Loading Profile...
Recent activity
Subscribe to this feed
kevin1 shared an idea in Get Satisfaction on September 21, 2008 08:27:
User created FAQsUsers of Get Satisfaction should be able to compile a list of frequently asked questions. I'm wondering why this feature isn't there already.
One of the key concepts of Get Satisfaction is for users of a particular program or company to help out other users. But, I find myself wasting a lot of time answering the same questions over and over again. This is why FAQs were originally created. I'd like to write an answer once and then point people to it. Yet, I still want to be able to update the answer if something changes.
1) There isn't any repository of *the best answers* to commonly asked questions.
2) I can't create a repository of my own. The best answers I came up with.
3) I can't create a directory of best answers others came up with.
4) Is "Share an Idea" supposed to be a user sharing a great idea with *other users*? Or if it's a feature request to the company making the product?
The same question can be asked in ten different ways and ten different threads are created. I know that somebody answered the question excellently before, but I can't remember in which thread. Get Satisfaction lacks the ability to search for the answer in a reply, so in order to give the best answer I have to scan through ten threads, again.
Get Satisfaction lacks advanced search capabilities. So even if I've answered the question before, I can't easily search through my own answers in order to find it so I can post a link to it.
Currently I can tag other's replies as useful, but where do I get a list of all the replies that I tagged as useful? And how do I search through that list for the reply I want?
The solution:
I propose that advanced users of Get Satisfaction be given the option to create a FAQ, (An advanced user may be someone who has posted more than 20 times, etc.) A FAQ is a fifth type of topic. Along with "Ask a question", "Share and idea", "Report a problem", and "Start a discussion" should be "Start an answer list". I think it should be purple too. :)
The FAQ would be different from a regular topic in a couple ways.
1) It would be a list of other topics. (question and answer and full discussion)
2) the main list (Q and A) is continually editable.
3) each topic in the FAQ expands (or links to) a regular Get Satisfaction discussion thread.
The FAQ would be a kind of collection of answers / best practices / known bugs or behavior. It would be a list of other topics. I imagine it would look like a lot like the current redesigned page for any product on Get Satisfaction; a list of topics that have a short, one or two sentence summary but can be expanded by clicking on it.
Each topic in the FAQ expands to reveal 1) the rest of the question and answer, 2) a further discussion about the topic. This would be a regular Get Satisfaction discussion thread. That way if something isn't clear, people can ask follow up questions. And the follow up questions can then be incorporated back into the answer. The thread starts out with the *original* answer in the top box. That way, continuity is kept. People can follow the conversation.
The question and answer part of any topic in the FAQ can be edited at any time. But the further discussion section would not be editable. Further, only the author of the FAQ can edit the FAQ. (or someone the author designates. For larger FAQs, the author may wish to have help. )
So, people can make their own lists of answers. Naturally, we need a way for the best list of answers to rise to the top. Others can tag a FAQ as useful or not.-
kevin1 replied on September 21, 2008 06:50 to the idea "Keybreeze and Sandy" in I want Sandy:
Keybreeze looks pretty cool. Ubiquity looks cooler though. It's an extension for FireFox that brings the power of Quicksilver.
Ubiquity is in beta and has some rough edges at the moment. Still, it's already showing strong promise. It lets you add new commands, formatting, etc.to FireFox. You can even subscribe to commands, etc. written by other people. These commands can do anything that a FireFox extension can do.
I wrote a script that you can use to send Sandy an email through the website. It's an update of an earlier script I wrote that creates a "bookmarklet". A "bookmarklet" is a bookmark that runs code internally. Mine is capable of sending Sandy an email with (or without) everything selected in the current page (and the url). The bookmarklet works without installing anything at all.
http://getsatisfaction.com/iwantsandy...
http://getsatisfaction.com/iwantsandy... -
kevin1 replied on September 17, 2008 23:54 to the question "move items from calendar to notes" in I want Sandy:
It is possible to change an item from an "Appointment" to a "Note". Just add or remove a date from the item. Then go to the http://iwantsandy.com/appointments page (or the Notes page) to verify that your item has been changed and now shows up (or doesn't) on the appropriate page.
Previously it wasn't possible to change items from one type to another. And this looks like a special case. As far as I know, this functionality was never announced.
However, I'm not sure this solves your problem. It's possible to change an item type after it's been created. It's not possible to specify what type of item you want Sandy to create. Anything with a time or date in it becomes an appointment. If you put the date in quotes, then Sandy will not treat it as a date, but that has other problems.
Currently, Sandy treats quoted text specially. She doesn't use it as a date, etc, but she does use the quoted text as part of the name of the new item.
This behavior is also new and as far as I know, it was never announced. -
kevin1 replied on September 17, 2008 23:12 to the discussion "Removing the 2-email limit" in I want Sandy:
I think Rael is right to be very concerned about email security.
The difference with Sandy and those other services nbr mentioned is that you can *do* a lot more with Sandy through email than those other services. I mean, what's the worst thing that could happen if Twittermail or Brightkite accepted unverified email? Does it have the potential of completely destroying your account? Deleting everything, or publishing all your sensitive information?
Sandy is directly controlled by email. It's a command line interface. You can look up anything, forget anything, add anything, update anything through email. I would be very concerned about letting anybody have my username and password. Well, with Sandy, email access is almost the same thing.
I think there should be a middle ground. Sandy should accept email from more addresses, but it shouldn't *execute* email from these addresses. Perhaps Sandy could simply say "Hey I received mail from blah blah. Do you want to add him as a friend?" or something like that.
Ideally, there should be a couple of different security settings for email addresses. If the "From:" address matches a certain address, then the security settings for that address should be applied. Maybe there could be 4 levels of access.
1) accept email from this address but don't execute it, simply ask me what to do.
2) allow this email address to *add new items* to my account (for tasks from the wife)
3) full access (allow adding new tasks, updating, lookup and forget. Also, allow them to subscribe to your feeds.)
4) default is to ignore the email. -
kevin1 replied on September 17, 2008 22:47 to the problem "Sandy appears to misunderstand commands at random (examples follow) - is there some explanation? This is maddening" in I want Sandy:
I wish there was a way for us (the users) to create a large FAQ here in Get Satisfaction. I've tried a couple times, but the results are less than satisfactory.
See the Search FAQ and the List FAQ. (Which is incorrectly named "How to use Lists") I wish I could edit these things.
http://getsatisfaction.com/iwantsandy...
http://getsatisfaction.com/iwantsandy...
That being said, the current documentation on the "remember +" power move is actually correct. (Although it could be better explained with more examples.)
Any a) text and b) dates in the first line are REPLACED by corresponding text and dates in subsequent lines. Any c) tags in the first line are ADDED to subsequent lines.
If you had added any TEXT in subsequent lines, you likely would have had an additional surprise.
r Faculty meeting @meeting @campus 12-1:30pm +
* September 3, 2008
* October 8, 2008 (professor Tang will be there)
Is interpreted as:
r Faculty meeting @meeting @campus September 3, 2008
r (professor Tang will be there) @meeting @campus October 8, 2008 -
kevin1 replied on September 17, 2008 22:30 to the problem "Error Experienced from Server" in I want Sandy:
I get this a lot. I request the digest or the list page and I get the error Bricoleur mentioned. Then I refresh the page and I get the error again. Then I refresh the page again and it works.
This has been going on for about a month, maybe more. Happened just a couple minutes ago too.
I haven't mentioned it because it hasn't been that big a deal to me. I refresh the page. Even twice, when it's necessary.
I figure Sandy's going through growing pains. And because she's *free* there's not a lot of money to pay for bigger and better servers / infrastructure yet.
Of course I could be wrong. This behavior could be the result of a particular bug. When you have limited horsepower, smaller bugs can make a big difference. -
kevin1 replied on September 17, 2008 20:49 to the idea "Firefox plug-in for Sandy" in I want Sandy:
Oops. Maybe I'm not being clear. I think that having Sandy integrated with FireFox is a great idea. Extremely useful. I'm just saying that integration probably won't be a proper "plugin" or "extension". There's a newer (and easier) way to do it.
I'm very interested in creating a Sandy add-on for FireFox. I just don't have the time or the knowledge to create a full extension.
The new projects make it possible to add on just a little bit to to the already existing extension / framework. Just a single function or two. Using simple javascript. Previously that wasn't possible. Developers had no choice but to create a full extension. Now the Ubiquity project (for FireFox) and Accelerators (for IE8) will provide an easy way to add a little bit of functionality to your browser. -
kevin1 replied on September 12, 2008 08:25 to the idea "Firefox plug-in for Sandy" in I want Sandy:
There probably won't ever be a Firefox plug-in or extension for Sandy, as there's now a better way to do it.
