Question about entering ridiculous dates.

classic Classic list List threaded Threaded
11 messages Options
Reply | Threaded
Open this post in threaded view
|

Question about entering ridiculous dates.

Peter Landgren
Hi list,

The other day I entered a date, but missed the first digit, meaning the person
became born 1000 years ealier. The parents had complete correct dates. Why is
there no check for obviously wrong dates? I know I can check this with
Tools->Verify Data. [I wrote my own genealogy program for DOS (in FORTH) 15
years ago and had very strict checking of all dates during date entering.
This prevented many errors.]

/Peter
--
Peter Landgren
Talken Hagen
671  94 Brunskog
SWEDEN

0570-530 21
070-635 4719
[hidden email]
skype:pgl4820

-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >> http://get.splunk.com/
_______________________________________________
Gramps-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-devel
Reply | Threaded
Open this post in threaded view
|

Re: Question about entering ridiculous dates.

Douglas S. Blank
On Sat, October 20, 2007 3:22 pm, Peter Landgren wrote:

> Hi list,
>
> The other day I entered a date, but missed the first digit, meaning the
> person
> became born 1000 years ealier. The parents had complete correct dates. Why
> is
> there no check for obviously wrong dates? I know I can check this with
> Tools->Verify Data. [I wrote my own genealogy program for DOS (in FORTH)
> 15
> years ago and had very strict checking of all dates during date entering.
> This prevented many errors.]

+1 on this idea! Also, it could use "probably alive" and related
information to give (at least) warnings:

"This person's date of birth is more than 20 years away from that of a
sibling. Are you sure this is correct?"

"This person's date of death is more than 110 years away from one of their
other events. Are you sure that this is correct?"

There is a tool that checks some of these values, but it would be great to
have them in the data entry interface. BTW, some of these numeric range
values I've begun to work them into user definable config keys. See
src/Utils.py:

    _MAX_AGE_PROB_ALIVE   = 110
    _MAX_SIB_AGE_DIFF     = 20
    _MIN_GENERATION_YEARS = 13
    _AVG_GENERATION_GAP   = 20

-Doug

> /Peter
> --
> Peter Landgren
> Talken Hagen
> 671  94 Brunskog
> SWEDEN
>
> 0570-530 21
> 070-635 4719
> [hidden email]
> skype:pgl4820
>
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Splunk Inc.
> Still grepping through log files to find problems?  Stop.
> Now Search log events and configuration files using AJAX and a browser.
> Download your FREE copy of Splunk now >> http://get.splunk.com/
> _______________________________________________
> Gramps-devel mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/gramps-devel
>


--
Douglas S. Blank
Associate Professor, Bryn Mawr College
http://cs.brynmawr.edu/~dblank/
Office: 610 526 6501

-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >> http://get.splunk.com/
_______________________________________________
Gramps-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-devel
Reply | Threaded
Open this post in threaded view
|

Re: Question about entering ridiculous dates.

Randolph Bentson
On Sun, Oct 21, 2007 at 05:31:54AM -0400, Douglas S. Blank wrote:
>
> +1 on this idea! Also, it could use "probably alive" and related
> information to give (at least) warnings:
>
> "This person's date of birth is more than 20 years away from that
> of a sibling. Are you sure this is correct?"

You may have intended this, but let's be explicit: "This person's date
of birth is more than 20 years away from that of the sibling closest
in age. Are you sure this is correct?"

--
Randolph Bentson
[hidden email]

-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >> http://get.splunk.com/
_______________________________________________
Gramps-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-devel
Reply | Threaded
Open this post in threaded view
|

Re: Question about entering ridiculous dates.

Don Allingham
My college roommate would have violated this rule. His nearest brother
was 25 years older.


On Thu, 2007-10-25 at 19:23 -0700, Randolph Bentson wrote:

> On Sun, Oct 21, 2007 at 05:31:54AM -0400, Douglas S. Blank wrote:
> >
> > +1 on this idea! Also, it could use "probably alive" and related
> > information to give (at least) warnings:
> >
> > "This person's date of birth is more than 20 years away from that
> > of a sibling. Are you sure this is correct?"
>
> You may have intended this, but let's be explicit: "This person's date
> of birth is more than 20 years away from that of the sibling closest
> in age. Are you sure this is correct?"
>


-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >> http://get.splunk.com/
_______________________________________________
Gramps-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-devel
Reply | Threaded
Open this post in threaded view
|

Re: Question about entering ridiculous dates.

