Organising photos

Thanks, Rasmus. Sorry I forgot to mention that addition to the wiki article in my previous post.

Maybe I have misunderstood something but I can’t see why we have to choose between albums (which are basically folders, aren’t they?) and hashtags.

I personally classify thousands of photos on my computer : I indeed use keywords to be able to reference them and easily find every photo related to #ocean or #family or anything. But these photos are also primarily grouped into folders that generally correspond to the event. Both ways are complementary, but grouping them into folders appears to me as the priority. It is also what make more sense when looking at somebody else’s pictures : if I want to see holiday pictures of my friend, I am not interested in viewing only #beach photos or only his #food ones : I would more likely want to have an overview, including all the places he visited and activities he did in a chronological order (and I am probably not interested in viewing again his #beach photos taken the year before in another place).

It would definitely be a good feature to associate hashtags to photos, but why not being able to group them into folders first?

Here is what would make sense to me:

  • having a folder named Uploaded pictures where all the pictures posted in my stream would be stored by default (keeping the link to the original post could be useful since sometimes the photo doesn’t mean anything without the initial comment)
  • the ability to create other folders (albums) such as Holidays in Italy where I would be able to store some photos
  • a system of hashtag that works as filter over all the folders. It would allow me to fetch all #beach pictures for example, and allow general referencing on Diaspora (in accordance with the privacy settings)

Hope I’m clear enough, and thank you for the wiki page!

The point of the previous vote is that hashtags will be used to organise photos (into ‘albums’ if you want to call them that). You would group them into ‘folders’ by associating hashtags with them. In your example, someone could then click on #holiday2015 (or a combination of #2015 and #holiday) to see all the photos from that holiday. If they wanted to see only the photos taken on the beach during that holiday, they could add the #beach tag as a filter. The photos would appear to the user to be in an album, but that album would be created by use of hashtags.

As Jonne Haß said above: the question is not whether to do this, the question is how to achieve it. The decision taken is to use Diaspora’s existing #tag system to organise photos into albums. This means the albums will be dynamic and more flexible: a photo can be part of more than one album, and if you want to move a photo from one album to another, just change the tags attached to it. (At least, that’s how I envision it working.) That’s why the outcome said ‘use Diaspora’s #tag system to organise photos rather than using Facebook-style static albums’.

What you have written in your previous post is pretty much the outcome of the previous vote. That’s why the new proposal is superfluous.

OK, understood. :wink: You also said previously that

one would need clear evidence that the previous decision is no longer applicable/appropriate for the project

I agree with you, and I tried to get a clear view of the reasons that lead to the decision. The problem is that in previous vote (and discussion), the only visible arguments are the one of people that voted against. What are the real advantages of such a system, only based on hashtags? Albums would then be dynamic, OK, but what is the advantage of that?

I can notice from the experience of new friends arriving on D* that the use of hashtags in posts is far from being intuitive. I fear that managing photos that way would be even more complicated. I can’t help thinking that it is absolutely essential for a tool such as D* to remain very user friendly. Then every complicated (or at least not usual) feature has to be supported by strong arguments. I can hardly see anything of that kind on this topic. I’m not pretending there are’nt, just saying they are not clear to me.

Note that there is a discussion on github preceding this one. There, you find the argument that tags&filtering combines better with aspects than a folder system.

Ok, thanks, I know that. I’m sorry to insist, but even there I still don’t see any clincher. The advantage of being able to put a picture in multiple albums is nice, but not sufficient to put aside the classic album way.

The advantage of being able to put a picture in multiple albums is nice, but not sufficient to put aside the classic album way.

That’s fine for you to feel that, but understand that the Diaspora community already agreed to do just that, and now all that’s needed is people to code it.

That’s sad we cannot have an argued discussion on that issue and keep things moving…

You can discuss, of course. But if you define “keep things moving” that the community must agree on your “albums are better than tags”, than you’re wrong. Do you think, we missed any argument of yours?

The advantage of being able to put a picture in multiple albums is nice, but not sufficient to put aside the classic album way.

Note that there are more arguments in the github discussion than just this one.

@rasmusfuhse “keep things moving” -> It seems strange to me to see something that has been voted a long time ago and still not implemented to remain definitively locked (“Diaspora community already agreed”). Diaspora community may have changed, and users expectancies too. So why don’t we poll again (not every month, but once a year on such an important topic seems reasonable to me)?

Then I am not exactly saying that “albums are better than tags”. I think personally that indeed they cannot replace them, and that they rather are complementary. So I would prefer to add them to the album feature. But that is my personal view, and thus not the most important. My point is that

it is absolutely essential for a tool such as D* to remain very user friendly. Then every complicated (or at least not usual) feature has to be supported by strong arguments.

And again, (sorry @dremodaris ), I don’t see such clinchers.

Whether they are clinchers is subjective, but the following quotes combined constitute, at least, an argument:

flaburgan

But if we give full album access to every new person in the aspect, will you remember that you give access to this album one year before? You know, the one with awesome birthday pics? Now, you add your mum to this aspect, and magically she discovers this album. This is typically the kind of stuff we want to avoid in diaspora*.

jhass

And that’s then gets confusing, two of your friends sitting together looking at an album and noticing that one can see photos the other can’t, in the same album. Therefore my suggestion to not implement albums, but a nice view for all photos that were shared with filters based on the hashtags in the associated posts.

So if you stick to your opinion, I think you have to answer these concerns.

On the first point : as far as I understand, the tag system would require privacy settings for any picture independently. It is also possible to have this fine tune for independent pictures sorted in folders, doesn’t it? Concerning privacy management, you proposed earlier a nice way to do the job, which seems applicable to both picture organisation systems (and could be extended to all posts, in my opinion).

