Future plans for Biostar
Istvan Albert
In this post I would like to lay out what I believe that the future of
Biostar should be.
For reference let me mention that the StackExchange company has
offered to take over Biostar and fold it into the StackExchange 2.0
platform (SE2). This solution has the advantage of freeing us from all
administrative responsibilities for Biostar at the same time the
company will take control over presentation, organization and inner
workings of the site. The content would still be free to distribute,
but user data would not.
Everyone that I have interacted with at SE2 were nice an
accommodating. I want to mention that I have a lot of respect for the
founders of the company, having been an avid reader of both 'Coding
Horror' as well as 'Joel on Software' for possibly more than a
decade. I do believe that individuals at SE2 would try to accommodate
the needs of our community as much as they could.
That being said there are many negative aspects of switching to SE2.
One is that it is irreversible, we don't know what the future holds,
how fortunes and fates change. While the data is openly accessible the
login information is not, therefore one can not recreate the same
experience from the downloadable data. At the same time we have to
recognize that features, layout and functioning of SE2 is
fundamentally shaped by a non-scientific community. And those needs
don't necessarily align with that of a scientific community.
In general I feel uneasy about the entire model of one company ruling
them all. Like Facebook or Twitter, SE seems aim to own all Q&A - and
that is something I feel it leads to a diluted experience for
everybody. There is nothing you and I can do about Facebook - it is
what it is - and is going from bad to worse - but you know what? there
could be something that you and I can do about Q&A in bioinformatics.
Instead of watching what other people do we could be shaping it
exactly the way it should be like, adding features and functionality
that best serves education and research.
The more I thought about I realized that I want to be part of the
experience of building a new environment that goes well beyond what
SE2 can do.
In the past two weeks I have put a tremendous amount of effort into
the code base that we worked on for about two months in the summer. I
believe that I have currently brought it to a level of quality that is
a superior alternative to switching to SE2. It is a simple, open,
python based platform, with zero installation (other than python) that
allows anyone to run their own version of the Biostar Q&A site in five
minutes or less.
Here is the current test site with a reduced data set (with the first
300 users/500 posts):
The system will recognize major OpenID providers and should merge your
account as long as your email is sent with the login. The official
codebase and documentation can be found at:
https://github.com/ialbert/biostar-central
Notable changes:
There are several major features that I have added, and my mind is
brimming with ideas. For example comments are full posts: these are
searchable and editable. See your user page for links to all your
posts. Moderators will be able to reclassify a post to become an
answer or comment as needed.
Behind the scenes a hierarchical data structure is in place that will
allow comments to be threaded (in a Reddit like model). Therefore the
flow of discussions will be greatly improved over the current model
that is strictly temporal.
A private notification system (not even admins can see it) keeps track
of user activity on posts you asked or contributed to. You can see
these notifcations on your account page and you can have individual
private RSS feed to your messages. Plust now we can implement any type
of RSS feed that is necessary.
Important
*********
Depending on the outcome of the public test I'd like to move to the
new platform starting December 9, 2011
So please join me in building this new endeavor with any effort: spot
a typo, report a broken link, come up with a new badge, suggest a new
feature, contribute documentation, css styling, code --- or just be
excited about it! It is going to be great!
In fact there is rare Gold badge on the line: Beta Tester - and it
will be automatically awarded, don't let that badge slip away ;-)
Chris Miller
two weeks? Should we feel free to post test questions and answers,
and generally kick the tires to try to shake out any bugs?
-Chris
Istvan Albert
> To be clear, are we treating this new site as a sandbox for the next
> two weeks? Should we feel free to post test questions and answers,
> and generally kick the tires to try to shake out any bugs?
yes, absolutely,
occasional wipes, resets and the site being periodically offline will
probably also occur
(I'll try to save the badge info all throughout, that's going to be
the data recovery test)
best,
Istvan
Brad Chapman
> The more I thought about I realized that I want to be part of the
> experience of building a new environment that goes well beyond what
> SE2 can do.
Has the final decision been made, or are we still open for discussion?
To be honest, I'm surprised by this change in direction; the community
responded overwhelmingly in favor of moving to StackExchange 2.
From my view, the major benefit is moving to a developed, widely used,
Q/A platform in SE2. While I don't doubt your coding abilities and the
strength of the code base, it is very difficult to sustain momentum
long term with a small contributor community. Projects, especially
side projects not directly related to research, can easily stagnate as
the initial excitement and interest wanes. Your biostar-central code
base already shows this pattern; it was well developed earlier this
year but has been lying dormant for 6 months.
