Save?

[Joshua Kerievsky]

Anyone here know when the Save metaphor came into existence in the software world?   Saving files, etc?  

It is interesting that on iPads/iPhones, for example, we don't save, we just type and the device remembers what we typed.  

Will "save" eventually go away?

--
best,
jk

--
Joshua Kerievsky
Founder, CEO
Industrial Logic, Inc.
Web: http://industriallogic.com
Twitter: @JoshuaKerievsky, @IndustrialLogic

Amplify Your Agility
Coaching | Training | Assessment | eLearning

[Uberto Barbini]

Very good point.
I never worked on mainframe, but in the appleII era, saving was
something you do inserting a floppy disk or a tape somewhere.
It was really "save my work".
In current app (office suites i.e.) we shuold have only a "title it"
and "export it" function. Saving the work should be automatic.
Actually it were like this in AmiPro (late '90) if I remember well... :)

cheers

Uberto

[Michael Feathers]

A while back, I was wondering whether IDEs could do without it:
http://www.artima.com/weblogs/viewpost.jsp?thread=137562

[George Dinwiddie]

Josh,

On 2/4/11 3:39 AM, Joshua Kerievsky wrote:
> Anyone here know when the Save metaphor came into existence in the
> software world? Saving files, etc?

It was certainly well entrenched in text editors in 1980. I suspect
that it goes back a decade or two before then.

I wonder if it's related to the old batch-screen terminals, which would
let you fill a bunch of fields locally and then send all the fields at
once back to the mainframe. Explicit submits make sense in a
low-bandwidth world.

> It is interesting that on iPads/iPhones, for example, we don't save, we
> just type and the device remembers what we typed.
>
> Will "save" eventually go away?

I think that infinite Undo is necessary when explicit Save is not
available. In many ways, Undo is superior, as it allows us to recover
from mistakes after the fact rather than predict when we're likely to
make them.

I'm not sure Save will go away, as there's often a desire to explicitly
control what is saved. I'm often frustrated by software that has no
explicit save, but therefore doesn't give me access to the means to
version my data.

The related metaphor, Save As, is another interesting one. It's a very
different flavor than Copy.

- George

--
----------------------------------------------------------------------
* George Dinwiddie * http://blog.gdinwiddie.com
Software Development http://www.idiacomputing.com
Consultant and Coach http://www.agilemaryland.org
----------------------------------------------------------------------

[Yves Hanoulle]

Isn't it comming from before?

 

Punch cards had to be saved/printed

and then entered in a computer.

 

Word has already for a long time autosave.

It feels more logic to me. When I write on paper it is autosaved. I have to take an action to delete what I write.

 

yes not being able to save when you want it is bad, loosing something you wanted to keep is worse.

 

 

 

2011/2/4 George Dinwiddie <li...@idiacomputing.com>

--
Yves Hanoulle 

Phone 00 32 46 43 38 32

Skype YvesHanoulle
Blog: www.Hanoulle.be
Agile Games: http://www.AgileGames.org 
Coaching Question Of the Day: http://twitter.com/Retroflection

[George Dinwiddie]

Yves,

On 2/4/11 5:45 AM, Yves Hanoulle wrote:
> yes not being able to save when you want it is bad, loosing something
> you wanted to keep is worse.

Yes, but autosave can cause lost data. I had a problem just the other
day where a program feature didn't operate the way I expected, and two
days worth of work disappeared from view. Luckily it reappeared after
some more fumbling around, but I was quite worried. The operation I had
done did not have an Undo, but had immediately saved the change I didn't
want.

[Yves Hanoulle]

Ahg, that is bad.

 

Rescue:

 

that would a file to save in a dropbox folder

that does autosync to the cloud and has history

 

y

2011/2/4 George Dinwiddie <li...@idiacomputing.com>

[Ward Cunningham]

I would assume explicit "save" dates back to the days of sequential storage. One had to plan out reads and writes because both effected intrinsic state, the position of the medium, that would be adjusted as a side-effect of the operations.

The touring machine abstraction preserves this assumption. The metaphor is deep.

Rotating storage, aka Disk, aka DASD (direct access storage device) open the door to other metaphors but "tape simulation" became the norm and remain today. Even a db query returns a "cursor" that is moved by side effect on read.

The idea that you can have work in progress that is not yet saved can be very reassuring. WyCash captured this notion by saving everything automatically but not "committing" work as a "transaction" until the user was happy with the entry. Non-transacted work could be "tested" in WyCash by exploring what-if reports. This human-facing value of pre-save review was captured in my Checks pattern language. (Aside: the user could unplug the computer at any point in this dialog and find no work-in-progress lost as a result.)

We would do well to separate the mechanical act of saving from the social act of committing. Saving should be made automatic while committing can benefit from more support. 

__________________

Ward Cunningham

503-432-5682

[Yves Hanoulle]

2011/2/4 Ward Cunningham <wa...@c2.com>

I would assume explicit "save" dates back to the days of sequential storage. One had to plan out reads and writes because both effected intrinsic state, the position of the medium, that would be adjusted as a side-effect of the operations.

The touring machine abstraction preserves this assumption. The metaphor is deep.

Rotating storage, aka Disk, aka DASD (direct access storage device) open the door to other metaphors but "tape simulation" became the norm and remain today. Even a db query returns a "cursor" that is moved by side effect on read.

