Hi,

On Tue, 2004-04-06 at 10:25, Roman Joost wrote: [snip good initiative]

9. Add an image grid and change the distance of the grid points. a) Does the user find the image grid function? b) Does the user find the image grid preferences in the preferences dialog?
c) Does the user find the correct option for changing grid preferences?

Finding the Default Image Grid preferences in the Preferences dialog is not needed for changing the distance of grid points in an image. This can be achieved using the Configure Image Grid dialog found in /Image/Configure Grid...

The Default Image Grid preferences is only used to create a default grid for new images and for images that doesn't have a grid (non-XCF files).

Sincerely, Brix

Hi Roman,

Roman Joost wrote:

Well, i think first a bit to the background of upcoming Usability Tests. I met Ellen Reitmayr at the GIMP 2.0 Release Party in Berlin and she is involved in some Usability studies for the KDE Project (with the whole company relevantive AG). After some discussions about the KDE Usability tests, we agreed, that some Usability tests for the GIMP would be nice for having feedback to the developers from some usability experts.

This is excellent news, and an excellent initiative.

I've added usability@gnome.org to the CC list for their feedback too, and left the entire mail intact.

Thanks for doing this, Roman, and I'm looking forward to seeing the results.

Cheers, Dave.

I met Ellen again and we put some tasks for people who are new to the GIMP and poeple who consider to change from photoshp to the GIMP. Ellen wrote the draft that you'll find below. It's considered as a basis for comments and suggestions to improve the usability of the GIMP interface. It should not test the functionality itself and issues already known as bugs. The test should be as effective as possible with the 4 users and the mostly common tasks.

The first test consists of 4 persons and deals only the 'all-day-use' with the Gimp (what is the handling of the application like, how good can one achieve a task etc.). Later, we want to conduct a more extensive test that will also consider the users' understanding of the common concepts, such as layers, paths etc. - and how the new user could be helped to learn these basic concepts (that would be nice to know for the gimp-help-2 module to get the manual as effective as possible for helping the user).

The GIMP is the SuSE package of the GIMP 2.0 with the german language enabled.

Tasks for the first test (all-day-usage; all of the are common tasks for all people, except the one where the indicated group is mentioned):

1. Create a rectangular selection and subtract the edge of the created selection.
a) Were the necessary modes found?
b) Does the user select the modi in the right sequence to achieve the goal. ('Replace a current selection', then use 'Subtract from the current selection')
c) Does the user know, that the starting point of the subtract selection has to be placed outside of the current selection and not inside the selection itself.

2. Create a selection with fixed width and height and a fixed aspect ratio.
a) Does the user find the right input field. b) Does the user understand 'Fixed Aspect Ratio' or the german label: 'festes Seitenverhaeltnis'.

3. Selection by color, fill an area with a pattern and cancel the selection.
a) Does the user find the right tool in the toolbox or in the menu? b) Does the user know how to fill the area with the patterns? c) How to cancel a currently activated selection?

4. Measure how long a distance between some image objects is. a) Does the user find the right tool in the toolbox or in the menu? b) Does the user find the value directly? c) Where does the user look for the value? At the statusbar or at the info window?

5. Clone an image object with the clone tool. a) Does the user find the right tool in the toolbox or in the menu? b) Does the user understand how to indicate the starting point for the clone tool?
c) Does the user choose the right brush from the brush browser?

6. Transform image objects. The photoshop user should create a layer and shear it with a value for XX pixels.
The new user should only shear the whole image. a) Does the user find the right tool in the toolbox or in the menu? b) Is the user able to use the grid for shearing?

7. Use the color picker to get the color info with sample merged enabled.
a) Does the user find the right tool in the toolbox or in the menu? b) Does the user know why the color which is shown in the image window is not the same which the color picker gets from the layer? c) Does he know how to find the checkbox to enable the 'Sample Merged' function?

8. Using the undo history. a) Does the user find the journal directly? b) Does the user understand, how to undo single steps?

9. Add an image grid and change the distance of the grid points. a) Does the user find the image grid function? b) Does the user find the image grid preferences in the preferences dialog?
c) Does the user find the correct option for changing grid preferences?

10. This task is only for photoshop professionals. They should use the path tool to create a cloud and stroke it with some patterns. a) Does the user find the path tool? b) Does he know how to use the path tool? c) Does he know how to stroke the created path with a pattern?

