OEWG 20090113: Difference between revisions

From GO Wiki
Jump to navigation Jump to search
(New page: OBO-Edit Working Group, Tuesday January 13th, 2008 9.30 a.m. PST <br> ==Action items carried over== ==Discussion items== ==Action Items==)
 
No edit summary
 
(7 intermediate revisions by one other user not shown)
Line 1: Line 1:
[[Category:OBO-Edit working group]]
OBO-Edit Working Group, Tuesday January 13th, 2008 9.30 a.m. PST <br>
OBO-Edit Working Group, Tuesday January 13th, 2008 9.30 a.m. PST <br>
Agenda/Chair: Amina<br>
Minutes: Jennifer
Attendees: Jen, Melissa, KarenC, DavidOS, Harold, Nomi and Amina.
===Meeting Agenda===
Bug fixes and discuss ones in progress<br>
Deadline for reporting testing status<br>
Demo: Ontology Generation Plugin<br>


==Action items carried over==
==Action items carried over==
*[[OBO-Edit_2.0_Testing_Tasks | Sign up and update Testing Tasks for OBO-Edit2.0]]
* Set deadline for testing
==Discussion items==
We have agreed that we will choose a component each and commit to try to test it comprehensively by next tuesday.
Jen will try to test all of her chosen components by tuesday.<br>
Karen: I am interested to test the animated version of the graph editor. When it start failing to open or collapse nodes it gets problematic. This happens after a short while of viewing and is a big problem.<br>
David: it is the same with me. When the display gets too big is craps out. Eventually it crashes the program.<br>
Karen: I do not open whole nodes, I just select the children I want to see.<br>
Jen: are you running the reasoner?<br>
Karen: I don't think so. I have manually open the paths all up.<br>
David: you can check whether it shows all paths to the root, and if you have the reasoner on and don't filter out all transient relationship then the graph gets too huge and explodes. It is often better not to have the reasoner as this stops it showing extra stuff.<br>
Karen: It is great for viewing and I like it. It is fine without the reasoner as long as you are happy to manually open up the nodes.However it rapidly maxes out and is overloaded and uncontrollable.<br>
Amina: This is one of the major drawbacks. Chris and I will look at this in February.<br>
Karen: I don't care if I can edit in this, I just want to view. I am happy with the tree editing. (Melissa agrees)<br>
John was keen on editing the graph but I don't care. I love that this lets me control what terms I see.
David: it's great for making figures to go in presentations.<br>
Amina: I have not worked on this yet, so it would be great if you could get a really solid way to show this bug.<br>
Jen: actually you just need to browse around,<br>
Karen: any series of node opening and closing will eventually mean that it stops responding all together and then you have to restart.<br>
Harold: I am testing the viewers.
I have an issue of the graph viewer v graphviz.
A link is missing in one that is there in the other.
David: could this be to do with trimming?<br>
Harold: It wasn't a complex graph. It just had a few branches.
Jen shows url.<br>
Harold will test all viewers for next Tuesday.<br>
David looks now and says that the bug Harold has submitted has a screenshot of the graph editor so it may be that he just needs to turn features on. (display to root)<br>
However, Jen has seen this bug in the graph viewer.<br>
David - XP tab - manual commit on xp tab is show stopper for David's testing.<br>
Amina says that the known bug on this was fixed some time ago.<br>
David: I try to commit and then the genus disappears, sometimes it all disappears
if multiple differential then bits randomly disappear.<br>
David: this bug is on priority 8 since Nov since so bad.
Amina: ah I was not aware of that, I will look at that today. I'm investigating a related Dbxref Library bug that is a manifestation of the same commit bug.<br>
David: I am actually testing a lot but not doing it comprehensively.<br>
What else XP could I test? I don't use many other parts.
Jen: Perhaps if Amina fixes that bad bug then you can get back to testing that one component as that is the one thing that we need fully tested to be up to 1.101 standard.<br>
Amina: I am signing up to fix and test.<br>
melissa: signed up to test OTE, though there is an editing freeze. Hopefully this will be lifted soon.<br>
Amina: All please test using b53 as this has fixes.<br>
Amina - filter and renderer section.<br>
disjoint/tree view:
Filtering out disjoints on load into tree view so that now two terms share a single disjoint relationship.
sequence ontology has this problem too. Terms are not shown if they are children of terms with disjoint.
Sequence ontology now works.<br>