There are some projects shaping up that should make developing simple add-ons a lot easier. Previously a full extension had to be written in order to add functionality to the browser. These new projects make it possible to write a few lines of script to add some functionality, instead of spending hours writing a full extension. It makes adding custom functionality much easier. A simple script is much easier and faster to write than a full extension.
Multiple add-ons done as scripts can use the same scripting framework and host. In FireFox, this scripting framework/host is implemented as the Ubiquity add-on. In Internet Explorer 8, they have the ability to have little scripts called "Accelerators". This is close to the same thing. (IE8 provides a scripting framework and can host script add-ons.)
As a small example of scripting, see the bookmark icon I helped create. It pops up a box for the user to fill in with new items for Sandy. It can also send to Sandy all the selected text on the current page, as well as the page URL.
http://getsatisfaction.com/iwantsandy... -
kevin1 replied on September 12, 2008 07:55 to the idea "Can we get a bookmarklet for Sandy?" in I want Sandy:
I've experimented with it and I still can't get it to do exactly what I want it to do.
I'm still having a problem sending a URL to Sandy in such a manner that Sandy will display the URL as a link with a given name, not as simple text. For now, that doesn't seem to be working very well. I've fallen back on simply having Sandy display the URL as text, since sending Sandy an html link causes other problems.
javascript:var msg=prompt('Write to Sandy:',document.getSelection()+' ');if(msg!=null){var win=window.open('http://iwantsandy.com/email/create?message[subject]=From bookmark&message[body]=Remember '+msg+' ('+location.href+')','_iwantsandy');} -
kevin1 started following the idea "Firefox plug-in for Sandy" in I want Sandy.
-
kevin1 replied on September 12, 2008 06:51 to the question "How can I remove tags from the "My Tags" area?" in I want Sandy:
The only way to remove a tag from your "tag cloud" is to make sure that no Sandy items use that tag. This means that each of the items that use the tag must be manually edited to remove the tag.
The fast way to remove a tag from the tag cloud is to click on the tag name in the tag cloud. Sandy will display a list of all items with that tag. You can then edit the items one at a time.
People have reported that tags in the tag cloud will not go away until ALL items with that tag are edited to remove the tag. This can include archived items. (I'm not sure if this problem still exists, it might have been changed.) I know that the default list of items with a particular tag **does not include archived items**. So finding archived items that use the tag you want to remove can be confusing.
To include archived items in your tag list, add "archived=1" to the url. This will cause those items using both the "archive" tag AND the tag you want to remove to also be displayed in the list. (In fact, it's probably better to add all three special types to the url. Including "done=1" and "past=1".)
Example:
http://iwantsandy.com/list?tags=offic...
See the Search FAQ for more info on searching.
http://getsatisfaction.com/iwantsandy... -
kevin1 replied on September 10, 2008 00:30 to the discussion "Jott+IWantSandy goes Pay-Only" in I want Sandy:
I set this up just the other day to test it out.
1) create a ReQall.com account.
Connect your phone to it. (Settings -> Phone access)
Store the ReQall phone number in your phone's address book.
Copy the RSS feed address by going to Settings -> Feeds
http://www.reqall.com/user/feeds/rss/...
2) create a Twitterfeed.com account.
This means that you need an OpenID account.
(Try myOpenID.com if you don't have one.)
3) create a new twitter feed.
Enter the Twitter account username and password.
Paste in the ReQall RSS feed address.
Update frequency: every 30 minutes.
Prefix each tweet with: "d s ". (That's: D space S space)
(Make sure to include the ending space.)
4) call ReQall and "Add" a new item (press 1).
Every half hour, Twitterfeed will pull a new version of the RSS feed from ReQall, check to see if anything is new, and then publish the new items to Twitter. Twitter will send a direct message to Sandy (because of the "d s " prefix).
So Sandy will receive a new Tweet from Twitter with your instructions.
Remember that because this workaround goes through Twitter, you're limited to 140 characters.
This hack is necessary because
1) ReQall doesn't have a "link" to Sandy like Jott does. It's not as easy to send Sandy a voice transcription of what you said.
2) TwitterFeed only feeds Twitter. (and laconica). It can't feed your Sandy account directly. You also can't set up Sandy to subscribe to the ReQall rss feed.
3) ReQall can send you new transcriptions by email. If it was easy for Sandy to accept emails, we could use this method and bypass Twitter.
It is possible to set up ReQall to send email to your email account ever time a new transcription is made. You could set up a filter in your mailbox to forward mail from ReQall to Sandy. However, you'd still have to tell Sandy to accept email from the ReQall email address. (The "From:" field remains set to reqall.com). As a result, it means that using ReQall email notification requires you to use the ReQall address as your second Sandy email address.
4) ReQall can send you new transcriptions by IM. If Sandy understood IM, you could have ReQall IM Sandy and skip the whole Twitter thing. -
kevin1 replied on September 09, 2008 01:42 to the question "Why can't I set and get reminders on my phone?" in I want Sandy:
Is it possible to just use text messaging? Yes. With some workarounds.
You should be able to register your phone with Sandy and she can send you text messages directly. (Warning: these might cost you $0.10 or $0.20 each depending on how your carrier charges you. You might want to get a texting plan.)
It's not possible to send Sandy text messages from your mobile phone. Even if you have it registered with Sandy. Instead you have to use a workaround. If you try to text message Sandy directly, your message will most likely be converted into an email by your carrier (i.e. AT&T) and Sandy will reject the email because she doesn't recognize the sender address. (i.e. 9095551234@att.com)
1) You can use Twitter from your mobile phone to send a direct message to Sandy.
something like: d s r appt with Miranda Wednesday at 5:15.
2) You can use Jott (or some other service) to transcribe your voice and send the message to Sandy.
3) find out what email address your text messages are sent from and register that address with Sandy.
There's lots more information on these topics. Do a search to find them. -
kevin1 replied on September 09, 2008 01:19 to the problem "Outlook 2007 problems...can't subscribe or import ICS file" in I want Sandy:
That's odd. Outlook 2007 should have solved those problems. Outlook 2003 (pre Vista) had a lot of problems. That's why your todo events without a defined date are excluded from all *.ics feeds. Because they make Outlook 2003 crash.
My best guess is that you've managed to turn off your private calendar feed somehow. Or, there's been a mistake and you're not using Outlook 2007 somehow.
Try this, open up the calendar file. (http://iwantsandy.com/xxx.../list.ics in a text editor and see what's in there. If it's html, then you've managed to turn off your calendar feed. If it's full of VCALENDAR stuff, then the problem's something else.
> BEGIN:VCALENDAR .... -
kevin1 replied on September 09, 2008 01:06 to the problem "Sandy still doesn't understand dollar signs" in I want Sandy:
Mike,
You're right and Sandy has a bug interpreting dollar signs. That being said, what was it that you desired to do? How much was Sandy off by? Also, I'll point out that Sandy had a change a couple of months ago and now titles notes with whatever is in quotations. (Instead of ignoring them completely.)
So it seems that, even if Sandy understood dollar signs correctly,
she would have interpreted
> r $56.47 Lunch D.C. with Donald at Station "6Sep" @expense
as
> 6Sep <tagged> expense
And the rest of the note would be available in the context.
Try moving the dollar amount from the front to the back of the sentence.
> r Lunch with Dan at the Station D.C. $56.79 "6Sep" @expense
comes out as
> 6Sep <tagged> expense</tagged></tagged> -
kevin1 replied on September 06, 2008 08:15 to the idea "Use of a FIELD:VALUE pattern" in I want Sandy:
Interesting idea. I agree with you about the contacts, contacts needs some work. I find that Sandy misinterprets what I write consistently. Even when I update a contact. http://getsatisfaction.com/iwantsandy...
Is your idea mainly focused on how Sandy manages contacts? Somebody else has suggested using a @contact tag to have Sandy enter a special mode where she reads the email. So that she'll look for personal information and add it to the contact. Would this work along well with your idea?
About the url and Bookmark Title. I can see your point. Although these are initialized, it's not easy to specify what they are.
Also, let me clarify one thing. Are you suggesting that Sandy be able to add arbitrary fields? In other words, would there be a small, limited number of named fields that Sandy would recognize, or is the number unlimited? If the number is unlimited, then I'm guessing that anything in front of a colon would be interpreted as a new field.
Also, have you thought about how to handle field names of more than one word? Like "Phone number", or "Office phone". How would that work? Do we have to be sure to put in a comma before every field name? What happens if it's not there?
Example:
Sandy, remember name: Dan Marif, address:1234 Big Box Road Topeka, Kansas, phone: 555-555-7890, office phone:555-555-1234 fax number: 333-444-5678 height: 6 feet weight: 200 lbs eye color: blue-green website: www.thegreatdan.com organization: Peppers unlimited -
kevin1 replied on September 05, 2008 09:45 to the discussion "I've made a ubiquity script to work with sandy" in I want Sandy:
Manatlan: good going. I had the same idea, to use Ubiquity to send email to Sandy.
I've been poking around in Ubiquity a little bit. Here's my take on how to send a new task to Sandy. I've added two commands, one sends a url, the other does a http post. I'm still trying to figure out how to open a new browser page to display the data result of the http post. (It results in an html file that tells you how Sandy interpreted your message.)
Currently my script assumes that you're already logged into Sandy.
http://12rhinos.com/mozilla/ubiquity -
kevin1 replied on September 05, 2008 09:06 to the idea "Sandy quick add firefox extension" in I want Sandy:
The hot project now is Mozilla Ubiquity. It's a command line that runs inside of FireFox and can be extensively customized. To do maps, emails, translations, inline editing of html documents, etc. Even the developer of Devo has abandoned that project in order to work on Ubiquity.
http://labs.mozilla.com/2008/08/intro... -
kevin1 replied on September 05, 2008 04:37 to the question "create ubiquity command to search within a domain" in Mozilla:
Blipo said: "... if you wanted to show a preview you'd need to do a GET and then parse the page."
This code does a preview. Besides, it's not overkill because this is exactly what the built in search commands do. I copied the code from there. (And then added a lot of comments to explain everything to Codexx.) -
kevin1 replied on September 05, 2008 02:53 to the question "create ubiquity command to search within a domain" in Mozilla:
Codexx: I found an example of how to do a jQuery GET and I've modified my function to try and do that. I'm pretty sure my code has some errors, but it's only a first try.
Type this into the address bar to view the Ubiquity source code. then look for "makeSearchCommand", or "jQuery".
chrome://ubiquity/content/builtincmds.js
Blipo: I'm not sure what you mean when you say that I don't have to do a jQuery GET. (For the execute function, not the preview function.) I thought that's exactly what you previously said I needed to do. Why the change of mind?
I understand that the results may need to be parsed for a preview command, and I've written sample code for that. However, I really don't want to do that right now. I would be happy simply opening up the results of the GET in it's own browser window.
(As an html file.)
So, to continuemy above example: I might leave it as-is. (if it works with the particular domain that Codexx wants) or I could change it to do an actual jQuery get, and then display the result somehow.
See below
function doSearchPreview( pblock, obj )
{
// all the code I wrote previously, except the last line.
//Utils.openUrlInBrowser(url);
var params = {}; // create an empty object
params[text_field_name] = searchTerm; // add a name-value pair to the params.
// Eventually we could make the params into a loop,
// and add all the input fields from the form to the params.
// For now, I just want to present the easy case of one input.
// The loop would be something like this:
// Add a name-value pair for every "input" field internal to the search box.
// for (int i=0; i<form.elements.length; i++)
// params[form.elements[i].name] = form.elements[i].value;
// The "get" function of jQuery looks like this:
// jQuery.get( url, params, callback_function, request_type);
jQuery( rootOfSite, params, function(data) {
// I would be happy to simply display the result in a new browser window.
// But I would probably need to change the request_type or "html" or "xml"
var numToDisplay = 3;
var result_set = data.responseData.results.splice( 0, numToDisplay );
pblock.innerHTML = CmdUtils.renderTemplate( {file:"Codexx-website-search.html"},
{results:result_set}
);
}, "json");
};
var new_ubiquity_command2 = {
name: "search-Codexx-website",
takes: {"Term": noun_arb_text},
execute: doSearch,
preview: doSearchPreview
};
CmdUtils.CreateCommand(new_ubiquity_command2};
I'm not sure that the "Codexx-site" even supports a "json" data result. That might need to be changed. (to "xml" or something else.)
And the "Codexx-website-search.html" file needs to be made, so that it will format the data correctly.
| next » « previous |
