by

Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Activity
  • 17:20

    mtholder on fish

    Add a function to determine if … Add is_extinct() members to Tax… Add flag on SunTreeNodeData to … and 84 more (compare)

  • Jul 07 17:45
    MonicaSteffi commented #1245
  • Jul 07 17:01
    mtholder commented #1245
  • Jul 06 06:39
    MonicaSteffi opened #1245
  • Jul 04 06:35
    MonicaSteffi opened #157
  • Jul 02 00:20
    jimallman opened #1244
  • Jul 01 23:36

    jimallman on development

    Show failure message if study n… Merge branch 'show-failed-study… (compare)

  • Jul 01 23:36

    jimallman on show-failed-study-fetch

    Show failure message if study n… (compare)

  • Jul 01 23:20

    bredelings on wsgi-py3

    Install libapache2-mod-wsgi-py3 Remove forwards to localhost:19… Make ws_wrapper the wsgi applic… (compare)

  • Jul 01 23:20

    bredelings on master

    Build otcetera with -DNDEBUG (compare)

  • Jul 01 23:04

    jimallman on development

    Don't keep repeating failed stu… Merge branch 'show-failed-study… (compare)

  • Jul 01 23:04

    jimallman on show-failed-study-fetch

    Don't keep repeating failed stu… (compare)

  • Jul 01 22:48

    jimallman on development

    Merge pull request #1243 from O… Recover from a failed study loo… Merge branch 'show-failed-study… (compare)

  • Jul 01 22:48

    jimallman on show-failed-study-fetch

    Recover from a failed study loo… (compare)

  • Jul 01 19:38

    bredelings on master

    Add summary so we can see what … Add some helper functions. Add dynamic program matrix clas… and 14 more (compare)

  • Jul 01 19:10
    bredelings closed #99
  • Jul 01 19:10

    bredelings on development

    Change the JSON for conflicting… (compare)

  • Jul 01 03:36

    jimallman on master

    Recent additions for production… (compare)

  • Jun 25 21:02

    snacktavish on show-multiple-witnesses

    (compare)

  • Jun 25 21:01

    snacktavish on master

    Improved conflict display w/ mu… typo typos and 3 more (compare)

Jim Allman
@jimallman
... and a goofy one on devtree:
Screen Shot 2020-06-18 at 7.36.19 PM.png
Benjamin Redelings
@bredelings
If something conflicts with Bacteria and Archea, it just conflicts with both.
Jim Allman
@jimallman
ah, my bad. what do think is sensible behavior here?
Benjamin Redelings
@bredelings
I think we should say conflicts with taxon1, taxon2, taxon3, ...
This is what it is doing on production.
However, on dev we have the opportunity to make three different links there.
Jim Allman
@jimallman
ok, and link to each?
Benjamin Redelings
@bredelings
Yeah.
On production, it will list three taxa, but only have one link which goes to the first one, which is sub-optimal.
Jim Allman
@jimallman
gotcha. that popup should expand to handle whatever we put in it
do you have a handy example of listing all three on production? i don't recall making that change here in the conflict viewer
(maybe devtree is on a weird branch)
Benjamin Redelings
@bredelings
@jimallman If you search on "Microchiroptera", it should take you to a list of trees that conflict with that taxon.
If you click on the second tree, it will show you a view with information about conflict against OTT.
Some of the orange branches have e.g. Conflicts with Tadarida + Molossus + Mormopterus
Basically I hacked the old conflict protocol that only reports one name to say that Tadarida + Molossus + Mormopterus is the name of the single example.
If you look at the new JSON response for conflict versus the old JSON response that I posted on the ot-private channel, it should make sense.
... and actually the JSON there shows a case with > 3 conflict witnesses.
Benjamin Redelings
@bredelings
@snacktavish BTW, I changed the conflict response JSON on devapi so that Jim could update the webapp to consume the new format.
I would expect the webapp to not-work until it is updated.
Jim Allman
@jimallman
awesome, thanks @bredelings
Jim Allman
@jimallman
ok, i believe we've got the desired behavior working now on devtree, see PR OpenTreeOfLife/opentree#1243
Benjamin Redelings
@bredelings
@jimallman That's great!
It looks good.
Benjamin Redelings
@bredelings
@jimallman Do we need to wait until we've merged the otc changes in order to deploy this on api?
I have fixed the bug that I mentioned on the call last week, but on a different branch, and so I'm looking at which order to deploy changes in.
Benjamin Redelings
@bredelings
The layout for 4 conflicting taxa is cool.
Jim Allman
@jimallman
i believe this will work with old or new JSON. at this point, it's very accommodating and will adapt to no witnesses, a bare id string, an array of id strings, and matched arrays of ids + names.
@bredelings ^
regarding the layout, the indentation is a fudge since we don't know whether prefix text is "Aligned with", "Resolves", or what... this could probably be measured and refined if it's an eyesore.
Screen Shot 2020-06-23 at 6.47.15 PM.png
Benjamin Redelings
@bredelings
I think that only "conflicts with" can have multiple witnesses, if that helps.
^ @jimallman
Jim Allman
@jimallman
ah, that makes sense. i could have sworn i'd seen counter-examples, but it might have just been single taxa with very long names:
Screen Shot 2020-06-24 at 1.45.13 AM.png
Benjamin Redelings
@bredelings
The ctrie fixes are now deployed on devtree. You can again type genus + SPACE to see the species.
Mark T. Holder
@mtholder
awesome. Thanks, @bredelings
Benjamin Redelings
@bredelings
I ended up merging the fixes to the development branch, and then rebasing the new-conflict on top of that in order to keep the new-conflict changes on devapi.
:-)
That way, we could deploy the development branch on production and call it master if everything works.
We could also merge the whole thing (including conflict protocol changes) to development and deploy on production, provided we also update the opentree app on production.
Benjamin Redelings
@bredelings
BTW I did all this with building + restarting otc by hand. I don't know if the germinator scripts are supposed to work...
Emily Jane McTavish
@snacktavish
very snazzy!
Benjamin Redelings
@bredelings
Oops... found another bug...
Benjamin Redelings
@bredelings
OK, bug fixed and deployed on devapi.
I've now added a new fuzzy-matching test, as well as a ctrie test to the otcetera test cases. But if people want to give the autocomplete bar on devtree some testing, that might be helpful.
BTW, I was thinking about adding a fuzzy prefix match. Right now we have fuzzy complete matches and exact prefix matches, but no fuzzy prefix match.
Benjamin Redelings
@bredelings
I'm also thinking about adding the ability to read in amendment files on startup.
It might be good to write the code to read them, since it could be used in multiple ways.