Douglas S. Blank
In reply to this post by Randolph Bentson
On Thu, October 25, 2007 10:23 pm, Randolph Bentson wrote:

> On Sun, Oct 21, 2007 at 05:31:54AM -0400, Douglas S. Blank wrote:
>>
>> +1 on this idea! Also, it could use "probably alive" and related
>> information to give (at least) warnings:
>>
>> "This person's date of birth is more than 20 years away from that
>> of a sibling. Are you sure this is correct?"
>
> You may have intended this, but let's be explicit: "This person's date
> of birth is more than 20 years away from that of the sibling closest
> in age. Are you sure this is correct?"

Actually, the way that the code is written in src/Utils.py it has the more
general meaning of being more than max-sib-age-diff away from any sibling.
I don't know if we need to make that fine-grained distinction of your
interpretation, but the code currently doesn't. Of course, it also isn't
used for such a warning on data entry either.

I hope we add something like these to GRAMPS 3.0. It would catch a lot of
(my) errors.

-Doug

> --
> Randolph Bentson
> [hidden email]
>
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Splunk Inc.
> Still grepping through log files to find problems?  Stop.
> Now Search log events and configuration files using AJAX and a browser.
> Download your FREE copy of Splunk now >> http://get.splunk.com/
> _______________________________________________
> Gramps-devel mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/gramps-devel
>


--
Douglas S. Blank
Associate Professor, Bryn Mawr College
http://cs.brynmawr.edu/~dblank/
Office: 610 526 6501

-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >> http://get.splunk.com/
_______________________________________________
Gramps-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-devel
Reply | Threaded
Open this post in threaded view
|

Re: Question about entering ridiculous dates.

Douglas S. Blank
In reply to this post by Don Allingham
On Thu, October 25, 2007 10:36 pm, Don Allingham wrote:
> My college roommate would have violated this rule. His nearest brother
> was 25 years older.

Two points: this would just be a warning, and it is also user-definable in
the config keys. I don't mind warnings when we get out near the end of the
tail of a distributions. A bunch of people ran some stats on their
databases, and there were just a few people with sib-age-diffs greater
than 20. However, depending on the century, it was more common.

-Doug

>
> On Thu, 2007-10-25 at 19:23 -0700, Randolph Bentson wrote:
>> On Sun, Oct 21, 2007 at 05:31:54AM -0400, Douglas S. Blank wrote:
>> >
>> > +1 on this idea! Also, it could use "probably alive" and related
>> > information to give (at least) warnings:
>> >
>> > "This person's date of birth is more than 20 years away from that
>> > of a sibling. Are you sure this is correct?"
>>
>> You may have intended this, but let's be explicit: "This person's date
>> of birth is more than 20 years away from that of the sibling closest
>> in age. Are you sure this is correct?"
>>
>


--
Douglas S. Blank
Associate Professor, Bryn Mawr College
http://cs.brynmawr.edu/~dblank/
Office: 610 526 6501

-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >> http://get.splunk.com/
_______________________________________________
Gramps-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-devel
Reply | Threaded
Open this post in threaded view
|

Re: Question about entering ridiculous dates.

Benny Malengier


2007/10/26, Douglas S. Blank <[hidden email]>:
On Thu, October 25, 2007 10:36 pm, Don Allingham wrote:
> My college roommate would have violated this rule. His nearest brother
> was 25 years older.

Two points: this would just be a warning, and it is also user-definable in
the config keys. I don't mind warnings when we get out near the end of the
tail of a distributions. A bunch of people ran some stats on their
databases, and there were just a few people with sib-age-diffs greater
than 20. However, depending on the century, it was more common.

-Doug

If it is a warning, please group them. On save we now have the warning on new person: no gender given. Cancel/Ok
You could make a general thing that checks can be added to on save. If you do that on EditPrimary, it would be available everywhere.

Note however that the philosophy of gramps has always been: people can enter whatever they like. For fields we can parse while typing we use the highlighted box (eg date, longitude), but still allow save.

Know also that many people hate 'intelligent' machines. We don't want clippy popping up with an annoying finger in the air every time.

Benny


-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >> http://get.splunk.com/
_______________________________________________
Gramps-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-devel
Reply | Threaded
Open this post in threaded view
|

Re: Question about entering ridiculous dates.

Douglas S. Blank
On Fri, October 26, 2007 3:53 am, Benny Malengier wrote:

> 2007/10/26, Douglas S. Blank <[hidden email]>:
>>
>> On Thu, October 25, 2007 10:36 pm, Don Allingham wrote:
>> > My college roommate would have violated this rule. His nearest brother
>> > was 25 years older.
>>
>> Two points: this would just be a warning, and it is also user-definable
>> in
>> the config keys. I don't mind warnings when we get out near the end of
>> the
>> tail of a distributions. A bunch of people ran some stats on their
>> databases, and there were just a few people with sib-age-diffs greater
>> than 20. However, depending on the century, it was more common.
>>
>> -Doug
>>
>
> If it is a warning, please group them. On save we now have the warning on
> new person: no gender given. Cancel/Ok
> You could make a general thing that checks can be added to on save. If you
> do that on EditPrimary, it would be available everywhere.
>
> Note however that the philosophy of gramps has always been: people can
> enter
> whatever they like. For fields we can parse while typing we use the
> highlighted box (eg date, longitude), but still allow save.
>
> Know also that many people hate 'intelligent' machines. We don't want
> clippy
> popping up with an annoying finger in the air every time.

You are thinking about this being far more intrusive than I was. I was
thinking that if the date turned say, yellow, that you could mouse-over or
you could press an Info icon (or maybe help) and it would tell you all the
reasons why that item is in a warning state.

But if you want something more Clippy like, we could make Grampy. He
wouldn't be very happy if you enter incorrect dates. :)

-Doug

>
> Benny


-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >> http://get.splunk.com/
_______________________________________________
Gramps-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-devel
Reply | Threaded
Open this post in threaded view
|

Re: Question about entering ridiculous dates.

Benny Malengier
A color indicator has following problems:
-/you cannot turn it of for a single instance. So you have a strange date and you know it is correct but it turns yellow. Would be annoying to always see the yellow anyway.
-/color blind people

A warning message on the other hand can be made to only be displayed if there is a change with what is already in the database. So you only get an indication when first entering it, if wrong, you correct it, if ok, you continue. Then you don't get the indication anymore.

Benny

2007/10/26, Douglas S. Blank <[hidden email]>:
On Fri, October 26, 2007 3:53 am, Benny Malengier wrote:

> 2007/10/26, Douglas S. Blank <[hidden email]>:
>>
>> On Thu, October 25, 2007 10:36 pm, Don Allingham wrote:
>> > My college roommate would have violated this rule. His nearest brother
>> > was 25 years older.
>>
>> Two points: this would just be a warning, and it is also user-definable
>> in
>> the config keys. I don't mind warnings when we get out near the end of
>> the
>> tail of a distributions. A bunch of people ran some stats on their
>> databases, and there were just a few people with sib-age-diffs greater
>> than 20. However, depending on the century, it was more common.
>>
>> -Doug
>>
>
> If it is a warning, please group them. On save we now have the warning on
> new person: no gender given. Cancel/Ok
> You could make a general thing that checks can be added to on save. If you
> do that on EditPrimary, it would be available everywhere.
>
> Note however that the philosophy of gramps has always been: people can
> enter
> whatever they like. For fields we can parse while typing we use the
> highlighted box (eg date, longitude), but still allow save.
>
> Know also that many people hate 'intelligent' machines. We don't want
> clippy
> popping up with an annoying finger in the air every time.

You are thinking about this being far more intrusive than I was. I was
thinking that if the date turned say, yellow, that you could mouse-over or
you could press an Info icon (or maybe help) and it would tell you all the
reasons why that item is in a warning state.

But if you want something more Clippy like, we could make Grampy. He
wouldn't be very happy if you enter incorrect dates. :)

-Doug

>
> Benny



-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >> http://get.splunk.com/
_______________________________________________
Gramps-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-devel
Reply | Threaded
Open this post in threaded view
|

Re: Question about entering ridiculous dates.

Douglas S. Blank
On Fri, October 26, 2007 6:45 am, Benny Malengier wrote:
> A color indicator has following problems:

We already use color as an indicator in gramps when you have entered a bad
date. Now you're suggesting a different method which you wouldn't see
until you try to save.

> -/you cannot turn it of for a single instance.

