search

GMOD / jbrowse

JBrowse 1, a full-featured genome browser built with JavaScript and HTML5. For JBrowse 2, see https://github.com/GMOD/jbrowse-components.
http://jbrowse.org
Other
461 stars 199 forks source link

NClist track loading unreliable under iframes #250

Closed rbuels closed 11 years ago

rbuels commented 11 years ago

On 04/22/2013 06:05 AM, Steffi Geisen wrote:>

maybe you have heard from that?! We let a user search in a gene database and provide links to a window which include 2 frames of Jbrowse from different species: http://bioinfo2.ugr.es/CpGislandEVO/show_evo.php?name=AAAS If I click for example on the rheMac link (Macaca mulatta, show http://bioinfo2.ugr.es/CpGislandEVO/JBrowseViewer.php?chromosome=chr11&region=50380016-50389838&assembly=rheMac2&chromosome_hg19=chr12&region_hg19=53698404-53718247&track_hg19=refGene&track=ensGene) first there is no gene shown in the ensGene track, although the browser points to the correct coordinates. If I do a refresh the page with F5 or open the data in a new window clicking on the rheMac title, the ensGene track shows the gene. This happens in other species as well, but only with ensGene and refGene tracks.

Do you have any idea what could be the problem here and how to fix this?

rbuels commented 11 years ago

Email interactions:

On 05/02/2013 02:39 AM, Steffi Geisen wrote:

I installed the newer testversion in the same directory (CpGislandEVOtest/JbrowseViewer). First I thought that it might have to do with the version 1.9(.1) compared to 1.8.1 we are still using in CpGislandEVO, but here it happened again:

http://bioinfo2.ugr.es/CpGislandEVOtest/JBrowseViewer.php?chromosome=chr6&region=18324824-18422480&assembly=panTro3&chromosome_hg19=chr6&region_hg19=18141925-18237777&track_hg19=refGene&track=ensGene

  • the human refGene track was not displayed, only after clicking F5.

2013/5/1 Robert Buels <rbuels@gmail.com mailto:rbuels@gmail.com>

You're right, the problem seems to be gone in that test
installation.  I can still see the problem in the original
installation, though.

Here, make another test installation, this time using the attached
minimized JBrowse build, and see if that one exhibits the problem.

Robert Buels
Lead Developer
JBrowse - http://jbrowse.org

On 05/01/2013 11:26 AM, Steffi Geisen wrote:

    Hi Robert,

    I uploaded it to a test directoy:
    http://bioinfo2.ugr.es/__CpGislandEVOtest/show_evo.php?__name=AAAS
    <http://bioinfo2.ugr.es/CpGislandEVOtest/show_evo.php?name=AAAS>
    and the only thing that is changed so far compared to the original
    installation is that the subdirectory which contains all the Jbrowse
    data - JbrowseViewer - has all rights (chmod 777) on all folders
    + files.

    Can you reproduce the error? At the moment the problem does not
    occur
    when I open a frame, even in the original installation... really
    strange! If it doesn't happen to you as well anymore I think we can
    close the issue? I will test it tomorrow again at the University on
    another computer to see if it can be reproduced there (mainly
    working
    from home, too).

    Have you heard from others that they have this problem or could
    it be a
    server configuration thing?

    Greetings to Cambridge,
    Steffi

    2013/4/30 Robert Buels <rbuels@gmail.com
    <mailto:rbuels@gmail.com> <mailto:rbuels@gmail.com
    <mailto:rbuels@gmail.com>>>

         Hi Steffi,

         I've spent quite a while working on this, and I haven't had
    any success.

         Would you be able to make another version of that
    JBrowseViewer.php
         page that displays the same data, but uses the JBrowse
    development
         version in the zipfile at
    http://jbrowse.org/test/____JBrowse-dev-dev.zip
    <http://jbrowse.org/test/__JBrowse-dev-dev.zip>
         <http://jbrowse.org/test/__JBrowse-dev-dev.zip
    <http://jbrowse.org/test/JBrowse-dev-dev.zip>>?  This version of

         JBrowse has non-compressed source code, and will help me
    drill down
         and find the problem.

         Robert Buels
         Lead Developer
         JBrowse - http://jbrowse.org

         On 04/22/2013 06:05 AM, Steffi Geisen wrote:

             Hey Robert,

             maybe you have heard from that?! We let a user search
    in a gene
             database
             and provide links to a window which include 2 frames of
    Jbrowse from
             different species:
    http://bioinfo2.ugr.es/____CpGislandEVO/show_evo.php?____name=AAAS
    <http://bioinfo2.ugr.es/__CpGislandEVO/show_evo.php?__name=AAAS>

    <http://bioinfo2.ugr.es/__CpGislandEVO/show_evo.php?__name=AAAS
    <http://bioinfo2.ugr.es/CpGislandEVO/show_evo.php?name=AAAS>>
             If I click for example on the rheMac link (Macaca
    mulatta, show

    <http://bioinfo2.ugr.es/____CpGislandEVO/JBrowseViewer.____php?chromosome=chr11&region=____50380016-50389838&assembly=____rheMac2&chromosome_hg19=chr12&____region_hg19=53698404-__53718247&__track_hg19=refGene&__track=__ensGene
    <http://bioinfo2.ugr.es/__CpGislandEVO/JBrowseViewer.__php?chromosome=chr11&region=__50380016-50389838&assembly=__rheMac2&chromosome_hg19=chr12&__region_hg19=53698404-53718247&__track_hg19=refGene&track=__ensGene>

    <http://bioinfo2.ugr.es/__CpGislandEVO/JBrowseViewer.__php?chromosome=chr11&region=__50380016-50389838&assembly=__rheMac2&chromosome_hg19=chr12&__region_hg19=53698404-53718247&__track_hg19=refGene&track=__ensGene
    <http://bioinfo2.ugr.es/CpGislandEVO/JBrowseViewer.php?chromosome=chr11&region=50380016-50389838&assembly=rheMac2&chromosome_hg19=chr12&region_hg19=53698404-53718247&track_hg19=refGene&track=ensGene>>>)

             first there is no gene shown in the ensGene track,
    although the
             browser
             points to the correct coordinates.
             If I do a refresh the page with F5 or open the data in
    a new window
             clicking on the rheMac title, the ensGene track shows
    the gene. This
             happens in other species as well, but only with ensGene and
             refGene tracks.

             Do you have any idea what could be the problem here and
    how to
             fix this?

             Thank you in advance,
             Steffi
rbuels commented 11 years ago

Still can't reproduce this. Intermittency: the bane of plumbers and developers everywhere.

rbuels commented 11 years ago

Moving this to the 2.0 milestone to revisit later.

rbuels commented 11 years ago

This was probably caused by the race condition that was fixed in 1.9.7