Spiking neural network model format
12 views
Skip to first unread message
Tom O'Connell
Nov 27, 2016, 4:23:57 AM11/27/16
to COMBINE discussions
Combine includes standardization as part of its mission, yet we offer both SpineML and NineML on the "Standards" section of the website. The two standards are ostensibly competitors, and I feel it conflicts with the goal of this initiative to not at least advertise ways in which they have meaningfully different advantages. Without some kind of justification for presenting both standards, I think we need to pick one to support.
Thoughts?
Thoughts?
Alexander Mazein
Nov 27, 2016, 6:09:48 AM11/27/16
to Tom O'Connell, COMBINE discussions
Hello,
From what I see on the website http://co.mbine.org/standards, none of the two is in the main list of the COMBINE standards. They are in the additional list of "Related standartization efforts" (efforts of interest for COMBINE, either as candidate standards, or similar efforts in different domains).
NeuroML is in the main list. Looks like these three languages are related.
COMBINE is a community effort, it would be logical to suggest the representatives of the three languages talk to each other? This discussion could result in one single improved standard hopefully.
http://software.incf.org/software/nineml/wiki/summary-of-nineml-spineml-teleconference-on-2013-06-14
Best regards,
Alexander
On Sun, Nov 27, 2016 at 10:23 AM, Tom O'Connell <tom.oc...@gmail.com> wrote:
Combine includes standardization as part of its mission, yet we offer both SpineML and NineML on the "Standards" section of the website. The two standards are ostensibly competitors, and I feel it conflicts with the goal of this initiative to not at least advertise ways in which they have meaningfully different advantages. Without some kind of justification for presenting both standards, I think we need to pick one to support.
Thoughts?
--
You received this message because you are subscribed to the Google Groups "COMBINE discussions" group.
To unsubscribe from this group and stop receiving emails from it, send an email to combine-discuss+unsubscribe@googlegroups.com.
To post to this group, send email to combine-discuss@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/combine-discuss/6d825643-2455-472d-a2a5-6512d1e84ccd%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
Padraig Gleeson
Nov 28, 2016, 12:17:45 PM11/28/16
to combine...@googlegroups.com
Hi,
Regarding NineML and SpineML, there has certainly been a lot of overlap between the two of those for various historical reasons, but now the main developers of SpineML are on the NineML specification committee: http://incf.github.io/nineml/committee/, and so it's quite possible the two will formally be merged at some point in the future. The focus of Nine/SpineML and NeuroML have been slightly different, but as you point out there are ways to map some of models specified in these languages to and from NeuroML. Andrew Davison, one of the main NineML developers is also an editor for NeuroML, and he and I are both authors on a paper on NineML due out some time in the future.
I agree that it shouldn't be the job of COMBINE to just pick a standard that it feels is the "right" one, but it's better to encourage best practices for what a standard/community should be and support any and all that live up to that and want to interact with others in the field.
And will there be a merger between NineML and NeuroML? That's estimated to happen about a month after the v0.1 release of the merged SBML-CellML specification ;-)
Regards,
Padraig
Regarding NineML and SpineML, there has certainly been a lot of overlap between the two of those for various historical reasons, but now the main developers of SpineML are on the NineML specification committee: http://incf.github.io/nineml/committee/, and so it's quite possible the two will formally be merged at some point in the future. The focus of Nine/SpineML and NeuroML have been slightly different, but as you point out there are ways to map some of models specified in these languages to and from NeuroML. Andrew Davison, one of the main NineML developers is also an editor for NeuroML, and he and I are both authors on a paper on NineML due out some time in the future.
I agree that it shouldn't be the job of COMBINE to just pick a standard that it feels is the "right" one, but it's better to encourage best practices for what a standard/community should be and support any and all that live up to that and want to interact with others in the field.
And will there be a merger between NineML and NeuroML? That's estimated to happen about a month after the v0.1 release of the merged SBML-CellML specification ;-)
Regards,
Padraig
To unsubscribe from this group and stop receiving emails from it, send an email to combine-discu...@googlegroups.com.
To post to this group, send email to combine...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/combine-discuss/CAOPOR24hOC40dRq4y65Eu82druPU6rS06VgHq9Vs_e8Yh0nrmw%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.
-- ----------------------------------------------------- Padraig Gleeson Room 321, Anatomy Building Department of Neuroscience, Physiology& Pharmacology University College London Gower Street London WC1E 6BT United Kingdom +44 207 679 3214 p.gl...@ucl.ac.uk -----------------------------------------------------
Tom O'Connell
Nov 28, 2016, 4:06:40 PM11/28/16
to COMBINE discussions
Hi all,
Thanks for your replies. I'm glad I asked, because it is good to hear that the efforts have a chance of merging in the near future. Unless the SBML-CellML merger is a joke for always being on the horizon? Either way, it was nice to see the points about merging some of these standards on the agenda you linked, Alexander.
I suppose it would be out of place for COMBINE to "pick" one, but what do you see as the main role of a standard if not to maximize interoperability by minimizing competing formats, and the additional development time required to support all of them?
Cheers,
Tom
Thanks for your replies. I'm glad I asked, because it is good to hear that the efforts have a chance of merging in the near future. Unless the SBML-CellML merger is a joke for always being on the horizon? Either way, it was nice to see the points about merging some of these standards on the agenda you linked, Alexander.
I suppose it would be out of place for COMBINE to "pick" one, but what do you see as the main role of a standard if not to maximize interoperability by minimizing competing formats, and the additional development time required to support all of them?
Cheers,
Tom
On Monday, November 28, 2016 at 9:17:45 AM UTC-8, p.gleeson wrote:
Hi,
Regarding NineML and SpineML, there has certainly been a lot of overlap between the two of those for various historical reasons, but now the main developers of SpineML are on the NineML specification committee: http://incf.github.io/nineml/committee/, and so it's quite possible the two will formally be merged at some point in the future. The focus of Nine/SpineML and NeuroML have been slightly different, but as you point out there are ways to map some of models specified in these languages to and from NeuroML. Andrew Davison, one of the main NineML developers is also an editor for NeuroML, and he and I are both authors on a paper on NineML due out some time in the future.
I agree that it shouldn't be the job of COMBINE to just pick a standard that it feels is the "right" one, but it's better to encourage best practices for what a standard/community should be and support any and all that live up to that and want to interact with others in the field.
And will there be a merger between NineML and NeuroML? That's estimated to happen about a month after the v0.1 release of the merged SBML-CellML specification ;-)
Regards,
Padraig
On 27/11/16 11:09, Alexander Mazein wrote:
Hello,
From what I see on the website http://co.mbine.org/standards, none of the two is in the main list of the COMBINE standards. They are in the additional list of "Related standartization efforts" (efforts of interest for COMBINE, either as candidate standards, or similar efforts in different domains).
NeuroML is in the main list. Looks like these three languages are related.
COMBINE is a community effort, it would be logical to suggest the representatives of the three languages talk to each other? This discussion could result in one single improved standard hopefully.http://software.incf.org/software/nineml/wiki/summary-of-nineml-spineml-teleconference-on-2013-06-14
Best regards,
Alexander
On Sun, Nov 27, 2016 at 10:23 AM, Tom O'Connell <tom.oc...@gmail.com> wrote:
Combine includes standardization as part of its mission, yet we offer both SpineML and NineML on the "Standards" section of the website. The two standards are ostensibly competitors, and I feel it conflicts with the goal of this initiative to not at least advertise ways in which they have meaningfully different advantages. Without some kind of justification for presenting both standards, I think we need to pick one to support.
Thoughts?
--
You received this message because you are subscribed to the Google Groups "COMBINE discussions" group.
To unsubscribe from this group and stop receiving emails from it, send an email to combine-discu...@googlegroups.com.
To post to this group, send email to combine...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/combine-discuss/6d825643-2455-472d-a2a5-6512d1e84ccd%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
--
You received this message because you are subscribed to the Google Groups "COMBINE discussions" group.
To unsubscribe from this group and stop receiving emails from it, send an email to combine-discu...@googlegroups.com.
To post to this group, send email to combine...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/combine-discuss/CAOPOR24hOC40dRq4y65Eu82druPU6rS06VgHq9Vs_e8Yh0nrmw%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.
Nicolas Le Novere
Nov 28, 2016, 4:22:39 PM11/28/16
to combine...@googlegroups.com
Hello,
A while ago, a list of criteria was drafted to delineate what could be considered as a candidate for becoming a core standard. This is in addition of the will to interoperate. Those are just guidelines, and not absolute rules. However, reading them might lead to thoughts on what a good community standard could be.
Best regards
http://co.mbine.org/Documents/criteria
On 28/11/16 21:06, Tom O'Connell wrote:
> Hi all,
>
> Thanks for your replies. I'm glad I asked, because it is good to hear that the efforts have a chance of merging in the near future. Unless the SBML-CellML merger is a joke for always being on the horizon? Either way, it was nice to see the points about merging some of these standards on the agenda you linked, Alexander.
>
> I suppose it would be out of place for COMBINE to "pick" one, but what do you see as the main role of a standard if not to maximize interoperability by minimizing competing formats, and the additional development time required to support all of them?
>
> Cheers,
> Tom
>
> On Monday, November 28, 2016 at 9:17:45 AM UTC-8, p.gleeson wrote:
>
> Hi,
>
> Regarding NineML and SpineML, there has certainly been a lot of overlap between the two of those for various historical reasons, but now the main developers of SpineML are on the NineML specification committee: http://incf.github.io/nineml/committee/ <http://incf.github.io/nineml/committee/>, and so it's quite possible the two will formally be merged at some point in the future. The focus of Nine/SpineML and NeuroML have been slightly different, but as you point out there are ways to map some of models specified in these languages to and from NeuroML <https://github.com/OpenSourceBrain/NineMLShowcase>. Andrew Davison, one of the main NineML developers is also an editor for NeuroML <https://neuroml.org/editors>, and he and I are both authors on a paper on NineML due out some time in the future.
>> http://software.incf.org/software/nineml/wiki/introduction-to-nineml <http://software.incf.org/software/nineml/wiki/introduction-to-nineml>
>> To post to this group, send email to combine...@googlegroups.com <javascript:>.
>> To view this discussion on the web, visit https://groups.google.com/d/msgid/combine-discuss/6d825643-2455-472d-a2a5-6512d1e84ccd%40googlegroups.com <https://groups.google.com/d/msgid/combine-discuss/6d825643-2455-472d-a2a5-6512d1e84ccd%40googlegroups.com?utm_medium=email&utm_source=footer>.
>> For more options, visit https://groups.google.com/d/optout <https://groups.google.com/d/optout>.
>> To post to this group, send email to combine...@googlegroups.com <javascript:>.
>> To view this discussion on the web, visit https://groups.google.com/d/msgid/combine-discuss/CAOPOR24hOC40dRq4y65Eu82druPU6rS06VgHq9Vs_e8Yh0nrmw%40mail.gmail.com <https://groups.google.com/d/msgid/combine-discuss/CAOPOR24hOC40dRq4y65Eu82druPU6rS06VgHq9Vs_e8Yh0nrmw%40mail.gmail.com?utm_medium=email&utm_source=footer>.
>> For more options, visit https://groups.google.com/d/optout <https://groups.google.com/d/optout>.
> -----------------------------------------------------
> To post to this group, send email to combine...@googlegroups.com <mailto:combine...@googlegroups.com>.
> To view this discussion on the web, visit https://groups.google.com/d/msgid/combine-discuss/e4085dc3-c856-45d1-ac6e-22ac413c96d9%40googlegroups.com <https://groups.google.com/d/msgid/combine-discuss/e4085dc3-c856-45d1-ac6e-22ac413c96d9%40googlegroups.com?utm_medium=email&utm_source=footer>.
