1 Topic by fleurdelys (edited by fleurdelys 2012-09-07 00:57:53)

  • Registered: 2012-08-25
  • Posts: 4

Topic: Several Remarks Regarding the Website in General

Hi all!

First of all, I should say many positive things about this site because there are plenty. Alas, good news are frequently pushed aside by the bad news and in the like manner it often happens with praise deserved. Therefore let me simply say: Congratulations for this site!

And no, it's not that there are any bad news, really. Rather, while using the Ath. website I encountered some minor difficulties or problems which I wasn't able to solve or overcome. I guess other users will have made similar experiences and thus it might be about time to mention some of them. Moreover, I would like to make some suggestions as to future components and/or rules of the site. I hope this will be of benefit for all of us (or at least serve as a basis for discussion).

First, problems:

(a) When marking a person as sculptor, I wasn't able to upload any artwork. In fact, clicking upon "Art" does not, on my computer, produce any result, and thus there is no way to upload anything (i.e., nothing like "add more" or similar appears as it does in the case of painters).

(b) When adding or editing a museum, this cannot be done if the devilish automatic device based on Google Maps cannot locate the address. Obviously, in some cases there is no clear or fixed address. There are some definitely erroneous or incomplete entries in the museum list that, however, cannot be changed because of the said problem.

(c) Some names in the alphabetical listing of names definitely do NOT appear in the place they should appear. I even checked, in a few cases, whether someone has misspelled a name in the rubric "Name used for listing", which sometimes was the case and sometimes not, so there seems to be a problem here after all. The main problem arising out of this fact is that while searching a person's name (e.g. because you want to add a new person and would like to check whether (s)he's already there) you can't rely on the alphabetical order. This, consequently, means that you might miss a person already listed and add it a second time. So there should be done something here.

These are some of the problems I encountered. It may well be that I didn't get it right or missed something important, anyhow. If so, please let me know or correct me. But maybe others noticed the same and so there is something not working properly after all.

Second, suggestions:

(a) It should be established once and for all whether paintings should be named and quoted by their original titles if known (i.e. in whatever language these are) and followed by an English translation, if feasible, in brackets. Or, on the contrary, whether all paintings should be named and quoted by English titles (and followed by the original title in brackets, where applicable). I don't mind the one or the other, but maybe it would be good to establish kind of a general rule here.

(b) Likewise, there is as far as I could see no general rule how to list "de", "di", "van" or "von" in the alphabetical listing of names. In general use, French and Germans disregard "de" and "von" for alphabetical listing (thus "von Stuck" would appear under "Stuck" and consequently under "S"), while the Dutch and Italians consider "Van" and "De/Di" part of the name (and consequently written with Capitals, in difference to the French "de" and the German "von"). So, "Van Gogh" appears correctly under "V". It would be useful to have a general rule in this case as well, not necessarily, but possibly, according to what I suggested.

(c) As far as I can see, there is no way of searching the site for places (towns, regions, countries) connecting to artworks. Obviously, many artworks are connected to places (Venice, Cairo, Yorkshire, whatever) and it would be nice to have a search function for a particular place connected to artworks. I thought this might be done not using tags but rather by inserting a further line in the "adding an artwork" form, something like "Place (town, region, country) connected to this artwork". Then you would be free to put "Spain, "Andalusia" or "Seville", for example, as you need. Filling out this line would be the work of those adding an artwork and could proceed along the same lines as indicating the owner.

(d) I couldn't find a way to enlist artworks whose maker is unknown like, in the case of paintings, "Italian school 17th century" or "Dutch painter in the manner of So-and-So" ... So what to do about anonymous artworks of this kind? Should there be an "anonymous" section in the listing of names where you could introduce "Dutch painter" or "Italian school, 17th century"? Maybe somebody already came up with a solution but I haven't found it on the website.

(e) Lastly, tags. Quite obviously, we need many more tags! The more tags the better, actually, for research purposes. Now, I sympathize with the policy of the moderators not to leave tags to the users or contributors. In fact this would lead to chaos as it does on many other webpages, because there would be no unity or any comprehensible system behind this wild tagging. So, let the editors provide for the tags, based on suggestions of the users. Said this, it seems to me that we need some kind of forum to propose certain tags (complete with trees of sub-tags) or to discuss which tags we would like to have. I think this is pretty important because only by tags, proberly used and placed, you can research a database of almost 70.000 entries. So please help here! Or come up with suggestions!

Thanks so much. In all modesty, these were the few things I had on my mind. I thought it worthwhile to put them under the eyes of you all. Please let me know what you think!

