Proposal talk:Camera equipment program/Archive 1

From Wikimedia Australia
Jump to navigation Jump to search

seconders comments

Following are some comments extracted from my mail as seconder.

I was initially concerned that the bar might be set a bit high but the existence of a number of potential recipients suggests it might be OK. We might need to clarify that "categories" on Commons are a lot looser than on WP and define "freely license".

If the grant is retrospective, do the "500 images using the new equipment" overlap with the images used to justify the grant? (IE, I buy a new camera, take 1000 photos with it and upload them with an appropriate free license, then apply for a grant. If I receive the grant, have I already fulfilled the "freely license 500 photographs that use the new equipment" requirement?)

--PeterJeremy 15:36, 6 August 2011 (EST)

I've revised "freely license" to indicate that they need to be uploaded to Wikimedia Commons, and they need to comply with the definition of "photograph" provided. I've also removed "use the new equipment" for the moment, as the new equipment may be a lens which may only be appropriate in limited photography opportunities.
My thinking was that the "500" images would retrospectively cover photographs already uploaded. I think the "500" number may be too high. John Vandenberg 16:04, 6 August 2011 (EST)

points system

The numbers may be ok if additional weight is given to quality/valued/featured images. I suggest:

  • +1 point per MAX(total num images, total distinct commons categories)
  • +1 bonus points per image used in content space for any Wikimedia project other than Commons
  • +2 bonus points per image assessed as Quality Image
  • +4 bonus points per image assessed as Valued Image
  • +10 bonus points per image assessed as Featured Picture at either Commons or en-Wikipedia

I think this is about the right trade off between quantity, quality, and usefulness. Certainly it will incentivise photographers to both put in the hard yards and improve our work.--99of9 23:03, 22 August 2011 (EST)

I like "MAX(total num images, total distinct commons categories)"
I think the "used in content space" bonus needs to include some minimum period of use - at least one month in opinion, in order to avoid rewarding edit wars. John Vandenberg 23:22, 22 August 2011 (EST)
Oops, I meant MIN. MIN insists that prolific photographers keep up the diversity and the categorization. I agree that long-term usage is the important usage, but am not sure what's best to measure. --99of9 23:35, 22 August 2011 (EST)

recipient uploading overseas photos

An overseas community member has emailed me indicating they are interested in the program. They can upload images of Australian topics, however they would not be using the equipment for Australian topics, because they are not in Australia. In order to not exclude overseas recipients, or Australians who go overseas on holidays, I think we should change

"Recipients of grants in this program must agree to upload 500 photographs (as defined above) to Wikimedia Commons within 24 months."

to

"Recipients of grants in this program must agree to upload 500 photographs to Wikimedia Commons within 24 months. These photographs should include photographs using the new equipment when appropriate, be properly described (as above), however they do not need to be Australian topics."

John Vandenberg 12:25, 27 August 2011 (EST)

Members discussion report

The mailing list discussion involved 12 members posting 33 responses to John's initial proposal. The methodology for examining the content was based on discourse analysis, as typically employed in grounded theory research: each email was broken into individual points, which were then listed. These were then categorised and grouped as necessary, with non-relevant points being moved to a separate list. If a point was made by multiple contributors it was tagged. The points in the categories were then analysed again, with a summary written and included here.

Discussion

Suggested clarifications

There were a number of recommendations to improve how certain issues are defined in the proposal. Specific concerns related to being clear by what is meant by "Categories", as they are used in a different manner on Commons compared to how they are employed on Wikipedia; to define clearly what is meant by "freely licensed"; and to make it clear in the proposal that the program is only going to be funded for a finite length of time. More generally, it was suggested that the aim of the program needs to be clearly articulated: in particular, is it intended to encourage the taking of new photos, read the inclusion of existing photographs, or encourage the use of photographs in Wikipedia to illustrate articles.

Criteria

Along with concerns about the specific criteria raised below, there were some more general issues raised about the use of criteria. One member raised the idea that listing specific criteria isn't the best approach, and that instead the rules should provide general (but fair and consistent) criteria, combined with a statement that the committee reserves the right not to fund contributors if the quality does not reach the required standard (referring to both image quality, subject matter and meta-data). This suggestion had the support of at least three other members.

Multiple subjects

A specific problem was raised with the possibility of two many photos being taken of a single subject. This engendered a degree of debate: on the one hand, the requirement that unique categories were required seems to allow for this, but on the other it was argued that having multiple images of one subject is valuable from the perspective of Commons and the project's aims, even if it is not as valuable from the perspective of Wikipedia. There was no clear consensus on this point, although it was generally agreed that 500 photos of a person's back lawn shouldn't qualify (noting the off-topic proviso listed below).

One approach raised was to list specifically needed images, to get people to focus on what was needed for Wikipedia, rather than a single topic. This had some support, and http://en.wikipedia.org/wiki/Category:Wikipedia_requested_photographs_in_Australia was identified as a possible source, but it was also argued that this would encourage people to be the first one to take the picture, rather than to take a good picture. (The issue of quality vs quantity was also raised in other areas of discussion, and is treated separately below). It was, however, suggested that a mashup could be generated, using geotags, to identify where desired subjects are located.

A related concern was the need for photographs of specific areas on Commons, such as sports photography. These areas may require specialised equipment, so a lowering of the criteria for people engaged in these areas might be worthwhile.

Quantity vs quality

A second specific area of concern was the issue of quantity vs quality. The proposal as written requires a specific number of photos to be taken, but doesn't speak to the quality of those works. This raised concerns, as it may be that a smaller number of high significance subjects are far more valuable than a large number of easy to take but less significant subjects, and the approach would lead to a systemic bias towards easy photos. It was also suggested that the program should be used as an incentive to improve photo quality, as well as to increase the overall number of photos available.

Thus there were a number of suggestions to counter this. The first was to make it clear that poor quality works can be rejected from the count. The second was to halve the numbers of photos required for each category, to allow for photographers who focus on producing a smaller number of better images. However, a more complex suggestion was to use a points scheme, which rewards images which have passed peer review, focusing on the Featured Image, Valued Image and Quality Image processes on Commons. It was identified that these models for peer review are imperfect, but they provided an independent means of rating photographic works.

Usage

The current proposal focus on the use of the images on Wikipedia. This was raised as a concern by some members. Most of the discussion regarded the difficulty of meeting the criteria, although the risk of encouraging gallery spam was also raised. Thus it was suggested that all projects be included, rather than just Wikipedia - use in Wikibooks, Wikispecies or other projects should count towards the total. It was also suggested that there is value in having images in Commons irrespective of their use in other projects, suggesting that there may not be a need to look at the use of the images outside of the one project.

Geotagging

Not all images can be geotagged, so the criteria cannot always be met. Furthermore, it was argued that this would preclude photographers who are unable to geotag images, due to lacking the required technical skills, in spite of the potential quality of their work.

Budget

It was asked if the budget, currently set at $10,000, would be sufficient if there were a large number of unexpected applicants. In response, it was suggested that the budget would be capped, and any additional applicants would therefore have to wait until the next round.

Other

Three issues were raised in addition to the above. The first was a concern that there was no way to ensure that recipients meet their obligations after receiving their funds: although the proposal requires them to upload more images, there is no means by which this can be enforced. The other question was whether or not people outside Australia would qualify - the example of an Australian now living in the UK was raised, as was the possibility of providing grants to people within the region who could, for example, take photos of sporting events featuring Australians but held outside the country.

Beyond that, discussion focused on how the money could be used by recipients. There was a belief expressed that just getting SLRs into the hands of contributors would result in an improvement, although it was also suggested that high-end point and shoot cameras would be good, as would the newer EVIL cameras.

Off-topic discussion

Most of the discussion was focused on the proposal, or issues closely related. However, it was noted that 'grasses' were a valid topic for photography, as there are over 1200 species of grass. There was also some discussion about encouraging photographs of sporting events, and it was suggested that the Chapter a) contact sporting bodies to request photographs where possible; try to organise media passes to events; and purchase a small set of specialised cameras and lenses to use at such events.

