packages, is in part to help disseminate the relatively lesser known softwares. Ones that we either ourselves make use of in our work and find useful or know well, and therefore consider worthy topics to be taught.

Btw... two more suggestions for the list in the same vein (for the survey):

appukuttan-shailesh commented 2 years ago

It is likely that such "specialized" topics might in fact not be very popular in our survey (as they are not well known), so an idea might even be to list them by purpose rather than solely by tool name.

E.g.

appukuttan-shailesh commented 2 years ago

I'm not currently sure what the advantage of doing it under our umbrella as opposed to doing it themselves is either

I think "smaller" tools/packages might not normally/frequently conduct their own training sessions. They might not be suitable for satellite workshops, and also I suppose there is an overhead for them in organizing such events which might deter them. With the Software WG "inviting" these groups, we are offering them a platform where they can simply come and teach online, without worrying about applying for calls etc.

IMO, ideally we would offer something more custom-tailored, e.g. something that teaches, say, bash and git along the way, but in the context of making research reproducible, publishable, etc.

I do like this idea of using simpler tools and putting them together to produce effective workflows.

kernfel commented 2 years ago

I'd be happy to repeat the bash tutorial. I do agree that more integrated sessions would be more useful, although I suspect it might be difficult to balance between teaching the basics and moving forward with the higher-level stuff, particularly if the tutorials are to be relatively short.

It is likely that such "specialized" topics might in fact not be very popular in our survey (as they are not well known), so an idea might even be to list them by purpose rather than solely by tool name.

For beginners, there's definitely a hurdle in finding out what tools are even available, what they do, and what makes them unique or particularly useful for certain applications. More descriptive survey choices definitely have my vote, but I wonder if there might not also be room for something like a signposting session where we briefly present a number of popularly requested tools -- or tasks/topics -- with the aim of helping people make sensible choices.

sanjayankur31 commented 2 years ago

Sounds good. Let's include them all in the survey but grouped according to function as suggested?

@appukuttan-shailesh : there's also a list on the OCNS website, but I'm pretty sure it is almost never used and is rarely updated:

https://www.cnsorg.org/index.php?option=com_content&view=article&id=243:software&catid=20:site-content&Itemid=157

(probably best to remove it XD)

There's also this list here: https://open-neuroscience.com/tag/computational-neuroscience/

sanjayankur31 commented 2 years ago

Seems like the current plan is to have an in-person meeting in Melb:

https://twitter.com/RenaudJolivet/status/1487053413842046980?s=20&t=4OJCCloB8w-_kHx1VDXEvg

I guess we can send the survey out as soon as the call for tutorials has gone out. @appukuttan-shailesh : any news on the calls?

appukuttan-shailesh commented 2 years ago

Actually it still has not been decided if we will have an on-site vs hybrid vs online conference? The twitter poll was intended more towards gauging the public opinion, and using it to help arrive at a decision. I think it will still take some time to confirm one or the other. The intention is to try for an in-person meeting, if circumstances allow for it.

The calls for tutorials should be out shortly.

sanjayankur31 commented 2 years ago

Ah, OK. Sounds good.

Will the calls be made after this has beem decided? I guess we have time before we send out the survey then?

ChristophMetzner commented 2 years ago

I believe they want to send out the calls for abstracts/tutorials/workshops asap but wait with the decision about the format til ~ end of March. But there were many emails, so I am not overly confident if I recall everything correctly.

Am So., 30. Jan. 2022 um 12:58 Uhr schrieb Ankur Sinha < @.***>:

Ah, OK. Sounds good.

Will the calls be made after this has beem decided? I guess we have time before we send out the survey then?

— Reply to this email directly, view it on GitHub https://github.com/OCNS/SoftwareWG/issues/57#issuecomment-1025127444, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABJ4AJDHCKWOZIFVT4243RLUYURUPANCNFSM5LY4VGRA . Triage notifications on the go with GitHub Mobile for iOS https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Android https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub.

You are receiving this because you are on a team that was mentioned.Message ID: @.***>

--

Dr. Christoph Metzner, Dipl.-Math., B.Sc.