11. Try to dock some tabs into the docks. a) Does the user understand that only tabs can be moved betweend docks and not the docks itself.
b) Does the user find the active area, where he can drop the tab to the dock.

Thats all of them. In hope it would be useful,

Greetings

Roman Joost (roman@bromeco.de) wrote:

Well, i think first a bit to the background of upcoming Usability Tests. I met Ellen Reitmayr at the GIMP 2.0 Release Party in Berlin and she is involved in some Usability studies for the KDE Project (with the whole company relevantive AG). After some discussions about the KDE Usability tests, we agreed, that some Usability tests for the GIMP would be nice for having feedback to the developers from some usability experts.

Great.

1. Create a rectangular selection and subtract the edge of the created selection.
a) Were the necessary modes found?
b) Does the user select the modi in the right sequence to achieve the goal. ('Replace a current selection', then use 'Subtract from the current selection')
c) Does the user know, that the starting point of the subtract selection has to be placed outside of the current selection and not inside the selection itself.

c) is not necessary. When pressing CTRL you can happily subtract stuff starting inside the current selection (otherwise it would be impossible to create holes in the selection)

Bye, Simon

Hi Simon (and Brix),

Simon Budig wrote:

Roman Joost (roman@bromeco.de) wrote:

1. Create a rectangular selection and subtract the edge of the created selection.
c) Does the user know, that the starting point of the subtract selection has to be placed outside of the current selection and not inside the selection itself.

c) is not necessary. When pressing CTRL you can happily subtract stuff starting inside the current selection (otherwise it would be impossible to create holes in the selection)

I know we're correcting the usability expectations with these comments, but somehow they have already highlighted some usability concerns - if the people doing the usability tests, in consultation with someone pretty familiar with the GIMP, isn't aware of these things, it might indicate that this is non-obvious (for example, the separation of the default grid and the grid in the current image is one which is not obvious, similarly, selection handling with modifier keys is not obvious).

If I understood correctly, the objective of this exercise is to see if people find the tool options, understand the icons, and understand what the changing cursor means.

Cheers,
Dave.

On Tue, 6 Apr 2004 10:25:07 +0200, Roman Joost wrote:

Well, i think first a bit to the background of upcoming Usability Tests. I met Ellen Reitmayr at the GIMP 2.0 Release Party in Berlin and she is involved in some Usability studies for the KDE Project (with the whole company relevantive AG). After some discussions about the KDE Usability tests, we agreed, that some Usability tests for the GIMP would be nice for having feedback to the developers from some usability experts.

As others have already said, this is an excellent initiative. Thanks!

Tasks for the first test (all-day-usage; all of the are common tasks for all people, except the one where the indicated group is mentioned):

[...]

The tests look good. There is another test that I have tried in the past with new users. The optional parts may be a bit more complex for new users, but the basic part could be useful for your test or for tuture tests:

Draw a red circle around some elements of a given image (objects or people's faces).

Optional parts: - ask for an oval or for a perfect circle, - thick circle or thin circle (1 pixel), - fully opaque or partially transparent red circle.

Things to observe: - Does the user find the right tool(s) to do the job? (selections or paths)?
- If not, then is a basic explanation sufficient to understand the concept? Or does the user need more guidance? (Basic explanation: "First select the shape that you want, then use Edit->Stroke Selection.")
- For perfect circles, does the user find the right modifier? - For partially transparent circles, how does the user do it? (using a separate layer, using brush transparency, etc.)

The option about 1-pixel thin circles was only relevant for gimp-1.2 and may not be useful for 2.0, thanks to the new stroke dialog.

-Raphaël

I just wrote:

The tests look good. There is another test that I have tried in the past with new users. The optional parts may be a bit more complex for new users, but the basic part could be useful for your test or for tuture tests:

s/tuture/future/ Maybe I was thinking of "torture" at the same time?

In the "things to observe", I could also add: - Does the user draw in the right color on the first attempt, or do they first get a black circle and then redo it in red?

-Raphaël

hi everybody,

first I quickly want to introduce myself: I am ellen, the one Roman met at the GIMP 2.0 release party. I am working for a small usability company in Berlin, and one of our major concern is to improve the usability of open source software. Well, so Roman and I decided to do some usability tests on the GIMP.

Raphael suggested to add another task to the usability test:

Draw a red circle around some elements of a given image (objects or people's faces).

Optional parts: - ask for an oval or for a perfect circle, - thick circle or thin circle (1 pixel), - fully opaque or partially transparent red circle.

