I recently posted some loose requirements for selecting my next online photo database size. The requirements generally included rationales, where it wasn’t obvious.
Stage Two: clean up the requirements, categorise, prioritise and number them.
Voila.
Category | Id | Requirement | Priority |
---|---|---|---|
General | GEN1 | Accessible by a typical web-browser | Critical |
GEN2 | Active development of new features | Important | |
Ownership | OWN1 | Ownership of photos is retained by me. | Critical |
OWN2 | Future-proofed against money running out: mine or yours | Critical | |
OWN3 | Ownership of meta-data is retained by me. | Important | |
Cataloguing | CAT1 | Photographs grouped into albums by event | Critical |
CAT2 | Photographs tagged with people | Critical | |
CAT3 | Photographs tagged with locations, objects, activities | Important | |
CAT4 | Performers tagged with real and stage names. | Important | |
CAT5 | Attribution of photographer details | Important | |
CAT6 | Attribution of copyright owner’s details | Important | |
CAT7 | Rating of photographs | Important | |
CAT8 | Sorting/Filtering by rating | Important | |
CAT9 | Albums grouped by type | Desirable | |
CAT10 | Albums grouped by date | Desirable | |
CAT11 | Areas or points of photograph tagged | Desirable | |
CAT12 | Simple contact management of subjects | Desirable | |
CAT13 | Hidden fields on contacts to distinguish like-named people | Desirable | |
CAT14 | Tagging of anonymous people to enable searching | Desirable | |
CAT15 | Corrections to names update everywhere | Desirable | |
CAT16 | Attribution of copyright details | Desirable | |
CAT17 | Control over (default) ordering | Desirable | |
Searching and Linking | S+L1 | Link to other photos with same tag within an album | Critical |
S+L2 | Associate URLs with subjects, that are displayed. | Important | |
S+L3 | Link to other photos with same tag across my albums | Important | |
S+L4 | Search of tags by keyword | Important | |
S+L5 | Cross-promotion of other albums and sites. | Important | |
S+L6 | Link to other photos with same tag across other photo sites | Desirable | |
S+L7 | Search of album names by keyword | Desirable | |
S+L8 | User-generatable URLs to search tags by keyword | Desirable | |
Community | COMM1 | Multi-user Tagging | Desirable |
COMM2 | Notifications of appearance in photos | Desirable | |
COMM3 | Comments permitted | Desirable | |
COMM4 | Notification of comments | Desirable | |
COMM5 | RSS or Atom Feeds for comments | Desirable | |
COMM6 | RSS or Atom Feeds for subjects | Desirable | |
COMM7 | RSS or Atom Feeds for new photos | Desirable | |
Photoquality | PQ1 | Web-quality images shall be displayed by default. | Critical |
PQ2 | Print-quality images shall be available. | Important | |
PQ3 | Automatically generated thumbnail and web-quality versions. | Important | |
PQ4 | Archive-quality images shall be stored. | Desirable | |
PQ5 | Custom thumbnails (e.g. choosing to crop over shrinking.) | Desirable | |
PQ6 | Support for short video | Desirable | |
PQ7 | Support for long video | Desirable | |
Performance | PERF1 | Quota > 0.5 TB, if any | Important |
PERF2 | Low-cost | Important | |
PERF3 | Fast response time | Important | |
PERF4 | Scale to thousands of tags | Important | |
PERF5 | < 1 minute face-time per photograph | Important | |
PERF6 | Free | Desirable | |
UI | UI1 | Forward/Backward navigation between photos in album. | Critical |
UI2 | Slideshows | Desirable | |
UI3 | Display of many thumbnails at once | Desirable | |
Privacy | PRIV1 | Their email address should never be published on the web. | Critical |
PRIV2 | Registration and logging in not required for general use. | Important | |
PRIV3 | Robust privacy features for photographs | Important | |
Workflow | WF1 | Hint to original location on my harddrive | Important |
WF2 | Auto-complete or partial search on tags during input | Important | |
WF3 | Read EXIF data from image | Desirable | |
WF4 | Support unpublished draft state | Desirable | |
Migration | MIGR1 | API to add photos | Critical |
MIGR2 | API to add tags | Critical | |
MIGR3 | Tags can be non-specific to areas of photo | Important | |
Metrics | METR1 | “How many visitors?” metric | Desirable |
METR2 | “How long does a visitor stay?†metric | Desirable |
Comment by Sunny Kalsi on June 14, 2008
Might be late to say this, but I’m surprised here that “Robust privacy features for photographs” is a single requirement, and fairly loosely termed.
should password protected photos (or groups of photos) be a requirement? classifying “groups” or other denominations so that allow you to separate photo ACLs?
