Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Repo info
Activity
  • 19:23

    jlpereira on development

    Fix (compare)

  • 19:01
    tmcelrath commented #1143
  • 19:01
    mjy closed #1143
  • 19:01

    mjy on development

    Fix #1143 Merge branch 'development' of g… (compare)

  • 19:00
    mjy commented #1143
  • 19:00

    mjy on page

    Revert "belongs_to require by d… backport fix for #1195. POTENT… Fix #1195 for real. POTENTIALL… and 4 more (compare)

  • 18:59
    jlpereira edited #1214
  • 18:58

    jlpereira on development

    Added recent button (compare)

  • 17:53
    mjy edited #1214
  • 17:52
    mjy edited #1214
  • 17:50
    mjy assigned #1214
  • 17:50
    mjy labeled #1214
  • 17:50
    mjy opened #1214
  • 17:48
    mjy labeled #1213
  • 17:48
    mjy labeled #1213
  • 17:48
    mjy assigned #1213
  • 17:48
    mjy opened #1213
  • 17:47
    mjy opened #1212
  • 17:47
    mjy assigned #1212
  • 17:47
    mjy labeled #1212
Matt
@mjy
yes
Hernán Lucas Pereira
@LocoDelAssembly
live
Hernán Lucas Pereira
@LocoDelAssembly
@/all production is about to be updated to 0.8.4.
Hernán Lucas Pereira
@LocoDelAssembly
@/all 0.8.4 ready to use.
Matt
@mjy
Thanks @LocoDelAssembly
Hernán Lucas Pereira
@LocoDelAssembly
@/all updating sandoxes.
Hernán Lucas Pereira
@LocoDelAssembly
@/all production is about to be updated to 0.8.5. Also sandboxes will be updated.
Hernán Lucas Pereira
@LocoDelAssembly
@/all 0.8.5 is live now.
Matt
@mjy
@LocoDelAssembly asside from deprecation warnings do you see any reason not to move to Rails6?
Seems like it might be a good time to roll dev into master, do a small bump, then start experimenting with 6 in sandboxes?
And another question @LocoDelAssembly - if I want the Page branch to start testing on Travis is it still OK to just add it to the test list in .travis.yml or do you have another mechanism that is preferred?
Matt
@mjy
Went ahead and added to .travis.yml, revert if bad idea.
Hernán Lucas Pereira
@LocoDelAssembly
Updated page branch. Actually it seems enough to declare you want the branch built on the .travis.yml of the same branch, no actual need to also do it on development.
We could experiment with rails 6, but are there plans to release new production version this week?
Matt
@mjy
Yes, we will update production this week.
We won't get Page in there for sure.
But there are various small bug fixes I think to roll in, and there might be a couple more to get done. I'd like to fix some matrix things if I can get Page responses more or less done.
Hernán Lucas Pereira
@LocoDelAssembly
rails6 did not like last merge... checking why
Hernán Lucas Pereira
@LocoDelAssembly
Did someone experienced this problem?
/home/hernan/.rvm/gems/ruby-2.6.3/gems/addressable-2.7.0/lib/addressable/idna/pure.rb:252: [BUG] Segmentation fault at 0x000000000000fff9
ruby 2.6.3p62 (2019-04-16 revision 67580) [x86_64-linux]