I think it's a good idea, thank you 8-)

I would like to start doing the tests as soon as possible, so if anybody has some more suggestions please let me know!

Have a nice day /ellen

On 6 Apr 2004, at 10:25, Roman Joost wrote:

Well, i think first a bit to the background of upcoming Usability Tests. I met Ellen Reitmayr at the GIMP 2.0 Release Party in Berlin and she is involved in some Usability studies for the KDE Project (with the whole company relevantive AG). After some discussions about the KDE Usability tests, we agreed, that some Usability tests for the GIMP would be nice for having feedback to the developers from some usability experts.

Since you're planning multiple tests, you may consider running a first test that is not very task driven, and in which the (few) tasks are relatively broad. The results from that test could then serve as the basis for the next test(s).

What kind of users will you be working with? How familiar are they with the testing platform? How familiar with photography? How familiar with digital imaging?

One of the Cinepaint developers once had a user write down his suggestions for improvement. These could perhaps also serve as a starting point for an exploration of the sort of problems a GIMP user might run into. I don't seem to be able to find a link to that discussion, though.

V St 07. 04. 2004 v 11:44 +0200 píše Ellen Reitmayr:

I would like to start doing the tests as soon as possible, so if anybody has some more suggestions please let me know!

One thing I'd like to see some usability testing is the GIMPs use of the toolbar for a global menu. I would expact users confused by the use of two menu sets in the interface - There is a FIle menu in the toolbar that makes it possible to open files, but not save them because it isn't related to a particular image window for example. This has following issues

* Having two sets of menus is confusing and doesn't exist in any other application apart from sodipodi which copied this behaviour from GIMP (this needs empirical data from the user testing).
* The toolbar needs to be wide enough to hold the menu. Especially true for some languages.
* You cannot hide the toolbar. It will always eat screen estate even though the tools are accessible with other means.

An alternative approach that might be nice to have and test is merging this menu into the image window and making the toolbar a regular dock just like any other. This exact interface exists in OpenOffice 1.1 for example. That would mean opening a new document window upon launching gimp (size and properties defined in the preferences).

There is a minor glitch to this. In OpenOffice the document window is enclosed in a placeholder window. When the user chooses to close the document and not quit, he can. When he chooses to open a new document, the old one, if it's untouched, closes so the new document gets placed where the old one was. If we choose not to use the placeholder window*, it may bring some problems:

* User launches GIMP, chooses not to start from scratch and chooses to close the image window. *plop* GIMP quits. * User launches GIMP, chooses not to start from scratch and selects "FIle>Open" after locating the right document, the old document/image window closes and new one appears. It may not appear at the same location because it's the WM's job to decide.

Since nobody implemented the alternative approach (or the two variants of it), it would be nice to at least see if the existing interface makes it hard for users to distinguish the image-related menus and the global one. A sample task may be

* Open an image and save it under a different filename

cheers

* which may or may not be difficult, I don't know

Jakub Steiner art designer

Novell, Inc.

Hi,

Since you're planning multiple tests, you may consider running a first test that is not very task driven, and in which the (few) tasks are relatively broad. The results from that test could then serve as the basis for the next test(s).

Actually, the first test serves as a basis for an article for the German LinuxUser magazine. My colleagues and I have a monthly article on usability issues, and next month's issue should be the Gimp 2.0. Apart from general improvements such as viusal feedback etc, we want to provide some tips how to handle difficult tasks (which means we have to find out which tasks are difficult!). That's why the test has to be a bit more specific.

Other issues, such as the toolbar problem mentioned by Jakub Steiner will be handled in the following tests. Please keep on making proposals, I'll note them and come back on you!

What kind of users will you be working with? How familiar are they with the testing platform? How familiar with photography? How familiar with digital imaging?

For the first test, we'll test two types of people (as it should serve only as a basis for the article, we will test four persons only): 2 people who know photoshop very well and/or work with it every day, and 2 people who have little or no experience with digital imaging.

One of the Cinepaint developers once had a user write down his suggestions for improvement. These could perhaps also serve as a starting point for an exploration of the sort of problems a GIMP user might run into. I don't seem to be able to find a link to that discussion, though.

If you find it, please send it to me. Currently, some KDE developers and my company are setting up a platform called openusability.org. It is thought to become a platform where open source projects (not KDE specific!) and usability experts should come together. If you are interested, we would like to welcome the Gimp project as one of the pilot projects. Apart from a forum, openusability.org will offer an opportunity to manage usability processes, provide space to file usability test reports (like the one you described above) and other informational resources, and discuss them.

have a nice day /ellen

On Thu, Apr 08, 2004 at 01:03:52PM +0200, Ellen Reitmayr wrote:

Hi,

Since you're planning multiple tests, you may consider running a first test that is not very task driven, and in which the (few) tasks are relatively broad. The results from that test could then serve as the basis for the next test(s).

Actually, the first test serves as a basis for an article for the German LinuxUser magazine. My colleagues and I have a monthly article on usability issues, and next month's issue should be the Gimp 2.0. Apart from general improvements such as viusal feedback etc, we want to provide some tips how to handle difficult tasks (which means we have to find out which tasks are difficult!). That's why the test has to be a bit more specific.

one of the least discussed or researched is probably how useable products that work on other operating systems is to linux users.

another way to say that is this: what about how usable photoshop is for someone who knows only TheGIMP?

TheGIMP is not photoshop. it was written so that linux people would not not have to use a different operating system and/or steal.

you make gimp race on a playing field it never was meant to be on and i dont see anyone testing the competitor on TheGIMP's home field.

what would the point of such a race be?

carol

Hi,

one of the least discussed or researched is probably how useable products that work on other operating systems is to linux users.

yes, that's true 8-)

TheGIMP is not photoshop. it was written so that linux people would not not have to use a different operating system and/or steal.

Of course the GIMP is not photoshop, and it is not our intention to make the GIMP a copy of photoshop! What we want to identify in this first test are the major difficulties somebody is faced when FIRST using the GIMP. That may be photoshop (or other imaging software) users, or newbies to digital imaging - no matter which operating system they use. That's why both groups are tested.

I think it would also be interesting to see how more experienced GIMP users handle the software - but for an initial feedback new users are more revealing. We are not trying to make the GIMP run on a field it was never meant for - except you think it was never meant to be used by new user 8-)

Greetings,
/ellen

On Thu, Apr 08, 2004 at 04:19:33PM +0200, Ellen Reitmayr wrote:

I think it would also be interesting to see how more experienced GIMP users handle the software - but for an initial feedback new users are more revealing. We are not trying to make the GIMP run on a field it was never meant for - except you think it was never meant to be used by new user 8-)

i had a heck of a time learning computer graphics. another problem with this test is that a seasoned photoshop user would not even think to use tools and methods that are what makes TheGIMP what it is.

i am sorry. i am really not being combative on purpose. it has been several years for me of watching the gimp develop to work on several operating systems and race apps that dont even work on linux. some how, some day, some one will think about the linux people also -- so let me keep pointing out how much the tests are very very uneven and actually sort of rude.

can someone please make a fair test, just once?

carol

ellen.reitmayr@relevantive.de (2004-04-08 at 1303.52 +0200):

For the first test, we'll test two types of people (as it should serve only as a basis for the article, we will test four persons only): 2 people who know photoshop very well and/or work with it every day, and 2 people who have little or no experience with digital imaging.

No gimp 1.x users at all?

GSR

Carol Spears writes:
> TheGIMP is not photoshop. it was written so that linux people would not > not have to use a different operating system

Was it? GIMP was originally written on the commercial Unixes of the time (Solaris, HP-UX etc), wasn't it? And it used Motif. And it still runs as well on most other UNIXes as it does on Linux, doesn't it?

> and/or steal.

As there aren't any commercial competition to the GIMP on Linux (or is there?), your statement doesn't really make sense, does it? True, GIMP being available on Windows might prevent Windows people from pirating Photoshop or PaintShopPro. But that hardly applies to Linux, what would they steal if there wasn't GIMP?

--tml

V ?t 08. 04. 2004 v 06:35 -0700 píše Carol Spears:

TheGIMP is not photoshop. it was written so that linux people would not not have to use a different operating system and/or steal.

Here you propose GIMP would be good for people who don't feel comfortable stealing (Photoshop)?

you make gimp race on a playing field it never was meant to be on and i dont see anyone testing the competitor on TheGIMP's home field.

what would the point of such a race be?

And here you're saying you don't like the comparsion to Photoshop? Gosh, if you're suggesting GIMP as an alternative you hardly can avoid comparsion. That's hypocrisy.

I'm one of the people who used Photoshop at the university and then started looking around for an affordable alternative at home. Trying to address issues that people may encounter doing the switch does make perfect sense.

cheers

Jakub Steiner art designer

Novell, Inc.