The problem of the two friends looking at your pictures is the same as those same friends looking at your whole profile : they will also notice they don’t see the same posts. I don’t understand why this problem is specific to pictures, and why sorting them using tags instead of albums would solve the problem. I see it as the normal consequence of using aspects.

I more or less agree with what you say, but then an individual picture has everything in common with a post, except that it has a unique “album” associated with it. Now, this uniqueness is more of a restriction than a feature, and if you drop it, then an album is no different from a tag. Perhaps in the context of photos, there could be a user interface that makes tags look a lot like albums, but I think tags are more flexible than albums in every respect, so I don’t see the added value of introducing a new organisational concept of albums.

Also, I disagree on your premise that tags are difficult to understand for users. GMail has them, Thunderbird has them, Wordpress blogs have them… It is a quite well-known concept by now.

Mmmh I’m not sure we have the same idea in mind. Let me describe what would make sense to me (very basic proposal actually):

  • having a folder named Uploaded pictures where all the pictures associated to a post would be stored by default. It would basically be the same content as the “Photos” item which already exists on people’s profiles (keeping the link to the original post could be useful since sometimes the photo doesn’t mean anything without the initial comment)
  • being able to create other folders such as Holidays in Italy, where I would be able to upload some photos (or move them from another folder). For each picture, I would be able to add a caption and some tags if I want to. My friends would be able to drop a comment.
  • a system of hashtags that works as filters over all the folders. It would allow me to fetch all #beach pictures in the folders I have access to, and allow general referencing on Diaspora (in accordance with the privacy settings as well)
  • individual privacy settings for every picture that are in Uploaded pictures (inherited from original post), and then one setting per created folder

Main differences with conventional posts are : we can upload several pictures at once, storing them in folders which are easy to find and to browse. These folders could also be a simple tool to manage privacy settings, easy to understand, and easy to check in one glance.

Counterproposal: We can upload several pictures at once, tagging them with tags that are easy to find and use as filters. These tags could also be a simple tool to manage privacy settings (make all photos with tag T available to aspect A), easy to understand, and easy to check in one glance. :wink:
Tags could do anything folders can do, and if necessary, tags could be made to look like folders by making some tweaks at the UI-level. This would be a second step with much less priority.
Of course folders can form hierarchies, but you could have tags of the form #tag:subtag:subsubtag.

Basically, I think folders can be built/mimicked based on tags, so that organizing photos using tags is a very sensible first step.

you could have tags of the form #tag:subtag:subsubtag

Ok, good point, I haven’t thought about it.

But sorry, I still think playing with tags is much less intuitive for most people (we have to keep in mind that people like you and me, spending time on Loomio and gitHub or mastering Markdown are not representative of the average population), and I’m afraid it would rapidly getting messy. With tags you can easily create mutually exclusive conditions without notice, for example.

Then how would we deal with big picture collections: I click on the profile of some friend of mine that has hundreds of pictures referenced with hundred of tags. How would things be displayed: a list of hundreds of tags and one has to pick some randomly? Or a page with hundreds of photo thumbnails and filters at the top I have to try one after another?

And what if I upload dozens of holiday pictures and forget to give them a common tag?

I think that system is very convenient to retrieve your own photos, since you know which tag refers to what (that’s the way I use for the pictures I take). But when you jump into somebody else’s collection you need a clear view of the hierarchy since you don’t have a clear idea of what you are looking for or which pictures are present.

Globulle, all your arguments can be used to say, that albums are unusable by the users.

How would things be displayed: a list of hundreds of tags and one has to pick some randomly?

I now just replace one word with another: How would things be displayed: a list of hundreds of albums and one has to pick some randomly?

And what if I upload dozens of holiday pictures and forget to give them a common tag?

And what if I upload dozens of holiday pictures and forget to [add] them [to] a common album?

Tags have two main advantages over albums:

a) They are more flexible, because a photo can have more than one tag. If you prefer having exactly one tag per photo (like it would be with albums), then you can use tags just like this.

b) Photos can inherit the tags from the postings, they are uploaded with. So even a user who never ever visited the dedicated photos-section in diaspora, but only posted stuff, might have a well tagged bunch of photos.

And now what if a user uploads trillions of photos and adds thousand of tags? Well, the same user might get into trouble with albums as well.

Of course you can implement albums quite nicely so that they work fine for most users. But the same applies to tags. You can display tags exactly like you might have displayed albums to the user.

and I’m afraid it would rapidly getting messy.

I agree that tags like #thisisasentencewithoutspaces:subtag look messy. Perhaps, especially in the context of photos, it would be worthwile to find a way to make tags look more appealing. However, I think much of this can be done at the UI level. E.g. we could allow tags like #{this is a sentence with spaces}:subtag and then make them look like colored boxes or so; both while typing and while viewing.

With tags you can easily create mutually exclusive conditions without notice, for example.

Could you rephrase?

a list of hundreds of tags and one has to pick some randomly?

I’d say the viewer could choose to sort them alphabetically or by prevalence. Perhaps the profile owner could also create a list of “favourite filters” or something like that.

I think that system is very convenient (…)

This argument is entirely applicable to albums.

Photos can inherit the tags from the postings, they are uploaded with.

I think this is an excellent idea.

@globulle

So why don’t we poll again (not every month, but once a year on such an important topic seems reasonable to me)?

Because unneeded bureaucracy isn’t helping anyone. If a lot of the community has changed, fresh discussion presents a new idea, a developer wants to implement it in an alternative way, then it will probably happen anyway - no need to poll just because a certain amount of time has passed. :slight_smile: