Forms for updating data -- Website experience -- Project #2

2 views
Skip to first unread message

Tariq Nisar Ahmed

unread,
Nov 18, 2006, 2:05:10 AM11/18/06
to prayinjam...@googlegroups.com

As salamu alaykum wa Rahmat Allah wa Barakatuhu,

Ma sha Allah, Basim, your portfolio shows a lot of experience with forms.  In sha Allah, you could help the project by improving how we collect information on new jamat locations, iqamah times, and corrections of either.

We do not have up and running databases yet, but we are collecting more data than is currently displayed on the web site.  Also, we need more flexible forms because the services listed in project #1 are data we are not yet collecting.

Whatever data is collected, we need forms that collect information in a way (1) that can be easily exported to a database(s), (2) that can be easily added to the existing site until we switch over to a database driven code, and (3) that least burdens the contributors.

One idea is to add the "jamat ID" to the data array, and let the info windows have two links -- one for updating jamat information and the other for iqamah information.  Then that infowindow would pass the jamat id to the proper form automatically and it would be submitted with the new data.

Another idea would allow a person to add iqamah information for multiple dates on the same form.

Jazak Allah khair for your help!

As salamu alaykum wa Rahmat Allah wa Barakatuhu,

Tariq Ahmed
prayinj...@gmail.com


Basim Mousilli

unread,
Nov 18, 2006, 4:28:01 AM11/18/06
to prayinjam...@googlegroups.com
Project #2 - I feel like this is where you guys need the most help.
 
Have you considered importing masjids into your website?
I once picked up a single flyer from USGH that lists 50 masijds/musallah in Houston with name, address, and phone. Consider this approach for Houston and other cities around the world.
 
Project #2 Questions:
Where does this form submit to?
What do you do with the data?
Looks like this is manual entry?
Does your host support php/mySQL?
On point (1), how do you plan to export to the database once its ready?
Is there a development/staging environment?
(a) Can you figure out a way to autopopulate map center coordinates into the form textboxes for latitude and longitude?
...or (b) Have them enter the address info and then have the latitude/longitude is generated? That part is most tricky for users...
 
 
Salam,
Basim

Tariq Nisar Ahmed

unread,
Nov 18, 2006, 7:08:51 AM11/18/06
to prayinjam...@googlegroups.com

As salamu alaykum wa Rahmat Allah wa Barakatuhu, 

Project #2 - I feel like this is where you guys need the most help.

 

Have you considered importing masjids into your website?
I once picked up a single flyer from USGH that lists 50 masijds/musallah in Houston with name, address, and phone. Consider this approach for Houston and other cities around the world.

Yes.  :-)  How would you import data from printed sources?  Would you use scanning software?  With the right resources -- a good scanner, good scanning software, and enough time -- that would be an excellent source.  There would still be accuracy issues -- I picked up a similar guide for Chicago while I was there.  The printed address for one of the largest jamat -- a building that holds 3,000 people at one salat during Eid -- was incorrect.  Another problem is geocoding -- discussed later in this message.  Time is still a factor here because none of the printed sources follow the same format/layout.

Some large lists of jamat do exist on-line -- I got all the New Zealand masajid from such a site.  And there is a truly massive MSA-ISNA list that is also really old -- but it would be easier to remove bad data points than to add them slowly by hand.  If I gave you the URL for that list could you retrieve the data from it?  Similarly, could you extract Iqamah times from on-line PDFs?

Another source of imported data is to request data from other databases.  So far I have only made one such request.  But I am working up the nerve to ask the really big databases for help: Islamic Finder, for one.  They can only say no, right?  But they have a ton of jamat.  I want to craft the right appeal so that they will be less inclined to say no, in sha Allah, and more inclined to take us to another level.  Sure, a lot of their data will be old.  But it will still be data.

 

Project #2 Questions:
Where does this form submit to?
What do you do with the data?

A .php file takes the form data and posts it to two files.  In one of the files the form data has been formatted by the php into code that can basically be pasted right into the existing jamat code.  The other file retains all the data that the form collects.  So far the form has worked pretty well, as a stopgap until we switch to a database driven site.  I will send you copies of the two files generated by the php code, as well as a copy of the php code.  The form source is available on-line.

Looks like this is manual entry?

Not sure I follow you here, but I assume you mean the html code that generates the map?  To a certain extent, yes.  But the code generated by the php file greatly simplifies updating the site.  Even when I am doing the data entry myself -- as I did for the New Zealand masajid -- I can add a masjid to the site in less than 5 minutes.  And if someone else did the data entry, adding the jamat or updating the iqamah times takes less than a minute.

If you are talking about the data entry of jamat and iqamah times, then yes.  Right now, the form and the php code are the only automated processes.  And when you are adding a lot of Iqamah data -- like the timetable from the Islamic Center of Bryan-College Station, for example -- you get really bored.  Really fast.  And you stay really bored till you have to switch to something else.  I have done about 4 months of data from that timetable.  In sha Allah, that data will be useful when we start testing the database.

Does your host support php/mySQL?

Alhamdolillah, yes.  Please direct those questions to Abdallah -- he knows much more about that than I do.

On point (1), how do you plan to export to the database once its ready?

There are a few options.  One is bite the bullet and re-enter the data.  It's just 118 jamat at this point.  Yes, a lot of the jamat have multiple Iqamah times.  But there are worse options.

Option 2 is not so bad, though.  The data from the second file -- the one that stores all the data generated by the form -- can be reformatted rather easily into a .csv file.  Then that could presumably be added or "parsed" -- which I have no idea how to do -- by someone with more expertise than I have.  There would still be about 30 jamat that would need to be hand entered.

Option 3 is relatively pain and "parse"-free -- all the data in the current site is stored in one array.  Surely the array could be copied into another set of code and that code could spit out a .csv, right?

Is there a development/staging environment?

Huh?  :-)

(a) Can you figure out a way to autopopulate map center coordinates into the form textboxes for latitude and longitude?

There are a number of ways to use a "geocoder" to determine latitude and longitude.  Relying solely on an automated process is the quickest such means, but yields the poorest quality data.  That is because free geocoders (like the one that Google provides for developers like us) are often inaccurate.  And the inaccuracies often are easily noticed on the map.

Plus, there are a limited number of countries for which free geocoders will work -- America and Australia work.  The United Kingdom does not.

Before the current form, the only 100% reliable means of obtaining coordinates was from handheld GPS devices.  A lot of the early data on the site is from such devices, and they are remarkably accurate.

The current form starts with a Google Map.  The coordinates of the center of the map are always displayed below the map.  Anyone who can center the map on the their jamat can provide accurate coordinates.

...or (b) Have them enter the address info and then have the latitude/longitude is generated? That part is most tricky for users...

If I follow you, see the previous response.

Salam,
Basim

Jazak Allah khair for the quick reply, and wa laykum alas salam wa Rahmat Allah wa Barakatuhu,

Tariq Ahmed
prayinj...@gmail.com


chakir kandil

unread,
Nov 18, 2006, 11:58:57 AM11/18/06
to prayinjam...@googlegroups.com

Salam,

For the first issue I think we can get lists from Islamic finder and similar sites, but for the moment I think we are doing just the ones submitted by individuals.  I’m not sure those lists are up to date. So if we want to do that, it’s better for each metropolitain area, someone should take pf contacting these massajids to make sure they still exist and the info is correct.

Salam

 

 


Basim Mousilli

unread,
Nov 18, 2006, 5:40:29 PM11/18/06
to prayinjam...@googlegroups.com
You can you a scanner with OCR (character recognition) abilitieis. Since the fylers are printed...let's say in Arial or Times New Roman font, it should bbe easy to pick up. Also, the import does not have to be scanned. You can manually type them in. That will give you a leg up on lead time instead of waiting for users to submit data. Or at least....the only dependency on ysers will be for them to submit iqama times....or hell, you folks can even split the lists and call the masjids/musallahs directly to obtain that info. Accuracy issues will always exist. It is challenging to overcome this, but you can institute user culture/discipline to "tame" the data and refine it over time, i.e. like wikipedia.com.
 
It looks like it is a very engaging task to collect iqama times due to the whimsical nature and lack of uniformity among masjids/musallahs. This will be an expensive data point to maintain - is the value worth the effort to maintain this changing data? I'm being the devil's advocate here because I know you are considering an ideally-automated system. Please enlighten me - couldn't we just put in prayer times and then let the users learn the iqama times by going to the masjid?
 
 
On 11/18/06, Tariq Nisar Ahmed <prayinj...@gmail.com> wrote:
Reply all
Reply to author
Forward
0 new messages