-
-
Notifications
You must be signed in to change notification settings - Fork 39
Feedback on 2.4.1 #20
Comments
Thanks again for all your help and suggestions David. To touch on each of your points individually...
Oh well; ya win some, ya lose some. I think as the bugs get worked out people will come back. Not terribly worried.
You totally did not have to do that. Your help and suggestions have been more than enough in return! But I appreciate it nevertheless 👍
I was going to pursue having stuff autopopulate, but unfortunately most WP sites will have trouble with CORS requests. I decided to not take the headache and just do it the way it is.
Good thoughts. The only issue with going about it that way is that you and everybody else would have to stay committed to having the plugin installed for the long haul. The moment somebody uninstalls the plugin, everything would fall apart. Having said that, I'm not totally opposed to the idea.
It certainly would. Good point.
I'll open a new issue for this. Thanks
Try adding So, for example: .abt_arrow_up {
border-color: transparent transparent magenta !important;
}
.abt_arrow_down {
border-color: magenta transparent transparent !important;
}
That selector is for the close button on mobile view. There's a little circle with an X on the top right that mobile users have to touch for the tooltip to close.
Thanks for letting me know. I believe I already opened an issue for this. It'll be fixed on the next update.
Yeah, I don't do anything that messes with the favicons in this plugin so it's likely unrelated.
Hmm.. Interesting. Any way you could send me a screencast of that?
This was fixed to my knowledge. Perhaps you're still using a cached version of one of the older plugin files? Thanks again! |
*Re Proposed Regenerate References feature: * A way to implement this kind of "regeneration" of citations and references that would not be a problem if people wanted to stop using the plugin could be as follows. As of right now, if you deactivate the plugin, you see [cite num ="x"] in the articles, and you see a references list at the end. This is alright, since those citations still make sense. So presumably this is a state a person can still delete the plugin in and consider it to be acceptable. So if the final state of a "regeneration" mode was just the same as this, then there would not be any additional irritation caused relative to the present state of deactivating the plugin. In other words, say you are editing a post that has inline citations in it (i.e. [cite num="x"], [cite num="y"], etc), and of course there is also a reference list at the end. Then say you clicked a "regenerate citations/references" button. Then an algorithm goes and does a bunch of switching things around. CItation 2 turns into citation 6, and citation 5 turns into citation whatever. And reference 4 is moved to number 9 on the list. And reference 23 is moved to number whatever on the list, etc. Okay, so the code does its thing, and now you are still left with a post that is no more cumbersome than it is at present, should the user want to stop using the plugin. In more other words...The idea I previously gave you of having only "inline references" would be totally scrapped. There would still be both inline citations and a reference list at the end , both of which would still be generated using the present smart biblio mode. But there would be an additional feature that if you click it, it does does crazy re-ordering of the references list and re-numbering of inline citations. Anyway, I've been thinking about the algorithm for how that might work. It's doable, but could be a real pain. If you get tempted, let me know and I will write up some pseudo code. Basically, it's feasibility would depend on being able to find specific instances of "[cite num= “i”]" within a post, for a given "i". If that's not possible then I think the only way to implement such a feature would have to involve pissing people off who wanted to stop using the plugin. Re: color of arrows) Adding !important, fixed it. Now I can color my arrows. Ideally I would also want a border for the arrows, which would seamlessly blend into the border of the rest of the tooltip. But that's no big deal, and it is fine the way it is. Re: Arrow alignment of tooltip on mobile: See this url: Re close button on mobile tooltips Re multiple PMID bugs: |
Okay, I figured out the issue with multiple PMIDs. It works fine, unless you input a PMID that does not exist. For example, 1234568 does not exist in pubmed. That's why I was getting the bugs. So maybe you need a try, catch block or something. |
I've thought about this a bit more as well and I think the way to pull it off would be to store a hidden But 2 issues come to mind with that:
A reasonable way to go about doing this I think would be to do it similar to how it's currently done if you have a smart-bib enabled where the list of citations in your list pops up and you can click on each one you want in that position. Open to suggestions though.
