search

ufal / clarin-dspace

clarin-dspace digital repository based on DSpace and LINDAT/CLARIN DSpace
http://lindat.cz
BSD 3-Clause "New" or "Revised" License
27 stars 18 forks source link

Access other versions of a resource #412

Closed stranak closed 6 years ago

stranak commented 8 years ago
stranak commented 7 years ago

I think the first point has been true for a while now.

Ansa211 commented 6 years ago

I am not sure about point 1, I just added a relation in one corpus and had to add the corresponding relation to the other corpus manually (so if it is added automatically, it does not happen immediately).

related:

stranak commented 6 years ago

I would say the full listing can’t (rather shouldn’t) show something that is not in the metadata. I.e. the full listing is showing exactly what is stored in the metadata and I think it should keep doing so.

BUT I think that showing titles of the replaced submissions is of course the right approach (probably with the PID). Just not here, but in the special “version browser” component that I hope is in development.

On 23 Nov 2017, at 14:12, Anša Vernerová notifications@github.com wrote:

I am not sure about point 1, I just added a relation in one corpus and had to add the corresponding relation to the other corpus manually (so if it is added automatically, it does not happen immediately).

Related: it would be nice if instead of showing the handle of the replaced/replacing submission, the box "This item is replaced by a newer submission" and the full listing of metadata would (also) show the title of the linked submission. https://user-images.githubusercontent.com/1833866/33174424-44c0f764-d058-11e7-96e8-4e954abc6982.png — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/ufal/clarin-dspace/issues/412#issuecomment-346614720, or mute the thread https://github.com/notifications/unsubscribe-auth/AAWpoor0SnZO-UexHVzRGVuELC8uiIloks5s5W8qgaJpZM4GilR2.

Ansa211 commented 6 years ago

But don't we want force the relations dc.relation.isreplacedby and dc.relation.replaces to be strict inversions of each other? That was sort of my expectation. From user's point of view, a single relation would then be enough (probably dc.relation.replaces, so that when adding the new submission, the user does not have to edit the metadata of any other submission).

BTW, I just added a handful of these relations for our own Prague corpora (PDT 2.5 < PDiT 1.0 < PDT 3.0 < PDiT 2.0); is there any way to strongly remind the person entering a submission to make sure they do not forget to add this information?

stranak commented 6 years ago

What we show users and what we store in metadata doesn’t need to be exactly same. I.e. we store ISO 639-3 language code, but we show users “Czech”. The full metadata listing is not a user view.

--p

On 23 Nov 2017, at 14:51, Anša Vernerová notifications@github.com wrote:

But don't we want force the relations dc.relation.isreplacedby and dc.relation.replaces to be strict inversions of each other? That was sort of my expectation. From user's point of view, a single relation would then be enough (probably dc.relation.replaces, so that when adding the new submission, the user does not have to edit the metadata of any other submission).

BTW, I just added a handful of these relations for our own Prague corpora (PDT 2.5 < PDiT 1.0 < PDT 3.0 < PDiT 2.0); is there any way to strongly remind the person entering a submission to make sure they do not forget to add this information?

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/ufal/clarin-dspace/issues/412#issuecomment-346623612, or mute the thread https://github.com/notifications/unsubscribe-auth/AAWpotEDJhDNi8j3807bfpuRNHGlgYjWks5s5XhVgaJpZM4GilR2.

kosarko commented 6 years ago

Yes the plan is to show names "of the versions" in the simple item view as a dropdown; the tricky part is that dc.relation.replaces or the handle/pid can lead to a different repository; you then don't know how to obtain the item name.

The automatic replaces/replaced by happens when you start a submission as a version of an older one, or when you fill that in during the submission (user's see both the handle and the title).

Not sure if we also bound the auto fill of replaces/replaced by to edit metadata; will check

stranak commented 6 years ago 
    For the replaces from a different repository I would display just the PID and some icon they it leads “away”. 

    --ps

On Thu, Nov 23, 2017 at 3:06 PM +0100, "Ondřej Košarko" notifications@github.com wrote:

Yes the plan is to show names "of the versions" in the simple item view as a dropdown; the tricky part is that dc.relation.replaces or the handle/pid can lead to a different repository; you then don't know how to obtain the item name.

The automatic replaces/replaced by happens when you start a submission as a version of an older one, or when you fill that in during the submission (user's see both the handle and the title).

Not sure if we also bound the auto fill of replaces/replaced by to edit metadata; will check

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or mute the thread.

Ansa211 commented 6 years ago

Is this correct?

PDT 3.0

PDT 2.5

kosarko commented 6 years ago

It is correct, in the sense of what a version is that we use.

PDT 2.5 is replaced by PDT 3.0 is replaced by PDiT 2.0. In this sense the newer resource always improves/extends the older one.

On the other hand PDT 3.0 replaces PDT 2.5 and PDiT 1.0. PDT 3.0 "merges" (that's the improve/extend part of version) them, that's why they are both shown as previous version of PDT 3.0. The two resources (PDT 2.5, PDiT 1.0) don't necessarily share a single line, so you cannot assume one is a version of the other

I did embed a comment about this into the code. The thing is that you follow the replacedby relation from the current resource; then you follow the replaces relation and join the two lists together. You don't switch the relation as you are visiting the related resources (from PDT 2.5 to PDiT 1.0 you get throught replacedby - PDT 3.0 - and replaces - PDiT 1.0 )

Does it make sense? Do you see there's no immediate relation between PDT 2.5 and PDiT 1.0?

Btw there was some debate whether PDiT 2.0 should even be replacing PDT 3.0; don't know what the status of PDiT 1.0 in this chain is. @stranak any comments?

Ansa211 commented 6 years ago

I see, I did not realize that PDT 2.5 and PDiT 1.0 were simultaneous rather than successive projects. I might have been confused by the wording on the pdt2.5 page, see below in bold.

The wording on the ufal.mff.cuni.cz/pdit and similar webpages is usually rather careful:

stranak commented 6 years ago

Well, maybe at some point we'll need to draw a tree there :-)

--p

On 19 Feb 2018, at 14:20, Anša Vernerová notifications@github.com wrote:

I see, I did not realize that PDT 2.5 and PDiT 1.0 were simultaneous rather than successive projects. I might have been confused by the wording on the pdt2.5 page, see below in bold.

The wording on the ufal.mff.cuni.cz/pdit and similar webpages is usually rather careful:

PDT2.0 http://ufal.mff.cuni.cz/pdt2.0/ -> Please note that new versions of this corpus have been published: PDT 3.0 (2013), PDiT 1.0 (2012), PDT 2.5 (2012). PDT 2.5 http://ufal.mff.cuni.cz/pdt2.5/ -> The Prague Dependency Treebank 2.5 annotates the same texts as the PDT 2.0. Please note that since the release of PDT 2.5 in 2012, new versions of the corpus have been published: PDT 3.0 (2013), PDiT 1.0 (2012) PDiT1.0 http://ufal.mff.cuni.cz/pdit -> Annotation of discourse relations is a project related to the Prague Dependency Treebank 2.5 (PDT; Bejček et al. 2011), which is a revised, updated and extended version of the Prague Dependency Treebank 2.0 (Hajič et al. 2006). Please note that an updated version of the corpus was published in December 2013 as Prague Dependency Treebank 3.0. Even more updated version (enriched by the annotation of secondary connectives) was published in December 2016 as Prague Discourse Treebank 2.0. PDT3.0 http://ufal.mff.cuni.cz/pdt3.0 -> The Prague Dependency Treebank 3.0 (PDT 3.0) annotates the same texts as the PDT 2.0 (Hajič et al. 2006), PDT 2.5 (Bejček et al. 2011), and the Prague Discourse Treebank 1.0 (PDiT 1.0, Poláková et al. 2012). Please note that an update of the annotation of discourse relations in PDT (newly enriched by the annotation of secondary connectives) was published in December 2016 as Prague Discourse Treebank 2.0. PDiT2.0 https://ufal.mff.cuni.cz/pdit2.0 -> It also contains a revised annotation of primary discourse connectives from the previous versions, PDiT 1.0 and PDT 3.0. ... Annotation of discourse relations is a project related to the Prague Dependency Treebank 3.0 (PDT; Bejček et al. 2013), which is a revised, updated and extended version of the Prague Dependency Treebank 2.5 (Bejček et al. 2011) and the Prague Dependency Treebank 2.0 (Hajič et al. 2006). — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/ufal/clarin-dspace/issues/412#issuecomment-366691261, or mute the thread https://github.com/notifications/unsubscribe-auth/AAWpoloM4ZViqSxg-MyIUr-Pn3JTYL4rks5tWXU4gaJpZM4GilR2.

Ansa211 commented 6 years ago

Not just a tree, but a directed acyclic* graph - PDiT 1.0 and PDT 2.5 have both a common ancestor (PDT2.0) and a common descendant (PDT3.0).

cyplas commented 5 years ago

Accessing other versions doesn't seem to work (the same) in different browsers. In chrome, it seems to work fine. But in firefox, some versions are not clickable while others take me to totally different links (e.g., see https://lindat.mff.cuni.cz/repository/xmlui/handle/11234/1-2515).

stranak commented 5 years ago

@cyplas There seems to be an issue for it already: #874 I have tested it in Safari and all works fine, so it is probably correct it is only Firefox.