Page 1 of 2

Paradigms

Posted: Tue Sep 05, 2006 4:09 pm
by laup
The discussion on highlighting raised basic questions about what attributes are desirable as Mellel goes forward. One paradigm is for a superb but seemingly light word processor, something like a modern-day version of Word 5.1, for those who remember it. "Modern-day" would include the very nice outliner pane, excellent rtf import/export, and an up-to-date GUI that exploits OS 10. A different paradigm might have future-Mellel be more of a brainstorming and/or collaboration tool, perhaps even for real-time on-line collaboraton. A third paradigm might me more along the Scrivener line, although there is some overlap. And so on. The multiple-language features could be part of any of these.

My own thoughts are, at present (idiosyncratic to me, of course):

1. Brainstorming is best done in conversation, at a whiteboard or its electronic equivalent, or on the proverbial back of an envelope. I don't see need for brainstorming-type notes to be incorporated in an emerging or near-final document. To the contrary, as ideas begin to come together, I tend to discard all such notes anyway. And, if they are electronic, they could just as easily be in email, PowerPoint, DevonThink, Word, or even Mellel. The notes may well contain some drawings from either PowerPoint or OmniGraffle, or from a scan of something done on paper. Others, obviously, have different styles.

2. Once I begin document preparation, outlining can still be very useful, but only if it is flexible--as in Mellel, but it doesn't really incorporate literally the raw creative notes. The flexibility needed would seem to me to be provided by the move-stuff-around aspects of Mellel's outline and the ability to add "Author Notes" as a separate stream parallel with real footnotes and/or endnotes.

3. Highlighting seems to me overrated, although I occasionally use it with Adobe Acrobat when reading pdfs. In taking notes on a Word document or a Mellel document, what's wrong with just selecting a passage and turning the text red? That's certainly easy to do in Mellel with the F keys (I have one of them set to make text red). Perhaps I'm missing the point.

In summary, I find myself leaning back to the first paradigm of seeking just an excellent modern word processor. In many respects, Mellel fits the bill: the outliner, the Cocoa interface, the flexibility on notes, performance, good import/export, integration with Bookends, and now the excellent find-and-replace function and XML.

Easy, however, it is not. My biggest quarrels with Mellel relate to the horrendously complicated style mechanism (even though I seem now to understand it decentlyl), the lack of relatively-defined styles, the lack of templates for what I want (making it necessary to create my own), and some inconsistency of GUI concept among styles, setup, notes, and lists. Others mention cross-referencing, more keyboard shortcuts, markup, and some other items. All things considered, I would put by far the biggest weight on having a simple-but-excellent mode, based probably on a richer set of custom templates and some tweaking of the interface for consistency.

Re: Paradigms

Posted: Tue Sep 05, 2006 6:05 pm
by Ori Redler
pkensildavis wrote:All things considered, I would put by far the biggest weight on having a simple-but-excellent mode, based probably on a richer set of custom templates and some tweaking of the interface for consistency.
Sorry for touching upon this last point in an off-topicish manner -- can you elaborate a little about "a richer set of custome templates"?

Ori

Re: Paradigms

Posted: Tue Sep 05, 2006 8:50 pm
by laup
Ori Redler wrote:
pkensildavis wrote:All things considered, I would put by far the biggest weight on having a simple-but-excellent mode, based probably on a richer set of custom templates and some tweaking of the interface for consistency.
Sorry for touching upon this last point in an off-topicish manner -- can you elaborate a little about "a richer set of custome templates"?

Ori
Ori,

A set might include templates for a book, a technical report, a technical memorandum, a technical journal article, and perhaps a few others. Each, however, would be fully developed--not merely a skeleton. For example, a book template should allow for parts, chapters, sections within chapters; it would have both footnotes and endnotes; it would have separate contents for chapters, tables, and figures; it would have styles for equations, quotations, bullets, etc. And, of course, tables and figures. Perhaps it would have an alternate language built it. It probably would use vanilla fonts, such as Times Roman.