Nicolas LE NOVERE, Babraham Institute, Babraham Campus Cambridge, CB22 3AT
Tel: +441223496433 Mob:+447833147074 n.len...@gmail.com
orcid.org//0000-0002-6309-7327 http://lenoverelab.org/perso/lenov/
Skype:n.lenovere twitter:@lenovere http://nlenov.wordpress.com/
A while ago, a list of criteria was drafted to delineate what could be considered as a candidate for becoming a core standard. This is in addition of the will to interoperate. Those are just guidelines, and not absolute rules. However, reading them might lead to thoughts on what a good community standard could be.
Best regards
http://co.mbine.org/Documents/criteria
On 28/11/16 21:06, Tom O'Connell wrote:
> Hi all,
>
> Thanks for your replies. I'm glad I asked, because it is good to hear that the efforts have a chance of merging in the near future. Unless the SBML-CellML merger is a joke for always being on the horizon? Either way, it was nice to see the points about merging some of these standards on the agenda you linked, Alexander.
>
> I suppose it would be out of place for COMBINE to "pick" one, but what do you see as the main role of a standard if not to maximize interoperability by minimizing competing formats, and the additional development time required to support all of them?
>
> Cheers,
> Tom
>
> On Monday, November 28, 2016 at 9:17:45 AM UTC-8, p.gleeson wrote:
>
> Hi,
>
>
> I agree that it shouldn't be the job of COMBINE to just pick a standard that it feels is the "right" one, but it's better to encourage best practices for what a standard/community should be and support any and all that live up to that and want to interact with others in the field.
>
> And will there be a merger between NineML and NeuroML? That's estimated to happen about a month after the v0.1 release of the merged SBML-CellML specification ;-)
>
> Regards,
> Padraig
>
>
>
>
> On 27/11/16 11:09, Alexander Mazein wrote:
>> Hello,
>>
>> From what I see on the website http://co.mbine.org/standards <http://co.mbine.org/standards>, none of the two is in the main list of the COMBINE standards. They are in the additional list of "Related standartization efforts" (efforts of interest for COMBINE, either as candidate standards, or similar efforts in different domains).
> I agree that it shouldn't be the job of COMBINE to just pick a standard that it feels is the "right" one, but it's better to encourage best practices for what a standard/community should be and support any and all that live up to that and want to interact with others in the field.
>
> And will there be a merger between NineML and NeuroML? That's estimated to happen about a month after the v0.1 release of the merged SBML-CellML specification ;-)
>
> Regards,
> Padraig
>
>
>
>
> On 27/11/16 11:09, Alexander Mazein wrote:
>> Hello,
>>
>>
>> NeuroML is in the main list. Looks like these three languages are related.
>> http://www.opensourcebrain.org/projects/ninemlshowcase <http://www.opensourcebrain.org/projects/ninemlshowcase>
>> NeuroML is in the main list. Looks like these three languages are related.
>> http://software.incf.org/software/nineml/wiki/introduction-to-nineml <http://software.incf.org/software/nineml/wiki/introduction-to-nineml>
>>
>> COMBINE is a community effort, it would be logical to suggest the representatives of the three languages talk to each other? This discussion could result in one single improved standard hopefully.
>> http://software.incf.org/software/nineml/wiki/summary-of-nineml-spineml-teleconference-on-2013-06-14 <http://software.incf.org/software/nineml/wiki/summary-of-nineml-spineml-teleconference-on-2013-06-14>
>> COMBINE is a community effort, it would be logical to suggest the representatives of the three languages talk to each other? This discussion could result in one single improved standard hopefully.
>>
>> Best regards,
>>
>> Alexander
>>
>>
>>
>> On Sun, Nov 27, 2016 at 10:23 AM, Tom O'Connell <tom.oc...@gmail.com <javascript:>> wrote:
>>
>> Combine includes standardization as part of its mission, yet we offer both SpineML and NineML on the "Standards" section of the website. The two standards are ostensibly competitors, and I feel it conflicts with the goal of this initiative to not at least advertise ways in which they have meaningfully different advantages. Without some kind of justification for presenting both standards, I think we need to pick one to support.
>>
>> Thoughts?
>> --
>> You received this message because you are subscribed to the Google Groups "COMBINE discussions" group.
>> To unsubscribe from this group and stop receiving emails from it, send an email to combine-discu...@googlegroups.com <javascript:>.
>> Best regards,
>>
>> Alexander
>>
>>
>>
>> On Sun, Nov 27, 2016 at 10:23 AM, Tom O'Connell <tom.oc...@gmail.com <javascript:>> wrote:
>>
>> Combine includes standardization as part of its mission, yet we offer both SpineML and NineML on the "Standards" section of the website. The two standards are ostensibly competitors, and I feel it conflicts with the goal of this initiative to not at least advertise ways in which they have meaningfully different advantages. Without some kind of justification for presenting both standards, I think we need to pick one to support.
>>
>> Thoughts?
>> --
>> You received this message because you are subscribed to the Google Groups "COMBINE discussions" group.
>> To post to this group, send email to combine...@googlegroups.com <javascript:>.
>> To view this discussion on the web, visit https://groups.google.com/d/msgid/combine-discuss/6d825643-2455-472d-a2a5-6512d1e84ccd%40googlegroups.com <https://groups.google.com/d/msgid/combine-discuss/6d825643-2455-472d-a2a5-6512d1e84ccd%40googlegroups.com?utm_medium=email&utm_source=footer>.
>> For more options, visit https://groups.google.com/d/optout <https://groups.google.com/d/optout>.
>>
>>
>> --
>> You received this message because you are subscribed to the Google Groups "COMBINE discussions" group.
>> To unsubscribe from this group and stop receiving emails from it, send an email to combine-discu...@googlegroups.com <javascript:>.
>>
>> --
>> You received this message because you are subscribed to the Google Groups "COMBINE discussions" group.
>> To post to this group, send email to combine...@googlegroups.com <javascript:>.
>> To view this discussion on the web, visit https://groups.google.com/d/msgid/combine-discuss/CAOPOR24hOC40dRq4y65Eu82druPU6rS06VgHq9Vs_e8Yh0nrmw%40mail.gmail.com <https://groups.google.com/d/msgid/combine-discuss/CAOPOR24hOC40dRq4y65Eu82druPU6rS06VgHq9Vs_e8Yh0nrmw%40mail.gmail.com?utm_medium=email&utm_source=footer>.
>> For more options, visit https://groups.google.com/d/optout <https://groups.google.com/d/optout>.
>
> --
>
> -----------------------------------------------------
> Padraig Gleeson
> Room 321, Anatomy Building
> Department of Neuroscience, Physiology& Pharmacology
> University College London
> Gower Street
> London WC1E 6BT
> United Kingdom
>
> +44 207 679 3214
> p.gl...@ucl.ac.uk <javascript:>
> --
>
> -----------------------------------------------------
> Padraig Gleeson
> Room 321, Anatomy Building
> Department of Neuroscience, Physiology& Pharmacology
> University College London
> Gower Street
> London WC1E 6BT
> United Kingdom
>
> +44 207 679 3214
> -----------------------------------------------------
>
> --
> You received this message because you are subscribed to the Google Groups "COMBINE discussions" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to combine-discu...@googlegroups.com <mailto:combine-discu...@googlegroups.com>.
> --
> You received this message because you are subscribed to the Google Groups "COMBINE discussions" group.
> To post to this group, send email to combine...@googlegroups.com <mailto:combine...@googlegroups.com>.
> To view this discussion on the web, visit https://groups.google.com/d/msgid/combine-discuss/e4085dc3-c856-45d1-ac6e-22ac413c96d9%40googlegroups.com <https://groups.google.com/d/msgid/combine-discuss/e4085dc3-c856-45d1-ac6e-22ac413c96d9%40googlegroups.com?utm_medium=email&utm_source=footer>.
Nicolas LE NOVERE, Babraham Institute, Babraham Campus Cambridge, CB22 3AT
Tel: +441223496433 Mob:+447833147074 n.len...@gmail.com
orcid.org//0000-0002-6309-7327 http://lenoverelab.org/perso/lenov/
Skype:n.lenovere twitter:@lenovere http://nlenov.wordpress.com/
Reply all
Reply to author
Forward
0 new messages