Summary After Usage for 3 Weeks

For all things Mellel

Moderators: Eyal Redler, redlers, Ori Redler

digimarkus
Already downloaded the guide
Posts: 27
Joined: Wed Sep 06, 2006 6:34 am

Re: Summary After Usage for 3 Weeks

Post by digimarkus »

Ori Redler wrote:...the problem is not the interface but the rather inflexibility of the way certain things work in Mellel.
I agree with Ori's point (if I understand it correctly). When I think about a UI there are at least two significant aspects to consider: the individual control elements and the procedural workflows. As a Mellel user I appreciate that for the most part each control element (i.e change font size) is simple and easy enough to learn. However, the sequencing of control elements within a writer's workflow don't alway seem that natural. This sense of awkwardness in the UI's workflow is partly do conventions that we have learned from other programs and also partly doing to the particular writing styles of Mellel's user base.

Now the trick to a great UI is not how easy it is to use, but how easy it is to learn and to remember what I have learned (without referencing a manual). IMHO, the Mellel UI is easy enough to use, but it's not always that easy to learn. As I already said, this is largely due to the fact that Mellel does some things in a ways that I have no prior learning to draw on. That's OK, if breaking with common conventions helps me to write, or managing the writing process in a better way. In fact, its a good thing, because then the software is actually help me learn to write better. An example of this is auto-titles. Learning this feature actually helped me to write in a more disciplined way. However, sometimes Mellel forces me into a workflow that seems to go against the way I want to write. I am not opposed to doing things in a different way from other software (that's why I use a Mac), however, my desire is to have software that fits my writing workflow closely.

So the challenge is not just designing a better UI, but better understanding people's writing workflows. The only real way to get this info is to watch how people use Mellel to write. Whenever, the software breaks a person's writing flow their is a place for improvement. For example, when I am writing in Mellel and insert a new autotitle the pop-up dialogue breaks my flow. My attention is focused on the place where I expect the autotitle to appear and then all the sudden I have to reorient myself to the a different context. In other words, its a bit jarring to my train of thought. So auto-titles are great (I love them) and iliminating the pop-up would just make my writing workflow a little bit smoother. Also, if I am writing (with my hands comfortably positioned over my keyboard) and I want to insert an auto-title I have to engage my mouse and open a pallette select a title option and then click insert (or use the right mouse button in context). Adding a keyboard command method would make this part of my writing workflow a lot smoother.

All this to say that Mellel is generally easy enough to use, but the more I use it the more I want it to better fit my writing workflows. It's like urban planning. You can put the sidewalks where you think they should go, or you can wait until people create pathways in the grass and then put the sidewalks there. The challenge for the Mellel developers is determining where most people want the sidewalks to go. Without a usability lab this insight can be hard to determine. These forums can be helpful in this regard, but people often jump to requesting a solution that they think with solve a workflow problem rather than to articulate how the current UI breaks their writing process.

Still, I am hopeful and I believe there is good evidence to support the fact that the Mellel developers truly seek to understand the writer's workflow and feature set needs. And in response they have continually worked to evolve Mellel into a simple, flexible and powerful tool from writing professionals. There is still much work to do, but I think it continues to move in the right direction.

- Mark
digimarkus
Already downloaded the guide
Posts: 27
Joined: Wed Sep 06, 2006 6:34 am

Re: Some ideas from another recent switcher

Post by digimarkus »

Danoz wrote:I suggest then, that from now on, this thread be used to discuss how we can address Ori's request that we practically implement this round table. I truly believe that this is an excellent opportunity given to us by Redlers and we should not squander it.
Nick
I agree and generally feel that you make good recommendations about how this might be achieved. One UI development technique that I have used in the past is a concept called proto-specing. The idea here is to create your UI concept in a rough visual form that illustrates a new UI idea (or the spec) and you then let a group of users give you feedback (does it meet your need? solve your problem?). Technically this can be done with flash mock-ups, quicktime movies of the app in development, light UIs with Applescript to fake functionality, or even screenshot sequence PDF documents. One nice thing about PDF mock-ups is that they are really easy to make and users can easily include comments, point out specific things and then send their feedback back to the developers.

- Mark
rickl
Got the auto-title mojo working
Posts: 17
Joined: Mon Nov 21, 2005 12:08 pm

Post by rickl »

I suggest then, that from now on, this thread be used to discuss how we can address Ori's request that we practically implement this round table. I truly believe that this is an excellent opportunity given to us by Redlers and we should not squander it.
I also agree. It would be nice to hear what Ori thinks of Nick's ideas.
rickl
Got the auto-title mojo working
Posts: 17
Joined: Mon Nov 21, 2005 12:08 pm

Post by rickl »

Or maybe we should ask on TidBits Talk for suggestions on how to do this? There are a lot of people there with diverse expertise.
Ori Redler
One of the boys
Posts: 342
Joined: Wed Oct 19, 2005 11:45 pm
Contact:

Re: Some ideas from another recent switcher

Post by Ori Redler »

Danoz wrote:I suggest then, that from now on, this thread be used to discuss how we can address Ori's request that we practically implement this round table. I truly believe that this is an excellent opportunity given to us by Redlers and we should not squander it.

Nick
I'm not technologically savvy -- to put this mildly -- in this kind of thing, but the two easy and viable options for discussions seem to be:
A. Discussing this in a form of a chat of sorts via iChat or maybe another tool which is more in-depth friendly. (suggestions would be welcome).
B. Discussing this in a forum format, but with some special accommodations (e.g., a sticky thread that stays on top for several days). That would have the advantage of allowing people to discuss things at length, but will be less "guided".

Either way, I think that the subject matter of each discussion should be narrow rather then wide, so that things can be more clearly defined (and thus be more valuable as a guide for action for us).
Ori Redler from RedleX
Timotheus
Knows everything, can prove it
Posts: 285
Joined: Thu Oct 20, 2005 4:36 pm

Post by Timotheus »

Something I regret very much is the lack of connection between the discussions on this forum and Mellel's actual evolution. In the past few years many people have made many sensible proposals without getting clear answers; and if the answers were clear and positive, they were rarely followed by concrete and positive actions. In the end, this lack of connection inevitably leads to apathy.

If other people do still have the energy for broad in depth-discussions about general topics, so much the better. I don't have it anymore. To be honest, I have lost faith in similar discussions. The only kind of discussion is still believe in are concrete, well defined discussions about concrete, well defined topics, which should be followed by concrete promises made by the developers.
Ori Redler
One of the boys
Posts: 342
Joined: Wed Oct 19, 2005 11:45 pm
Contact:

Post by Ori Redler »

Timotheus wrote:Something I regret very much is the lack of connection between the discussions on this forum and Mellel's actual evolution. In the past few years many people have made many sensible proposals without getting clear answers; and if the answers were clear and positive, they were rarely followed by concrete and positive actions. In the end, this lack of connection inevitably leads to apathy.

If other people do still have the energy for broad in depth-discussions about general topics, so much the better. I don't have it anymore. To be honest, I have lost faith in similar discussions. The only kind of discussion is still believe in are concrete, well defined discussions about concrete, well defined topics, which should be followed by concrete promises made by the developers.
Timotheus, I agree with you. The points for discussion should be well defined and preferably attached in some way or the other to:
A. Something we plan to implement in the near future.
B. Something that is doable within a given, limited time frame.
Ori Redler from RedleX
NC
New to all this
Posts: 1
Joined: Tue Sep 26, 2006 4:22 am

Post by NC »

It's very heartening to see this discussion.

I'd bet a small amount of money that I was the first user to publish a book written entirely in Mellel, and I retain a great fondness for it. But later releases simply became too complex for me to use. My needs are pretty simple (and perhaps I'm a little impatient) but Mellel's implementation of styles just baffles me. I look at it, and my eyes go strange.

Previous to my discovery of Mellel, each of my books had been written in Word, and it was to Word I returned -- not without reluctance, or a sense of defeat,. But I understand Word; I understand its templates, its outliner, its views.

Mellel is an application for professional writers. That's what I am. But Melle makes me feel dumb. Which I'm not. And I don't like my applications to do that.

I know Mellel inspires great passion, and as you can see, I feel residual loyalty to it. But a great many professional writers need great organisational tools BECAUSE THEY'RE DISORGANISED. Mellel requires a user-discipline I just can't give it. I suspect I am not alone.

It would be great to go back to Mellel. All it needs is some tools for people like me, people who want to organise and write -- not necessarily in that order; much of my outlining is done retrospectively -- and who are easily intimidated by complex and rigid interfaces.

There are, I guess you'll know, quite a number of us....
Timotheus
Knows everything, can prove it
Posts: 285
Joined: Thu Oct 20, 2005 4:36 pm

Post by Timotheus »

@ Ori: Good to hear that! Because the posts on this forum are full of useful, sensible, concrete and in my opinion sometimes even really important suggestions which have never got not even the beginning of an answer.

Just to give a concrete example: some months ago I made a series of short, very concrete proposals with regard to footnotes / endnotes; others subsequently added theirs. But from the side of the developers … silence, absolute silence!

See: http://forum.redlers.com/viewtopic.php? ... =footnotes

And this is just one example out of many dozens one could easily give. When this happens often (and it has happened often in recent times!) it spoils the appetite for making proposals, for contributing to the discussion, for making Mellel "insanely great", as somebody else recently said on this forum.

In my opinion, the implementation of major features should always be preceded by a thorough discussion on this forum. And the developers themselves should take the initiative to this discussion, which at a certain moment should lead to concrete decisions and to concrete and public committments.

We all know that presently major features like indexing and cross-referencing are being implemented. But I haven't seen as yet any general discussion about these topics. About cross-referencing there have been of course many threads, but not a general discussion leading to concrete decisions. And the discussion about indexing until now has been rather meagre. And that's a pity, because indexes are a really important part of a scientific publication, and can be implemented in many different ways.

Moreover, perhaps the time has come to divide the threads on this forum into sections. This would make the discussion much better surveyable, and would help to limit the proliferation of threads about identical or very similar topics.
nvalvo
Read the guide, knows everything
Posts: 50
Joined: Mon Nov 14, 2005 10:08 am
Location: The Train between Davis and San Francisco

Post by nvalvo »

Ori said:
A. Discussing this in a form of a chat of sorts via iChat or maybe another tool which is more in-depth friendly. (suggestions would be welcome).
B. Discussing this in a forum format, but with some special accommodations (e.g., a sticky thread that stays on top for several days). That would have the advantage of allowing people to discuss things at length, but will be less "guided".
Umm, how about both? Option A is obviously a lot more scheduling (given, for example, that I am in San Francisco and the Redlers are in Israel, right? A time difference of something like ten hours?) than I am ready to do, but some people are obviously keen to.

I wouldn't mind writing up some interface ideas, however. I've been itching to play with FlySketch... :wink: Let there be a forum!
verma
Knows everything, can prove it
Posts: 171
Joined: Thu Oct 20, 2005 6:00 am

Post by verma »

Timotheus wrote: In my opinion, the implementation of major features should always be preceded by a thorough discussion on this forum. And the developers themselves should take the initiative to this discussion, which at a certain moment should lead to concrete decisions and to concrete and public committments.

We all know that presently major features like indexing and cross-referencing are being implemented. But I haven't seen as yet any general discussion about these topics. About cross-referencing there have been of course many threads, but not a general discussion leading to concrete decisions. And the discussion about indexing until now has been rather meagre. And that's a pity, because indexes are a really important part of a scientific publication, and can be implemented in many different ways.
I'm all for openness and grassroot democracy, but not when the starting points are fundamentally inequal.