Proposed changes

Based on the discussion, the following changes are proposed:

  1. Clarify that the "least two subject categories" are on Commons
  2. Note the end date of the program
  3. Clarify what constitutes a free license
  4. Extend the current introduction, stating the intent of the program.
  5. Add to the conditions that the final decision will be at the discretion of the committee, and that in some instances people who do not meet the specific technical requirements, but do meet other conditions, may be included in the program, and that the committee will take into account the technical quality of the submissions when considering funding.
  6. Provide a points system, whereby (for example) one Quality Image is worth 5 normal images, one Valued Image is worth 10, and one Featured Image is worth 15 normal images.
  7. Change syntax from "must have been in use on a Wikipedia page for three months" to "must have been in use in content space on a Wikimedia Foundation project for three months".
  8. Add "where applicable" to the geotagging requirement.

External to the program:

  1. Work at creating a more accurate and useful list of required photos that are provided by the current category on Wikipedia.
  2. Look into the sports suggestions - in particular, look into getting media passes for events.

Bilby 23:26, 27 August 2011 (EST)

Thanks Adam for this overview. I think I have integrated most of those recommendations into the proposal.
Re 2. I can't see any concerns about an "end date of the program". The program needs to have a life of at least two years in order to be useful. I have clarified that the budget for next year will be subject to review of the committee after the first financial year. The life of the program beyond that is two committee's away. If the program isn't working by then, I have no doubt that someone will can it.
Re 6. I don't see any consensus on a points system. I agree a points system is wanted and needed, but the weightings to be used isnt agreed upon. I would rather that we start with the bar high, and lower it carefully, rather than need to raise it as that would undermine the stated aim of allowing contributors to work towards a clear goal. I've restructured this section of the proposal a bit so that the reports will include multiple factors, including the technical conditions listed, and VI, QI and FP, and allows the committee to select a weighting algorithm after public community consultation. The committee recently approved Proposal:Non-member participation, which will mean we can have a discussion of the algorithm here on the wiki with broader input and greater transparency than is possible on our private members mailing list. Weighting quality is an interesting problem that I hope will attract a lot of feedback.
Finally Re 8. I have yet to see an example of where geotagging at an accuracy of 10km2 is not applicable or not feasible. It may not always be high value metadata to us, here and now, however it could be of vital importance to a researcher in 100 years. If a grant application includes a high percentage of photographs which are difficult to geotag to that granularity, and the committee is of the opinion that there isn't much value in geotags being added, we can waive that technical requirement.
John Vandenberg 00:04, 4 September 2011 (EST)
Geotagging makes sense for some subjects, such as landscapes, architecture, or events. It doesn't make sense for things such as studio work, where it would be irrelevant (and potentially problematic, if you are effectively geotagging your place of residence, given that the "within 10km rule" allows room there). I've added "where applicable" to allow for these, otherwise we're saying to people who do studio projects that they aren't in the running, even though we are willing to include them in practise. - Bilby 19:12, 15 September 2011 (EST)
If studio work is at home, or is otherwise too identifying, that is a valid reason for not geotagging. Thanks. I would prefer for this to be an exception rather than incorporated into the text, but provided it is understood that this is not a 'soft' requirement than I am happy with your amendment. John Vandenberg 19:35, 15 September 2011 (EST)

A couple of issues

I was overseas with no meaningful internet access when the proposal was lodged, and hadn't had time until last night to properly review it. When I did, I found some areas of moderate to serious concern, and in conversation with a friend who is a photographer who looked at the proposal from that angle, discovered a few more. I'll put the two critical changes I'll suggest first:

  • For the 1,000, the amount should also be 1,000, not 2,000. Otherwise, we're communicating to our more prolific photographers that their contribution is not as valuable as a smaller contribution. $1,000 in real terms is a fairly small amount of money in terms of what it can get you camera-wise; we shouldn't set the bar unreasonably high.
  • The line "Committee members are expected to initiate and/or participate in deletion discussions on Commons if the applicants images do not appear to clearly fall within Commons policies, such as commons:COM:SCOPE." should be entirely removed from the proposal as, in my view, it conflicts with the role of the committee which is to *encourage* good work. I would actually argue (and I note this binds myself as a committee member!) that we should not only not initiate such debates on Commons, but should abstain from them entirely if or when they arise.

