Broken links to ontologies #257
Comments
|
I have an SVN checkout of the VTO releases. The files are too big for GitHub. I'll try to get these online elsewhere as soon as possible. |
|
Thanks @digitalneoplasm we're on it @balhoff if it's individual file size limits you're hitting you could try modularizing, e.g. ontodev/robot#39 |
|
For now I put VTO releases in S3. Pull request is #259. |
|
Here are the purls still failing. Is anyone chasing this up? chmo - should be https://raw.githubusercontent.com/rsc-ontologies/rsc-cmo/master/chmo.owl |
|
I am working on OGG… From: simonjupp [mailto:notifications@github.com] Here are the purls still failing. Is anyone chasing this up? chmo - should be https://raw.githubusercontent.com/rsc-ontologies/rsc-cmo/master/chmo.owl — Electronic Mail is not secure, may not be read every day, and should not be used for urgent or sensitive issues |
|
Hi Oliver and all, Thanks for letting me know. Yes. I am responsible for the OGSF. Thanks, On Mon, Jun 13, 2016 at 9:20 AM, He, Oliver yongqunh@med.umich.edu wrote:
|
selewis
added
bug
website
|
Hey! xl.owl Is there a way I can get an up to date dump as a zip file including all of them? Is tried chasing down the ontologies using their project links, but failed with the first three attempts, so I though I might as well ask here. |
|
Hi, we recently made a change in how we distribute some of the ontologies: OBOFoundry/purl.obolibrary.org#261 The change went successfully, but it uncovered a hidden issue. For many ontologies the central build was failing, and what was being retrieved may have been a stale copy. This mostly applies to ontologies that are semi-abandoned. However, there are active ontologies like ddanat that are failing, and we will attend to these first (OBOFoundry/purl.obolibrary.org#266). I note your list contains some ontologies that are not officially registered with OBO - for example OGI. In some cases you have the case wrong (e.g DINTO should be dinto) OBO-tech people: This is all documented in http://obofoundry.org/faq/what-is-the-build-field.html The way to vet this is to look at the log output from the simple obo build job For convenience, this can be done as in this repo. The most common cause of error here is a source obo file that does not declare an ontology tag. We must inject this, see: |
|
But really we should be pushing more groups to take control of their builds. Unfortuately a lot of important ontologies are released by their maintainers as simple obo files made with OE |
|
Hey Chris, thanks for that. The DINTO download button for example does not even point to a file: It points to a GitHub repo home (https://github.com/labda/DINTO). And with respect to OGI, perhaps you are thinking about an ontology with the same id, but there is definitely something here: http://www.obofoundry.org/ontology/ogi.html Just to be plain, is there a chance I can get these 28 ontologies somehow with a bit or URL hacking (trying the old berkeleybob etc), or is it best I wait until the issues are resolved? I need as much of the OBO foundry as I can get until Wednesday next week. What I dont have by then, I will simply ignore for the experiment. Cheers. |
|
OGI - ah yes, I was getting confused with OGG. OK, if you want a quick hack by wednesday, then for any failed PURL, try getting it from the old http://www.berkeleybop.org/ontologies/ |
|
Perfect, thanks! |
|
@cmungall http://ci.monarchinitiative.org/view/dev/job/load-scigraph-ontology-on-dev/lastBuild/console |
|
Hey, here a revised list of the few in question. (Sorry for the typos before, I copied them from an unreadable download log of chrome) xl.owl - id: xl, name: Cross-linker reagents ontology, home: http://www.psidev.info/groups/controlled-vocabularies |
|
ceph and caro fixed (it looks like the CARO that was being resolved was "caro1". We made "caro2" some time ago but caro1 was never replaced: obophenotype/caro#13 -- in fixing this issue I just used caro2 as we have a github repo for this. See OBOFoundry/purl.obolibrary.org@2df5887) |
|
Update: there are now only 24 ontologies with broken PURLs: out of those, 13 can be found on the old berkleybob pages, but the purl points to a NoSuchKey Error message. The remaining 11 have no listing on the old berkleybob pages, and mostly fail because of googlecode. However, at least four (dinto, cheminf, ogsf, pdro) should be super easliy fixable by redirecting the purl to github (they all have github repos). xl clearly does not exist (not even a OBO home page) and should be removed from the foundry web site listing (or at least made obsolete). Berkleybob (13): All have NoSuchKey errors, but are available from the old berkleybob pages. Not on berkleybob / PURL does not resolve 404 (11): |
|
On 5 Dec 2016, at 3:55, matentzn wrote:
xl - 404, <http://obofoundry.org/ontology/xl.html> \- DELETE FROM
LISTING
I have found the source of this oddity
ontologies/xmlmod.md:
```
id: xl
label: XL
description: A structured controlled vocabulary for cross-linking
reagents used with proteomics mass spectrometry.
```
the fileame should always match the idspace. We need a check for this
btw, it's berkeleybop.org (not bob)
|
|
I have updated the Transmission Process and Symptom Ontology URLs to their GitHub sites. |
Fixes #266 See also dictyBase/migration-data#6 and #257
|
Still broken links as of today: rnao |
|
goodness, sep is still on googlecode, I thought all the PSI ontologies
had been moved across, I'll follow up, and look into rnao too
…On 28 Mar 2017, at 3:29, matentzn wrote:
Still broken links as of today:
rnao
sep
--
You are receiving this because you were mentioned.
Reply to this email directly or view it on GitHub:
#257 (comment)
|
|
I produced another OBO dump today. Out of 154 downloadable ontologies on the 24th April 2017, The following twelve were not: Old berkleybob references (8) Still hosted on Googlecode (4) |
|
cc @graybeal |
|
I encountered broken PURLs for: cro: http://purl.obolibrary.org/obo/cro.owl ... which didn't seem to be mentioned yet in comments above. Also, xl is marked as fixed in a comment above, but the xl page results in a 404 error, and the PURL is broken: http://purl.obolibrary.org/obo/xl.owl |
|
I produced another OBO Foundry dump today! 404 - Not found (2): Berkleybob, no such key (6), (all same as 7 months ago) Googlecode (4), (all same as 7 months ago) Works manually
|
As of a few minutes ago there are broken links to the following ontologies:
aero, cheminf, chmo, cmf, cvdo, ido, mirano, ncithesaurus, ogg, ogi, ogms, ogsf, rxno, and vto.
It seems the authors have not updated their purls since the google code shutdown. There are versions of aero, ido, ogg, ogi, ogms, ogsf, and vto available on BioPortal. cheminf, chmo, and rxno are available on their respective github pages. I was unable to find cvdo, mirano, ncithesaurus
For ncithesaurus the issue is more critical - http://www.obofoundry.org/ontology/ncithesaurus.html does not seem to exist. Its website (https://cabig.nci.nih.gov/concepts/EVS/) is also defunct.
The text was updated successfully, but these errors were encountered: