Options
getImageEXIF
cmccrack
Registered Users Posts: 6 Beginner grinner
I didn't find it reported anywhere in the forum, so I guess I'll post it.
When I use getImageEXIF I get an invalid user fault even where other functions using the same sessionID work. I assume this is one which has not been completely debugged yet.
When I use getImageEXIF I get an invalid user fault even where other functions using the same sessionID work. I assume this is one which has not been completely debugged yet.
0
Comments
For the record: I didn't try that API myself. But I got a strong feeling that most of the so called "New API" is, mmmh, not entirely functional (Hi, Don!:-).
I burned my fingers on a few and decided to wait until SM's R&D actually does some decent QA work on them, because it was simply a waist of my time - bugs are reported, practically zero feedback, no fixes - so, what's the point?
I posted the list of "working" APIs (whcih I personally tried and found working) somewhere in mid-January in this forum, you can search for it.
One more thing: in case if you're using anonymous login - it never worked. May be there was a hidden fix, but nobody I know was ever able to do anything with this login type.
Not sure if it helps, but that's it..
Cheers!
i'm getting the same result !!
it's driving me nuts, all the recent calls I have implemented...none of them work.
David
SmugMug API Developer
My Photos
I was so excited to see that new API.. So I jumped in, wrapped few new calls, added to the UI, - bang, dang, nothing works...
I guess we're glad to the at least old stuff is still working..
My friends keep telling me to do flckr - their set of API seems to be far more exntensive and stable.. it would be a bliss.. But I have to finish what I started here first... Too many SE users already, can't simply leave them: 5,500+ hits and counting...
But it IS frustrating, I agree 100%..:cry
Cheers!
It's such a catch-22 for smugmug. They seem to be doing everything they can to expand and improve their core set of services and therefore, their API improvements seem to be set on the back burner.
Its unfortunate because there are many developers who are anxious to be able to develop some neat tools.
Having a solid API and an excited developer base could quickly put smugmug on par with some of the third party tools that flickr offers.
My opinion is that if concentration is placed on API development, smugmug has greatly increased the number of developers "working" for them. This could lead to some very innovative solutions.
Aaron Christy
http://www.surfacedamage.com/
... and I don't even mind to collaborate on doing this API for them. Outsourcing can be a pretty good deal sometimes.
Imagine you have a solid set of API which covers everything you can do via UI (and hopefully, which IS used by the UI, so API comes first and it DOES work), existing, as in flckr, in RPC, SOAP, etc..
But it's SM's call.
All we can do is to ask, offer - and wait.. wait... wait...
That's the way it should work, but unfortunately (and I am as guilty as the next person on this), I tend to start development from the top down -- in other words, I am so excited about the user interface and providing something of value to the end user, that I start there and work my way to the data level.
What really needs to happen is exactly what you mentioned -- build a useful API and then build all functionality on top of it. (That is not to say that SM is not doing this already -- I don't really know much about their architecture. Perhaps they are using the API but simply have not exposed all of the methods?)
For a public, fee-based system (like SM), this is a scary proposition because what it means is that the founders lose some level of control. Users are building applications using only the basic services they provide. That's a risk, but it also has very, very huge potential.
Users can develop features, functionality, and ultimately provide more useful services to the end users than a small team of internal SM developers may not have time to do.
Take for instance, your Smugmug Explorer. I signed up and cancelled my smugmug account the very same day. The file size of the pictures I take means I could only upload like 50 pictures at a time using the standard uploader. This killed the deal. After I cancelled, Chris pointed me to Smugmug Explorer and I signed right back up.
My point is that a good API = useful third party applications = more exposure = retained customers.
In my opinion, it is a risk worth taking.
Aaron Christy
http://www.surfacedamage.com/
There is nothing wrong with UI-first approach - as long as you make it. Competition is fierce, and the one who got "there" sooner might as well control the market, or at least has a sizeable piece of it.
The things are quite different if you wanna *stay* on that market. Pretty face is nice, but you gotta have a workhorse also.
And in SM case, I don't even think the risk is that big. What can you possibly do with a 4 level deep tree structure (I'm not talking about the sturcture change)?
On each level:
- get the list
- get the children
- add item
- remove item
- rename the item
- get item's properties
- change item's properties.