Building good communities that can sustain themselves long term is a
hard, unsolved, problem. My vote is to focus our energy there and
build off the good work the StackExchange developers have already done
on the infrastructure.
Brad
Peter Cock
On Monday, November 28, 2011, Brad Chapman <chap...@50mail.com> wrote:
> Istvan;
>
>> The more I thought about I realized that I want to be part of the
>> experience of building a new environment that goes well beyond what
>> SE2 can do.
>
> Has the final decision been made, or are we still open for discussion?
> To be honest, I'm surprised by this change in direction; the community
> responded overwhelmingly in favor of moving to StackExchange 2.
transferring user's posts etc copyright from the user to SE.
> From my view, the major benefit is moving to a developed, widely used,
> Q/A platform in SE2. While I don't doubt your coding abilities and the
> strength of the code base, it is very difficult to sustain momentum
> long term with a small contributor community. Projects, especially
> side projects not directly related to research, can easily stagnate as
> the initial excitement and interest wanes. Your biostar-central code
> base already shows this pattern; it was well developed earlier this
> year but has been lying dormant for 6 months.
>
> Building good communities that can sustain themselves long term is a
> hard, unsolved, problem. My vote is to focus our energy there and
> build off the good work the StackExchange developers have already done
> on the infrastructure.
> Brad
would this code base be useful outside BioStar for other
Q&A communities wary of the closed SE model? If so,
there is scope for a wider development community.
It won't happen overnight though.
Peter
Sean Davis
Simon
for the community when I don't feel that views have been adequately
canvassed, or, if expressed, fully taken into account. It seemed to me
that the weight of opinion was in favour of the move to SE. Why are
these valid views from senior members of this community being ignored?
I think the cross-pollination effect of being part of the larger Stack
Exchange community would be really valuable to BioStar. I also think
the hard-working members of this community, who give freely and
generously of their spare time should not be asked to beta-test
software that addresses a problem I would consider solved (the
community building is the hard part, not the software that supports
the community). Whether a badge is on offer or not. I thought
bioinformaticians had moved beyond constantly reinventing the wheel.
It appears not in this case.
> But not unanimously - some of us balked at the idea of
> transferring user's posts etc copyright from the user to SE.
Do I really 'own' my content any way, if Istvan can decide
unilaterally what is going to happen to it (within the framework in
which it currently exists)? There is no option open to me, as a member
of the community, to take my content and offer it to Stack Exchange
(where it would be CC-BY anyway).
It is unlikely that a unanimous decision is going to be reached on
this one, so surely the majority opinion of the community should be
followed here.
> On the other hand (while I accept Brad's concerns are valid),
> would this code base be useful outside BioStar for other
> Q&A communities wary of the closed SE model? If so,
> there is scope for a wider development community.
> It won't happen overnight though.
OSQA already exists (http://www.osqa.net/) and has a modicum of
traction. Why do we need more than one open-source implementation of a
Q&A engine for the community to arrange themselves around?
At the end of the day, as Brad mentioned, this is a side project for
those involved, attention will wax and wane, depending on whether it
is grant season, or the time of year for teaching, and according to a
thousand other factors and distractions. At least with Stack Exchange
the core fundamentals of the engine would be taken care of by a
company whose existence depends on that technology working for the
members of their community. The content (which is what is paramount
here, I couldn't give 2 hoots about my "reputation") would be covered
by an open licence, guaranteeing its availability. I sincerely urge a
rethink of this strategy.
SJC
Leonor Palmeira
this is really great news and I am fully supportive of the effort
you've put in this development!
I feel this is really the way to go for the community at large, and am
glad you chose this path :-)
How can we contribute?
Best,
Leonor.
Michael Kuhn
in were in favor of moving to SE2. As Simon said, if Istvan is
unilaterally deciding to decline the SE2 offer, then it doesn't make
any difference for the community if it's the SE network or Istvan who
call the shots. Therefore, as I've stated before, I urge Istvan to
conduct an election, to determine the will of the community. (I
proposed to either have a reputation cut-off for voting, or weight
votes by reputation.)
Being part of the SE network, along with StackOverflow and, soon,
Biology, would make for easy shifting of questions between the sites.
A community-driven site should be trust-worthy: I rather put my time
and effort into a proven solution that is someone's main focus, than a
side-project.
On Twitter, Egon Willighagen suggested "bioinformatics / biology
extensions... auto-text mining, auto tagging of protein, DNA, ..." as
potential advantages for a home-brew solution. (*) Of course, these
are just examples, but I think the capabilities of the SE network for
Q&A are just fine. Istvan wants to make something "well beyond what
SE2 can do" -- but it this really what the community wants? I think
it's better to stick with a simple model, than to attempt to do too
many things at once.
~Michael
* http://twitter.com/egonwillighagen/status/141060408801902592
Daniel Swan
I was under the impression that by and large the group here was in
favour of moving to SE2, therefore I am disappointed that this
discussion has recurred. I thought we had reached a consensus opinion,
so I would like to lend my support to the views expressed by Simon and
Michael.
regards,
Dan
--
Daniel Swan || dan.swan[at]gmail.com || http://eridanus.net/ ||
http://twitter.com/d_swan
"Reality is that which, when you stop believing in it, doesn't go away."
(Philip K. Dick - How to Build a Universe)
Casey Bergman
Best regards,
Casey
Istvan Albert
thanks for everyone's input. I have read them all but I am unable to
respond to everyone in turn.
First and foremost there is a valid concern as to whether my interest
in the endeavor will run out and we're left with nothing. That's not
going to happen. I won't list all my web contributions that I wrote,
there are many and varied, affecting probably tens of thousands of
people - from my first job ever of porting the entire Movielens
recommendation system from perl to Java (Movielens
http://www.movielens.com), to the LionDB data storage system
(http://liondb.org) to Galaxy that I initially designed and wrote with
the help of James Taylor (http://www.usegalayx.org/) - I am almost
certain that the majority of you have already ran services on code or
frameworks that I have originally written or designed. I an very
devoted towards creating sites that stand the test of time in every
aspect, from installation, maintenance and usability perspectives.
What I want to create is not a simple SE clone like all others out
there, but a site that is specifically designed for disseminating
science and research.
I do think that most of you that advocate going to SE are little bit
overly optimistic in how that will turn out. Or perhaps I am a bit
pessimistic. But it is a valid choice and I will fully support your
choice of doing anything that you wish with your own content!
Not only that, I want to create the software that will allow anyone to
easily (!) run their own content in any way they see fit! And that is
the *real choice* and *real freedom* - one that you will never have
again once you go to SE. I think that SE gives you the illusion of
freedom - just as facebook is free - until the "bills" come!
Peter Cock reminded me that we must to follow some rules when
disseminating content. First there will be a datadump that contains
all the posts that were added after the licensing change took place.
For posts that were created before this time there will be an option
on the new site that asks for the user to opt into re-licensing their
content. Finally on the new site there will be an opt-in option that
would ask the user if they want their personally identifiable
information included in the datadump (email or OpenID). Having this
information would allow anyone to create an exact replica of BioStar
(I want to point out again that this will never be allowed under the
new SE licensing! You would never be able to run another complete Q&A
site off of SE data dump!).
An important detail I want to mention here: BioStar started out as a
private site run by a single individual that owns everything on the
site. As the owner I could have already monetarily profited in many
ways: running ads, spinning it off to a private and closed software
platform (etc), and I had a few requests to do so). Maybe you don't
realize this but just about any other bioinformatics site makes money
off of you any time you visit. You put an answer up there, and with
that you just put money in someone's pocket, not only that you don't
really have an option of taking your content out and running it on
your own.
That something that I am profoundly against, and everything that I
propose above is to cut this problem off at its head. I want to make
money off of BioStar - but not as the owner but as someone that
demonstrates their skills via BioStar - and everyone else should be
able to do just that.
I would hate to see anyone go but I will help everyone individually or
as a group to do so.
Also I want to apologize to anyone that feels that events took a turn
they did not expect. I don't think everyone will agree and if you
don't agree please consider what I wrote above as an explanation of
why I am choosing this path.
I absolutely agree that that everyone that contributed to the site
should have a say in what happens with the content - but if one's
choice is that I someone else has to give up some of their rights,
then voting can not be the way to decide the matter. This has nothing
to do with democracy. Thankfully, I might say so. One can make bad
decisions in a democratic way, and when that decision is irreversible
then I think we need to rethink it.
Please consider that perhaps real freedom of content is more important
than certain type of conveniences.
best regards,
--
Istvan Albert
Associate Professor, Bioinformatics
Pennsylvania State University
http://www.personal.psu.edu/iua1/
Istvan Albert
ouch, make that: http://movielens.org
Michael Schubert
I agree with copyright transfer on the SE2 option being a major point
of concern, if not outright impossible due to every Biostar member
that ever posted a question, answer, or comment would have to opt-in.
Peter raised this point very well on the last discussion.
In your explanation, you stress the point of easily setting up an own
instance of the QA platform. While this looks good on paper and the
data being freely available (as in cc-by-sa) is crucial, it's also a
point of concern: dilution of content and users might damage the
community more than it helps, which in combination with your soloist
decision is unfortunately a possibility. Please provide details of
your negotiations with Stackexchange and why you now think it's a
definite no-go (compared to an open mind before). I have the
impression that the moderator election was a major point for you,
whose importance I disagree with.
The other thing is long-term development and maintainance. I see that
you (and maybe some others) are committed to doing this now, and I see
that you have experience in getting projects started. It might work.
The point is, it might also be abandoned due to a lack of time from
people whose primary interest becomes something else. I am not happy
to take this chance.
Best,
Michael
Brent Pedersen
As I seem to be in the minority, I'll add my 0.02 and write out what I
understand about the discussion...
I was/am hesitant to move to the SE2 platform due to licensing and
ownership issues.
On quick look, I think that the new site looks great and gives anyone
interested the
opportunity to submit changes/improvements.
It seems that one major source of disappointment for those opposed is
the ability to
integrate better with other SE2 sites? Is there any way to improve
that in Istvan's code-base?
Another concern was that Istvan might lose interest in the project. To
me, it doesn't seem that this is more of an issue than having BioStar
on stackexchange where we have no guarantee that they'll continue
making improvements and we're limited to the
functionality they decide to provide. Do I misunderstand the concern here?
The test site seems very functional to me on a brief look. Is there
some missing functionality?
thanks,
-Brent
Istvan Albert
> decision is unfortunately a possibility. Please provide details of
> your negotiations with Stackexchange and why you now think it's a
> definite no-go (compared to an open mind before).
Hi Michael,
The examples are numerous. For example here is what SE considers
attribution-required policy:
http://blog.stackoverflow.com/2010/08/defending-attribution-required/
Now consider just how many hoops you would need to jump through to
reuse the content published on a SE site.
We can all see that SE mandates that all content created by a third
party has to be attributed as theirs. You have to attribute it to SE
first then to the author second. Moreover you have to attribute it to
the main domain, provide a direct link and they even regulate what the
link may or may not contain.
Now let's think about it, is it even possible to create a site with SE
content? I don't think so - basically all your content would have to
be doubly linked to SE.
This is not what I would call an open data distribution policy - and
in fact we are going the opposite way in scientific publishing.
In my opinion the only person any content must be attributed to should
be the person(s) that created the content - and that's it! How one
chooses to cite this person is up to them, could be Biostar, or a blog
or gist, github code etc,
best,
Istvan
jestevez
> We can all see that SE mandates that all content created by a third
> party has to be attributed as theirs. You have to attribute it to SE
> first then to the author second. Moreover you have to attribute it to
> the main domain, provide a direct link and they even regulate what the
> link may or may not contain.
IANAL, but the text of the license [1] regarding attribution seems
reasonably
clear:
1. The name of the author and the site/journal
2. The title of the work
3. The URI, "to the extent reasonably practicable"
Perhaps I misunderstand the nature of your objection to the CC-wiki
license.
However, given that the license is perpetual, the concept of
'ownership' doesn't seem
particularly applicable.
Best regards,
James Estevez
[1]: http://creativecommons.org/licenses/by-sa/3.0/legalcode
Mary Mangan
Is there a place to collect bug reports? I have run into some differences.
Istvan Albert
> IANAL, but the text of the license [1] regarding attribution seems
> reasonably
> clear:
>
> 1. The name of the author and the site/journal
> 2. The title of the work
> 3. The URI, "to the extent reasonably practicable"
Don't you think that the SE attribution rules are substantially more stringent?
There are rules that dictate how you are linking and where you must
link to, You have to link to question, you have to link to author
profile, visually indicate that the content comes from SE, you may not
shorten the link etc.
Here is an example, take some content that comes from an SE site, does
linking to it via a reference like this [1] (with the references at
the end) visually indicate that the content comes from SE? How about
the other links, where do those go?
Also don't you think an answer to a Biostar question should be
governed by a little less stringent rules than a scientific
publication? Rules that allow people to actually easily reuse this
content without tying them up with all kinds regulations.
best,
Istvan
Istvan Albert
> I just love to test software.
Sounds like a dream to any software developer.
> Is there a place to collect bug reports? I have run into some differences.
We collect issues here:
https://github.com/ialbert/biostar-central/issues
thanks a lot,
Istvan
neilfws
James Estevez
On Nov 28, 12:54 pm, Istvan Albert <i...@psu.edu> wrote:
>Don't you think that the SE attribution rules are substantially more stringent?