The organization I work for has such a set of Word templates and, despite being imperfect, they are a great time-saver. Further, even for people like me who never quite like templates as they are, it is easy to make the necessary changes because they are changes "on the margin." In Mellel, changes would be more difficult to make because the styles are not defined or definable relative to a base style, but it still would not be so bad. In contrast, building a Mellel style set from scratch, as I have recently done, took a lot of effort and I'm sure I will continue to discover small problems for quite awhile.

The current tutorials have lots of good material, but also take considerable time; they are not substitutes for templates. The particular current templates I looked at did not seem to be fully developed, much less guerilla proofed. Others may be. Where they are incomplete, it wouldn't take much further effort by a Mellel expert to add the additional features. Suggestions are cheap, but I'd suggest that Mellel should polish up a set of templates, perhaps starting with the ones on the website, and treat them as part of the basic package. Rather like Apple does it with Pages, but in more depth as discussed above.

Changing tone, Mellel is really neat; congratulations on what you've accompished so far and how well you're exploiting the technology.

Posted: Wed Sep 06, 2006 6:17 am
by zoul
That seems like a good idea to me. It could even be done as a contest (“design the best template for Mellel and win a week with the Redlers by the sea”), but from my viewpoint there is one substantial catch: the fonts. I do not remember what fonts are installed with OS X, but every good-looking thesis or book or whatever deserves good fonts. I’m not implying the fonts shipped with OS X are bad, I just doubt there is enough of them to make a variety of high-quality templates.

Posted: Thu Sep 07, 2006 2:20 pm
by nicka
Replying to the original post:
I always assumed the aim for Mellel is to be a word processor that doesn't suck. That means it has to have support for modern standards: Unicode, OpenType, XML, and some kind of regular expression find and replace -- all now in place. It also has to be utterly stable and reliable, which it has been from the beginning.
Then it has to have the features that support creation of complex long documents for books and dissertations: numbering, note-streams, bibliography capabilities, table of contents, cross-referencing and indexing. These are about half done, with the rest promised soon.
And then it needs features for drafting, editing and collaborative working: outlining, full-screen mode, highlighting, version tracking and others. We have a bit of this, and have been told that the rest is the next phase after long document features. (I don't care so much about this stuff, because I tend to draft and proofread on paper, but others obviously use these facilities and miss them in Mellel.)

I don't think Word is the inspiration, so much as classic Nisus Writer and FrameMaker, both of which in their own way didn't suck (for their era).
I believe Word 5 couldn't do a lot of the things I've listed, and while modern Word can do almost everything there it's buggy and ungainly and its support for Unicode and OpenType is only partial.

In a way, it's a bit strange that there isn't already a word processor with all these features. They are all obvious and have all been implemented by one program or another, just not all together. Presumably it just takes a lot of development time and effort. At any rate it looks like Mellel will be the first.

Posted: Thu Sep 07, 2006 2:30 pm
by nicka
I do not remember what fonts are installed with OS X, but every good-looking thesis or book or whatever deserves good fonts. I’m not implying the fonts shipped with OS X are bad, I just doubt there is enough of them to make a variety of high-quality templates.
The fonts with the OS are OK but they don't show off Mellel's typographic abilities.
Perhaps someone should design a template that uses Cardo or Junicode so it can have section titles in small caps. There would have to be instructions on the first page of the template to download and install the required fonts, though.

Posted: Thu Sep 07, 2006 3:10 pm
by daiyi
nicka wrote:Replying to the original post:


I don't think Word is the inspiration, so much as classic Nisus Writer and FrameMaker, both of which in their own way didn't suck (for their era).
I believe Word 5 couldn't do a lot of the things I've listed, and while modern Word can do almost everything there it's buggy and ungainly and its support for Unicode and OpenType is only partial.
Not to mention that Word on an Intel machine runs much slower now under Rosetta and eats up huge chunks of memory. No future VBA support for Word, either. Mellel is very nimble and stable, as always.