Postdoctoral Research Fellow

Neural Information Processing Group

Institute of Software Engineering and Theoretical Computer Science

Technische Universität Berlin

Visiting Research Fellow and Visiting Lecturer

Biocomputation Research Group

School of Physics, Engineering and Computer Science

University of Hertfordshire

Visiting Lecturer

University of Applied Sciences Berlin

E-Mail: @.***

tnowotny commented 2 years ago

Hi … yes, we are aiming to send out calls fairly soon. Likely tutorials/workshops as one call together first and call for abstracts soon after. The format of the meeting will be decided later but even the date of that decision is yet to be determined. Best, thomas

-- Prof Thomas Nowotny Head of AI Research Group School of Engineering and Informatics University of Sussex Falmer, Brighton BN1 9QJ, UK Phone: +44-1273-678593 FAX: +44-1273-877873

From: Christoph Metzner @.> Date: Sunday, 30 January 2022 at 17:57 To: OCNS/SoftwareWG @.> Cc: Thomas Nowotny @.>, Team mention @.> Subject: Re: [OCNS/SoftwareWG] Draft of software tutorial survey form (Issue #57) I believe they want to send out the calls for abstracts/tutorials/workshops asap but wait with the decision about the format til ~ end of March. But there were many emails, so I am not overly confident if I recall everything correctly.

Am So., 30. Jan. 2022 um 12:58 Uhr schrieb Ankur Sinha < @.***>:

Ah, OK. Sounds good.

Will the calls be made after this has beem decided? I guess we have time before we send out the survey then?

— Reply to this email directly, view it on GitHub https://github.com/OCNS/SoftwareWG/issues/57#issuecomment-1025127444<https://github.com/OCNS/SoftwareWG/issues/57#issuecomment-1025127444>, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABJ4AJDHCKWOZIFVT4243RLUYURUPANCNFSM5LY4VGRA<https://github.com/notifications/unsubscribe-auth/ABJ4AJDHCKWOZIFVT4243RLUYURUPANCNFSM5LY4VGRA> . Triage notifications on the go with GitHub Mobile for iOS https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675<https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub<https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>.

You are receiving this because you are on a team that was mentioned.Message ID: @.***>

--

Dr. Christoph Metzner, Dipl.-Math., B.Sc.

Postdoctoral Research Fellow

Neural Information Processing Group

Institute of Software Engineering and Theoretical Computer Science

Technische Universität Berlin

Visiting Research Fellow and Visiting Lecturer

Biocomputation Research Group

School of Physics, Engineering and Computer Science

University of Hertfordshire

Visiting Lecturer

University of Applied Sciences Berlin

E-Mail: @.***

— Reply to this email directly, view it on GitHubhttps://github.com/OCNS/SoftwareWG/issues/57#issuecomment-1025195541, or unsubscribehttps://github.com/notifications/unsubscribe-auth/ABJZIV5MRTIHCFOBIIHPC53UYV3YXANCNFSM5LY4VGRA. Triage notifications on the go with GitHub Mobile for iOShttps://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Androidhttps://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub. You are receiving this because you are on a team that was mentioned.Message ID: @.***>

sanjayankur31 commented 2 years ago

Sounds good, we'll fair out the survey form so it's ready to go.

sanjayankur31 commented 2 years ago

@appukuttan-shailesh : should we tweak the survey to use categories of tools? (as you suggested in your comment)?

appukuttan-shailesh commented 2 years ago

Yes, I think that would be more useful. I might be able to restructure the survey early next week.

p.s. I would unfortunately not be able to attend our upcoming meeting (I'm away Feb 14-16), but will go through the minutes once back.

sanjayankur31 commented 2 years ago

Sounds good. I need to set up a ticket for the next meeting too, and put up the logs from the last one. I'll do it this week.

appukuttan-shailesh commented 2 years ago

Apologies for the delay. I have now updated the survey form, and tried to frame the options around topics (rather than tools):

https://docs.google.com/forms/d/e/1FAIpQLSd1DYK0ordTZJU04yc20ou2pVeRUcgHKYtUnNtaE-gYbyVkLg/viewform?usp=sf_link

Edit access: https://docs.google.com/forms/d/1iu8XX4WyXquD3ESH0N9akC_soK1TzlugAE6BQ_Z8joM/edit?usp=sharing

@sanjayankur31 : I haven't added the list of simulators (from this comment). They probably will need to be listed individually, but also had doubts if they were slightly "heavy" topics requiring more time. (I have the same concern for "deep learning frameworks"). Feel free to update as required.

