gmod-ajax — JBrowse list

[Gmod-ajax] Similie

[Aanensen, D. <d.a...@im...>]

Hi,
=20
have you by any chance seen MIT's Similie project -   Timeline widget?=20
=20
http://simile.mit.edu/timeline/
=20
It struck me that this could be utilised in a gmod-ajax type fashion but =
with the timescales converted to sequence co-ordinates for displaying =
sequence annotation.
=20
this is a very quick (and apologetically shoddy)  thrown-together =
example that I have done for parsing an EMBL file into XML and using the =
'days' in the timeline to display CDS positions based on annotated =
sequence start and end points -=20
=20
http://www.mlst.net/chr/chr.html
=20
click on the gene to view homolog details from a database (probably best =
to ignore the details!).=20
=20
the scale labels  need changing and the iframe at the bottom needs =
replacing with a DIV and results AJAX'd in but you can see the kind of =
thing I mean.
=20
The scales can quite easily be changed, various levels added and custom =
graphics added (bioperl glyphs for eg) but I don't know how well it =
would scale.
=20
one of their examples is here - =
http://simile.mit.edu/timeline/examples/dinosaurs/dinosaurs2.html=20
=20
all the best

David
=20
=20

________________________________

From: gmo...@li... on behalf of Steve Taylor
Sent: Fri 01/06/2007 14:11
To: Mitch Skinner
Cc: gmo...@li...
Subject: Re: [Gmod-ajax] Openlazlo



Hi,

>
>>perhaps a Google Maps-type widget is lurking out there that's
>>exactly what we need (though I doubt any available toolkit can do the
>>scary things that Mitch is currently trying with tracks built entirely
>>out of DIVs)
>
>
> I've been experimenting with this over the last few weeks, which is =
part
> of the reason I've been so quiet lately.  This is still in an =
extremely
> sketchy state, but there's enough there to get the idea:
> http://genome.biowiki.org/test/divbrowser/genome2.html
>
> The features are exons from D. melanogaster chromosome 3R.  There are
> about 16,000, of which Firefox is able to show a few thousand on my
> hardware before it starts to bog down.  If you zoom out beyond about 3
> megabases the zooming gets to be quite slow, but that's above the =
point
> where we're currently switching to a feature density histogram for =
that
> track, anyway.
>
> There's still some low-hanging fruit performance-wise, so I'm =
optimistic
> that it'll scale to a full set of tracks, especially if we're smart
> about interval search and attaching and detaching things from the DOM.
> And adjusting the threshold where we switch to a density histogram
> should accommodate hardware differences.
>
> The rationale for using divs is:
> 1. reduces server side processing/storage for image tiles
> 2. allows for a nice client-side feature creation/editing UI
> 3. zooming is cool, but it's more than just eye candy--it helps you =
keep
> an intuitive sense of where you are
> 4. client can cache an entire chromosome worth of features (at least =
for
> some tracks)--given where firefox and WHAT-WG are going this may =
enable
> offline use in the future.
> 5. more cross-browser than SVG (I've been testing in firefox and IE =
6/7
> and I'll be trying for Safari as well)
>

Hey that is really cool! Reminds me of the Integrated Genome Browser =
zooming functionality.

I would be surprised if OpenLazlo could do the hardcore stuff but was =
wondering if it would be worth looking into using the components for =
other parts of the interface such as popups, menus, search
forms, sidebars, annotation editing etc.

Cheers,

Steve
------------------------------------------------------------------
Medical Sciences Division
Weatherall Institute of Molecular Medicine/Sir William Dunn School
Oxford University

-------------------------------------------------------------------------=

This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/
_______________________________________________
Gmod-ajax mailing list
Gmo...@li...
https://lists.sourceforge.net/lists/listinfo/gmod-ajax

Re: [Gmod-ajax] Openlazlo

[Michele C. <mc...@br...>]

All right then I'll bite and show my ignorance.

What's a div?

Nice demo though - I like the swoopy zooming.

On May 31, 2007, at 4:15 PM, Mitch Skinner wrote:

> On Thu, 2007-05-31 at 11:13 -0700, Ian Holmes wrote:
>> perhaps a Google Maps-type widget is lurking out there that's
>> exactly what we need (though I doubt any available toolkit can do the
>> scary things that Mitch is currently trying with tracks built  
>> entirely
>> out of DIVs)
>
> I've been experimenting with this over the last few weeks, which is  
> part
> of the reason I've been so quiet lately.  This is still in an  
> extremely
> sketchy state, but there's enough there to get the idea:
> http://genome.biowiki.org/test/divbrowser/genome2.html
>
> The features are exons from D. melanogaster chromosome 3R.  There are
> about 16,000, of which Firefox is able to show a few thousand on my
> hardware before it starts to bog down.  If you zoom out beyond about 3
> megabases the zooming gets to be quite slow, but that's above the  
> point
> where we're currently switching to a feature density histogram for  
> that
> track, anyway.
>
> There's still some low-hanging fruit performance-wise, so I'm  
> optimistic
> that it'll scale to a full set of tracks, especially if we're smart
> about interval search and attaching and detaching things from the DOM.
> And adjusting the threshold where we switch to a density histogram
> should accommodate hardware differences.
>
> The rationale for using divs is:
> 1. reduces server side processing/storage for image tiles
> 2. allows for a nice client-side feature creation/editing UI
> 3. zooming is cool, but it's more than just eye candy--it helps you  
> keep
> an intuitive sense of where you are
> 4. client can cache an entire chromosome worth of features (at  
> least for
> some tracks)--given where firefox and WHAT-WG are going this may  
> enable
> offline use in the future.
> 5. more cross-browser than SVG (I've been testing in firefox and IE  
> 6/7
> and I'll be trying for Safari as well)
>
> Mitch
>
>
> ---------------------------------------------------------------------- 
> ---
> This SF.net email is sponsored by DB2 Express
> Download DB2 Express C - the FREE version of DB2 express and take
> control of your XML. No limits. Just data. Click to get it now.
> http://sourceforge.net/powerbar/db2/
> _______________________________________________
> Gmod-ajax mailing list
> Gmo...@li...
> https://lists.sourceforge.net/lists/listinfo/gmod-ajax

Re: [Gmod-ajax] Openlazlo

[Steve T. <st...@mo...>]

Hi,

> 
>>perhaps a Google Maps-type widget is lurking out there that's 
>>exactly what we need (though I doubt any available toolkit can do the 
>>scary things that Mitch is currently trying with tracks built entirely 
>>out of DIVs)
> 
> 
> I've been experimenting with this over the last few weeks, which is part
> of the reason I've been so quiet lately.  This is still in an extremely
> sketchy state, but there's enough there to get the idea:
> http://genome.biowiki.org/test/divbrowser/genome2.html
> 
> The features are exons from D. melanogaster chromosome 3R.  There are
> about 16,000, of which Firefox is able to show a few thousand on my
> hardware before it starts to bog down.  If you zoom out beyond about 3
> megabases the zooming gets to be quite slow, but that's above the point
> where we're currently switching to a feature density histogram for that
> track, anyway.
> 
> There's still some low-hanging fruit performance-wise, so I'm optimistic
> that it'll scale to a full set of tracks, especially if we're smart
> about interval search and attaching and detaching things from the DOM.
> And adjusting the threshold where we switch to a density histogram
> should accommodate hardware differences.
> 
> The rationale for using divs is:
> 1. reduces server side processing/storage for image tiles
> 2. allows for a nice client-side feature creation/editing UI
> 3. zooming is cool, but it's more than just eye candy--it helps you keep
> an intuitive sense of where you are
> 4. client can cache an entire chromosome worth of features (at least for
> some tracks)--given where firefox and WHAT-WG are going this may enable
> offline use in the future.
> 5. more cross-browser than SVG (I've been testing in firefox and IE 6/7
> and I'll be trying for Safari as well)
> 

Hey that is really cool! Reminds me of the Integrated Genome Browser zooming functionality.

I would be surprised if OpenLazlo could do the hardcore stuff but was wondering if it would be worth looking into using the components for other parts of the interface such as popups, menus, search 
forms, sidebars, annotation editing etc.

Cheers,

Steve
------------------------------------------------------------------
Medical Sciences Division
Weatherall Institute of Molecular Medicine/Sir William Dunn School
Oxford University

Re: [Gmod-ajax] Openlazlo

[Mitch S. <mit...@be...>]

On Thu, 2007-05-31 at 11:13 -0700, Ian Holmes wrote:
> perhaps a Google Maps-type widget is lurking out there that's 
> exactly what we need (though I doubt any available toolkit can do the 
> scary things that Mitch is currently trying with tracks built entirely 
> out of DIVs)

I've been experimenting with this over the last few weeks, which is part
of the reason I've been so quiet lately.  This is still in an extremely
sketchy state, but there's enough there to get the idea:
http://genome.biowiki.org/test/divbrowser/genome2.html

The features are exons from D. melanogaster chromosome 3R.  There are
about 16,000, of which Firefox is able to show a few thousand on my
hardware before it starts to bog down.  If you zoom out beyond about 3
megabases the zooming gets to be quite slow, but that's above the point
where we're currently switching to a feature density histogram for that
track, anyway.