The developers of Mellel have written the code, know what is possible, and have made their own product (which we can choose to buy or not). I'm just a user, who on a blue Monday might have an idea - a halfbaked one at that - but at the end of the day, the product is theirs, and so are, inevitably, decisions that have to be made for which they need to take into account all aspects of development.

I agree that there are could be different (but not that many) ways to implement cross-referencing, and it's my personal top priority. But I trust the Mellel folks to come up with a practical, flexible, extendible solution which I will test with fearless determination...

A user suggestion box could be a good idea - giving the Redlers all creative freedom to do what they want with it. But I wouldn't see it fit to have decisive power. We all have different ways and likes, and the consultation rounds would probably end up taking the average route. That, imho, would be a bad choice.

But I'd second something like an open suggestion box, for free.

Mine would be, on the topic of cross-referencing:
- a system using labels that are determined much like the auto-title insertion pop-up window
- these labels should be "invisible" but with the option of making them visible
- a label search option in Find&Replace
- a crossreference palette that allows the user to define how the various labels (a bit like the doc. variables) are to be treated in the final print version (numbers, letters, character styles ...).

My 2 cents.

best,
v.
zoul
Knows everything, can prove it
Posts: 120
Joined: Thu Aug 10, 2006 1:48 pm
Location: Boskovice, Czech Republic
Contact:

Post by zoul »

verma wrote:I'm all for openness and grassroot democracy, but not when the starting points are fundamentally inequal.
Exactly. We would all agree that users are a good source of comments, since they are the ones who use Mellel all the time (I use it about six hours a day myself). But there are hundreds of users in this forum and with every hundred more it is going to be harder to get the sound ideas through.

I am afraid that discussing every feature that is going to be implemented in the forum would be hard, because this alone could easily slurp all the time of a single developer, which is not worth the trouble.

The least we could do is to create a new top-level thread, so that the development-related discussion does not mix with ordinary Mellel troubleshooting. Then we IMHO need a way to “formalize” (sorry, no idea of a better word right now) the suggestion process so that the developers do not have to spend too much time digging through the threads and can simply review some discussion results that most of the interested users agree on.

[Just an idea, of course.]
laup
Knows everything, can prove it
Posts: 311
Joined: Tue Aug 22, 2006 4:13 am
Location: Topanga, California

Chunks

Post by laup »

If designers lay out a good strawman, along with rationale, then it can be very efficient and productive to elicit comments and suggestions. One "session" might go a long way.

Focus is good. My priority topic would be interface. A great deal could be done with reasonable effort to unify the interface paradigms and make them intuitive, while using only the current bag of tricks and not changing any fundamentals of design. In a session on this it would be out of bounds to talk about desired new features, nice to haves, new ways of doing things, etc.

I doubt that a "session" on cross referencing is needed because the designers probably have that problem nailed (although it may be easy or hard). That may be true also for front matter (e.g., table of contents). However, perhaps a pre-beta review would be useful. I don't know.

A "session" on collaboration-related options might be useful to Redlers because differences of opinion would be aired and might be illuminating. I would think that there are lots of dangers in this realm (note that Word crashes and corruptions worsen with collaboration).

The "styles" issue is the most knotty. I can imagine two sessions: one to air specific critiques rather than general laments, the second to air proposed or alternative approaches. And, of course, there is alpha and beta testing.

How best to use the special mechanism of forum, iChat, or whatever is not clear to me. It's a fascinating issue. In the old days, we believed that the key was brilliant design and good testing; these days, there is much more interest than previously in distributed problem-solving. There is no question about that having great potential power, but it also has costs and serious potential drawbacks.
Paul
joewiz
Knows everything, can prove it
Posts: 199
Joined: Sun Oct 23, 2005 9:42 pm

Post by joewiz »

(This comment doesn't deal with the netconferencing proposed in some posts above.)