Tagging @OCNS/software-wg

p.s. I will be a bit occupied the coming few weeks. Would be happy if anyone would like to take the lead on this.

kernfel commented 2 years ago

I'd maybe add a generic "Simulators" checkbox with a free-form follow-up for respondents to specify which simulator they're interested in. Also, since there are experts from at least a few simulators in this WG, maybe a "how to choose your simulator" kind of session would be feasible?

sanjayankur31 commented 2 years ago

I've made some changes to the survey now.

So, it gives us general information on peoples' use of software, and also information on what tools people are interested in using.

The idea of the survey is also to make people aware of all the software out there, so I've added sections for data analysis and computational modelling, and I've listed lots of software there.

How does this look? Please add tools that I've missed there---there really are lots of tools, but I generally sort of skipped lots of fmri tools etc. because we're sending this out to more comp-neuro folks, and they can just add them in the free-text boxes.

sanjayankur31 commented 2 years ago

Forgot: tet's timebox this. So, edits/discussion welcome till the coming Monday (7th March), when we aim to send it out to the usual mailing lists. That should give us some responses before our monthly WG meeting on the 14th at 1700 UTC so that we can discuss what we want to do for tutorials etc.

tnowotny commented 2 years ago

from where I sit I am almost certain that we can offer online "satellite tutorials" before the conference.

sanjayankur31 commented 2 years ago

Thank you. That'll be great. I'll try to confirm this with the tutorials chair ASAP.

mstimberg commented 2 years ago

An unrelated question/comment: do we really want to ask for name (even a required field) and institution? I think if we do this, we are collecting personal data, need to take care of GDPR-related stuff, yada yada. Don't think that's worth it. Maybe remove the name (and mention on top that the survey is anonymous), and replace the question for institution by something that asks for the type of institution (e.g academic vs. private)?

sanjayankur31 commented 2 years ago

An unrelated question/comment: do we really want to ask for name (even a required field) and institution? I think if we do this, we are collecting personal data, need to take care of GDPR-related stuff, yada yada. Don't think that's worth it. Maybe remove the name (and mention on top that the survey is anonymous), and replace the question for institution by something that asks for the type of institution (e.g academic vs. private)?

+1,

kernfel commented 2 years ago

In the interest of keeping answering as un-annoying as possible, and of reducing horizontal scrolling, I'd favor dropping the first checkbox column ("Don't know").

sanjayankur31 commented 2 years ago

Done, dropped the column in all sections now.

@OCNS/software-wg : could everyone maybe please submit the form once? It'll get us started, and any issue will get caught before we send it out on Monday too. It should really only take 2 minutes.

https://forms.gle/NXWv2CwtWBSB9BXv9

appukuttan-shailesh commented 2 years ago

Have made a dummy submission now. Made a few minor changes: Ebrains -> EBRAINS Contribute to -> Contributed to (undo if the intention was different)

Also, was wondering if we should drop the option "Would like to contribute to"... (i) to reduce the # options, (ii) wasn't certain if this info provided value to us. On similar lines (i.e. ii), was thinking if we should also drop the "contributed to" option, so that we simply keep the same three options (Know but have not used, Have used, Would like to learn/use) in sections 3, 4 as in section 2.

sanjayankur31 commented 2 years ago

I'd leave the contribution related options in. The idea is to also gauge if people help maintain the tools they use, or if theyre interested in doing so. (If people are, we can try to organise onboarding sessions to help people with contributing to the development of tools too, in addition to usage tutorials etc.)

-- Thanks,

Ankur

(Sent from mobile device)

sanjayankur31 commented 2 years ago

I submitted the form too, took less than 2 minutes. I'll send it out this afternoon, and share on social media etc.

sanjayankur31 commented 2 years ago

Sent out to MLs now. On twitter here, please retweet etc:

https://twitter.com/sanjay_ankur/status/1500761811024588805?s=20&t=j-M1Sbm1psVxtTCOpccpjg

@appukuttan-shailesh : I think we should turn off edit access on the form now (because the edit link is visible here in this ticket)?

appukuttan-shailesh commented 2 years ago

@sanjayankur31 : Thanks Ankur. I have now disabled the edit access by URL, and have added you (your UCL email address) as a collaborator for this form.