Closed sidkoul closed 11 years ago
Hi, Guys -- Well, I was about halfway through, but the site went down. Is it the nightly update? Anyhow, will resume tomorrow morning...Elizabeth
On Mon, Nov 26, 2012 at 4:45 PM, Sid Koul notifications@github.com wrote:
For the most part, issue #284https://github.com/newfs/gobotany-app/issues/284(Family and Genera thumbnails for the dkey) works well. However, there are a few places where it could be improved. Non Representative Species
Sometimes the app presents users with a species that is _not_representative of the group. This causes the botanists to scratch their head in confusion.
For example, Group 4: http://gobotany-dev.herokuapp.com/dkey/group-4/ Woody angiosperms with opposite or whorled leaves includes Petasites japonicushttp://gobotany-dev.herokuapp.com/species/petasites/japonicus/to illustrate the Asteraceae http://gobotany-dev.herokuapp.com/family/asteraceae/ family.
While that species is in the family, it is not a woody angiosperm. Using Species with More Images
To avoid displaying the Image not available yet message to the user, we could choose species with a greater number of images. For example, Group 7: Herbaceous angiosperms with superior ovaries and zygomorphic flowers
http://gobotany-dev.herokuapp.com/dkey/group-7/ Suggested Solution
As noted, most times (at last count, there were 185 families in the Flora) the displayed thumbnails work well. Talking with Elizabeth, she'd like to designate maybe 20-40 custom species that illustrate a group.
I've asked her to take a stab at this in a spreadsheet: group, family, spreadsheet
— Reply to this email directly or view it on GitHubhttps://github.com/newfs/gobotany-app/issues/304.
Sid, I seem to remember that you were planning on moving the update to very late at night — 3am or something like that? Were you able to go into your integration app on Heroku and change the scheduler time? If it's catching Elizabeth in mid-stream, then it probably needs to be later.
Hi, Everybody: I sent the list of image changes a couple of days ago...so I climbed out of the stream and dried off and am now warm and toasty. Let me know if you need clarification or anything else. Cheers, Elizabeth
On Thu, Nov 29, 2012 at 4:25 PM, Brandon Rhodes notifications@github.comwrote:
Sid, I seem to remember that you were planning on moving the update to very late at night — 3am or something like that? Were you able to go into your integration app on Heroku and change the scheduler time? If it's catching Elizabeth in mid-stream, then it probably needs to be later.
— Reply to this email directly or view it on GitHubhttps://github.com/newfs/gobotany-app/issues/304#issuecomment-10867117.
I just checked, the scheduler refreshes the demo server at 00:00 UTC which is 7pm in plain speak. I confess though, I may have been manually refreshing the dev server earlier in the day.
Elizabeth wrote: Looks like you've made great progress simplifying the image types for dichotomous key! Here is a list of the images that should be used for certain groups to better illustrate the characters of each group (and that have better image coverage). I'll try to attach it as an excel spreadsheet, but haven't had much success with attachments at Github. So I'll also copy the fields here:
Group Family OldSpecies NewSpecies
3 Araceae Lemna valdiviana Peltandra virginica
3 Cyperaceae Cyperus retrorsus Cyperus strigosus
3 Juncaceae Juncus inflexus Juncus gerardii
3 Juncaginaceae Triglochin maritima Triglochin gaspensis
3 Smilacaceae Smilax rotundifolia Smilax herbacea
4 Amaranthaceae Atriplex glabriuscula Salicornia ambigua
4 Asteraceae Petasites japonicus Artemisia pontica
4 Hypericaceae Hypericum gentianoides Hypericum prolificum
4 Ranunculaceae Aquilegia vulgaris Xanthorhiza simplicissima
4 Scrophulariaceae Verbascum thapsus Buddleia davidii
5 Ranunculaceae Aquilegia vulgaris Xanthorhiza simplicissima
5 Apiaceae Osmorhiza depauperata Aralia elata
5 Asteraceae Petasites japonicus Iva frutescens
5 Berberidaceae Caulophyllum giganteum Berberis thunbergii
5 Cistaceae Lechea tenuifolia Hudsonia tomentosa
5 Fabaceae Tephrosia virginiana Cercis canadensis
5 Malvaceae Pseudabutilon stuckertii Tilia americana
5 Polygonaceae Polygonum buxiforme Fallopia baldschuanica
5 Rosaceae Sanguisorba canadensis Rosa virginiana
5 Solanaceae Physalis grisea Lycium barbarum
5 Salicaceae Salix occidentalis Salix bebbiana
6 ALL OKAY
7 Phrymaceae Mimulus alatus Mimulus guttatus
8 Lamiaceae Dracocephalum thymiflorum Scutellaria lateriflora
8 Nymphaeaceae Brasenia schreberi Nymphaea odorata
8 Malvaceae Pseudabutilon stuckertii Abutilon theophrasti
9 Phrymaceae Mimulus alatus Mimulus guttatus
9 Lamiaceae Dracocephalum thymiflorum Scutellaria lateriflora
10 Nymphaeaceae Brasenia schreberi Nymphaea odorata
10 Malvaceae Pseudabutilon stuckertii Abutilon theophrasti
I've also copied Sid's newfs email so he at least will get the attachment. Let me know if you have any problems seeing these or have additional questions. Best, Elizabeth
Sid, does this mean that groups 1 and 2 are also ALL OKAY
?
Sid, this should really be part of the data import. Eventually, of course, it should be a table in the app that's maintained in the app, but since we destroy the app database currently with each reload I think this has to become a CSV file?
For now I will write an import routine and the other logic this needs so that it pulls from a CSV file on my local disk. Then I'll email you the CSV for inclusion in your periodic data update.
There! I am done coding for this until you incorporate the CSV, and am emailing you the CSV file right now, Sid!
Should I stick this in the zip file that's uploaded to S3 every night?
Yes! Please do.
gobotany-csv-exports-2012-12-05.zip on the S3 bucket contains the dkey_illustrative_species.csv file you emailed me.
Sid, I think we're done! Re-open if the new import step runs into any problems, or the feature does not seem to be working tomorrow.
For the most part, issue #284 (Family and Genera thumbnails for the dkey) works well. However, there are a few places where it could be improved.
Non Representative Species
Sometimes the app presents users with a species that is not representative of the group. This causes the botanists to scratch their head in confusion.
For example, Group 4: Woody angiosperms with opposite or whorled leaves includes Petasites japonicus to illustrate the Asteraceae family.
While that species is in the family, it is not a woody angiosperm.
Using Species with More Images
To avoid displaying the Image not available yet message to the user, we could choose species with a greater number of images. For example, Group 7: Herbaceous angiosperms with superior ovaries and zygomorphic flowers
http://gobotany-dev.herokuapp.com/dkey/group-7/
Suggested Solution
As noted, most times (at last count, there were 185 families in the Flora) the displayed thumbnails work well. Talking with Elizabeth, she'd like to designate maybe 20-40 custom species that illustrate a group.
I've asked her to take a stab at this in a spreadsheet: group, family, species.