 |
Update
Now the only necessary parameter in the feed's URL is the ID of the photoset (the user ID is no longer necessary).
Posted 32 months ago.
(
permalink
)
|
|
I have now written a Greasmonkey script that will insert RSS feed links into photoset pages, just as flickr does for group pools.
Read more above!
Posted 31 months ago.
(
permalink
)
|
 |
Nice! Great for those who have constantly growing sets.
Posted 31 months ago.
(
permalink
)
|
 |
Seriously - thanks for this! Everyday greasemonkey makes my flickr experience more enjoyable!
Posted 31 months ago.
(
permalink
)
|
|
I like that this one has both a greasemonkey script and a "real world web" element. Greasemonkey is great for us... but seeing/subscribing to rss feeds related to my stream is something I'd love to encourage... and not just among greasemonkey users.
Posted 31 months ago.
(
permalink
)
|
 |
@ alto
The things you are doing with "pipes" & "greasemonkey" are amazing.
Now for some open questions...
I went to yahoo pipes and created a pipe that included all of the groups that I like to keep track of.
After doing so I noticed that there are many photos that are cross posted in multiple groups.
(example)
a photo of a "goth" person dressed in "black" being posted in a "goth" photo pool and also posted in the "black" photo pool.
To keep this from occurring, I used a "unique" operator, and filtered the non-unique items by "item.title".
This has worked very well for me so far.
My question is...
Am I filtering out images that are different, but have the same "item.title"?
If this is the case, is there a better way to filter out cross posted images without removing unique images using the same "item.title"?
Also I'd like to know, if I add a new "group pool photo feed" to the same pipe and rebuild the pipe, do I have to resubscribe to that feed in my feed reader?
Thanks in advance to anyone who can help me with this.
Posted 31 months ago.
(
permalink
)
|
|
clickykbd, squarefrog, doorman_665, thanks & you are all welcome.
doorman, I have published a pipe that will take your group OMPL file and aggregate all the feed discussions together. Then I modified it to work for group pools, but did not pursue it so I left it unpublished!
I had encountered your problem, and tackled it as follows. If you look at the items in the RSS of a flickr group pool, the link is something like http://www.flickr.com/photos/dwardu/457094861/in/pool-96035807@N00. Remove the final /in/pool-96035807@N00 part with a Regex replace operation, and then the same photos appearing in different pools will have the same link value, to which you may then apply the Unique filter to that and solve your first problem.
I'm not sure if I understand your second query, but in order to not need to resubscribe to a pipe's feed, you should do the modifications inside the same pipe. Also, if you have inputs you must leave them with the same name (though not necessarily the same prompts) as these appear as parts of the RSS feed's URL as query parameters. Remove an input only if you have made a modification that renders it useless (e.g. before you were asking for username and userid, but now you are just asking for userid because you are getting the name via an API call). If you add a new input, keep in mind that the default value you will give it will be the one that will apply for anyone who had subscribed to the pipe's feed before your modification i.e. whose feed URL to which they are subscribed does not include your new parameter.
Posted 31 months ago.
(
permalink
)
|
|
Thank you Alto for the quick reply
I have been very busy and have not been online to respond.
I can see where using my group OPML file to create the feed would work better.
If I subscribe to a new group, will the pipe update automatically through that same OPML file call.
I think I understand, a "Regex replace operation" removes the /in/pool-######@N00 part of the photo URL, leaving just the base portion of the link.
Then filtering those results with the unique filter, leaves only one call to each photo in the output.
I think I will start over with a new pipe, now that I understand a little better.
This is my very poor first pipe attempt.
I think the next one will be much better..... I hope ;-)
Thank you again for your help
Originally posted 31 months ago.
(
permalink
)
doorman_665 edited this topic 31 months ago.
|
|
doorman, you are very welcome :)
No problem.
I posted my reply to you in the Group OPML Y! Pipe's topic…
Posted 31 months ago.
(
permalink
)
|
 |
Hello. Nice tricks with the pipes! But, is this suppose to work still? It seems that no photos are piped from any given set. I have tried your exemple: pipes.yahoo.com/pipes/pipe.run?_id=ggi_H7jn2xGoWGzY6UjTQA...