Also vaguely surprised there’s no migration requirements for an API to pull the photos back out again.
Comment by Julian on June 15, 2008
Sunny,
If these were requirements written to implement a new system, I would indeed need to think through different options for privacy, and nominate the one that I wanted.
However, I am just using them to judge existing solutions. I have a fairly open mind about how the privacy should work.
Password-protected albums could work; that was the direction I was expecting to go with my software. Tracking groups could be done in many ways. Flickr has the concept of a “pass” that could work. FaceBook has searchable tags can only be added by friends, and which can be revoked by the subject. It also supports group permissions on albums.
I don’t have a strong opinion which suggests that there is One True Way. I’ll accept any system that passes muster.
Similarly, an API to pull data back is one way to implement the ownership requirements. Another might be to make my photo folders on my hard-drive the master-copy. Then the images never need to be “pulled back out” over an API as they are always kept where I can see them.
Comment by Sunny Kalsi on August 11, 2008
You might also want to consider My Picture Town by Nikon. I have no idea if it’s good or not, but just happened to see that it existed. It seems to have a bunch of photo editing stuff on it.
Comment by Richard Atkins on December 2, 2008
What’s the summary count of criticals/important/desirables here? If you had to pick a weight between 0 and 10 for each requirement, would you? Could you be bothered?
Comment by Julian on December 2, 2008
Richard,
If I needed to show my bosses and auditors that I had been completely fair in my selection criteria, I would consider coming up with some absolute weightings. Given there are prices associated with some of the solutions, coming up with a dollar value for each requirement would be the economic-rationalist way to value them.
I’ve certainly done that before.
However, I don’t think that is required here. Ultimately, I will be using this tool for pleasure, and the real dollar costs are relatively small, so the subjective element deserves a greater prominence.
I haven’t been stating up-front what the entire process will be. That way. it doesn’t look as back when I change my mind and just do whatever I want! However, I will go out on a limb now and explain my current plan.
I plan to use the initial evaluations to cull the choices down to a shortlist of 3 or 4 solutions, which I will then briefly look at again.
I expect the final choice will have a significant subjective choice, rather than just a weighted sum of the Requirements Met.
That said, I am unlikely to include any solutions on the shortlist that failed any of the critical criteria, without a very good reason/rationalisation.
My goal is to have the new web-site operational prior to purchasing a new camera in January, prior to the next circus festival.
Ideally, I will have migrated my existing circus festival-related photos across by early February.
I was planning to get a secondhand slide scanner in a similar time-frame.
Of course, that means I need a new hard-drive (cluster) to store all the new data, and a new CPU and new RAM to handle all the image processing that entails. I’ll need a new motherboard to handle all that. A new monitor would be nice too, which means a new graphics card to drive it.
Suddenly PERF6 is starting to sound like a critical requirement!
Comment by Richard Atkins on December 5, 2008
That January deadline is looming awfully close. I hope you finish in time!
I agree that you should not do a detailed evaluation of a tool with any failed critical requirements – they wouldn’t be critical if they were optional. I’m not entirely sure that your arguments using weightings are quite right – there’s probably an argument for not using them in any situation, and just going with three or so “neededness” criteria as you’ve done.
Just answering my first question above: there are 11/24/30 requirements in each category.
I guess I’m looking for a quick way to see how many of each category exist, and how many in each category each tool met/failed. For example, how far away from what you need is your custom photo DB software? After importing into Excel (paste special as text!) and doing some Auto-filtering, I can see it meets 10/11 Criticals, 22/24 Importants, and 12/30 Desirables. That doesn’t seem bad at all (nor does it seem surprising, since you wrote it to meet your needs), but it fails a critical requirement, so no need to evaluate it further (as if you’d need to).
How about Gallery2? 11/11, 17-18/24, 15-18/30 met (since there’s Y? and N? results here). Sounds well worth pursuing a more thorough study of some of its plugins. If this isn’t how you feel about the tool (that subjective choice you mention), maybe there’s some missing requirements here: easy to install, stable, refined/polished. And given that your photo db will grow — probably beyond anyones desire to recreate collections, tagsets and so on — and quite likely outlive your tool choice, you should probably also look at requirements for migrating out of the tool too.
This is the sort of quick reference results I’d like to see.