The idea that you can have work in progress that is not yet saved can be very reassuring. WyCash captured this notion by saving everything automatically but not "committing" work as a "transaction" until the user was happy with the entry. Non-transacted work could be "tested" in WyCash by exploring what-if reports. This human-facing value of pre-save review was captured in my Checks pattern language. (Aside: the user could unplug the computer at any point in this dialog and find no work-in-progress lost as a result.)

We would do well to separate the mechanical act of saving from the social act of committing. Saving should be made automatic while committing can benefit from more support. 

 

I like that

what would be a good methaphor for that?

 

__________________

Ward Cunningham

503-432-5682

On Feb 4, 2011, at 12:39 AM, Joshua Kerievsky wrote:

Anyone here know when the Save metaphor came into existence in the software world?   Saving files, etc?  

It is interesting that on iPads/iPhones, for example, we don't save, we just type and the device remembers what we typed.  

Will "save" eventually go away?

--
best,
jk

--
Joshua Kerievsky
Founder, CEO
Industrial Logic, Inc.
Web: http://industriallogic.com
Twitter: @JoshuaKerievsky, @IndustrialLogic

Amplify Your Agility
Coaching | Training | Assessment | eLearning

[toddb]

On Feb 5, 5:35 am, Ward Cunningham <w...@c2.com> wrote:

> We would do well to separate the mechanical act of saving from the social act of committing. Saving should be made automatic while committing can benefit from more support.
>

I like the distinction. Interestingly, if we take committing to be a
social act then perhaps we already have tons of support that we might
be careful about because it has social effects. In separating it from
the mechanical act of saving we often create histories and versions;
this isn't a necessary result of the split but seems to have give rise
to it nonetheless. For example, source control commits create versions
that we can review, track, alter and report on. This information is
public (to those who have access). This information can be used
usefully - I do it myself in coaching when I am interested in the rate
of check in (rate and quantity). But it can also be used to other ends
that I am less clear about being positive - I've seen enough misguided
management initiatives driven from these stats. So it raises questions
about the public and private nature of the information about, in this
case, our checkins. I actually don't want all my check in history
available to everyone/anyone - whether it be the quality of an
increment (particularly lack of) or even the time of day. Ultimately,
I realise that my coding product is public but often I want to keep my
process private. Because of these problems, I am likely therefore to
use the commit support from a DVCS (eg GIT) so I can protect my
personal checkins in a personal fork which hides them in the public
master.

Now, don't get me started on track changes in Word ;-)

--todd

[toddb]

> It is interesting that on iPads/iPhones, for example, we don't save, we just
> type and the device remembers what we typed.
>
> Will "save" eventually go away?
>

This designer is hoping so! http://robmientjes.nl/blog/the-save-metaphor

[Joshua Kerievsky]

On Fri, Feb 4, 2011 at 11:35 AM, Ward Cunningham <wa...@c2.com> wrote:

The idea that you can have work in progress that is not yet saved can be very reassuring.

Yes, it would encourage experimentation without the fear of loss or strict control.  

In general, I think we want our work to be "safe" rather than "saved."  

Telling a computer to "save" is primitive.  Let the computer do the work, so we are never stressed out about losing work.  We ought to pay close attention to anything that elevates our stress, at it may be sign that the wrong metaphor is at work. 

WyCash captured this notion by saving everything automatically but not "committing" work as a "transaction" until the user was happy with the entry. Non-transacted work could be "tested" in WyCash by exploring what-if reports. This human-facing value of pre-save review was captured in my Checks pattern language. (Aside: the user could unplug the computer at any point in this dialog and find no work-in-progress lost as a result.)

That is a nice nuanced example.  For those who haven't ever read Checks, the pattern Ward is referring to is called Deferred Validation (http://c2.com/ppr/checks.html --> Deferred Validation

We would do well to separate the mechanical act of saving from the social act of committing. Saving should be made automatic while committing can benefit from more support. 

Well said Ward.

-- 

[edm...@lia.ufc.br]

I agree with the thesis that Save is a primitive (hardware) instruction
and that Saving, as a bad abstration, produce stress.
To the general user, the visual feedback of seeing what you typed is
enough. The computer (hardware and software) shoud take care of the
rest...

Edmundo Andrade

[Joshua Bloom]

I found this hanging in an office and thought it was an appropriate touch of humor/insight about saving...

[George Dinwiddie]

On 2/9/11 11:56 AM, Joshua Bloom wrote:
> I found this hanging in an office and thought it was an appropriate
> touch of humor/insight about saving...

Yes, I understand the terror. Certainly software should not lump
changes it made automatically (such as updating the Current Date field)
with those the user made explicitly.

It should also not prompt for saving when there are no changes made.
More than once I save unintended changes in Lotus 1-2-3 because it
*always* asked if it should save, and clicking 'Y' had become a routine
habit.

- George

>
>
>
>
> --jabe

>
> On Feb 4, 2011, at 1:39 AM, Joshua Kerievsky wrote:
>
>> Anyone here know when the Save metaphor came into existence in the
>> software world? Saving files, etc?
>>
>> It is interesting that on iPads/iPhones, for example, we don't save,
>> we just type and the device remembers what we typed.
>>
>> Will "save" eventually go away?
>>
>> --
>> best,
>> jk
>>
>> --
>> Joshua Kerievsky
>> Founder, CEO
>> Industrial Logic, Inc.

>> Web: http://industriallogic.com <http://industriallogic.com/>

>> Twitter: @JoshuaKerievsky, @IndustrialLogic
>>
>> Amplify Your Agility
>> Coaching | Training | Assessment | eLearning
>

--