Posted: Sat Sep 09, 2006 5:14 pm
by Ori Redler
zoul wrote:That seems like a good idea to me. It could even be done as a contest (“design the best template for Mellel and win a week with the Redlers by the sea”)
Well, we want to encourage people to participate, not scare them away.... :P

Re: Paradigms

Posted: Sat Sep 09, 2006 5:19 pm
by Ori Redler
pkensildavis wrote: Ori,

A set might include templates for a book, a technical report, a technical memorandum, a technical journal article, and perhaps a few others. Each, however, would be fully developed--not merely a skeleton. For example, a book template should allow for parts, chapters, sections within chapters; it would have both footnotes and endnotes; it would have separate contents for chapters, tables, and figures; it would have styles for equations, quotations, bullets, etc. And, of course, tables and figures. Perhaps it would have an alternate language built it. It probably would use vanilla fonts, such as Times Roman.
Since I do not know everything (just found out about this inherent limitation) we'd have to use either templates for Word or for another WP as a model, or a finished work that we can "break" into such a template.

Anyone with one of those (or others required) is more than welcome to send them to me.

Posted: Sat Sep 09, 2006 7:09 pm
by BreakNeckRidge
With the recent awesome feature additions, the next big project Mellel should tackle is an interface design overhaul. It seems like many of the abilities that people are asking for can be accomplished using the powerful features already present in Mellel, if only the interface made it easier to gain quick understanding of how to use them.

Posted: Sat Sep 09, 2006 8:30 pm
by zoul
BreakNeckRidge wrote:With the recent awesome feature additions, the next big project Mellel should tackle is an interface design overhaul.
I think the overall interface is very good, but there are numerous glitches:

• Keyboard shortcuts. Setting shortcuts via System Preferences is not really a solution, because it is not intuitive [at least not for me], it is slow and depends on menu labels. Keyboard shortcuts are a must for routine work. Once I saw an application where the user could set shortcuts simply by hovering over a menu item and pressing the correct shortcut, now that was something. [But maybe such solution has its downsides, too?]

• Menus. The Edit menu is quite big now: How about the “Special characters” item? Doesn’t it rather belong to the Insert menu? Does anybody really use the Edit>Clear path instead of hitting Delete? Then there is the Layout menu [that I do not use at all] with only two submenus, that feels weird.

• Unused variables and titles cluttering screen. Maybe this is just obsession of mine, but I do not like at all when I open the Insert>Document variables menu (or some other place with document variables) and get a list of 15 variables I do not use. [The same goes for the Preferences view with variables, the Auto-title palette, the level list in the Auto-title flow setup screen or character style variation names.] Maybe the list could always grow one item larger when I use the last item? (I set document variable name #3 and #4 appears in the list.)

• Does anybody use the Default character/paragraph/… style combo boxes? (This is not a rhetorical question :) Wouldn’t it be easier to always use the first style as default and let the user reorder the styles by dragging?

• Why does not the Onscreen background preferences option change the background of an existing open document?

• It is not possible to move the main window using the bottom border; it works with all the others brushed metal [or err… plastic :)] applications (Safari, iTunes, …).

• Some dialogs use in-place name editors (e.g. the document variable names), others use dialogs (e.g. autotitle flow setup). This should be consistent, I prefer in-place editors.

These are some of the things that I do not like personally, I do not say all of them are bugs.

Posted: Sat Sep 09, 2006 9:17 pm
by D'Espice
The problem I'm beginning to see is that everybody has his or her own expectations of what a word processor should be either capable of, or look like, or behave like, etc. pp.

As for the interface, I'm loving it. Being a person that gets easily distracted (see, I should be writing but am instead posting here... so long boss, article won't be finished on monday for sure) I find the fact that there are no icons in Mellel incredible. No visual distraction means I can focus on the other distractions more easily... anyhow, please do not add any icons or graphics whatsoever to that interface. Actually, I don't even use the Mellel but the Compact theme - I definitely prefer that to the first one, not because I dislike the brushed metal interface but simply because it's easier on the eyes.

My personal requirements regarding a word processor are quite minimalist - a simple interface, performance and reliability, cross-referencing, outlining, page numbers, some kind of literature database integration, fullscreen view, support for more than one language per document, and in general, easy to use. So far, Mellel has fulfilled my needs perfectly, except for the cross-referencing part, which however has been promised to come, and that's fine for me. The key is actually performance and reliability, especially when working on long documents. I do not need any templates for I prefer to create my own templates from scratch. I do not need any special fonts or characters since I have never been able to figure out why I should use these instead of the default fonts that are easy on the eyes. I barely need tables or images.
And that's the point, everyone has certain needs and requirements, and gues what happens if you try to find a compromise between these interests... that's right, you end up with a bloated version of Word (aka OpenOffice), and I surely hope that is not what neither the users nor the Redlers have in mind.
The only thing that I would love to see is a cross-referenced table of contents like Word offers it. I find the solution of first deleting the old TOC, and then creating a new one quite weird, although I did get used to that rather quickly. Now that they have added the fullscreen feature, this is the last thing remaining on my list, and it's not even that urgent since I know it's coming, and meanwhile manage to get along somehow.

Posted: Sun Sep 10, 2006 6:17 am
by BreakNeckRidge
D'Espice wrote:As for the interface, I'm loving it. Being a person that gets easily distracted (see, I should be writing but am instead posting here... so long boss, article won't be finished on monday for sure) I find the fact that there are no icons in Mellel incredible. No visual distraction means I can focus on the other distractions more easily... anyhow, please do not add any icons or graphics whatsoever to that interface. Actually, I don't even use the Mellel but the Compact theme - I definitely prefer that to the first one, not because I dislike the brushed metal interface but simply because it's easier on the eyes.
You misunderstood me, I agree with you 100% on the no icons, no visual distraction thing, but that doesn't mean there isn't still a lot of interface that you still have to interact with, and that's the interface that I'm talking about. For instance, the interface for the style system borders on incomprehensible to the beginner and even limits what the experts are able to do with it because it's so hard to get a full understanding of all of the power that the style system imparts. If they could come up with a new interface paradigm for the style system that's already in place then everyone would get so much more use out of it.

Posted: Sun Sep 10, 2006 2:46 pm
by FA1
BreakNeckRidge wrote:For instance, the interface for the style system borders on incomprehensible to the beginner and even limits what the experts are able to do with it because it's so hard to get a full understanding of all of the power that the style system imparts. If they could come up with a new interface paradigm for the style system that's already in place then everyone would get so much more use out of it.
As a beginner, styles are intimidating, although I haven't devoted enough time to learning them yet. Perhaps another tutorial (or more than one) that walks you through styles would be a good idea, even if it is somewhat repetitive with what is already in the beginner tutorial.

I would add that part of the apparent complexity comes from never really using styles properly in Word, and now wanting to do so. I suspect I'm not the only one. But more experienced users can attest to the difficulty that might be unique to Mellel's system.

Posted: Sun Sep 10, 2006 6:51 pm
by verma
FA1 wrote: I would add that part of the apparent complexity comes from never really using styles properly in Word, and now wanting to do so. I suspect I'm not the only one. But more experienced users can attest to the difficulty that might be unique to Mellel's system.
For me, Mellel's style system is complex yet simple. The complexity resides in its flexibility (pages, characters and paragraphs), the simplicity in its logical approach (style sets - a family of styles that belong together).

A style set avoids having endless miles of customised style settings under the Paragraph/Character/Layout menu, and it enables you to make stylesheet-design rather simple.

I'm not in favour of prebuilt (Mellel-made) templates - that's not a developer's job. And no pre-built template will suit my needs (read: the needs of the publishers I write for): there is no such thing as a uniform set of templates that will suit everyone. Consequence: you'll have to do it yourself. And though it's time-consuming, it is a good way to learn how the programme works.

And there's always the fabulous Mellel Guide, or the beginner's tutorial (use the Mellel format so you can immediately practice hands-on). Both are excellent reference works.