David: what about the tree editor?<br>
Amina: This is now fixed, will add the sourceforge ticket ids to the wiki.<br>
Fixed:
- [ 2184288 ] filtered save fails when using descendant filter<br>
- [ 2412063 ] OTE display issue with multiple disjoint roots<br>


Ontology generation Demo: (as shown in GO meeting by Michael Shroder)<br>
Name recommendations: Pubmed  term definition generation tool<br>
This gets information from pubmed<br>
It will be in the next release and there are docs. Please have a play and come up with ideas for a name.
Also mention if anything is missing from the docs.<br>


Graphviz viewer:
Can Jen remove the menu for selected to root and selected only and minimal selected graph?<br>
Melissa: this was for my ontology which has two roots and one has a lot of non-transitive relationships to the other. This was originally to filter out non-transitives.<br>
These features never actually worked.<br>
David: this was a really hacky fix though. Much better to use the graph editor or tree editor and specify by name the relationships that we want to see.
So: Yes you can comment this out.<br>


==Discussion items==
The save gif in Graphviz produces corrupted files. Amina volunteers to look at this and says if she does not fix it in two weeks we can comment out the feature.<br>
Jen mentions that this may be a graphviz bug.<br>


Amina mentions that it would be good to report bugs like this that we are so used to that we have forgotten even to report them.<br>
Since testing and fixing is a cycle we should set a 2 week intensive testing deadline and FEBRUARY 28th as final deadline. That gives us a month before the next GO meeting to consider the release at that point.<br>
Meeting minutes rota:
We will fill everyone in in random order.<br>
Please if you find that you are busy on a given day then write to others and swap yourself in advance of the meeting.<br>






==Action Items==
==Action Items==
Deadline for testing: January 31st
Editors dedicate more time - other members pick a component every week.
Report testing progress the following on Tuesday Jan 20th<br>
Jen: All sections I signed up for<br>
Melissa: Ontology Tree Editor<br>
KarenC: Graph Editor w/ graphics<br>
DavidOS: xps and related topics<br>
Amina testing and bugs -<br>
- Filters and Renderers section.<br>
- Cross products tab in Text Editor<br>
- Manual and auto-commit<br>
- Graphviz viewer save image gif format<br>
[[OBO-Edit_Working_Group_Meeting_Agenda_and_Minutes|Back to OEWG Meeting agenda and minutes List]]

Latest revision as of 16:39, 30 June 2014

OBO-Edit Working Group, Tuesday January 13th, 2008 9.30 a.m. PST

Agenda/Chair: Amina
Minutes: Jennifer

Attendees: Jen, Melissa, KarenC, DavidOS, Harold, Nomi and Amina.

Meeting Agenda

Bug fixes and discuss ones in progress
Deadline for reporting testing status
Demo: Ontology Generation Plugin

Action items carried over

Discussion items

We have agreed that we will choose a component each and commit to try to test it comprehensively by next tuesday.

Jen will try to test all of her chosen components by tuesday.
Karen: I am interested to test the animated version of the graph editor. When it start failing to open or collapse nodes it gets problematic. This happens after a short while of viewing and is a big problem.
David: it is the same with me. When the display gets too big is craps out. Eventually it crashes the program.
Karen: I do not open whole nodes, I just select the children I want to see.
Jen: are you running the reasoner?
Karen: I don't think so. I have manually open the paths all up.
David: you can check whether it shows all paths to the root, and if you have the reasoner on and don't filter out all transient relationship then the graph gets too huge and explodes. It is often better not to have the reasoner as this stops it showing extra stuff.
Karen: It is great for viewing and I like it. It is fine without the reasoner as long as you are happy to manually open up the nodes.However it rapidly maxes out and is overloaded and uncontrollable.
Amina: This is one of the major drawbacks. Chris and I will look at this in February.
Karen: I don't care if I can edit in this, I just want to view. I am happy with the tree editing. (Melissa agrees)
John was keen on editing the graph but I don't care. I love that this lets me control what terms I see. David: it's great for making figures to go in presentations.
Amina: I have not worked on this yet, so it would be great if you could get a really solid way to show this bug.
Jen: actually you just need to browse around,
Karen: any series of node opening and closing will eventually mean that it stops responding all together and then you have to restart.

Harold: I am testing the viewers. I have an issue of the graph viewer v graphviz. A link is missing in one that is there in the other. David: could this be to do with trimming?
Harold: It wasn't a complex graph. It just had a few branches. Jen shows url.
Harold will test all viewers for next Tuesday.
David looks now and says that the bug Harold has submitted has a screenshot of the graph editor so it may be that he just needs to turn features on. (display to root)
However, Jen has seen this bug in the graph viewer.

David - XP tab - manual commit on xp tab is show stopper for David's testing.
Amina says that the known bug on this was fixed some time ago.
David: I try to commit and then the genus disappears, sometimes it all disappears if multiple differential then bits randomly disappear.
David: this bug is on priority 8 since Nov since so bad. Amina: ah I was not aware of that, I will look at that today. I'm investigating a related Dbxref Library bug that is a manifestation of the same commit bug.
David: I am actually testing a lot but not doing it comprehensively.
What else XP could I test? I don't use many other parts. Jen: Perhaps if Amina fixes that bad bug then you can get back to testing that one component as that is the one thing that we need fully tested to be up to 1.101 standard.
Amina: I am signing up to fix and test.
melissa: signed up to test OTE, though there is an editing freeze. Hopefully this will be lifted soon.
Amina: All please test using b53 as this has fixes.

Amina - filter and renderer section.
disjoint/tree view: Filtering out disjoints on load into tree view so that now two terms share a single disjoint relationship. sequence ontology has this problem too. Terms are not shown if they are children of terms with disjoint. Sequence ontology now works.

David: what about the tree editor?
Amina: This is now fixed, will add the sourceforge ticket ids to the wiki.
Fixed: - [ 2184288 ] filtered save fails when using descendant filter
- [ 2412063 ] OTE display issue with multiple disjoint roots

Ontology generation Demo: (as shown in GO meeting by Michael Shroder)
Name recommendations: Pubmed term definition generation tool
This gets information from pubmed
It will be in the next release and there are docs. Please have a play and come up with ideas for a name. Also mention if anything is missing from the docs.

Graphviz viewer: Can Jen remove the menu for selected to root and selected only and minimal selected graph?
Melissa: this was for my ontology which has two roots and one has a lot of non-transitive relationships to the other. This was originally to filter out non-transitives.
These features never actually worked.
David: this was a really hacky fix though. Much better to use the graph editor or tree editor and specify by name the relationships that we want to see. So: Yes you can comment this out.

The save gif in Graphviz produces corrupted files. Amina volunteers to look at this and says if she does not fix it in two weeks we can comment out the feature.
Jen mentions that this may be a graphviz bug.

Amina mentions that it would be good to report bugs like this that we are so used to that we have forgotten even to report them.
Since testing and fixing is a cycle we should set a 2 week intensive testing deadline and FEBRUARY 28th as final deadline. That gives us a month before the next GO meeting to consider the release at that point.
Meeting minutes rota: We will fill everyone in in random order.
Please if you find that you are busy on a given day then write to others and swap yourself in advance of the meeting.


Action Items

Deadline for testing: January 31st Editors dedicate more time - other members pick a component every week.

Report testing progress the following on Tuesday Jan 20th
Jen: All sections I signed up for
Melissa: Ontology Tree Editor
KarenC: Graph Editor w/ graphics
DavidOS: xps and related topics
Amina testing and bugs -
- Filters and Renderers section.
- Cross products tab in Text Editor
- Manual and auto-commit
- Graphviz viewer save image gif format


Back to OEWG Meeting agenda and minutes List