There's still some low-hanging fruit performance-wise, so I'm optimistic
that it'll scale to a full set of tracks, especially if we're smart
about interval search and attaching and detaching things from the DOM.
And adjusting the threshold where we switch to a density histogram
should accommodate hardware differences.

The rationale for using divs is:
1. reduces server side processing/storage for image tiles
2. allows for a nice client-side feature creation/editing UI
3. zooming is cool, but it's more than just eye candy--it helps you keep
an intuitive sense of where you are
4. client can cache an entire chromosome worth of features (at least for
some tracks)--given where firefox and WHAT-WG are going this may enable
offline use in the future.
5. more cross-browser than SVG (I've been testing in firefox and IE 6/7
and I'll be trying for Safari as well)

Mitch

Re: [Gmod-ajax] Openlazlo

[Ian H. <ih...@be...>]

I did look at OpenLaszlo very cursorily a few months back and concluded 
that we were doing too much custom stuff with the UI to take much 
advantage of it. Same as with Google Web Toolkit. I don't think that a 
genome browser breaks down neatly into the standard set of widgets that 
are used for most Ajax apps. But that is not a very well-informed 
opinion; perhaps a Google Maps-type widget is lurking out there that's 
exactly what we need (though I doubt any available toolkit can do the 
scary things that Mitch is currently trying with tracks built entirely 
out of DIVs)


I.


Steve Taylor wrote:
> Hi,
> 
> Has anyone seen this? Basically XML and Javascript that's converted to Flash/DHTML, so may ultimately solve the dreaded browser incompatibility problems.
> 
> http://www.openlaszlo.org/
> 
> There are some very cool demos and I have seen one bioinformatics app so far (http://www.chromhome.org/chromhome/)
> 

Re: [Gmod-ajax] Openlazlo

[Seth C. <sjc...@lb...>]

I took a look at this recently and there still seems to be some real
compatibility issues, especially if you don't want to use the Flash
engine (apparently the DHTML stuff is very weak). It looks like it still
depends on whether or not you're in the Flash is evil camp.

-Seth


On Thu, 2007-05-31 at 17:20 +0100, Steve Taylor wrote:
> Hi,
> 
> Has anyone seen this? Basically XML and Javascript that's converted to Flash/DHTML, so may ultimately solve the dreaded browser incompatibility problems.
> 
> http://www.openlaszlo.org/
> 
> There are some very cool demos and I have seen one bioinformatics app so far (http://www.chromhome.org/chromhome/)
> 

Re: [Gmod-ajax] Openlazlo

[Eric J. <e-...@no...>]

Hi Michele,
> The question is - is it a toy  or can you do really fancy things with
> it?
>
> All the demos I've seen aren't really doing any heavy lifting.
>

I saw an email program (similar to outlook) with all the drag and
drop/multiple mailbox functionality that looked really great. It was
written by the company that currently maintains OpenLaszlo (of course
they didn't show the code!)  But for me the prospect of what could be
done was pretty amazing.  They also seemed like they were really
trying to get people to use it, so they might be interested in demoing
some stuff privately.

Re: [Gmod-ajax] Openlazlo

[Michele C. <mc...@br...>]

On May 31, 2007, at 12:20 PM, Steve Taylor wrote:

> Hi,
>
> Has anyone seen this? Basically XML and Javascript that's converted  
> to Flash/DHTML, so may ultimately solve the dreaded browser  
> incompatibility problems.
>
> http://www.openlaszlo.org/
>
> There are some very cool demos and I have seen one bioinformatics  
> app so far (http://www.chromhome.org/chromhome/)


Yeah saw this this morning and it does look very cool.

The question is - is it a toy  or can you do really fancy things with  
it?

All the demos I've seen aren't really doing any heavy lifting.

Re: [Gmod-ajax] Openlazlo

[Eric J. <e-...@no...>]

I saw a live demo of this at a local meeting.  It was very impressive
indeed.  It looks like the real deal to me and they have a number of
production quality apps written with it.  They did some simple
examples and apart from having to write 'code' in XML it even seemed
relatively straightforward (albeit for a simple app).

