Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Repo info
Activity
  • May 24 05:32
    dependabot[bot] labeled #25
  • May 24 05:32
    dependabot[bot] opened #25
  • May 24 05:32

    dependabot[bot] on bundler

    Bump nokogiri from 1.13.5 to 1.… (compare)

  • May 23 11:03

    translatewiki on main

    Localisation updates from https… (compare)

  • May 20 19:20

    JRice on main

    Updating notes on BuildIdentifi… (compare)

  • May 20 17:54

    JRice on main

    Hopefully fixes missing media t… Merge branch 'main' of github.c… (compare)

  • May 19 17:39

    JRice on production

    Removed #check_for_duplicate_* … (compare)

  • May 19 17:36

    JRice on main

    Removed #check_for_duplicate_* … Merge branch 'main' of github.c… (compare)

  • May 19 11:05

    translatewiki on main

    Localisation updates from https… (compare)

  • May 19 10:25

    JRice on main

    Bump Nokogiri Merge branch 'main' of github.c… (compare)

  • May 19 10:22

    dependabot[bot] on bundler

    (compare)

  • May 19 10:22

    JRice on main

    Bump nokogiri from 1.13.2 to 1.… Merge pull request #24 from EOL… (compare)

  • May 19 10:22
    JRice closed #24
  • May 19 03:33

    dependabot[bot] on bundler

    (compare)

  • May 19 03:33
    dependabot[bot] closed #23
  • May 19 03:33
    dependabot[bot] commented #23
  • May 19 03:33
    dependabot[bot] labeled #24
  • May 19 03:33
    dependabot[bot] opened #24
  • May 19 03:33

    dependabot[bot] on bundler

    Bump nokogiri from 1.13.2 to 1.… (compare)

  • May 18 15:46

    JRice on production

    Remove old refs to Link class (… (compare)

Jen Hammock
@jhammock
It’s probably worth regenerating the autogen text. The Stolonifera example entails extra mysteries, but there will be less dramatic out-of-date text on many pages. @JRice I daresay this belonged to MV, but as I recall, he needed you to intervene to get these updates run. Can you track down the process?
Jeremy Rice
@JRice
Yes, I will do. ...but... probably not right now; I am about to do the car thing, just waiting on a final call. Should be any second, actually.
...But will look into this tonight!
Jen Hammock
@jhammock
no urgent rush
Good luck!
Jeremy Rice
@JRice
Thanks.
Katja Schulz
@KatjaSchulz
@JRice Turns out the identifier map on opendata already points to the url you provided, so it automatically gets the latest version. I just double-checked to make sure and Wikidata is in there. I just need to add it to the list of included classification providers in the resource description.
Jeremy Rice
@JRice
Oh, nice! That's conveeeeeeenient. I'll update my notes to make sure I don't misrepresent that again, sorry. :S
Jeremy Rice
@JRice

Yes, I will do. ...but... probably not right now; I am about to do the car thing, just waiting on a final call. Should be any second, actually.

I did not get a chance to do this tonight. I should have time tomorrow morning, though.

Jeremy Rice
@JRice

It’s probably worth regenerating the autogen text.

So the code here is, unsurprisingly, really complicated. ...but I do not see any evidence at all that this is stored in the database. The only place this appears to be cached is in memory (memcached), and I'm 99% sure I have cleared the cache (a few times, even) since DH2.1 has been put in place.

...So the problem must be in the generation of the summary itself. Which, as mentioned ... is complicated. I'll have to investigate on Monday.

Jeremy Rice
@JRice
Ahh, so I think this problem is simple: there's a limit on the number of children that the summary lists.
That's currently set to 3.
It's not hard to change (at all), but it would require a redeploy/restart to take effect.
@KatjaSchulz - Were there any any other problems with that autogen text?
Oh! Shoot, I just looked at the calendar and saw that I have a dentist appointment today; I'll be out from 10:30 until 12:30 or so.
Katja Schulz
@KatjaSchulz
@JRice The rest of the Stolonifera autogen text looked good to me, and I haven't seen any major problems elsewhere. However, I haven't done a thorough review of autogen texts. If you'd like us to do that before you get to work on this issue, let us know and we'll spend some time poking around specifically in that area,
Jeremy Rice
@JRice
Mostly I just wanted to verify that it was using V2.1 and didn't need "re-generation," but really I don't see anything in the code that stores this text other than by normal memcached caches.
...Which is perhaps not ideal, IMO (it's part of what makes the first page load so slow), but... at least it seems to be working! :D
Jen Hammock
@jhammock
There’s supposed to be a limit on the children listed in the summary. Those are IMMEDIATE children.
Current problems in the autogen:
The first sentence should refer to two different ancestral landmark groups
On https://eol.org/pages/46545576, it is the COUNTS of species, genera, and familes that are off
Jeremy Rice
@JRice
Ahh, yes. That makes sense.
Jen Hammock
@jhammock
:+1:
Jeremy Rice
@JRice

On https://eol.org/pages/46545576, it is the COUNTS of species, genera, and familes that are off

This is now fixed...

The first sentence should refer to two different ancestral landmark groups

This must be an actual bug.

Jeremy Rice
@JRice
Yeah, there's nothing in the code that checks one against the other.
...What should the behaviour here be? If they are the same, should I omit the "family" clause? (This is easier than:) Should I change the non-family one to be... the next landmark up? I'm not sure. :S
(FWIW I agree that this is really awkward and we should do something. Omitting the family would be very easy, so I'm leaning toward that.) :)
...Another option is that the "other landmark" could just make sure that it never picks a family, I suppose. That wouldn't be too hard to add, but still more complicated than removing the family if it's the same string.
Jen Hammock
@jhammock
I'm pretty sure this changed with the new DH, so my guess is that families used to appear in the breadcrumb without actually having the landmark tag they currently have, and that changed with DH2. Katja knows best which two ancestors she wanted to show, and I daresay ultimately we can restore the full behaviour, but I approve of removing the family in the meantime
Jeremy Rice
@JRice
It could have been that choice we made between "minimal" and "abbreviated," then!
I went with "minimal." I could try "Abbreviated" and see if that changes things?
It doesn't cost much to try, I will go ahead.
Katja Schulz
@KatjaSchulz
It's been such a long time since we put together the taxonomy sentence. I honestly don't remember the specifics. However, the current sentence "Tamias dorsalis (Cliff Chipmunk) is a species of rodents in the family Sciuridae." looks good to me. I also checked a few pages in groups with differently structured ancestries, and everything has looked good so far.
Jeremy Rice
@JRice
...Then I guess it was supposed to be "abbreviated" after all! :D
(I did change it and clear cache)
Katja Schulz
@KatjaSchulz
And the families are still there in the breadcrumbs. Didn't we lose those before when we switched to abbreviated? Anyway, all is well that ends well.
Jeremy Rice
@JRice
I never actually cleared the cache when I first tried abbreviated. :S
Jeremy Rice
@JRice
I'm not expecting to be "fully available" tomorrow. My son is getting a cough, so he'll be staying home again, and my wife has to go back to work. I'll try to do what I can, but it does get pretty hard to focus when I'm home alone with him and the puppy. :S
Jen Hammock
@jhammock
understood
Jen Hammock
@jhammock
@JRice it took me awhile to notice, but Tom O. aked if EOL was down around 11am. Anything in the logs?
Jeremy Rice
@JRice
No, actually.
Jen Hammock
@jhammock
it may have been at his end, then. I didn’t notice anything myself, though I could easily have missed it
Jeremy Rice
@JRice

I'm not expecting to be "fully available" tomorrow. My son is getting a cough, so he'll be staying home again, and my wife has to go back to work. I'll try to do what I can, but it does get pretty hard to focus when I'm home alone with him and the puppy. :S

Same again today. ...tomorrow she's off, though, so even if Evan is still sick, I'll be working.

Jen Hammock
@jhammock
Poor kid. Lucky dog, though! More people around to play with...
Jeremy Rice
@JRice
For scratches, at least. He's not particularly high-energy today. :)
Jen Hammock
@jhammock
scratches are important too
Jeremy Rice
@JRice
The dog certainly thinks so!