Ward Photo Directory Template

So you have the BIG idea that the Church or community needs to develop. Discuss that idea here. Maybe you just want to make a suggestion on a new forum topic. Let us know.
russellhltn
Community Administrator
Posts: 34416
Joined: Sat Jan 20, 2007 2:53 pm
Location: U.S.

#21

Post by russellhltn »

pwhiting wrote:I'm interested in comments on the following:
Comments from a general kabitzer, not a stake holder:

4. I'm not sure as there is any way to distinguish this now, but I think it would make thinks a lot simpler for the users if the first child started at Name3. So single parent households would have a blank Name2.

It's probably outside of the scope, but I think some people would like to see HT/VT export so as to support PH/RS functions/applications. I think more leaders have PDAs then Planners now.
AArnott
New Member
Posts: 15
Joined: Sun Feb 04, 2007 4:25 pm

Free photo directory PDF generator

#22

Post by AArnott »

I was recently in a BYU student ward. While I was there serving as executive secretary, I wrote a simple wizard-like program to generate the ward photo directory. You just copy and paste from MLS, point it at the directory of photos, and away it goes!

You can download it and read more about it at my blog: http://cs.nerdbank.net/blogs/darnit/arc ... nutes.aspx
bartbarker
New Member
Posts: 15
Joined: Fri Jan 26, 2007 7:40 pm

#23

Post by bartbarker »

pwhiting wrote:I'm interested in comments on the following:

1. mark the last name with / / when it appears in the name* column. The last name is included when it is different than the head of household last name - should it just be included all the time?

It think it's important to be consistent.
2. each row is variable length, depending upon the number of children. We only list name1 and name2 in the column headers. We could add more column headers (name3, name4, ...) or we could go through the list once, figure out the maximum number of family members for this csv, and then create that number of column headers. I realize the second approach is more robust, but I'm trying to keep the code supporting this as simple as possible, and doing two passes through the ward list, one to figure out the maximum family size and one to print everything out seemed like more effort than it was worth.

I vote for more robust. Let's make the feature work as well as it can.
3. split the name and email into two distinct columns. Currently they are combined, it is trivial to split them. The column headers would be name1, email1, name2, email2,

This is the most important change. I strongly suggest you do it.
4. distinguish between parents and children - currently name2 contains information about the first child in single parent households.

I vote for this, too. Name2 would be blank for a single head of household. Once again, this provides consistency in output and makes the file more useful. Also, each column should have a heading: name3, name4, etc.
russellhltn
Community Administrator
Posts: 34416
Joined: Sat Jan 20, 2007 2:53 pm
Location: U.S.

#24

Post by russellhltn »

pwhiting wrote:2. each row is variable length, depending upon the number of children. We only list name1 and name2 in the column headers. We could add more column headers (name3, name4, ...) or we could go through the list once, figure out the maximum number of family members for this csv, and then create that number of column headers. I realize the second approach is more robust, but I'm trying to keep the code supporting this as simple as possible, and doing two passes through the ward list, one to figure out the maximum family size and one to print everything out seemed like more effort than it was worth.
Is there a limit in MLS at to how many children there can be? I'm wondering about the idea of just going with a fixed number (say 20) children. That way CSV users can plan ahead of time how many columns to allot for children. It also allows you in the future to add columns after the children for backward compatibility.
User avatar
thedqs
Community Moderators
Posts: 1042
Joined: Wed Jan 24, 2007 8:53 am
Location: Redmond, WA
Contact:

#25

Post by thedqs »

I believe this was suppose to dynamic to each ward. So if Brother Jones had the most kids of 12 there would be 12 columns for kids for everyone. So Any parsing program would read the column heading and import appropriately.
- David
User avatar
epoulin-p40
New Member
Posts: 22
Joined: Tue Jan 30, 2007 8:28 am
Location: Bowmanville, Ontario, Canada
Contact:

No variable length rows

#26

Post by epoulin-p40 »

pwhiting wrote: 2. each row is variable length, depending upon the number of children. We only list name1 and name2 in the column headers. We could add more column headers (name3, name4, ...) or we could go through the list once, figure out the maximum number of family members for this csv, and then create that number of column headers. I realize the second approach is more robust, but I'm trying to keep the code supporting this as simple as possible, and doing two passes through the ward list, one to figure out the maximum family size and one to print everything out seemed like more effort than it was worth.
Its generally bad practice to create output of variable length. Its much better to write code to a standard output length. Rather can you put the variable length fields (number of children) into 1 column as a delimited string. For example, you put the children into name2 as "name 1; name 2; name3" and if you want to add email addresses you further delimit such as "name 1|email 1; name 2|email 2; name 3|email 3"
Eric Poulin
Bowmanville, Ontario, Canada
bartbarker
New Member
Posts: 15
Joined: Fri Jan 26, 2007 7:40 pm

#27

Post by bartbarker »

epoulin wrote:
pwhiting wrote:For example, you put the children into name2 as "name 1; name 2; name3" and if you want to add email addresses you further delimit such as "name 1|email 1; name 2|email 2; name 3|email 3"
I still think email addresses should be in their own fields.
User avatar
epoulin-p40
New Member
Posts: 22
Joined: Tue Jan 30, 2007 8:28 am
Location: Bowmanville, Ontario, Canada
Contact:

#28

Post by epoulin-p40 »

bbarker wrote:
epoulin wrote: I still think email addresses should be in their own fields.
I agree with that. Email should be its own field, but variable row lengths are bad. In that case, you can delimit as name1="name1;name2;name3" name2="email1;email2;email3"
Eric Poulin
Bowmanville, Ontario, Canada
User avatar
thedqs
Community Moderators
Posts: 1042
Joined: Wed Jan 24, 2007 8:53 am
Location: Redmond, WA
Contact:

#29

Post by thedqs »

You mean variable field count, since variable field length means that each field uses a delimiter to seperate the columns, fixed field length is where if the name is too short (Doe, John for example) you would fill the rest of the name field with 28 spaces to reach the proper length.

Though it does force you to program more robustly (You have to read in the heading line and figure out how many columns there are and what they represent) it is a lot nicer then having to always have space for 20 or so children for the largest family in the church. And since the CVS can change in the future, this also makes it so that you don't have to update your input scheme later on.
- David
User avatar
epoulin-p40
New Member
Posts: 22
Joined: Tue Jan 30, 2007 8:28 am
Location: Bowmanville, Ontario, Canada
Contact:

Semantics of CSVs

#30

Post by epoulin-p40 »

thedqs wrote:You mean variable field count, since variable field length means that each field uses a delimiter to seperate the columns, fixed field length is where if the name is too short (Doe, John for example) you would fill the rest of the name field with 28 spaces to reach the proper length.

Though it does force you to program more robustly (You have to read in the heading line and figure out how many columns there are and what they represent) it is a lot nicer then having to always have space for 20 or so children for the largest family in the church. And since the CVS can change in the future, this also makes it so that you don't have to update your input scheme later on.
Right I was referring to variable field counts. CSV doesn't have a column length issues, the column can be as long as needed, just waiting for the next comma. In general its better to have a fixed count of columns, because if the code consuming the CSV has to start parsing the column names, its not really possible for code to understand the semantics of column names. However, code can understand semantics of delimited field, which is not parsing column names. Of course if there is a published spec, either way is achievable, but depending on code to parse column header names just feels dangerous.
Eric Poulin
Bowmanville, Ontario, Canada
Locked

Return to “Ideas & Suggestions”