On 5/31/07, Hilmar Lapp <hl...@gm...> wrote:
> Are they still using XUL? Interesting - there was some hype 3 years
> ago around the company, but apparently the closed-source business
> concept didn't work out?
>
>         -hilmar
>
> On May 31, 2007, at 12:20 PM, Steve Taylor wrote:
>
> > Hi,
> >
> > Has anyone seen this? Basically XML and Javascript that's converted
> > to Flash/DHTML, so may ultimately solve the dreaded browser
> > incompatibility problems.
> >
> > http://www.openlaszlo.org/
> >
> > There are some very cool demos and I have seen one bioinformatics
> > app so far (http://www.chromhome.org/chromhome/)
> >
> > --
> > Steve
> > ------------------------------------------------------------------
> > Medical Sciences Division
> > Weatherall Institute of Molecular Medicine/Sir William Dunn School
> > Oxford University
> >
> > ----------------------------------------------------------------------
> > ---
> > This SF.net email is sponsored by DB2 Express
> > Download DB2 Express C - the FREE version of DB2 express and take
> > control of your XML. No limits. Just data. Click to get it now.
> > http://sourceforge.net/powerbar/db2/
> > _______________________________________________
> > Gmod-ajax mailing list
> > Gmo...@li...
> > https://lists.sourceforge.net/lists/listinfo/gmod-ajax
>
> --
> ===========================================================
> : Hilmar Lapp  -:-  Durham, NC  -:-  hlapp at gmx dot net :
> ===========================================================
>
>
>
>
>
>
> -------------------------------------------------------------------------
> This SF.net email is sponsored by DB2 Express
> Download DB2 Express C - the FREE version of DB2 express and take
> control of your XML. No limits. Just data. Click to get it now.
> http://sourceforge.net/powerbar/db2/
> _______________________________________________
> Gmod-ajax mailing list
> Gmo...@li...
> https://lists.sourceforge.net/lists/listinfo/gmod-ajax
>

Re: [Gmod-ajax] Openlazlo

[Hilmar L. <hl...@gm...>]

Are they still using XUL? Interesting - there was some hype 3 years  
ago around the company, but apparently the closed-source business  
concept didn't work out?

	-hilmar

On May 31, 2007, at 12:20 PM, Steve Taylor wrote:

> Hi,
>
> Has anyone seen this? Basically XML and Javascript that's converted  
> to Flash/DHTML, so may ultimately solve the dreaded browser  
> incompatibility problems.
>
> http://www.openlaszlo.org/
>
> There are some very cool demos and I have seen one bioinformatics  
> app so far (http://www.chromhome.org/chromhome/)
>
> -- 
> Steve
> ------------------------------------------------------------------
> Medical Sciences Division
> Weatherall Institute of Molecular Medicine/Sir William Dunn School
> Oxford University
>
> ---------------------------------------------------------------------- 
> ---
> This SF.net email is sponsored by DB2 Express
> Download DB2 Express C - the FREE version of DB2 express and take
> control of your XML. No limits. Just data. Click to get it now.
> http://sourceforge.net/powerbar/db2/
> _______________________________________________
> Gmod-ajax mailing list
> Gmo...@li...
> https://lists.sourceforge.net/lists/listinfo/gmod-ajax

-- 
===========================================================
: Hilmar Lapp  -:-  Durham, NC  -:-  hlapp at gmx dot net :
===========================================================

[Gmod-ajax] Openlazlo

[Steve T. <st...@mo...>]

Hi,

Has anyone seen this? Basically XML and Javascript that's converted to Flash/DHTML, so may ultimately solve the dreaded browser incompatibility problems.

http://www.openlaszlo.org/

There are some very cool demos and I have seen one bioinformatics app so far (http://www.chromhome.org/chromhome/)

-- 
Steve
------------------------------------------------------------------
Medical Sciences Division
Weatherall Institute of Molecular Medicine/Sir William Dunn School
Oxford University

[Gmod-ajax] job opening

[Andrew U. <and...@gm...>]

Hi everyone...

On a note somewhat related to this list, we have a job opening for a
sys admin/programmer at the Holmes lab... please pass it on to anyone
who might be interested.  The complete details are here:

http://biowiki.org/SysAdminJob

Cheers,
Andrew

Re: [Gmod-ajax] Pfam tree widget

[Ian H. <ih...@be...>]

I guess it's based on the Yahoo UI tree widget

http://developer.yahoo.com/yui/examples/treeview/customicons.html?mode=dist

Ian Holmes wrote:
> I've just been playing with the new Pfam tree widget. Seems quite nice
> 
> http://www.sanger.ac.uk/catalyst/pfam/family?acc=PF00516
> 
> -------------------------------------------------------------------------
> This SF.net email is sponsored by DB2 Express
> Download DB2 Express C - the FREE version of DB2 express and take
> control of your XML. No limits. Just data. Click to get it now.
> http://sourceforge.net/powerbar/db2/
> _______________________________________________
> Gmod-ajax mailing list
> Gmo...@li...
> https://lists.sourceforge.net/lists/listinfo/gmod-ajax

[Gmod-ajax] Pfam tree widget

[Ian H. <ih...@be...>]

I've just been playing with the new Pfam tree widget. Seems quite nice

http://www.sanger.ac.uk/catalyst/pfam/family?acc=PF00516

Re: [Gmod-ajax] AJAX GBrowse poster feedback from RECOMB

[Ian H. <ih...@be...>]

I'd add speed: UCSC's rendering engine is all in C and very fast.

However I think Suzi's point about the data is the winner, by a long 
chalk. I hope one day we can offer that much data at a genome wiki site.

Hilmar, I'd say you're absolutely right about mashups. I don't know if 
there's actually anything stopping our browser from being used as a 
component of a larger page right now. In any case this doesn't seem hard 
to do, given our general approach, and we hope to demo this before too long.


Hilmar Lapp wrote:
> On May 2, 2007, at 9:47 AM, Suzanna Lewis wrote:
> 
>>> I think the biggest objection to Java (that I haven't seen anyone
>>> mention in this thread yet) is that it's been tried.  Given your
>>> list of
>>> links, it looks like people have already implemented genome browsing
>>> using Java Web Start, and still UCSC seems to be the browser everyone
>>> uses.
>> Yes, but I think that we also have to consider that it might just be
>> the data. If you're looking for actual genomic data they've made it
>> really easy for people to add their own. The data is the most
>> comprehensive and up-to-date set available, with one-stop shopping.
>> The browser is not the primary consideration here.
> 
> I'd second that, and would add ease of use (incl. speed and minimal  
> system requirements on the client's end), and ease of adding one's  
> own data, and downloading data.
> 
> Less obvious possibly but what's also great about the UCSC browser is  
> that it's actually very easy to save the image and stick it into a  
> PPT presentation. You may laugh about this, but the researchers I was  
> working with were doing this all the time (and I was too when I  
> worked on particular loci).
> 
> I'd also argue that in the future (and possibly already) most use of  
> a genome browser will be made through aggregated content or mash-ups.  
> That's one of the cool things about Google maps - it's easy to stick  
> it into your own page while it remains interactive yet doesn't  
> disturb or control the rest of the page.
> 
> Just my one cent ...
> 
> 	-hilmar

Re: [Gmod-ajax] AJAX GBrowse poster feedback from RECOMB

[Hilmar L. <hl...@gm...>]

On May 2, 2007, at 9:47 AM, Suzanna Lewis wrote:

>> I think the biggest objection to Java (that I haven't seen anyone
>> mention in this thread yet) is that it's been tried.  Given your
>> list of
>> links, it looks like people have already implemented genome browsing
>> using Java Web Start, and still UCSC seems to be the browser everyone
>> uses.
>
> Yes, but I think that we also have to consider that it might just be
> the data. If you're looking for actual genomic data they've made it
> really easy for people to add their own. The data is the most
> comprehensive and up-to-date set available, with one-stop shopping.
> The browser is not the primary consideration here.

I'd second that, and would add ease of use (incl. speed and minimal  
system requirements on the client's end), and ease of adding one's  
own data, and downloading data.

Less obvious possibly but what's also great about the UCSC browser is  
that it's actually very easy to save the image and stick it into a  
PPT presentation. You may laugh about this, but the researchers I was  
working with were doing this all the time (and I was too when I  
worked on particular loci).

I'd also argue that in the future (and possibly already) most use of  
a genome browser will be made through aggregated content or mash-ups.  
That's one of the cool things about Google maps - it's easy to stick  
it into your own page while it remains interactive yet doesn't  
disturb or control the rest of the page.

Just my one cent ...

	-hilmar
-- 
===========================================================
: Hilmar Lapp  -:-  Durham, NC  -:-  hlapp at gmx dot net :
===========================================================

Re: [Gmod-ajax] AJAX GBrowse poster feedback from RECOMB

[Suzanna L. <su...@be...>]

On Apr 30, 2007, at 1:10 PM, Mitch Skinner wrote:

> Hi Ann!  Congratulations on your professorship.
>
> On Mon, 2007-04-30 at 13:11 -0500, Ann Loraine wrote:
>> Did anybody mention Java Web Start as an alternative method of
>> deploying java applications?
>
> I've had a nice experience using Web Start applications before, and it
> does solve some of the problems with applets.
>
> I think the biggest objection to Java (that I haven't seen anyone
> mention in this thread yet) is that it's been tried.  Given your  
> list of
> links, it looks like people have already implemented genome browsing
> using Java Web Start, and still UCSC seems to be the browser everyone
> uses.