On Fri, Apr 09, 2004 at 01:15:41PM +0200, Roman Joost wrote:

On Thu, Apr 08, 2004 at 08:03:57AM -0700, Carol Spears wrote:

On Thu, Apr 08, 2004 at 04:19:33PM +0200, Ellen Reitmayr wrote:

can someone please make a fair test, just once?

What will be a fair test in your mind?

a test that resembles a demonstration more than a test.

carol

On Fri, Apr 09, 2004 at 05:57:17AM +0000, Tor Lillqvist wrote:

Carol Spears writes:
> TheGIMP is not photoshop. it was written so that linux people would not > not have to use a different operating system

Was it? GIMP was originally written on the commercial Unixes of the time (Solaris, HP-UX etc), wasn't it? And it used Motif. And it still runs as well on most other UNIXes as it does on Linux, doesn't it?

okay, this is the reason that i used it. i learned of its superiority. like tigert thought that you didnt use the gimp until after you used the irc, i guess i did not think that much about gimp without gtk. there is a couple of emails about when gimp development was started.

http://groups.google.com/groups?selm=petm-2907952329370001%40charnley.hip.berkeley.edu&output=gplain

http://groups.google.com/groups?selm=3ua2hd%24bbv%40agate.berkeley.edu&output=gplain

heh, google.

and/or steal.

As there aren't any commercial competition to the GIMP on Linux (or is there?), your statement doesn't really make sense, does it? True, GIMP being available on Windows might prevent Windows people from pirating Photoshop or PaintShopPro. But that hardly applies to Linux, what would they steal if there wasn't GIMP?

well, they decided not to steal motif anymore or something.

carol

On Fri, Apr 09, 2004 at 01:10:56PM +0200, Jakub Steiner wrote:

V ?t 08. 04. 2004 v 06:35 -0700 p??e Carol Spears:

TheGIMP is not photoshop. it was written so that linux people would not not have to use a different operating system and/or steal.

i am quoting spencer with my own reasons. http://groups.google.com/groups?selm=petm-2907952329370001%40charnley.hip.berkeley.edu&output=gplain

the actual word here is "akin". probably you should use google to determine this just for consistency.

Here you propose GIMP would be good for people who don't feel comfortable stealing (Photoshop)?

you make gimp race on a playing field it never was meant to be on and i dont see anyone testing the competitor on TheGIMP's home field.

what would the point of such a race be?

And here you're saying you don't like the comparsion to Photoshop? Gosh, if you're suggesting GIMP as an alternative you hardly can avoid comparsion. That's hypocrisy.

i spent years telling photoshop people to give gimp a chance. i am a blithering idiot by now. thanks for pointing that out.

I'm one of the people who used Photoshop at the university and then started looking around for an affordable alternative at home. Trying to address issues that people may encounter doing the switch does make perfect sense.

i used photoshop at the university also. it was the only way i could get the scanner to work there. i still had to learn about scanning also.

carol

Carol Spears wrote:

On Fri, Apr 09, 2004 at 01:15:41PM +0200, Roman Joost wrote:

On Thu, Apr 08, 2004 at 08:03:57AM -0700, Carol Spears wrote:

On Thu, Apr 08, 2004 at 04:19:33PM +0200, Ellen Reitmayr wrote:

can someone please make a fair test, just once?

What will be a fair test in your mind?

a test that resembles a demonstration more than a test.

Can you provide an example?

Michael

On Fri, 9 Apr 2004, Carol Spears wrote:

On Fri, Apr 09, 2004 at 05:57:17AM +0000, Tor Lillqvist wrote:

and/or steal.

As there aren't any commercial competition to the GIMP on Linux (or is there?), your statement doesn't really make sense, does it? True, GIMP being available on Windows might prevent Windows people from pirating Photoshop or PaintShopPro. But that hardly applies to Linux, what would they steal if there wasn't GIMP?

well, they decided not to steal motif anymore or something.

Just a note: illegal/illegitimate/unallowed copying of intangible media is not "stealing" or "theft" or "piracy", either legally or morally. It is simply that - "copying". Stealing actually means a tangible property was taken away without permission. Copying intangible media may or may not be OK, but nonetheless is not "stealing", because the original copy is left perfectly OK.

Regards,

Shlomi Fish

---------------------------------------------------------------------- Shlomi Fish shlomif@vipe.technion.ac.il Home Page: http://shlomif.il.eu.org/

You are banished! You are banished! You are banished!

Hey? I'm just kidding!