-- Control frame information -----------------------------------------------
c:0067 p:0093 s:0360 e:000358 BLOCK  /home/hernan/.rvm/gems/ruby-2.6.3/gems/addressable-2.7.0/lib/addressable/idna/pure.rb:252 [FINISH]
c:0066 p:---- s:0355 e:000354 CFUNC  :each
c:0065 p:0010 s:0351 e:000350 METHOD /home/hernan/.rvm/gems/ruby-2.6.3/gems/addressable-2.7.0/lib/addressable/idna/pure.rb:245
c:0064 p:0038 s:0340 e:000337 METHOD /home/hernan/.rvm/gems/ruby-2.6.3/gems/addressable-2.7.0/lib/addressable/idna/pure.rb:121
c:0063 p:0307 s:0332 e:000330 METHOD /home/hernan/.rvm/gems/ruby-2.6.3/gems/addressable-2.7.0/lib/addressable/uri.rb:566
c:0062 p:0115 s:0321 e:000320 METHOD /home/hernan/.rvm/gems/ruby-2.6.3/gems/addressable-2.7.0/lib/addressable/uri.rb:1126
c:0061 p:0024 s:0316 e:000315 METHOD /home/hernan/.rvm/gems/ruby-2.6.3/gems/addressable-2.7.0/lib/addressable/uri.rb:850
c:0060 p:0012 s:0312 e:000311 METHOD /home/hernan/.rvm/gems/ruby-2.6.3/gems/webmock-3.7.6/lib/webmock/util/uri.rb:37
c:0059 p:0016 s:0303 e:000302 METHOD /home/hernan/.rvm/gems/ruby-2.6.3/gems/webmock-3.7.6/lib/webmock/request_pattern.rb:143
c:0058 p:0060 s:0298 e:000297 METHOD /home/hernan/.rvm/gems/ruby-2.6.3/gems/webmock-3.7.6/lib/webmock/request_pattern.rb:37
c:0057 p:0011 s:0292 e:000291 BLOCK  /home/hernan/.rvm/gems/ruby-2.6.3/gems/webmock-3.7.6/lib/webmock/stub_registry.rb:58 [FINISH]
c:0056 p:---- s:0288 e:000287 IFUNC 
c:0055 p:---- s:0285 e:000284 CFUNC  :each
c:0054 p:---- s:0282 e:000281 CFUNC  :detect
(while running rspec)
Matt
@mjy
I have had 2 seg-faults in the past couple days, but I don't remember where I experienced them. Typically they indicate real problems with Ruby and or a compile bit of some gem.
Matt
@mjy
Another issue- I suspect these: PG::TRDeadlockDetected: ERROR: deadlock detected are common enough with taxon names that we should catch and return a proper response in the iterm.
Seems like we could resolve with more precise cache building I suspect.
Matt
@mjy
Another issue- Does anyone see these: 'ERROR Errno::ECONNRESET: Connection reset by peer @ io_fillbuf - fd:19 ' in the webbrick log?
Hernán Lucas Pereira
@LocoDelAssembly
image.png
How that identification history should be entered in TW? Comprenhensive digitization lets you define a single TypeMaterial, and if you create more than one from Data>TypeMaterial you get a soft validation error ("More than one primary type associated with the taxon").

Another issue- Does anyone see these: 'ERROR Errno::ECONNRESET: Connection reset by peer @ io_fillbuf - fd:19 ' in the webbrick log?

Never seen that. But were you working with anything triggering external API calls (e.g. Serrano)?

Matt
@mjy
No, this is just all the time on web-brick
Matt
@mjy
Re the type question:
Matt
@mjy
  • Goto glaberrimum, select syntype, and this specimen
  • Goto vulgare, select neotype, choose "Existing", find and add this specimen
I believe that is all you have to do.
It is a neotype of cuticulare via synonymy, so we can infer that I suppose.
Citations can be attached to the TypeMaterial record itself I think.
Though they might just go to the nomenclature directly, depends on what is being asserted.
Matt
@mjy
This use case should get described on taxonworks_doc by @proceps ;)
Matt
@mjy
So, SpeciesFileGroup/taxonworks#1195 is going to require a major change.
There is a conflict with 'url', attribute, we render it generically for all object in the metadata partial, and it's an attribute of source bibtex.
There are two solutions, neither fun: 1) we update bibtex model to include a virtual param bibtex_url which is mapped back/forth in the client and model as needed; or 2) we nest all metadata attributes in json responses in metadata: {}.
I suppose 3: we can pass a param to exclude :url param from metadata rendering ...
or 4) we can rename url param in metadata ( I think this)
... then update lots of js
Matt
@mjy
I'm going start a new branch, and go for 4. @jlpereira this will mean updating a lot of things from url to url_for I think (i.e. name it the same as the helper method).
Matt
@mjy
OK. New branch url_for is pushed. All interface elements that referenced urlas it comes from the metadata response must be updated to reference url_for.
This is a fairly high priority as Source updates are overweriting URLs.
Matt
@mjy
Taking a stab at the big rename now.
Matt
@mjy
Well- maybe not big rename. I can't actually find any confirmed instances. I'm going to merge to Page and figure it out with @jlpereira .
Matt
@mjy
I have pushed the fix for 1195, it may be breaking ofr smaller things, but it appears not to be a systemic break.
meh-
now pushed (development)
To re-iterate, this changes the RESTful response for things like /otus/123.json and /taxon_names.123.json by updating the shared partial (don't confuse with the annotations/metadata endpoint) app/views/shared/data/all/_metadata.json.jbuilder to use url_for instead of url. So it does change everywhere that partial is used, which is most SHOW member endpoints.