Which results in the following :
<rss version="2.0">
<channel>
<title>Flickr photoset</title>
<description>
More info at www.flickr.com/groups/flickrhacks/discuss/72157600063281311/
</description>
<link>
pipes.yahoo.com/pipes/pipe.info?_id=ggi_H7jn2xGoWGzY6UjTQA
</link>
<pubDate>Tue, 29 May 2007 07:34:50 -0700</pubDate>
<generator>http://pipes.yahoo.com/pipes/</generator>
</channel>
</rss>
<!--
m8.pipes.re3.yahoo.com uncompressed/chunked Tue May 29 07:34:24 PDT 2007 -->
Posted 30 months ago.
(
permalink
)
|
|
Thanks!
Though… it works for me!
Try again… Yahoo! Pipes exhibits these sort of glitches at times…
Actually it is not much of a problem if you use an online RSS reader, if it checks a feed and finds nothing, but finds something when it checks it again… you wouldn't even know about the glitch.
Posted 30 months ago.
(
permalink
)
|
|
Yes! I didn't worked a few times then after a few reloads it started to work out. The unreliability of the Yahoo Pipes is a bit unfortunate since I wanted to display a photoset on my blog.
Besides, the rss output differs a bit from the "original" where the media:content and media:title elements are missing. I think some tools (such as Word Press Flickr widget) logic is based on such elements. I ll investigate both the output and the pipe itself and keep you posted.
Thanks for the response.
Posted 30 months ago.
(
permalink
)
|
|
Thank you.
Have a look at this post about media:* elements.
Posted 30 months ago.
(
permalink
)
|
|
I created an alternative version of the pipe that encloses the photo and a thumbnail in every item (via media:* elements).
I modified the innermost subpipe flickr.photosets.getPhotos so that it creates a photourl field containing the URLs for different sizes.
Then in the top level pipe I inserted a "Rename" module before the "Pipe output" that renames the URLs to the appropriate media:* elements.
The pipe you see is a clone of the original. I have not added the functionality to the main implementation (so that I can test it a bit first).
I chose the "medium" as the media:content, and the "thumbnail" as the media:thumbnail. If you want to change these, clone my pipe, and change them from the drop-down field lists in the final "Rename" module.
Ideally I would have just created a new pipe, embedded the original "media-less" pipe, recreated the input fields, and added the "Rename" module. However for some reason this did not work, although it should have, and it usually does… so I just cloned the complete pipe.
Posted 30 months ago.
(
permalink
)
|
|
I had a look at the code of the generated RSS feed… and it's not working well! Must be a temporary problem with Pipes.
Posted 30 months ago.
(
permalink
)
|
|
Recently, some people asked me whether it is be possible to develop a Yahoo! Pipe that accessed the Flickr API without publicizing the Flickr API key. I replied that this would not be possible.
Thinking about it, I realized that it is not entirely true. There is a way to do it, at least if the Flickr user is logged in. In that case, there exists a javascript variable called global_magisterLudi that contains an API key. The pipe could be modified so that it accepts the Flickr API key as a parameter, and e.g. a javascript bookmarklet could be written that reads the value in that variable and passes it to the Pipe. As a simple demo, I edited the "Flickr photoset" pipe so that it takes the API key as a parameter. Then a bookmark (or favorite) containing the following URL could be written: javascript:location.href=
"http://pipes.yahoo.com/pipes/pipe.info
?num=500&set=72157594465647188&_id=njnTPfMi3BGfhxOVn0artA&_run=1&api_key="
+global_magisterLudi on one line. It works.
The downside is that API keys obtained via this method will expire quickly, so it makes no sense to, say, subscribe to such a Pipe's RSS feed.
Originally posted 29 months ago.
(
permalink
)
alto maltés edited this topic 28 months ago.
|
|
yeah great hack, but non-sensical... as feeds are intended to be read in readers, many of which are run on other domains, which will certainly not have access to this variable.
I had the thought to host my api-key externally and deliver it to the pipe in an get-data module. This is only pseudo private, as anyone could still get to it by viewing the pipe in debug mode, but certainly more hidden than just defaulting it into an input element.
You example is interesting though because it would allow for authenticated calls if I understand it correctly?
Originally posted 29 months ago.
(
permalink
)
clickykbd edited this topic 29 months ago.
|
|
"You example is interesting though because it would allow for authenticated calls if I understand it correctly?"
Yes, though you would need to access other variables for the other necessary authentication parameters (auth_token, api_sig) (which is trivial).
"yeah great hack, but non-sensical... as feeds are intended to be read in readers, many of which are run on other domains, which will certainly not have access to this variable."