Yes, but I think that we also have to consider that it might just be  
the data. If you're looking for actual genomic data they've made it  
really easy for people to add their own. The data is the most  
comprehensive and up-to-date set available, with one-stop shopping.  
The browser is not the primary consideration here.

>
> Admittedly, my direct experience with non-computational biologists is
> mostly limited to human and mouse researchers, but talking with people
> at the RECOMB poster session reinforced my sense that UCSC is the main
> one.
>
> So while the language question does deserve consideration, I think the
> more important question is, why is UCSC so popular?  It can't be 100%
> inertia.  The Java drawbacks that people have mentioned are all good
> points, and I'd add that java app(let)s are often not entirely a  
> part of
> the web, in that it takes extra effort to make them bookmarkable and
> linkable and to follow links from within them.  And the network  
> effects
> you get from being part of the web are pretty huge.  Which reminds me
> that we need to do a better job of linking out and being bookmarkable
> and linkable.
>
> The Java Web Start approach seems even more non-web than applets.   
> Is it
> even possible to have an html link that takes you to a specific region
> in a Java Web Start-using genome browser?
>
> The volume of data that UCSC makes available is also a pretty big
> strength.
>
> Also, maybe this is a cheap shot, but on the PlantIGB page I see this:
> ================
> Click here if your computer has at least 1 gig of memory.
> Click here if your computer has at least 786 meg of memory.
> Click here if your computer has at least 512 meg of memory.
> ================
> which is very much not the face I'd like to present to the
> non-computer-savvy biologists out there.
>
> Regards,
> Mitch
>
>
> ---------------------------------------------------------------------- 
> ---
> This SF.net email is sponsored by DB2 Express
> Download DB2 Express C - the FREE version of DB2 express and take
> control of your XML. No limits. Just data. Click to get it now.
> http://sourceforge.net/powerbar/db2/
> _______________________________________________
> Gmod-ajax mailing list
> Gmo...@li...
> https://lists.sourceforge.net/lists/listinfo/gmod-ajax
>

[Gmod-ajax] Feature upload

[Mitch S. <mit...@be...>]

Feature uploading is finally going on genome.biowiki.org for dmel.  This 
is very alpha code, so you can expect it to eat your data and be offline 
now and then, but it's definitely ready for people to try out.

I'll be committing the code to CVS as soon as I figure out how it should 
be laid out relative to the existing code.

Mitch

Re: [Gmod-ajax] AJAX GBrowse poster feedback from RECOMB

[Ian H. <ih...@be...>]

My impression was that Java moved from clients to servers around 1996 ;-)

I have started to see some hardcore bioinformatics applications written 
in Java. Ben Paten's PECAN alignment program is written in Java which is 
surely a watershed in the "they-said-it-couldn't-be-done" tradition.

However I agree with most everything people have said against it as a 
language for the web client. The Java virtual machine lacks a close 
interface to the web browser. That is why AJAX is winning out with its 
paradigm of HTML-based UIs, asynchronous operations & increased emphasis 
on client-server messaging.

All ultimately somewhat academic as this particular project is committed 
to Ajax, and while it is useful to have a stock reply to the "why 
javascript" question, the best answer is really our demo site.



