austintimeexchange / oscurrency forked from insoshi/insoshi

that's great! pull requests are welcome and you are right that the environments.rb needs to be cleaned up to specify all the required gems. if there is a particular feature you want to contribute, then rich and i would grant commit rights for that.

fyi, i created a new ubuntu slice from scratch last week while i was at lonestar rubyconf to test a clean install. to get rake gems:install to work, i only needed to specify the json and feed normalizer gems:

http://github.com/austintimeexchange/oscurrency/commit/09df7ec148f0ceea32edcc4a7babed2806952dcf

let me know if you find other parts of the install that are ambiguous or need fixing (#oscurrency on freenode) and now that you have commit rights, feel free to directly submit fixes into a new branch.

I'm setting up a friend who wants to get involved on some front end work and I noticed that rake won't run without the oauth gem installed. This includes running rake gems:install. I haven't found where rake is requiring that so I just installed it manually. It would be nice to fix, as these bootstrapping deps can creep into a Rails project and discourage new deployments.

yes, i get:
Could not find RubyGem oauth (>= 0.2.1)

i also installed it manually. vendor/plugins/oauth-plugin/init.rb has the following line:

gem 'oauth', '>=0.2.1'

i see you added it here:

http://github.com/austintimeexchange/oscurrency/commit/529b44648c7c8613cad7f2e289b73425c96a12ac

that should work.

sure, i'll add you as a committer. please create a new branch for these.

that's fine with me.

is this a reproducible error? if so, can you tell me what steps you did to get this?

I clicked on the link you sent with your last email... just reproduced it by going to the timebank home page and clicking Signup.

what browser/version and operating system are you using?

Using Firefox 3.5.3 on Windows XP

thanks. i'll try to reproduce it with that in a bit.

i tried to reproduce with firefox 3.5.3 and XP and couldn't. maybe with a screenshot, i can better guess what might be happening.

although i did notice the check box text isn't displaying right with this configuration. do you get the same where the text is underneath the checkbox instead of next to it?

Yeah, text is underneath the checkbox instead of next to it. I'll try to take a screenshot and sent it to you.

the fix for the checkbox has been checked in. i don't know what's causing zipcode freakiness. since the mouse cursor is not in the zipcode field in the screenshot, the only thing i can think of is maybe some browser add-on or some other software is trying to prepopulate the field. one thing we can do to rule out the possibility that it is a bug in oscurrency is do view source on the web page and see if the zipcode input field has the value attribute set to your email address. if it does not, then it is almost certainly not a bug with oscurrency. if it does, then i would assume there is a bug.

Here's the source, you're right the value is not set -

However, can we set the value to number instead of text? May be that will prevent the browser from populating it with irrelevant info.

"

<label for="zipcode">Zipcode</label>

<input id="person_zipcode" name="person[zipcode]" size="30" type="text" />

"

Sorry, the source code doesn't seem to what to show up eight in the comments :(

Hi Tom, is it possible to set the zipcode field value to number instead of text? It might prevent the funky email address from showing up there.

Zipcode

I'd be fine with it filtering out all HTML and showing it as text, so that people with non-HTML mail readers (mutt, emacs, some cell phones) can read them. There are libraries for many languages which strip out HTML, or one could even have a filter that shells out and runs it through "lynx -dump" :-)

clarification: a deleted offer is removed from activity stream. in this case, the offer existed but number available was empty. need to validate the number available field for the offer.

can you remind me what the pagination problem is (i can't find our previous discussion)? it looks like the proposed change is to sort by first letter instead of alphabetically which seems to be a step backwards. does the current sort by name not work on postgresql for some reason?

let's go ahead and do it even though we lose the alphabetical per first letter group so we can get postgresql working. we can investigate other alternatives later.

this person probably did not specify a proper zipcode or address. the default zipcode for someone who has not specified a location is next to area 51 which is probably why you're getting the funky distance. :)

Could the distance code use neighborhood instead, since the neighborhoods are mapped to zips? Or, as a quick hack for now, could it simply not report distance at all if the location is the "magic number" of Area 51? The aliens from Planet Zod would be thankful.

if someone doesn't choose a neighborhood, it will be just like if they didn't choose a zipcode, so there is no advantage gained by using neighborhood. i like the second option better, that if their zipcode is left at 89001, then the distance will not be reported. (people who live in the 89001 zipcode will need to specify address information for the distance calculation to be computed) alternatively, we can make the zipcode empty by default, which might be better.

I agree; zero might make a better default than 89001. And if the aliens (or active-duty Air Force personnel stationed at Area 51) ever want to join our timebank, then treating 89001 as a magic number might get in the way. :-) Have a great holiday!

Hmm, howzabout:

diff --git a/app/views/reqs/_req.html.erb b/app/views/reqs/_req.html.erb
index c83e04a..afd1769 100644
--- a/app/views/reqs/_req.html.erb
+++ b/app/views/reqs/_req.html.erb
@@ -26,7 +26,7 @@
     | <%= link_to "Delete", req, :confirm => 'Are you sure?', :method => :delete unless (req.has_commitment? || req.has_approved? ) %>
     <% end -%>
     

-    <% if !current_person.address.nil? && !req.person.address.nil? && !current_person?(req.person) -%>
+    <% if !current_person.address.nil? && current_person.address.zipcode_plus_4  != 89001 && !req.person.address.nil? && req.person.address.zipcode_plus_4  != 89001  && !current_person?(req.person) -%>
       <%= number_with_precision(req.person.address.distance_to(current_person.address), 1).to_s %> miles away
     <% end -%>