As I said, it does not make sense to subscribe to the RSS feed of a pipe that has been passed a Flickr API key of such a volatile nature. However, RSS is not the only useful rendering of a Yahoo! Pipe's output! e.g. the HTML rendering is useful too, and another possible use of this method could be to compose a URL pointing to a Google Map that superimposes a KML generated by a Y! Pipe (that takes a Flickr API key as an input parameter).
Therefore, since there are other useful applications of the hack that do not involve RSS, what is non-sensical is subscribing to an RSS feed of a so-"hacked" Pipe, and not the hack per se.
Posted 29 months ago.
(
permalink
)
|
|
"I had the thought to host my api-key externally and deliver it to the pipe in an get-data module. This is only pseudo private, as anyone could still get to it by viewing the pipe in debug mode, but certainly more hidden than just defaulting it into an input element."
If one is to hardcode the Flickr API key inside the pipe, I would not set it as the default value in an input parameter, but rather embed it inside the pipe itself, as I have done for all my pipes until now.
I have also thought of your option, but if anyone wanted to "steal your API key", it would be trivial to do it. Hence, personally I prefer to hardcode it inside the pipe, thus keeping things simple (and avoiding a further GET, for-each, etc.).
Posted 29 months ago.
(
permalink
)
|
 |
OK, how about the other half of this? IE, a feed for a particular user's sets? Basically, a feed for www.flickr.com/photos/*username*/sets/ ...
Posted 28 months ago.
(
permalink
)
|
|
I had implemented this, but forgot to publish it! I published it now :)
Posted 28 months ago.
(
permalink
)
|
|
Perfect! Thanks.
Posted 28 months ago.
(
permalink
)
|
 |
I don't know why, I just discovered that now, it's awesome!
Posted 28 months ago.
(
permalink
)
|
|
Hehehe :)
It's been here for a while!
I like this hack.
Posted 28 months ago.
(
permalink
)
|
|
it would be nice to have a way to automatically create a feedburner feed for each set. I am not sure if there is a nice API/entry point to feedburner... will have to check.
Posted 28 months ago.
(
permalink
)
|
|
It's here. It is called via a POST.
It would be very cool to enhance this script so that if the set's owner is viewing the page, instead of the RSS feed icon and "Subscribe to photos from the * set", he gets the FeedBurner flame icon and "Burn the feed of photos from the * set with FeedBurner". That would automatically pass the set title too!
What would be even cooler would be if the script can get back the response and add the "Subscribe to this feed" link pointing to the newly-burnt feed automatically to the description ;)
Originally posted 28 months ago.
(
permalink
)
alto maltés edited this topic 28 months ago.
|
|
I tested it (quickly using a simple HTML form… I've never POSTed from a GM script yet), and it works fine!
What is also nice, given that you need to pass to API the FeedBurner username and password, is that you may call the API over HTTPS.
Posted 28 months ago.
(
permalink
)
|
 |
So I love this hack, exemplifies everything cool about small pieces/loosely joined. Even makes me feel kind of guilty about this.
But I've added set feeds to the site. I'll be adding links pointing to them soon, but wanted to get your all feedback on the feature first.
Base feed urls is: api.flickr.com/services/feeds/photoset.gne
Takes all the standard Flickr feed arguments (lang, format, etc), plus
* set - the id of the set
* nsid - the nsid of the set owner
Feeds are sorted by the date the photo was added to the set, though the data for this only goes back 2-3 weeks. (so photos added before that are all treated as happening back when dinosaurs roamed the Earth, 1970 or so)
And just like all feeds on Flickr only public photos are included.
Thanks
Posted 28 months ago.
(
permalink
)
|
|
Hi kellan!
I checked out the feed… great :)
My feedback:
– If set IDs are unique, isn't the NSID parameter redundant, and hence unnecessary? (e.g. knowing the set ID, a 3rd party app would have to make an extra call to get the correct NSID)
– I see that the feed lacks an icon… how about the user's buddy icon, or maybe the set's primary photo square thumbnail?