Mitch Skinner wrote:
> Hi Ann!  Congratulations on your professorship.
> 
> On Mon, 2007-04-30 at 13:11 -0500, Ann Loraine wrote:
>> Did anybody mention Java Web Start as an alternative method of
>> deploying java applications?
> 
> I've had a nice experience using Web Start applications before, and it
> does solve some of the problems with applets.
> 
> I think the biggest objection to Java (that I haven't seen anyone
> mention in this thread yet) is that it's been tried.  Given your list of
> links, it looks like people have already implemented genome browsing
> using Java Web Start, and still UCSC seems to be the browser everyone
> uses.  
> 
> Admittedly, my direct experience with non-computational biologists is
> mostly limited to human and mouse researchers, but talking with people
> at the RECOMB poster session reinforced my sense that UCSC is the main
> one.
> 
> So while the language question does deserve consideration, I think the
> more important question is, why is UCSC so popular?  It can't be 100%
> inertia.  The Java drawbacks that people have mentioned are all good
> points, and I'd add that java app(let)s are often not entirely a part of
> the web, in that it takes extra effort to make them bookmarkable and
> linkable and to follow links from within them.  And the network effects
> you get from being part of the web are pretty huge.  Which reminds me
> that we need to do a better job of linking out and being bookmarkable
> and linkable.
> 
> The Java Web Start approach seems even more non-web than applets.  Is it
> even possible to have an html link that takes you to a specific region
> in a Java Web Start-using genome browser?
> 
> The volume of data that UCSC makes available is also a pretty big
> strength.
> 
> Also, maybe this is a cheap shot, but on the PlantIGB page I see this:
> ================
> Click here if your computer has at least 1 gig of memory.
> Click here if your computer has at least 786 meg of memory.
> Click here if your computer has at least 512 meg of memory.
> ================
> which is very much not the face I'd like to present to the
> non-computer-savvy biologists out there.
> 
> Regards,
> Mitch
> 
> 
> -------------------------------------------------------------------------
> This SF.net email is sponsored by DB2 Express
> Download DB2 Express C - the FREE version of DB2 express and take
> control of your XML. No limits. Just data. Click to get it now.
> http://sourceforge.net/powerbar/db2/
> _______________________________________________
> Gmod-ajax mailing list
> Gmo...@li...
> https://lists.sourceforge.net/lists/listinfo/gmod-ajax

Re: [Gmod-ajax] AJAX GBrowse poster feedback from RECOMB

[Mitch S. <mit...@be...>]

Hi Ann!  Congratulations on your professorship.

On Mon, 2007-04-30 at 13:11 -0500, Ann Loraine wrote:
> Did anybody mention Java Web Start as an alternative method of
> deploying java applications?

I've had a nice experience using Web Start applications before, and it
does solve some of the problems with applets.

I think the biggest objection to Java (that I haven't seen anyone
mention in this thread yet) is that it's been tried.  Given your list of
links, it looks like people have already implemented genome browsing
using Java Web Start, and still UCSC seems to be the browser everyone
uses.  

Admittedly, my direct experience with non-computational biologists is
mostly limited to human and mouse researchers, but talking with people
at the RECOMB poster session reinforced my sense that UCSC is the main
one.

So while the language question does deserve consideration, I think the
more important question is, why is UCSC so popular?  It can't be 100%
inertia.  The Java drawbacks that people have mentioned are all good
points, and I'd add that java app(let)s are often not entirely a part of
the web, in that it takes extra effort to make them bookmarkable and
linkable and to follow links from within them.  And the network effects
you get from being part of the web are pretty huge.  Which reminds me
that we need to do a better job of linking out and being bookmarkable
and linkable.

The Java Web Start approach seems even more non-web than applets.  Is it
even possible to have an html link that takes you to a specific region
in a Java Web Start-using genome browser?

The volume of data that UCSC makes available is also a pretty big
strength.

Also, maybe this is a cheap shot, but on the PlantIGB page I see this:
================
Click here if your computer has at least 1 gig of memory.
Click here if your computer has at least 786 meg of memory.
Click here if your computer has at least 512 meg of memory.
================
which is very much not the face I'd like to present to the
non-computer-savvy biologists out there.

Regards,
Mitch

Re: [Gmod-ajax] AJAX GBrowse poster feedback from RECOMB

[Ann L. <alo...@gm...>]

Hello,

I just joined the list and saw the interesting discussion on java applets,
javascript, etc.

Did anybody mention Java Web Start as an alternative method of deploying
java applications?

I like it a lot, myself. I've found it to be quite hassle-free - very easy
to set up. You just put your signed, compiled 'jar' files in a web
directory, write a jnlp file, and deploy.

Here are some examples from bioinformatics:

http://www.ccgb.umn.edu/software/java/apps/TableView/
http://www.jalview.org/
http://www.bioviz.org/plant_igb/
http://www.affymetrix.com/support/developer/tools/download_igb.affx

The wikipedia entry has more links:

http://en.wikipedia.org/wiki/Java_Web_Start

My experience with java applets from several years ago was that it was
always very difficult to get the applets to run properly in every user's
browser. The simple applications were not hard, but for more complex
programs with a lot of functionality, it was very difficult to get
everything to work as expected. After having that experience, working with
Java Web Start has been quite fun. It's so much easier!

-Ann

-- 
Ann Loraine
Assistant Professor
University of Alabama at Birmingham
http://www.transvar.org
205-996-4155

Re: [Gmod-ajax] AJAX GBrowse poster feedback from RECOMB

[Hilmar L. <hl...@gm...>]

On Apr 23, 2007, at 10:50 AM, Andrew Uzilov wrote:

> I realized I can't really come up with a convincing reason AGAINST  
> Java.