Yeah, it won't be possible to add a separate border color on the arrows due to the fact that the arrows themselves are borders.
I think that's an issue with the mobiletest site and not reality. On both my personal cell phone (iphone 5s) and on chrome devtools mobile mode it looks appropriate
Yeah, I think the tester is flawed. See my above screenshot for what the close button looks like.
I've got a few of those, but it's difficult because there are three separate async functions that happen in sequence per PMID. I'll look into better scrutinizing the error handling in future updates. 👍 |
Just brainstorming cases that might be helpful to think about, whatever the method may be:
And if the user while writing a post can't see the actual citations/references they have inputted, but just see some mysterious ID, then that might make the whole thing not worth it. But I think this is not necessarily an issue if I understood your suggestion correctly. If you are adding some kind of span tags, then if those would still be invisible to a website reader upon deactivating the plugin, then maybe that would not piss plugin users off should they want to stop using the plugin. Food for thought: What operations would a user have to do manually right now if he wanted to make changes to his citations/references? Or say he wanted to try and manually make the post look like the citations are in ascending order? And does the plugin accommodate for these operations already?:
Say a new feature took those citations 4,5,6,6,1,4,3 and automatically renumbered/reordered them. What would the result look like. Well, it would look like 1,2,3,3,4,1,6, if I am not mistaken. Note that 1 appears twice in a somewhat non-elegant way. But this has nothing to do with a lack of the plugin feature in this example. This is just inevitable owing to the way the user wrote the article. Regarding your popup idea, and user selecting where they want citations: I am not sure that is a good idea. That is basically asking a user to manually indicate optimal numbering of citations as best as they are able to. A robust algorithm would automatically do this better than a user would do manually in any case. And I think a robust algorithm would still be needed to accomodate such manual input anyway. Could be wrong though. Maybe some kind of compromise is an option. Like enable three basic operations the user can do through popup windows:
Na, I don't think implementing such basic manual operations would be any easier than implementing a completely automated thing. I'm just thinking out loud. Nothing in this post is necessarily important. |
@metallikat36 Interesting ideas. I think I may have figured out how to go about doing it, but the execution will likely take a little while. In short, I think I might be able to add a sidebar meta box that can collect a list of references and then interop between there and the post editor. The only way to really go about doing that is to completely ditch shortcodes and add html on the fly. I haven't totally thought this through, but I'd imagine it would be possible to reverse engineer the editor's interactive Each reference would have to be communicated to the editor using some type of unique ID and then, when the user chooses to generate a bibliography, the program could just apply a simple sort function based on reference occurrence. Having said that, this will take a while. Fair warning. |
Understood. Plugin in its present state is awesome and totally workable - in fact the only workable plugin out there at present that I know of. So if this all gets too frustrating, no worries if we just scrap the whole idea. |
@metallikat36 Closing this issue for now. Let me know if I forgot about any of your suggestions on this next update (if so, open a new issue) 👍 |
Okay, so I have tested the plugin and here is all my feedback:
.abt_arrow_up { border-color: transparent transparent magenta; } .abt_arrow_down { border-color: magenta transparent transparent; }
https://www.browserstack.com/start
The "Live" tab lets you test live browsers live. The tooltips do not seem to appear on IE still.
a) When I am not in smart-biblio mode, and I add 2 PMIDs, it seems to work.
b) When I am not in smart-biblio mode and I add 3 PMIDs, it just makes the loading-wheel spin forever
c) when I am in smart-biblio mode, and I add 2 PMIDs, sometimes it works, and sometimes it does nothing at all.
d) when I am in smart-biblio mode and I add 3 PMIDs, it adds 1 inline citation number (should add three) and does not add nothing to the references list.
Holy guacamole. I hope all this feedback is considered helpful information rather than a pain in the ***!
David
The text was updated successfully, but these errors were encountered: