These are chat archives for biojs/biojs

24th
Apr 2016
Ayush Sharma
@ayusharma
Apr 24 2016 04:08
Screen_Shot_2016-04-24_at_9.33.24_AM.png

@01remi I will and thank you so much for your motivation. As I know, GSoC's selection procedure based on contribution, code & proposal. These three things to be measured and greenify already said on march 27 : stay active on github.

  1. So the point is that the person who never commit a single line of code in BioJS got a chance.

  2. This year google introduced the feature of mentor review before the submission and the selected candidate itself came before a day of proposal deadline and asks that "Could he still get involved ?". As google already had given time for involvement with organisation after the org announcement.

  3. And above conversation I had with one of mentor

It just made me think that how can I work on a project for which I have not submitted proposal. @01remi so there is a no point of vision and proposal.
So anyone who thinks there is something false involve, Please raise their voice. Thanks :)
Sebastian Wilzbach
@wilzbach
Apr 24 2016 09:08
@ayusharma I wasn't involved in the decision, so I can't give any further information, but you should definitely ping @jessica-jordan and @goldbergtatyana. Speaking about contributions: yes while not officially required by Google, it is a defacto criteria for most organizations. I personally am just sad that the most important project (web components) didn't got selected :/
Awasum Yannick
@awasum
Apr 24 2016 09:08
Gsoc has become so compettitive
its difficult to select students based only on their proposals and design intellects.
have to show dat u can do the job, its the nature of Gsoc these days.
Sebastian Wilzbach
@wilzbach
Apr 24 2016 09:25
@ayusharma one huge problem in the selection of the projects is that the mentors favor there personal projects, so they try to push these projects in the gsoc conversation of an organization. Especially this year with biojs there are some projects on which the mentors directly depend on their company/university/Institute life and would have to do themselves otherwise or hire someone.
Another problem is that mentors favor students they already are familiar with. In 2014 we had one case where a student was picked by a mentor, because they were working at the same institute (in fact even the same group and project). In this case I was pretty angry because he got his usual salary during the entire gsoc and they outcome was "suboptimal" too.
Jessica Jordan
@jessica-jordan
Apr 24 2016 11:22

Hi @ayusharma First things first: We were very happy to have received your very strong proposal and were quite impressed by your active contribution on github so far. Still, the fact that GSoC is highly competitive holds true: For a couple of the projects we had a very hard time finding a suitable candidate because the number of very great proposals was just too high for our number of project slots provided. Unfortunately, this also meant that we had to make tough choices and even dismiss proposals from very strong applications in favor of sometimes just equally strong ones.

To answer your question: In the guidelines of the GSoC selection process we are asked to rate students on their final proposal to evaluate their fit for a project, but it is also highly recommended to take both interaction on our community channels and active contribution to either our code base or to other projects that the student is involved with, into consideration. It is therefore true, that your active contribution played a major part in the ranking of your application. Still, active contribution is just a part of the overall rating of the application which is why we found ourselves deciding between several very similarly strong proposals to the project idea you have applied for. Unfortunately out of these similarly strong proposals we had to pick one single strong application and had to dismiss all other ones, including yours.

Appunni M
@appunnicer13
Apr 24 2016 17:31
Hey I wanted to work on web components, but i felt too much pressure to apply through gsoc , so I waited until this Gsoc selection is over . I want to work without pressure , if any mentor is there who can guide me . i will be able to make contribution to web components
Sebastian Wilzbach
@wilzbach
Apr 24 2016 17:32
@appunnicer13 I happy to answer all the questions you have, I guess you already know the biojs3 repo?
Appunni M
@appunnicer13
Apr 24 2016 17:32
I have gone through google polymer and web components
yes , I have gone through cytoscape
I am not understanding how you are planning to convert the existing functionalities to web components
is there any projects i can refer to , to understand how this has been done ?
Sebastian Wilzbach
@wilzbach
Apr 24 2016 17:35
Polymer! = web components
Appunni M
@appunnicer13
Apr 24 2016 17:35
and which among the projects are candidates for web component integrations , are we planning to write code from scratch ?
Sebastian Wilzbach
@wilzbach
Apr 24 2016 17:35
You can map most functionality with html tags
Appunni M
@appunnicer13
Apr 24 2016 17:35
i meant polymer
Sebastian Wilzbach
@wilzbach
Apr 24 2016 17:36
And for everything else you can still expose a Javascript api
Appunni M
@appunnicer13
Apr 24 2016 17:36
but for starters i didn't see any projects written with raw web components either
polymer is a wrapper i suppose
Sebastian Wilzbach
@wilzbach
Apr 24 2016 17:36
We have some sample components in the biojs repo
Yeah I personally don't like it much
Read issue 8 there
It should explain most
Ayush Sharma
@ayusharma
Apr 24 2016 17:37
Could it be helpful http://webcomponents.org/ ?
for the guidelines and other standards .
Appunni M
@appunnicer13
Apr 24 2016 17:38
thanks ayush , i went through those along with polymer
i did make some websites with polymer
last month , so as to get a grip
Miguel Pignatelli
@emepyc
Apr 24 2016 17:41
Hi @ayusharma, I don't expect you to understand or accept the final decision we have made. I can understand your frustration but disclosing a private conversation here is certainly not acceptable. Regarding the final decision about the project you have applied for, we decided to go for the candidate with expertise in areas relevant to the project. This decision was agreed between all the mentors. If you had any doubt about the decision we made, you should have asked us instead of ranting here. I think I've been responsive during the whole process.
Sebastian Wilzbach
@wilzbach
Apr 24 2016 17:42
@appunnicer13 biojs/biojs3#4
I meant issue 4
Appunni M
@appunnicer13
Apr 24 2016 17:42
i saw issue 8 had link to it
Sebastian Wilzbach
@wilzbach
Apr 24 2016 17:43
:)
There are then also links to get simple components, but I hope the show the concept
*they
Appunni M
@appunnicer13
Apr 24 2016 17:44
https://github.com/ebi-uniprot/biojs-vis-sequence so this is done through web components ?
so the plan is to create simple components and build bigger components over it right ?
so I wanted to ask where do I start in case iIwant to cover a basic component to get an idea on how to proceed
I want*
Miguel Pignatelli
@emepyc
Apr 24 2016 17:46
Regarding web components integration, we shouldn't try to start addressing #4, I would prefer to have a more practical approach, like making it easier to use web components in the registry (less requirements on browserify, allow html imports in snippets, etc)
Sebastian Wilzbach
@wilzbach
Apr 24 2016 17:47
Nope it's still written with browserify
@appunnicer13 the plan was to start with simple components to figure out all the issues
Appunni M
@appunnicer13
Apr 24 2016 17:51
i hope I am not disturbing others with my doubt here , is there any other means where i can clarify my doubts ?
I mean platform
Miguel Pignatelli
@emepyc
Apr 24 2016 17:52
@appunnicer13, we should target first support for custom elements, that is the most matured standard
here or our slack channel are the best places to discuss this
also, the final goal is not to have all the biojs as web components, but to allow web components in biojs
Sebastian Wilzbach
@wilzbach
Apr 24 2016 17:54
@emepyc I don't agree - final target is to be like polymer, but for Biological components
Appunni M
@appunnicer13
Apr 24 2016 17:55
So it will be better idea if we take a small component first finish it
we can take the smallest one and reproduce the functionality with polymer
any inputs ?
Appunni M
@appunnicer13
Apr 24 2016 18:00
This message was deleted
Miguel Pignatelli
@emepyc
Apr 24 2016 18:10
@wilzbach, the less restrictions you impose, the better. The developers should choose their preferred dependencies. Even "safe" ones (like browserify a couple of years ago), are becoming obsolete now
Sebastian Wilzbach
@wilzbach
Apr 24 2016 18:11
@emepyc I think we have seen that this lousy coupled approach doesn't work either :/
Miguel Pignatelli
@emepyc
Apr 24 2016 18:12
what makes you think it doesn't work? I think one of the main success points of biojs2 is to go towards that freedom
even now, one of the main criticism I hear about biojs2 is having to use browserify, parcel.li, etc.
many are moving away from browserify, even the msa component uses now webpack IIRC
Appunni M
@appunnicer13
Apr 24 2016 18:23
@emepyc what is the difference between choosing custom elements over polymer integration , since polymer any way is packed with custom elements isn't polymer a broader approach ? even with polymer being a requirement i see a choice for developer to continue with polymer or not
Miguel Pignatelli
@emepyc
Apr 24 2016 18:27
My point is that we should support components written in polymer or using other wc libraries or even just the new apis directly. Biojs shouldn't be opinionated about what deps the developer should use. If we want to create additional libraries on top of web components that is perfectly fine, but we shouldn't restrict users
But as you can see, there is still no consensus in the community about this :-) @wilzbach we should attend the next conf call and discuss this with the rest of the community
Appunni M
@appunnicer13
Apr 24 2016 18:30
so polymer is never a a requirement on developer level , polymer is wrapped on the component , i don't think you require polymer to load the web component , like a developer having the choice of selecting his tools we can have choice of using polymer , polymer is not a requirement on further development its a dependency on the existing components
you can use web components and polymer together and it would still work without any trouble
Sebastian Wilzbach
@wilzbach
Apr 24 2016 18:31
@emepyc yep we shouldn't build something like polymer, but biojs components should have web components as lowest common denominator - otherwise the name biojs is has lost entirely its meaning
i am particularly worried about polymer because it makes integration of non-polymer code hard
Appunni M
@appunnicer13
Apr 24 2016 18:32
how so ?
Sebastian Wilzbach
@wilzbach
Apr 24 2016 18:33
@emepyc I am actually a bit sorry, didn't want to start a huge discussion here and my focus changed a lot over the last year, so I really want to restrict myself from getting involved
@appunnicer13 it makes a lot of assumptions like 2-way data sync that might not be needed at all, but are provided by polymer
of course you can use the tags, but you can't extend them without following the polymer way
Appunni M
@appunnicer13
Apr 24 2016 18:34
you can
consider them like an html tag
Sebastian Wilzbach
@wilzbach
Apr 24 2016 18:35
yep and the thing is they are not
html tags just work :)
Appunni M
@appunnicer13
Apr 24 2016 18:35
but once standards are set , they will be
Sebastian Wilzbach
@wilzbach
Apr 24 2016 18:36
yes, but polymer luckily will never be
don't mix polymer and and web components - they are two entirely different things
Appunni M
@appunnicer13
Apr 24 2016 18:36
I am not I see polymer like a wrapper over web components .
CustomElements
HTMLImports
ShadowDOM
NodeBind
TemplateBinding
WeakMap
observe-js
i think polymer is a wrapper for all of this
so we are wrapping the idea of using polymer and going back to web components ?
Sebastian Wilzbach
@wilzbach
Apr 24 2016 18:40
@appunnicer13 have you communicate with a non-polymer element to a polymer element?
standard should be web components, not sth. the polymer devs define
otherwise we end up with the same problem
Appunni M
@appunnicer13
Apr 24 2016 18:41
i had a problem with scoping when i was using polymer , that's the only problem I faced . I am still a beginner with polymer may be that is the problem
custom elements are a standard now i think . May be we can approach that, i feel that the requirement is custom element at the end of the day
Sebastian Wilzbach
@wilzbach
Apr 24 2016 18:44
I think you are still confusing me
We are all for custom elements, they are the standard - every framework on top might be useful, but is neither needed nor required
Appunni M
@appunnicer13
Apr 24 2016 18:46
okay , so web components is on top of custom elements ?
i have never heard of the " custom elements" before I might have missed some thing somewhere , please tell me which among them are based on which ,
is web components based on custom elements ?
Jessica Jordan
@jessica-jordan
Apr 24 2016 19:05
@appunnicer13 If you have time on upcoming Wednesday, please feel free to join our regular technical call to get into the discussion of BioJS and the integration of web components-based packages.

Hi @/all ,

this is a reminder that our next technical call regarding the development and roadmap on BioJS will be on Wednesday, April 27th at 16:00 GMT.

If you haven't already done so, please add me to your list of Skype contacts and I will be happy to add you to the call: jessicamm12

You can find a summary of our findings and recent developments initiated by the last call on which we can further discuss here:
https://groups.google.com/forum/#!topic/biojavascript/Oh8ndG9JSu0

Looking forward to talk to you soon!

Cheers,
Jessica

Tim Ruffles
@timruffles
Apr 24 2016 19:11
@appunnicer13 custom elements are one part of the web component 'stack'. You can use them on their own. They're also the 'public API' - i.e, it's what consumers of web components actually see.
like most of web-components, it was bit stuck in vendor discussions, but it's moving quite fast now -> https://lists.webkit.org/pipermail/webkit-dev/2016-March/027995.html