In my experience the main issues with applets are load-time and  
browser-specific event handling quirks that at first seem minor but  
turn out to be major headaches if one of the main goals is UI  
interactivity (which it seems to be here ...)

I agree with the stylesheet disconnect issue raised - that's clearly  
undesirable.

As for the some of the other arguments, such as printing or speed, in  
my experience they apply as much to JS or other embedded scripting  
languages for that matter (just compare a Google Map on the screen  
and once you print it, for example; you have full control over  
printing from an applet with a custom-written print handler). An old  
machine that slows down an applet will have at least as much issue  
with executing the JS.

	-hilmar
-- 
===========================================================
: Hilmar Lapp  -:-  Durham, NC  -:-  hlapp at gmx dot net :
===========================================================

Re: [Gmod-ajax] AJAX GBrowse poster feedback from RECOMB

[Steve T. <st...@mo...>]

Hi,

> On Mon, 2007-04-23 at 13:28 -0700, Ian Holmes wrote:
> 
>>>learned if there's a good reason for it.  But why not Java applets?
>>>The only two reasons we came up with are (1) long load time, and (2)
>>>need to install Java, but that's not hard...
>>
>>Interesting. I'd like to know what others think of this.
> 
> 
> I can think of a few!
> 
> I've done stuff before with Java applets as web 2.0 type thingys, and my
> recollection is that it was very nasty. One still vivid nasty part is
> that there was no good way to have the Java applet and JavaScript
> communicate with each other, thereby trapping what you're doing on one
> side or the other. There were classes and objects to get around this,
> but the implementations were buggy and platform dependent (LiveConnect,
> etc.).
> 
> Another big ding against applets was presentation/layout issues. For
> example, font resizing and layout won't work with the rest of the
> browser, making styles very hard to use and control. Also, printing just
> gave me some truncation and looked pretty bad when I tried it with an
> applet from Sun's Java site (the spreadsheet applet) and cut-and-paste
> didn't work at all. Scraping data would also be impossible.
> 
> It's hard to just stick with the tangibles: things were very...fiddly.
> It may have all changed since I last did this years ago, but doing a
> cursory google search right now it looks to be in more or less the same
> state as I remember. 
> 
> It's  hard to have the browser as a platform when it's actually Java.
> 

Haven't posted for a while, but I just thought I would say well done to getting AJAX GBrowse this far. It just gets better and better!

I absolutely agree with all these comments. I hate the language wars that seem to accompany certain discussions in bioinformatics. At the end of the day, users want functionality and I have seen
far too many projects go down the 'we must rewrite it in Java' route to know that gains can be slight/non-existent for a massive amount of recoding. Applets are often slow to load and run (let's not 
forget some biologists still have older machines out there). There are some good small applets out there but developers who can program fast and nice looking applets I would say are few and far 
between. The fact is you can get considerable bang for your buck with the latest web technologies - providing a great user experience and allowing developers to code in what they want to at the back end.

Keep up the good work - in the language of your choice!

Steve
------------------------------------------------------------------
Medical Sciences Division
Weatherall Institute of Molecular Medicine/Sir William Dunn School
Oxford University

Re: [Gmod-ajax] AJAX GBrowse poster feedback from RECOMB

[Andrew U. <and...@gm...>]

On 4/23/07, Andrew Uzilov <and...@gm...> wrote:
> > > less-than-reusable fashion.  A certain Bay Area company (should I
> > > mention who?  I don't know what the etiquette rules are...)
> >
> > Unless they asked you not to mention them (which would be odd at an open
> > poster session), I don't see why not.
> >
> > > "Pragmatic Ajax" book)... then they invited us to demo our version for
> > > them sometime.
> >
> > Can you pass on contact details and I'll try to arrange a date?
>
> The company is Genentech.  They've been brewing a lot of in-house
> stuff apparently.

Oh yeah... the man is Thomas Wu (tw...@ge...)... he seemed pretty
excited about the browser.

Re: [Gmod-ajax] AJAX GBrowse poster feedback from RECOMB

[Andrew U. <and...@gm...>]

> > less-than-reusable fashion.  A certain Bay Area company (should I
> > mention who?  I don't know what the etiquette rules are...)
>
> Unless they asked you not to mention them (which would be odd at an open
> poster session), I don't see why not.
>
> > "Pragmatic Ajax" book)... then they invited us to demo our version for
> > them sometime.
>
> Can you pass on contact details and I'll try to arrange a date?

The company is Genentech.  They've been brewing a lot of in-house
stuff apparently.


Andrew