Fleurdelys

  • Registered: 2009-11-28
  • Posts: 33

Re: Several Remarks Regarding the Website in General

Hello Fleurdelys! What a wonderfully thought-out set of notes and suggestions.  I'll answer them below; sorry if it's a long post, but there is much to discuss here.

It's good to have the context of how the site is coded. I started the site almost 12 years ago, and during that whole time, it's been a "one man operation" in terms of both funding (paying for the monthly hosting costs) and programming/design. I'd like to think that I've gotten a bit better over the years, but it's still one person applying their skills wherever needed at a given time. That's not intended as whining, but just to explain both the pace and the sometimes idiosyncratic nature of what I choose to work on.

Having said that, I want to rely on users/members of the site as much as possible when deciding where to apply my time. And someone like yourself who is adding so much content goes right to the top of the list, especially when you're able to put your concerns/wish list so clearly.

I think you and I should work together to spec out what "the right answer" is for these and then I can knock them down one-by-one. Some initial thoughts below.

1. Can't add sculpture: I've been slowed down by copyright concerns. It's well-established that representational photographs of two-dimensional artworks have no claim on copyright. This means that we can add those images (which originate as photographs) to our database without violating someone's copyright. With sculpture, there's much more of a case to be made that artistic choices such as angle and lighting make those photographs copyright-able. To be safe, I can only think of three ways to add images of sculpture:

a. Photos that are so old that they're public domain, which is fairly rare.
b. Photos where the uploader explicitly grants us use, typically because it's their own photo.
c. Line drawings of the sculpture that are themselves two-dimensional and public domain.

The best solution I can think of now is to open up the Add Artwork function on sculptor pages, and then add a prominent section in the instructions about copyright, and police the submissions for copyright issues. What are your thoughts?

As an aside, my goal for the next couple of years is that the site will grow in a number of ways that will make us an attractive proposition for museums to work with:
a. Our growing collection, thanks to folks like you
b. A growing viewership, currently up to about 45,000 people per month. I think when we get to 1M per year, that will spark some interest.
c. General professionalism, functionality, and polish of the site, but in particular a set of "tools" that will appeal to museums and museum professionals.

I think the sculpture issue and some others will be truly solved when we recruit museums to put their content on The Athenaeum, which may be a tall order. But it should be the goal!

2. Problems with Google Maps widget when adding museums: You got me here. I have played with it a few times to make the address check optional through some back end logic, to no avail. I think the answer here is to add a user setting to explicitly ignore mapping in order to get the record in. This is probably one of the easier items. I could possibly have that done by Monday.

3. Names not in order, and treatment of "of" in the last name: There are several factors here. Some thoughts:
a. When adding a person, it doesn't force the editor to add the "sorting name". I have been meaning to have a script auto-generate a starting sort name, which can then be edited, but will at least ensure that something goes in there.
b. Because the text used for sorting is "last_name,first_name" and because people put middle names and prefixes in different places, there can be problems. Separating this from cultural approaches to names, getting consistency here is difficult. More on that in a second.
c. As you say, there is not consistency on use of de/De di/Di/ van/Van etc.

Names are so complex (and we haven't even gotten to things like Chinese "last-name-first" names, or Arab names with many family names in the middle) that it's a bit hard to set automatic rules, and they do tend to connect to cultural context. Even answering a question such as "What constitutes a last name?" can be difficult. The overall design question is - how much of this *can* be automated, and how much is going to have to be done through instructions to the editors, and careful review?

I suggest we start a new thread on this one, and start working through rules one-by-one to see what can be done either by baked-in rules in the code, or through a "user guide" or "style manual".

4. Preferred title of paintings: My wife is an art historian and I we've discussed this one a few times. Thus far my preference has been to use the "most common" title, which in some cases is the non-English name (think Primavera by Botticelli). I should also note that we get a significant non-English audience on the site, and a big job one day is to internationalise the site as much as possible. Maybe there's a mid-term and then long-term solution? If we were going to take a crack at it now, I'd say that we should:

a. Default to the title in English, because 90% of our viewership is English-default, and probably many others understand English on some level.
b. Maybe add better management of "alternate titles" - ability to list many alternate titles per artwork, ability to indicate the language of a title on a per-title basis, so that one day a user could set the language to Russian (say), and they would see title in Russian where available.

The challenge is that the art world itself vacillates on this one. Therefore I am not sure there is an absolute "correct" approach to implement. Our goal should be usability - making it easy for users to find and recognise artworks. This probably needs more careful thought on the best set of rules.

5. Place connected to artwork: That you suggest this shows that you "get" where the site is trying to go. I do intend to do this, and almost exactly the way you note. The hold-up here is getting the geographic underpinnings in place, which is a huge job in itself.  The job of getting thousands (millions probably) of places into the database and making sure they are correct is HUGE. I don't want to ask users to slowly and meticulously do this, so I'm trying to do it by database scripts using public sources like the Geonames database: http://www.geonames.org/  However, there is a lot of "screwy" stuff in the data. For example, even at the level of first-level administrative units of countries (provinces/states/oblasts etc), the data is not complete or accurate.

I think that storing/encoding/standardising geography is such a big task that we'll eventually want a corps of geography buffs working here to get it right. My near-term goal is to get Nations, then Administrative Level 1 divisions, then major Cities in place. When that gets 90% done, adding the code to associate artworks will be very simple. Honestly though, I think this may take a few months.

Thoughts?

6. Unknown or non-specific artists: If you click "Add an Artwork" from the front page, you should be able to add an artwork with no artist information at all - this will cause it to be fled under "Unknown" . For other generics, I think we should treat them generally as person records, but perhaps with an extra tag that the "person" is a generic type (i.e., "Dutch painter" etc). Would like your thoughts on the "perfect" functionality here, as there are a few ways to go.

7. More tags! Yes, we do need more! They are limited in this way precisely to get us to discuss which ones to add, and when. I tend to disagree that tags will be the only way to sort through so much art - the whole site is intended to let you sort through it in many ways (by year, museum, artist, etc). Our goal in the next year is to upgrade the database again, which will give us some cool functions for "full text search", and then to build a custom search module that will replace the Google search we have now. So in future on the top right of most pages will be a box where you can type search terms and it will bring up a nicely formatted list of artworks, artists, museums, all with custom display and in separate lists.

So let's separate tags as ways to search for "subjects not contained in the name" from "searching by artwork/artist/museum name". Then the question is what tags are necessary. I feel that a lot of sites load up a big list that they find online and then they have tags for *everything*. I'd rather ours are added a few at a time, as people indicate they're needed. But I'm open to other opinions! Which ones do you think we need?


Phew. Sorry for all of that, but I love that you're thinking so carefully about this stuff, and it helps crystallise the plans for future development. I'd love for rocsdad or Irene to weigh in.

Could you rank these in your order of importance?

Thanks,
Chris

  • Registered: 2012-08-25
  • Posts: 4

Re: Several Remarks Regarding the Website in General

Hi Chris,

I am positively flabbergasted as to your detailed and quick reply! Thanks so much! And may I now add an apology, albeit belatedly, that I wrote my concerns and suggestions in such a straightforward and dashing manner while still being a novice in your project. Least that can be said for an excuse, I hope, is that finding my ways around your site as a newcomer inevitably led me to "discover" the points mentioned ...

At the moment, I need some hours to think about all the points you raised and will try to come up with some sort of ranking as to what should be tackled in the first place. I also would like to have the opinion of other much-involved members. In any case, your remarks and info have been extremely helpful, thanks again. I much appreciate the work you have done over the years!

As said, I'll reply more fully later today. Hope somebody else joins in, too.
Fleurdelys

4 Reply by fleurdelys (edited by fleurdelys 2012-09-18 06:41:46)

  • Registered: 2012-08-25
  • Posts: 4

Re: Several Remarks Regarding the Website in General

Hi Chris!
Finally I found time to put down some of my thoughts. Thanks again for your detailed information on all those points which was very helpful.
You asked me, justly, to establish some kind of ranking which things are to be tackled first and with a certain urgency. I try to do so in the following, putting the less urgent things first.

(i) Names and spellings
Less important and pressing, to my mind, is everything regarding names and spellings. First of all, mistakes or errors can be corrected online. Second, you are right, there will be no general rule of how to spell certain names in practice. And of course names in less familiar languages (and alphabets) present problems of their own. So maybe one could leave it that for the time being. (Apart from that, I still noticed that correct entries don't show in their correct alphabetical position, e.g. "Baum" after "Baumgras" ... I think this depends on the program being used; I have no clear idea how it works).

(ii) Google widget
Equally secondary seems to me the Google widget; I just mentioned it because I found it somewhat, let's say, annoying. Lately, it didn't give me any problems and anyhow the contributors are bound to use the "edit museum" function only rarely I presume because the vast majority of entries is correct. However, is the widget really needed? Couldn't it be just enough to enter the address in the relevant field? What's the advantage of the Google map device? But as said, this doesn't seem an important point after all.