Problem: This forum is great for troubleshooting but bad for feature requests. Too often, forum posts here contain a litany of requests that (a) have been mentioned before, (b) are combined with other tangential issues that become the topic of conversation, or (c) are brilliant but are poorly titled and go neglected. The forum is active but is very disorganized, and discrete feature requests/ideas/bugs are hard to track. Users feel their passionate contributions go unheard. The collective wisdom and passion is not being properly or productively channelled.

Q: What tool could best channel the wisdom and passion of users while making the myriad of requests and suggestions manageable and quantifiable for the Redlers?

A: The best example of a well-oiled machine like this that I've seen is the 'Requests and Bugs' section of Hog Bay Software's website: http://www.hogbaysoftware.com/requests

Q: How does this system work?

A: Users submit a feature request or a bug report. Each registered user can (1) comment on the request and/or (2) vote on the request using a 5-star system (5 star=top priority, 1 star=not interested). The website lets you sort requests by #s of votes and # of stars. If a request has been duplicated, vigilant users or the developer closes the request and points people to the original. The developer can add comments, can mark requests as closed when they've been implemented, or can mark them as 'won't fix.' This gives users very clear feedback about each request. Requests can be renamed to better reflect the content of the post, or changes in the conversation (i.e. if a post is originally titled: 'Aaargh!', but actually talks about a printing bug, it can be renamed 'Margins incorrect when using A4 size paper').

Q: How would it work here?

A: We'd have a single request about Cross Referencing. Comments would go on and on for pages about the best way to implement it, etc. Any new forum posts about cross referencing would be met with a cry of 'Look at http://...., and add your vote and ideas there!' Votes would grow and grow. The Redlers would see that many more people want Cross Referencing than Auto Correct (just an example). They could gauge the complexity of implementation against userbase desire for it. Once a feature is implemented, it's marked as closed. When someone notices a bug, a new post goes up, generating votes, etc.

Q: No more forum?

A: No, there is a forum too: http://www.hogbaysoftware.com/forum/mori . People can post long rants about how they'll never be able to use the program until features x, y, and z are implemented. But that doesn't constitute a feature request. So it may well go ignored until the user listens to all of the people saying, "Make a feature request".

Summary: This system enforces a degree of discipline on the users - requests have to be discrete. Debate goes on within these discrete areas. The voting system (the real innovation, I believe) shows how popular *each* request *really* is, regardless of how passionate or well-spoken the author is. The developers also have to do their part, marking requests as 'done' or 'won't fix', and occasionally giving their opinions or remarks about what the difficulty is in implementing a feature. Overall, this system has led Hog Bay to become very democratic and responsive to (but not locked into) users' requests.

Techie Note: Hog Bay built this forum/request system around Drupal. I'm not an insider on their web development, but I have the feeling that Hog Bay would share their methods about plugins and add-ons.
Timotheus
Knows everything, can prove it
Posts: 285
Joined: Thu Oct 20, 2005 4:36 pm

Post by Timotheus »

@ verma and zoul: please don't put into my mouth things I never said nor thought! I did not dispute at all the Redlers' fundamental right to shape Mellel as they think best, nor did I propose to discuss "every feature that is going to be implemented in the forum".

@ joewiz: sounds indeed very democratic; yet a very bad strategy as far as Mellel is concerned. Mellel, indeed, has a very concrete goal to achieve: it wants to be or to become undisputedly the best choice for academic writers, first-class journalists and their similes. In order to achieve this goal it must absolutely have, for instance:

- undisputedly the best footnotes / endnotes implementation
- undisputedly the best cross-referencing
- undisputedly the best indexing
- undisputedly the best outlining
- undisputedly the best text generation tools
- undisputedly the best handling of languages like Hebrew, Arabic, Greek, Chinese, Japanese
- undisputedly the best handling of tables, rtf, doc
- undisputedly (etc.)

That is Mellel's only reason of existence; and that is my Mellel. If Mellel will not become this, it will not interest me anymore. It's as simple as that.
Post Reply