Other points raised:

  • The limitation on past small grant recipients is petty and should be dropped, given the budget available for the scheme. I would be happy to support any of our three photography-related past requesters to get the full amount, and the fact it affects so few people suggests that it's not a good general term.
  • A good photographer may take 25 photos of the same thing, all of which are entirely within Com:Scope but the majority of which will never end up in any other article or project. Therefore I believe that the level of WP/WM usage should be lowered to 10%.
  • The category requirement may lead to category spam - it may be necessary in order to meet the requirements as written to create *500 categories*, many of which are entirely unnecessary, in order to fulfil the requirements. We should not encourage this, and whatever a category metric was seeking to achieve is unclear anyway. Furthermore, many specific-topic photographs would belong to only one category - for example, photos of a suburb, or of a particular species of tree.
  • Some of the requirements appear to be unauditable. Who is going to have the time to sit down and figure out whether someone has met the requirements? If a committee member or a bot makes a mistake, how are disputes resolved?
  • Adam acceptably addressed a point I had re geotagging of non-fixed subjects, such as human subjects.
  • The requirement following the grant regarding 500 in 24 months is unauditable, and in some cases functionally impossible. The $200 grant would not buy a substantial piece of equipment and it would be impossible to prove one had used a particular memory card in the phone for instance. The incentive to "take the money and run" is high.

I hope some or all of these can be accommodated, particularly the first two. I broadly support Adam's suggestions above, as well. Andrew Owens 00:00, 28 August 2011 (EST)

Regarding your first critical problem, the nature of the "Large equipment grant" is a long term investment which works in the favour of our prolific photographers, and the "Small equipment grant" is immediate assistance available to most of our prolific photographers. The "Small equipment grant" is meant to be easy to obtain, however it is only provided once. All of our prolific photographers should be able to obtain the "Small equipment grant" readily by adding a bit of missing metadata. i.e. of their 1000+ uploaded photographs, 500 should easily met the requirements with only a little bit of effort, and they can receive the first $500 quickly. People have receipts which are aging as we speak; lets get some money to them quickly. The "Large equipment grant" will require more effort on the part of our prolific photographers to bring their back-catalogue of uploads up to the metadata standards set by this program, and our less prolific contributors will need to upload more photographs that they haven't uploaded due to lack of incentive. Once a contributor has 2,000 compliant photographs (or 2,000 points once we have a weighting system) they can receive the second $500 (they have now received $1,000). Every subsequent 2,000 photographs allows them to purchase equipment to the value of $1,000 paid for by Wikimedia Australia. This grant is not intended to 'value' or 'reward' people; it is designed to provide assistance in the purchase of a piece of equipment that wears out with use. How frequently they are granted money to upgrade their equipment will depend on how frequently they use their equipment to provide free media to Wikimedia. It isn't currently designed to support purchase or renewal of high end equipment, as high end equipment needs to be tied to high quality results, and we need a more complex system for that. It is costed on moderate equipment replacement needs. If you browse through commons:Category:User categories, you will see many people who would be receiving $1000 roughly every two years if their chapter had a program like this. John Vandenberg 01:37, 4 September 2011 (EST)
Regarding your second critical problem, I agree, and I have changed this "evaluation" to be primarily a community task. The problem it was intended for is clear violations of policy, and our committee members regularly initiate and participate in discussions about clear violations of policy. However, all of our serious contenders for this grant are very respected members of the Wikimedia community, and quite often are admins on Commons or Wikipedia. John Vandenberg 01:37, 4 September 2011 (EST)
Regarding the categorisation, I have tightened up the wording on this criteria. Ultimately the Commons community will manage any category spam problems. In the SLQ upload, we have had to create thousands of categories, because they were missing. The commons community doesn't mind this; it encourages it. The problems arise when people create stupidly narrow categories like "photographs of a man hanging upside down from a tree in the Brisbane Botanical Gardens". John Vandenberg 01:37, 4 September 2011 (EST)
Regarding the condition that the recipient uploads another 500 photographs in 24 months that follow, these would be managed in the same way that we manage obtaining reports/results for small grants. The $200 and $500 are still quite small amounts, comparable to the small grants. For the large grant, which will go to our prolific and long-term contributors, I don't think we need to worry about them doing a runner on us frequently. Occasionally one might walk away from Wikimedia, but the reason will be more substantial than money, and we'll be more concerned about the loss of a valued member rather than the 500 more photos that we'll never see. We could add a caveat to this condition that it can be waived or extended at any time by the committee if the grant recipient informs the committee of a change of circumstances since they accepted the grant. John Vandenberg 01:37, 4 September 2011 (EST)