>There are rules that dictate how you are linking and where you must
>link to, You have to link to question, you have to link to author
>profile, visually indicate that the content comes from SE, you may not
>shorten the link etc.
I don't agree that the attribution rules of the _licence_ are
substantially more
stringent than any other CC-BY-SA website (e.g. Wikipedia [1]). SE is,
of
course, free to specify that any attribution must be in 24 point Times
New
Roman with specified `rel` and `title` link relations, just as the
licensee is free to
ignore those 'requirements' and comply with the letter and spirit of
CC-BY-SA.
Again, I don't agree that assignment of copyright is an issue. Indeed,
SE is
free to seed any SE2 Biology site with Biostar content [2], just as we
are
free to scrape and import any question tagged with `[bioinformatics]`
from
stats.stackexchange.com. The content is free to move in either
direction with a
statement of attribution (probably less than 250 characters in length)
in
perpetuity precisely because of CC-BY-SA, so I don't see how there are
any
'rights' to be lost.
In any event, if I understood your earlier email correctly, the issue
is moot. I'm
happy to help any way I can.
Best regards,
James Estevez
[1]: http://en.wikipedia.org/wiki/Wikipedia:Reusing_Wikipedia_content#Re-use_of_text_under_Creative_Commons_Attribute_Share-Alike
[2]: https://twitter.com/#!/codinghorror/status/50168599121362945
Michael Kuhn
Twitter:
Khader: "I prefer SE2 migration. Thought the #BioStar community
reached a consensus to embrace SE2 irrespective of its restrictions/
issues." https://twitter.com/#!/kshameer/status/141272640357801984
Pierre: "as long as the platform runs as before, I don't see any
problem" https://twitter.com/#!/yokofakun/status/141171670802108416
~m
Egon Willighagen
On Sun, Nov 27, 2011 at 8:24 PM, Istvan Albert <istvan...@gmail.com> wrote:
> Everyone that I have interacted with at SE2 were nice an
> accommodating. I want to mention that I have a lot of respect for the
> founders of the company, having been an avid reader of both 'Coding
> Horror' as well as 'Joel on Software' for possibly more than a
> decade. I do believe that individuals at SE2 would try to accommodate
> the needs of our community as much as they could.
The Blue Obelisk eXchange used to be on SE1. We were not as big as
BioStar and still aren't. SE was not interesting to them, and the cost
was immense. So, we moved to the Shapado platform. Not perfect, but
free, Open, and a well-defined way to move to an alternative.
I was surprised by this email, like many others, but unlike many it
seems, I welcome the suggested route. Choice is good, and a custom
platform provides more choice. Yes, it surely comes at the expense
that someone needs to maintain the platform, but if someone is
offering to do that, who am I to question that?
I have been one of the BioStar users who were not convinced yet about
a move to SE2, and outed the obstacle that every single user would
have to reassign copyright to SE. We all know how well
copyright-assignment to publishers has worked out for science, so kind
of shocked that few people shared that concern.
Another aspect is that with the combined python expertise in the
BioStar community, I can see Istvan's envisioned future of a more
platform that ties into science dissemination come true... I don't see
SE ever be interested in that.
> The more I thought about I realized that I want to be part of the
> experience of building a new environment that goes well beyond what
> SE2 can do.
Indeed. How long have many of us wished to FriendFeed icons for
science related content feeds, like CiteULike, Mendeley... that would
be a trivial change any of us could make. But, the platform did not
allow us, despite requests with FriendFeed developers... whether you
like it or not, there is no money in science (unless you have a
lock-in as with publishing in 'journals').
> Important
> *********
>
> Depending on the outcome of the public test I'd like to move to the
> new platform starting December 9, 2011
That is pretty soon, and I suggest to take a bit more time...
> In fact there is rare Gold badge on the line: Beta Tester - and it
> will be automatically awarded, don't let that badge slip away ;-)
One bug report I have is that logging in with my OpenID account
created a *new* account, rather than logging in to my previous
account. So, I now got a split identity on BioStar beta:
http://test.biostars.org/user/show/1/
http://test.biostars.org/user/show/71/
Egon
--
Dr E.L. Willighagen
Postdoctoral Researcher
Institutet för miljömedicin
Karolinska Institutet (http://ki.se/imm)
Homepage: http://egonw.github.com/
LinkedIn: http://se.linkedin.com/in/egonw
Blog: http://chem-bla-ics.blogspot.com/
PubList: http://www.citeulike.org/user/egonw/tag/papers