That's essentially it. There may be some "batch" performance optimized methods, but they are separate issue.Let me ask you - how hard can it be to implement and surface those 7 methods on ANY possible underlying data storage schema?
One day? One week? One month? More?
OK, first API was published when, in September 2004? It's March 2005 already, 6 months have passed. And what do we have? The same few September APIs (slightly updated, I give Don that:-).
All the "new" stuff only clutters the help section, but nobody managed to get it work so far.. God knows, people tried.. Python, apple thingie, .NET, Win32...
I was lucky enough to start early, so at least I know which method works.
If a new developer joins the game - how on Earth would he be able to tell a good one from bad one?
I very much understand guys' frustration level. Have it all designed, coded, wrapped, surfaced into UI, all that work - only to find out that the declared method does not work no matter what you do. Let alone to hear at least an acknowledgment of the fact that it does not work, so we can stop waisting our time in those futile attempts and do something useful until we get another green light...
I personally dropped the idea of trying the new APIs until I hear a confirmation that somebody got it working. Or until somebody from SM tells us that they actually did fix their stuff, so at least we get a hope of changes.
Whoooo-saaa :-)
getImageEXIF
Returns empty response
Life lost its meaning
Cheers!
Yeah, you are probably correct. I guess by "risk" I was refering to the fact that as soon as the API cat is out of the bag, it is very difficult to shove it back in. Any software or services that are created by third parties are just that ... third parties and smugmug no longer has true control over them.
For instance, it would be relatively easy to create a template style web application that totally replaces the smugmug UI on my personal site. This would take away one customization feature that people pay and upgrade to with smugmug.
But, in my mind, that is the beauty of it! See what kinds of things people are doing and embrace it.
The API as it is (and I'm not even talking about the broken methods) is the very, very base functionality. Performance-wise, it becomes very unwieldy to do relatively simple requests. For instance, if I want to return the the last five pictures I've uploaded, I have to make the assumption that the newest picture is in the newest album (the one with the highest ID).
Otherwise, I have to get all my albums, get all of the images within each album, and then get image details for every image. That is an enormous (and unfortunate) performance bottleneck. And let's face it, that's very discouraging, because without even knowing the database structure of SM, I can almost guarantee that their database has a field (or related field) for each image stating when it was uploaded. It would be a no-brainer SQL query that could be wrapped by the API, but it simply does not exist (yet).
Smugmug has the potential to be great (and I really want it to be great), but I feel like it's getting slowly eclipsed by other services.
Hey, while we're on the topic -- I would love, absolutely love -- the ability for smugmug explorer to have the ability to accept some type of plugin. I use Picasa for organizing all of my photos. Although it doesn't have an open API, most of the customizations I do to each picture (such as rotation, etc) is stored in an .ini file within each folder of pictures. I would love to be able to upload all of my pictures from smugmug explorer and then loop through the Picasa .ini file and perform any rotations on my uploaded images, set captions, and keywords, etc. I'm not sure how many other Picasa users there are out there, but man, would that be grand!
Aaron Christy
http://www.surfacedamage.com/
It really depends on *what* API are you surfacing. For the sake of this discussion, look at those "Magic Seven" I mentioned before. What can *possibly* go wrong if SM surfaces them? It's identical to what their own server does via web-based UI. far less than that - no search capability, no access to pro sales, no nothing, just basic management stuff.
That too. But even with existing *working* API it's possible to do a few things (don't ask:-) which may essentially undermine certain $-related issues.
(Baldy, don't worry, I'm not gonna do anything stupid just because I can:-)
605% :-)
I was asking (at least two times) for a possiblity to provide some, read-only, limited and controlled SQL access (not direct sql, simply certain clauses of SELECT, so nobody can kiil server by doing a couple of full outer joins on the whole DB).
Guess what was the answer..:-)
Well, you're absolutely right, there was none..:-))
Hard to say. I still like it a lot. I especially like the whole API thing.
But, to tell you the truth, the initial excitement is fading.. It's next to impossible to be a 3d party vendor without having good connection with primary R&D, which is, sad to say, virtually non-existing..
I will very much likely do it - just not now. No bandwidth left whatsoever.. What I can promise is when I start getting to it, I'll make sure to round table the ideas..:-)
Cheers!