Of course you can; we already do. A bad date is red during data entry, but
not afterwards.

> -/color blind people

Many color blind people just can't distinguish between colors that others
can.

> A warning message on the other hand can be made to only be displayed if
> there is a change with what is already in the database. So you only get an
> indication when first entering it, if wrong, you correct it, if ok, you
> continue. Then you don't get the indication anymore.

We could give a warning too if you wish. I'm just trying to avoid Grampy. :)

-Doug


> Benny
>
> 2007/10/26, Douglas S. Blank <[hidden email]>:
>>
>> On Fri, October 26, 2007 3:53 am, Benny Malengier wrote:
>> > 2007/10/26, Douglas S. Blank <[hidden email]>:
>> >>
>> >> On Thu, October 25, 2007 10:36 pm, Don Allingham wrote:
>> >> > My college roommate would have violated this rule. His nearest
>> brother
>> >> > was 25 years older.
>> >>
>> >> Two points: this would just be a warning, and it is also
>> user-definable
>> >> in
>> >> the config keys. I don't mind warnings when we get out near the end
>> of
>> >> the
>> >> tail of a distributions. A bunch of people ran some stats on their
>> >> databases, and there were just a few people with sib-age-diffs
>> greater
>> >> than 20. However, depending on the century, it was more common.
>> >>
>> >> -Doug
>> >>
>> >
>> > If it is a warning, please group them. On save we now have the warning
>> on
>> > new person: no gender given. Cancel/Ok
>> > You could make a general thing that checks can be added to on save. If
>> you
>> > do that on EditPrimary, it would be available everywhere.
>> >
>> > Note however that the philosophy of gramps has always been: people can
>> > enter
>> > whatever they like. For fields we can parse while typing we use the
>> > highlighted box (eg date, longitude), but still allow save.
>> >
>> > Know also that many people hate 'intelligent' machines. We don't want
>> > clippy
>> > popping up with an annoying finger in the air every time.
>>
>> You are thinking about this being far more intrusive than I was. I was
>> thinking that if the date turned say, yellow, that you could mouse-over
>> or
>> you could press an Info icon (or maybe help) and it would tell you all
>> the
>> reasons why that item is in a warning state.
>>
>> But if you want something more Clippy like, we could make Grampy. He
>> wouldn't be very happy if you enter incorrect dates. :)
>>
>> -Doug
>>
>> >
>> > Benny
>>
>>
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Splunk Inc.
> Still grepping through log files to find problems?  Stop.
> Now Search log events and configuration files using AJAX and a browser.
> Download your FREE copy of Splunk now >>
> http://get.splunk.com/_______________________________________________
> Gramps-devel mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/gramps-devel
>


--
Douglas S. Blank
Associate Professor, Bryn Mawr College
http://cs.brynmawr.edu/~dblank/
Office: 610 526 6501

-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >> http://get.splunk.com/
_______________________________________________
Gramps-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-devel
Reply | Threaded
Open this post in threaded view
|

Re: Question about entering ridiculous dates.

Benny Malengier


2007/10/26, Douglas S. Blank <[hidden email]>:
On Fri, October 26, 2007 6:45 am, Benny Malengier wrote:
> A color indicator has following problems:

We already use color as an indicator in gramps when you have entered a bad
date. Now you're suggesting a different method which you wouldn't see
until you try to save.

We give color warning AND an icon.

> -/you cannot turn it of for a single instance.

Of course you can; we already do. A bad date is red during data entry, but
not afterwards.

Hmm, I always considered that a bug. I even thought of looking at the code to 'correct' this. Apparently you see this as a feature :-)
I would like to see it red when opening the editor, as a bad date (not interpretable by the software) can cause all sorts of problems in genealogy. Unreadable dates on documents are better give the estimated tag in my opinion with something you can work with, and a note perhaps with what you see on the document.

> -/color blind people

Many color blind people just can't distinguish between colors that others
can.

Which does not mean we have to leave them in the cold. Of course, yellow and an icon is an option.

> A warning message on the other hand can be made to only be displayed if
> there is a change with what is already in the database. So you only get an
> indication when first entering it, if wrong, you correct it, if ok, you
> continue. Then you don't get the indication anymore.

We could give a warning too if you wish. I'm just trying to avoid Grampy. :)

-Doug



-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >> http://get.splunk.com/
_______________________________________________
Gramps-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gramps-devel