(iii) Titles of artworks
Likewise the mode of entering the titles of artworks: Yes, maybe best adopt the pragmatic stance here. As the overwhelming use on the website is to put the English version as only or at least as the privileged title, we probably should go on doing so. As for my part, I put the original language version, if known, in brackets after the English title (and if it's not too long). If, as you say, non-English titles have become so famous as to be not ignored ("Primavera"), so we can use them as well as primary titles. However, if we would like to preserve the original titles besides the English ones, maybe an easy solution would be adding an "original title" line to fill out in the "add artwork" form. It could then appear in small grey letters under the English heading same as the "aka ...." indication. What about that?

(iv) Sculptures
I also thank you for your explanation of why pictures of sculptures are not, so far, permitted for upload. I didn't know the respective copyright problems (different from two-dimensional artworks) and fully see your point. Yes, I guess, this must be solved before opening the site to sculptures (a pity, though). So this point is taken from the list for now.

Now the important stuff: Anonymous artists, geographical indications and tags.

(i) Anonymous artists
I think the best functionality to have here would be to have a separated "Anonymous" section, parallel to the persons's section. Hierarchically, it would be equivalent to a letter-section of the persons's register. It could appear as a list of its own after the "Z" for persons. The list in itself could be arranged alphabetically. Maybe something like this (and similar to how the persons' list do appear):

Denomination            Timespan        Nationality

Anonymous            17th century        Dutch
Anonymous            17th century         Italian
Follower of N.N.                 18th century        English
School of N.N.            16th century        German

and so on. We should only separate between "anonymous" and "follower/school of" in order to keep the list orderly (that is, no further differentations like "landscapist" or "portrait painter"). The list would then also be limited, because how many entries can there be? Every century and a number of nationalities for the "Anonymous" entries, likewise a probably limited number of "follower / school of ...". Could this be done that way?

(ii) Geographical indications
Yes, you underline that there are already various search functions (by artists, museum, years etc.) But obviously, and I surely don't need to tell you that, many key searches will always be according to places and/or subject matters (well, places being a certain kind of subject). Of course it's good to check what a museum possesses or the output of an individual artist, but art history (and many related studies) would like to search for, say, "church interiors" or "beach views" or "female oriental portraits" or "views of Marseille" and so on. (At least I tend to search mainly for those kind of themes). Without tags and/or geographical indications you cannot do it. So this, I think, remains of prime importance.
I am aware that it's difficult to do in practice. If individual contributors are allowed to enter place names etc., chances are that we'll have a plethora of spellings for one place and so on; result being complete chaos (same as tags). So  I agree with you in providing a pre-set list of names based e.g., as you suggested, on geonames.org. I looked at that website and found it in principle idoneous (although it is indeed very specific ...). You say "My near-term goal is to get Nations, then Administrative Level 1 divisions, then major Cities in place", and this seems to be the best step for the moment. I think it even can be done without too many difficulties; most difficult will prove the fact that it's a huge piece of work to put it together. I could help in that, and I think others would contribute as well. Maybe the first goal should be to have a list of "countries (subcategory:) regions (2. subcategory:) towns" in the same manner as the museum list.

(iii) Tags.
Of course, if the full search functions will be installed in the way you mentioned, tags will be less important. But still: Not every portrait painting is called "Portrait" in the title (or the accompanying text), so the general full text search function would not find it.
So, in a way I think tags are still good, because if properly managed they will be better than the results of an automatic text search (it's a bit like in books: preparing an index manually, so to speak, will result in a much better result than an index compiled by an automatic indexing program ...). 
So I am much in favour of tags in general. Maybe we should ask the contributors to submit their ideas for tags in a special thread and then discuss them. At the end we would have, hopefully, a set of tags for a certain theme. Tags could be structured in hierarchical trees with in all three levels. Level 1: General theme - level 2: first specification - level 3: second specification. Artworks might be tagged with more than one tag from one tree.

In any case, I keep on thinking on all these points. I am also looking forward what you will come up with in the future. Also, if other members will join in and share their thoughts.

For the moment, all the best
Fleurdelys

Ah well, P.S.
What about the lists? I noticed that there is the possibility of lists, but you cannot access them from the main page, only from the pages of single artworks. Could not the lists be made profitable? E.g. a list "Views of Venice" would for the time being suffice for not having geographical indications. Also, a list "Mosques" or "Oriental portraits: Children" could take the role of tags. If this is a way to go, please give me your thoughts on this!

Another thing I forgot above: Level 2 or 3 tags in one tree could appear in other tag-tree hierarchies as well, and vice versa. E.g. "Mosques" could appear in a tag-tree which centres exclusively on architecture and buildings. I do not know how this could be organized on the internal level, as I am not enough proficient in programming !!