– It is a good idea that feeds are sorted by the date the photo was added to the set, but it would be very useful for 3rd party apps (e.g. Pipes) that hook onto the feed if that date could be provided as an extra element inside every item. If you don't find a suitable RSS/Atom/DC element to describe that, you could always invent a new flickr xmlns…
– Why did you change the flickr tag: URI convention? I would have expected the feed and item IDs to be respectively in the forms tag:flickr.com,2005:/set/72157594465647188
tag:flickr.com,2005:/set/72157600818404318/photo/808104061 In fact that's how I had implemented them in the Pipe…
Once we're at it, why not add a num parameter to all Flickr feeds to control the number of items in the feed? If unspecified, it would default to 20…
I'm surprised you only started storing "date added to set" some weeks ago! Setting them to to Unix time 0 is not a bad solution, but if would be a good idea if the internal sorting algorithm sorted primarily on the "date added to set", and then secondarily on the upload date, as it will be very likely that the ordering of the unannotated photos will agree with the date they were added to the set, as I believe most images are added to their sets on upload.
The license element is cool.
I hope my feedback is of any help to you. Thank you for asking!
Well done.
Posted 28 months ago.
(
permalink
)
|
|
In the meantime I will start preparing for my Pipe's and script's funeral… :)
Posted 28 months ago.
(
permalink
)
|
|
kellan, I was thinking… actually, why doesn't Flickr power its own feeds with Yahoo! Pipes!?
Posted 28 months ago.
(
permalink
)
|
|
hey alto,
thanks for the feedback.
If set IDs are unique, isn't the NSID parameter redundant, and hence unnecessary? (e.g. knowing the set ID, a 3rd party app would have to make an extra call to get the correct NSID)
Actually the API method which allows us to find a set just by ID is a hack for backwards compatibility. Set ids are globally unique, but should be thought of as only unique within the namespace of the set owner. (because thats how we store them)
I see that the feed lacks an icon
Good catch, I'll add one.
would be very useful for 3rd party apps (e.g. Pipes) that hook onto the feed if [the date the photo was added to a set] could be provided as an extra element inside every item
Good suggestion, let me look at it. (not like we don't have a history of making up our own feed fields)
Why did you change the flickr tag: URI convention?
Just what our feed generation library spit out. I blame computers. :)
I'm surprised you only started storing "date added to set" some weeks ago!
We tend to not store data we don't use, at scale even small numbers start to add up. As sets have a manual ordering element...
secondarily on the upload date
Hmmm, definitely not worth doing the join, I'll take a look at doing it in script space.
Thanks for the feedback!
Oh. And because its significantly cheaper for us to operate on our own data, Pipes flexiblity comes with costs.
Originally posted 28 months ago.
(
permalink
)
kellan (staff) edited this topic 28 months ago.
|
|
You're welcome.
I see.
Great.
Great.
Hehe.
I see.
Great.
You're welcome.
True.
Have fun :)
Posted 28 months ago.
(
permalink
)
|
|
So, hey kellan - are we going to get a feed (with icon) for all the users sets? I'd like to be able to point less savvy members of the family to just a feed icon at the bottom of this page: www.flickr.com/photos/*username*/sets/ and then when I add a new set, they get it in a reader.
Posted 27 months ago.
(
permalink
)
|
 |
If you want to hide the API key, you could use a simple PHP or equivalent server script (ASP.NET, CGI, etc) to get the pipes feed and pass in the API key. Your own server script would then hide the API key from other users. If you wanted to call your own script from Javascript / Ajax, etc, (whcih publicly advertises your URL in your HTML source), then anyone nosey wouldn't see the API Key. You could also employ caching to speed things up.
For example in ASP.NET:
myserver.com/feeds/myflickrsets.aspx
Source:
string strResult = "";
string myAPI_Key = "4f4jh354j35j3jg3j4g5j3g3g45g3jh4";
string url = "http://pipesURI/params?xxx&api_key=" + myAPI_Key
System.Net.WebRequest objRequest = System.Net.HttpWebRequest.Create(strURL);
System.Net.WebResponse objResponse = objRequest.GetResponse();
using (StreamReader sr =
new StreamReader(objResponse.GetResponseStream()))
{
strResult = sr.ReadToEnd();
sr.Close();
}
Response.Clear();
Response.ContentType = "text/xml";
Response.ContentEncoding = Encoding.UTF8;
// Could cache this as well.
Response.Write(strResult);
Posted 25 months ago.
(
permalink
)
|
|
Or even easier is to take the API key in as text input, and then hide it behind a Feedburner feed. i.e. paste the full RSS feed from Pipes into Feedburner and bunr it. That way, external users don't see your API key (assuming you don't leave it embedded in the Pipes source.
I've used this technique to build a "Life Log": feeds.feedburner.com/JuntoLifelog
Originally posted 25 months ago.
(
permalink
)
Junto edited this topic 25 months ago.
|
 |
Is anyone else getting an error now with the greasemonkey script? I believe Flickr has changed the underlying structure of the set page.
document.evaluate("//div[following-sibling::div[1][@class='Footer']]", document, null, XPathResult.FIRST_ORDERED_NODE_TYPE, null).singleNodeValue has no properties
Line 58
I made the following edit to the script to get it to work now...
document.evaluate("//div[following-sibling::div[1][@class='FooterWrapper']]", document, null, XPathResult.FIRST_ORDERED_NODE_TYPE,
just change line 58 from Footer to FooterWrapper...
Originally posted 25 months ago.
(
permalink
)
www.jookwarrior.com edited this topic 25 months ago.
|
|
jookwarrior, thanks!
I updated the script to reflect the change.
I changed it to document.getElementById('Main').appendChild(feeds_div);… I think it will be more stable like that… + an XPath less… :)
Thanks again!
Posted 25 months ago.
(
permalink
)
|
|
junto, if one is going to involve a 3rd party website, then yours is probably the best idea to hide the API key.
Thanks.
Your Feedburner solution is good, but only for pipes that take no parameters.
Originally posted 25 months ago.
(
permalink
)
alto maltés edited this topic 25 months ago.
|
|
sweet! and no problem!
this script with the yahoo pipes (I changed that too slightly) makes my life so much easier for my weblog posts...I couple it with a RSS tool I made here...
www.ocf.berkeley.edu/~andychou/rss.php
to create my layout for my website. Imagestation closed down so I had to move over to Flickr completely for my image hosting needs.
THANK YOU! :-)
Originally posted 25 months ago.
(
permalink
)
www.jookwarrior.com edited this topic 25 months ago.
|
|
Cool!
You're welcome! :)
Posted 25 months ago.
(
permalink
)
|
 |
i tried using your cool script to get FLICKR SET feeds ...i first got them using the greasmonkey script (latest one you have up) then i burned that feed in FEEDBURNER.....and i couldn't get XML Flash Slideshow to pull in those feeds....this is what the programmers at XML Flash Slideshow told me:
"it appears that the first feed does not conform to Media RSS specifications. It only includes basic RSS information. (The image URL is embedded in the description, not a standalone media:content tag)."
**********************this is my original email to their tech support:
ok this is strange but i have one feed from feedburner working and no matter what i try i can't get others working .....the only difference between the RSS feeds is that the ORIGINATE differently one is a standard FLICKR feed burned at FEEDBURNER and one is a yahoo pipes feed to pull the FLICKR sets
but check this out this does not work
www.scottbarnes.tv/proxy.php?url=http://feeds.feedburner....
and this one does!
www.scottbarnes.tv/proxy.php?url=http://feeds.feedburner....
why?? ....this is the only change if i swap out the second one that works with the first one the slideshow gives me an error 'IMAGE FAILED TO LOAD'
i've even tried using a different cross domain proxy script but same thing....i've spent a couple hours wrestling with this and i give up any help would be appreciated very much!
thanks
***************
can you help me get the SET feeds to conform to Media RSS specifications?
thanks for your work ! makes Flickr much more powerful to be able to pull in sets with RSS this way!
Posted 25 months ago.
(
permalink
)
|
|
Ok, I'll add the media elements when I have some time…
Posted 25 months ago.
(
permalink
)
|
|
thank you so much!
will you publish a paypal address to send you donations?
Posted 25 months ago.
(
permalink
)
|
|
scottbarnesbeauty, hehe :)
I updated it to include the media elements.
(While I was at it, I restructured the whole Pipe… hope I didn't break something else!)
Give it a go… and tell me if your app works. I hope it supports the media:group tag…
Posted 25 months ago.
(
permalink
)
|
|
See details here…
Posted 25 months ago.
(
permalink
)
|
 |
I love you!!! You are saving my life and you don't even know it! Thanks man! :) :) I've been playing with your pipe and I'm stunned, I just found out this is exactly what I've been searching all over the net for a week now. Now I'll be able to sleep at night!
I only have two questions: how do I embed a "pubdate" for each picture in the feed (based on "lastupdate" or ideally on "datetaken")? and is there a way to include the caption (description) from the flckr picture too?
Thanks for your help!
Posted 25 months ago.
(
permalink
)
|
|
aofc, hehehe :)
To get the "last_update" or "date_taken", you must add these to the values already being passed from inside the Pipe to flickr.photosets.getPhotos via the "extras" parameter.
With "last_update", it's easy as the Flickr API returns it as a Unix time, so you just have to copy that value to a field named "y:published.utime".
For "date_taken", which is not formatted as a Unix time, you may embed this Date Item pipe into a "Loop", pass to it the date_taken returned by the Flickr API, and "assign first result" to "y:published".
The reason why I didn't include a pubDate is that I thought that some RSS readers might then try to sort the items by that pubDate, rather than leaving the order of the set as intended by the set owner (though I'm not so sure on how readers are actually ordering them). Ideally the pubDate would be the "date added to set", but this is not available through API.
Of course, yes, there is a way to include the description. I did not do it for performance reasons. As it stands, the Pipe makes a single call to flickr.photosets.getPhotos, whose response contains all the information necessary to build the current feed. However, to get the description, the Pipe would have to call flickr.photos.getInfo for every photo! so the worst case would be 500 calls! I think it would make the pipe unnecessarily slow. But of course, if you only access 20 photos from the set, then it shouldn't be too much of a problem.
Actually, it normally makes sense for a feed to contain a small number, like 20, of photos. However, it makes sense when those are the most recent 20. Since (I think) most people who have large sets do not manually reorder them frequently, the photos they "add to set" are added to the end of the set. My current Pipe implementation accesses the first number of entries, so (for large sets) accessing the first 20 would be useless.
Of course there is a way to access the last 20 photos, but it is not so straightforward as is the current implementation, and I have not yet dedicated time towards doing it.
Haha! That's a long post… and I'm not going to re-read it before I post… so forgive any redundancies it might contain :)
Originally posted 25 months ago.
(
permalink
)
alto maltés edited this topic 25 months ago.
|
|
Thanks a lot! Thanks to your explanations I was able to get exactly what I wanted, even though I am a complete newbie with pipes. I was able to get captions as I wanted, but there's still a minor problem: my captions are in French and while they are fetched correctly from flickr.photos.getInfo, they get messed up when "regexing" to include them with the picture in the "description" field. They end up not being displayed as Unicode, puting "é" instead of "é" for instance. Any hints on how to resolve this?
As for the pipe being too slow because of the numerous calls to flickr.photos.getInfo, it is not that bad actually, and since this pipe is intented for internal purpose, running once every X hours, it shouldn't put too much of a strain on flickr's api.
Posted 25 months ago.
(
permalink
)
|
|
Great.
Yes, I know… my username always becomes alto maltés, but I have decided to let Pipes resolve the bug, rather than wasting time with it myself.
However, you could do one thing… I have never tried it… but I think it should work… if the description contains html, then before doing anything, how about string-replacing accented characters with their equivalent HTML entity, e.g. replace every occurence of á with á and ç with ç and etc.
It might work!
Posted 25 months ago.
(
permalink
)
|
|
Ok, so it's a bug and not some option I have to specify? Alright, I figured out a way to have it work. Thanks a million!!!
Posted 25 months ago.
(
permalink
)
|
|
Great.
Can you share it?
Posted 25 months ago.
(
permalink
)
|
 |
Hello, I was looking for a way to return photosets in rss and this seems to be the magic trick to do just that. Excellent!
You wrote on your page about the recent update...
"However, in the current implementation, if the set owner does not permit the downloading of the original version, media:content/@url will contain an invalid URL. "
However, for some reason I am still getting the incorrect url returned. The url that is returned looks like the following...
http://farm3.static.flickr.com/12345/1234567890__o.
the correct path to the image would be something like this...
http://farm3.static.flickr.com/12345/1234567890_a21364fbf_o.jpg
I set the "allow download" to public, then back to private and back to public again but, the results are the same.
Am I doing something wrong or is this something you can look at on your end?
Either way, I apprectiate your efforts on this. It is really great.
Thanks
Posted 25 months ago.
(
permalink
)
|
|
If it returns that, then it means that for some reason, the Flickr API is not returning the "original secret" and "original_format" for that photo.
May you post a real link please, so I can do some testing? Thanks!
Posted 25 months ago.
(
permalink
)
|
|
Sure, and thanks for the quick response!
Here is a real link.
farm3.static.flickr.com/2274/1878333263_a212960fbf_o.jpg
the pipes rss feed is returning this..
farm3.static.flickr.com/2274/1878333263__o.
but this for the thumbnail(which is correct)..
farm3.static.flickr.com/2274/1878333263_a5b5835f64_s.jpg
One question, does it take a while for the photoset rss feed to get updated? FOr instance, I delete or add a photo to the set but the changes do not seem to be taking right away.
I am new to all of this so, forgive me if this is common knowledge.
Thanks again!!!
Originally posted 25 months ago.
(
permalink
)
www.scrude.com edited this topic 25 months ago.
|
|
I should probably have mentioned that, in the description section, the url is being returned correctly...
farm3.static.flickr.com/2274/1878333263_a5b5835f64_m.jpg
Thanks for your help.
Posted 25 months ago.
(
permalink
)
|
|
Ok… so it appears that even though I pass original_format to the extra parameter of flickr.photosets.getPhotos, I still do not get the original format and secret, even when for the same photo, and with the same API key (i.e. same permissions), I do get the original format and secret if I call flickr.photos.getSizes (which doesn't make sense, it must be an API bug)… so that's what I did, for every single photo in the set… thus defying the performance argument I brought in my previous post… but I don't care, as long as it works.
Posted 25 months ago.
(
permalink
)
|
|
Yes, it could take a while for your change in a set to be reflected in a pipe, as Pipes employs some form of caching (i.e. when it asks a question to Flickr, it "remembers" the answer for some time, to prevent bothering Flickr by asking it the same question over and over again).
Posted 25 months ago.
(
permalink
)
|
|
The RSS feed has been updated (you don't need to change anything).
It now fully supports "media" elements and the "enclosure" element.
Posted 25 months ago.
(
permalink
)
|
|
Absoutely Brilliant!!
Works perfectly now. Thank You for addressing this so quickly. You are truely a PRO, just like it says next to your username.
What a great sollution.
Ok, now I just have one more question. Is there a way to set a parameter to get the url of the desired size? The reason I ask is because getting the Original returned is not always optimal when your needing a smaller image. I was thinking there could be a "size" parameter in the pipe string, since you are using flickr.photos.getSizes anyway.
Love this. And thank you again!
Posted 25 months ago.
(
permalink
)
|
|
Hehehe.
Check it out now!
Enter the desired side length for the enclosure, and it will choose the "best" available version :) i.e. the version of the photo in question that is publicly available whose area is closest to the area of a square with the specified side length. The parameter is called enc_size.
I think this is better than choosing "Small", "Medium", "Large", etc. as you would have to know what these meant, whereas as I put it, you don't need to know any of this, but just specify the target size in pixels you would like.
I set the default to 10000, so that it will choose the largest available (most probably the original) if enc_size is left unspecified (as it is for those who are already subscribed to sets' feeds). If you wanted e.g. to choose the smallest picture available, you would set it to 0.
What do you think? :)
Originally posted 25 months ago.
(
permalink
)
alto maltés edited this topic 25 months ago.
|
|
Alto,
You're a genius! This is exactly what I wanted, only better!!! This is so much more useful than flickr default feed. I don't think I'll use anything but this.
Great Great Job! Very impressive.
Am I correct that, if I remove the &set=72157602056923859 variable from the string that it will return the feed from the default root of the users account? In other words, can this be used on the root account, as well as the set? I may be wrong but, I don't think the default feed that is available from Flickr allows you to specify size. It certainly doesn't handle it as well as you have.
Very very nice.
Posted 25 months ago.
(
permalink
)
|
|
Hehe. I'm glad you find it useful.
But no, if you leave the set parameter out, it will just not work.
By "default root" I supposes you mean photostream… yes it would be a good idea to have all feeds be generated from the same pipe, just by changing the parameters you pass… but that will be another time.
Cheers.
Posted 25 months ago.
(
permalink
)
|
|
Alto, I just added a single regex to convert "é" to "& eacute;" in my captions and it only converts the first instance of the letter. Oddly enough, it is sufficient to trick the pipe into outputting text in the correct format, and all other accentuated characters are correctly displayed. Of course it's only a temporary fix, but until they correct this it works and I'm happy. :)
Posted 25 months ago.
(
permalink
)
|
 |
Is there anyway to see a set's RSS Feed without a Greasemonkey script? (Which freaks me out and is way beyond my technical know-how?) LOL
Posted 24 months ago.
(
permalink
)
|
|
Sure :)
(I just updated the description above to make this clear)
Originally posted 24 months ago.
(
permalink
)
alto maltés edited this topic 24 months ago.
|
 |
I'm new to pipes, but I'd love to see a rendition of this one that pulls all the photos from a group, rather than a set.
I saw Alto Maltes' 'flickr.groups.pools.getPhotos' pipe, but it doesn't produce any thumbnail images. Is it possible to achieve results for the groups pipe that look like the results for the photosets pipe?
Posted 23 months ago.
(
permalink
)
|
|
@ateencreative, note that the Pipes listed on the Flickr! Pipes front page (which I have now transferred here) are intended to be used as sub-pipes i.e. as modules embedded in other Pipes, and not as stand-alone implementations (I have now added a note pointinig this out).
It was pretty simple to modify the Flickr user photostream Pipe to make it work for groups.
Here's the result
Originally posted 23 months ago.
(
permalink
)
alto maltés edited this topic 23 months ago.
|
 |
hm. i used your tool to create a feed url for a flickr photoset. it worked fine (thanks for developing it, btw, it's really useful!) until today... today i don't get anything...
have a look at the feed:
pipes.yahoo.com/pipes/pipe.run?_id=ggi_H7jn2xGoWGzY6UjTQA...
am i the only one with this problem?
comments would be appreciated! :)
Posted 23 months ago.
(
permalink
)
|
|
@stokedstephi, it must be a internal problem in Y! Pipes. The subpipe calling flickr.photosets.getPhotos works well when I test it by itself (so it's not a communication problem between Y! Pipes and Flickr), but inside the top-level Pipe gives the following error in the debug window: error fetching ine.pipes.vip.re3.yahoo.com:80/cgi-bin/rsspipes/dispatch?…
_out=PERL&_data=%04%07%041234%04%04%04%08%03%07…
%00%00%00%0AQlicense%2Cdate_upload%2Cdate_taken%2Cowner_name…
%2Cicon_server%2Coriginal_format%2Clast_update%06%00%00%00…
extras%08%81%04%00%00%00page%06%F4%01%00%00%08%00%00%00…
per_page%0A%16KL6ErB6H3BGYL5obYEsBXw%03%00%00%00…
_id%0A%1172157603256745425%0B%00%00%00…
photoset_id%0A+fe88fe5e5eb1905a208336c07fbe22a2%07%00%00%00a…
pi_key%0A%0Ephotoset.photo…
%05%00%00%00_path&_action=run&_in=PERL (500 Can't connect to engine.pipes.vip.re3.yahoo.com:8080 (connect: Connection refused) at /home/y/lib/perl5/site_perl/5.8/LWP/Protocol/http10.pm line 37. ) I've posted it on the Y! Pipes message board as a bug. Thanks.
Originally posted 23 months ago.
(
permalink
)
alto maltés edited this topic 23 months ago.
|
|
thanks a lot! i'll check the Y! pipes message board and this group here for updates on this issue!
Posted 23 months ago.
(
permalink
)
|
|
@stokedstephi, the Y! Pipes team have fixed the bug, and your feed is now working.
Posted 23 months ago.
(
permalink
)
|
|
amazing! thanks a lot and also thanks to the Y! pipes team!
Posted 23 months ago.
(
permalink
)
|
|
@mortimer?, I had never got down to writing a script to communicate with the FeedBurner API… but have a look at this.
Posted 23 months ago.
(
permalink
)
|
|
@stokedstephi, you are welcome :)
Posted 23 months ago.
(
permalink
)
|
 |
You don't need this anymore, Flickr provides it.
Posted 22 months ago.
(
permalink
)
|
|
Nicolas Hoizey, yes, I’ve just realised! Cool.
I won’t kill it however. I’m currently extending this script to do… stay tuned ;)
Posted 22 months ago.
(
permalink
)
|
|
Teasing... ;-)
Posted 21 months ago.
(
permalink
)
|
 |
clever...
Posted 13 months ago.
(
permalink
)
|
Would you like to comment?
Sign up for a free account, or sign in (if you're already a member).
|
Flickr Hacks /
Discuss
Feed – Subscribe to Flickr Hacks discussion threads
|
Deutsch
|
English
|
Español
|
Français
|
