Some comments regarding feature requests

Feature requests, and in-depth discussions of features and the way Mellel works

Moderators: Eyal Redler, redlers, Ori Redler

rpcameron
Knows everything, can prove it
Posts: 980
Joined: Wed Oct 26, 2005 12:48 am
Location: IE, CA, USA

Some comments regarding feature requests

Post by rpcameron »

This post is meant to be food–for–thought. I am aware that it will induce flames, and many other users will disagree with my perspective. I am writing this to allow for a place where users can discuss their views on the way that Mellel handles particular features, or to see what ideas may be borrowed (but not lifted completely) from other applications. (Completely lifting a feature from another product and implementing it in an identical manner would be a travesty, as Mellel functions differently from most every other word processor.)

This forum has been host to many requests that Mellel handle things differently than it does, for no better reason than the requester was used to the method implemented by another program (i.e., Word, Nisus Writer, FrameMaker, et al.). I can understand a user’s desire to use a best–of–breed application (such as Mellel, IMNSHO) to complete their tasks. However, I find it ridiculous and annoying that users constantly wish for the way in which the issue was handled in another program.

(I admit that when we upgrade our computers we’d like to see the applications we use evolve with the machines we use. But sometimes that is not the case, as is evidenced by the lack of features in Mellel that many were accustomed to in Nisus Writer and Adobe PageMaker. Yet, no one required you to change your workflow to incorporate Mellel. Nisus Writer and PageMaker still work exceptionally well in MacOS 9, just as you used them years ago.)

If a feature is missing from Mellel, then I can fault no one for wishing for its inclusion in a future version. However, expecting the implementation of said feature to work as it did in other software is a bit naïve. Features we have come to love and rely upon in Mellel—Auto–Titles, the style system with variations, note streams, &c.—do not work the same in any other word processor I know of; why would it be expected then that when Mellel implements a new feature it will work as it did before?

FrameMaker used a completely different paradigm and approach to documents: each page was made of floating frames that could be moved and linked to one another. Mellel is a stream–based word processor that uses no frames whatsoever. While free–floating frames may be useful for some situations, I can see absolute need for them in a word processor. (Text wrapping around images is much-requested, but this can be accomplished with a stream.)

If a frame–based workflow is what is required, then there are several options that are available. One is to return to FrameMaker. Just because it is no longer in production does not mean that you cannot continue to use it. Another option is to look at other frame-based word processors, such as KWord or Pages. (I know that Pages is not wholly frame-based, but it can operate that way.) A third option is to switch to a formal desktop publishing (DTP) and/or layout application. QuarkXpress and Adobe InDesign are two excellent applications; Scribus is an open–source DTP solution as well.

While I sorely miss some features from Nisus Writer (such as bookmarks, cross–references, indexing and glossary, among others), I understood that when I moved to OS X and Mellel that I would need to either find workarounds. (Interestingly, my switch from Word to Mellel was far easier because I tried to maintain a structured writing flow in Word, and tied my styles to structural elements; this is achieved far more easily in Mellel than it was in Word.)

Likewise, just as many users wish for features that I feel are more suited to a layout application, there appears to be an equal number of users requesting Mellel work in the same manner as a text editor. If you are looking for a text editor, let me recommend TextMate, Smultron, BBEdit or TextWrangler. Those are all excellent programs geared specifically for working with plain text files. They are also incredibly good at manipulating and working with large amounts of unstyled text.

Just as Mellel is not a layout program, it is also not a text editor. Mellel is a word processor. It aids in the crafting of beautiful documents. To me, this means that it ought to handle the creation of documents that are predominantly textual in nature. Images and/or tables may be added to aid in the document, but it is still a program that helps a user create beautiful texts.

Perhaps I am a bit too pedantic, but I do not want to see Mellel suffer the same fate that Word has. (Remember, Word started out as a word processor on the Macintosh before it ever made it to Windows.) As I said, there are many features that I would still like to see implemented in Mellel. But I would rather wait and have them done properly, and in keeping with the spirit of the rest of the application, than to be just lifted from another application.

I know this is a bit long–winded, but I hope I’ve also provided some material for others to think about…
— Robert Cameron
nicka
Knows everything, can prove it
Posts: 677
Joined: Thu Oct 20, 2005 2:55 pm
Location: Oslo
Contact:

Post by nicka »

No flames from me. I agree that Mellel is -- and I expect will continue to be -- intended to make possible the creation of (long, structured, possibly multi-script) textual documents.

I certainly agree that it shouldn't have a feature just because some other application has it. So 'Word has feature X, therefore Mellel must have feature X' is not a compelling argument. But I doubt anyone would think it was.

I think generally when someone posts a feature request, that person thinks that that feature would make Mellel more useful in its core areas. If they say that Word, or FrameMaker or Nisus Writer or Pages (...) has the feature, or an approximation to it, that does not mean necessarily that they think Mellel should in general be more like one of those applications. Rather it may be a proof of concept ('Look, it is possible, KWord has it') or a way to avoid giving lengthy descriptions ('it could work like THAT'), or even a kind of exhortation ('Mellel is supposed to be the best for this kind of work, but THIS application has this one great feature Mellel should steal').

Conversely, it doesn't make sense to say simply 'feature Y is a text-editor feature' or 'feature Z is a DTP feature' and end the debate there. If so, Mellel would not have regular-expression find on the one hand, or advanced typography and columns on the other.

The question for any feature, surely, is whether that feature would help the kind of people who use Mellel to produce the kinds of predominantly text documents they want to produce, in the formats they need to produce them in.

That goes for really minor proposed changes, like extending the shortcut keys available for styles, or for major conceptual ones, like, say, the idea that one day Mellel's streams of text could be arranged inside frames. I can't see any alternative to taking each idea on its merits, regardless of where it comes from.
Timotheus
Knows everything, can prove it
Posts: 285
Joined: Thu Oct 20, 2005 4:36 pm

Post by Timotheus »

Thanks, Patrick, for mentioning KWord and Scribus! I didn't know either of them, and they probably offer some possibilities that perfectly fit into my workflow, and which I dearly miss in wordprocessors: for instance, the possibility to place a text and its translation on opposite pages, or the possibility to place a commentary at the bottom of the page without tying it to the text by notes.
simifilm
Knows everything, can prove it
Posts: 93
Joined: Fri Oct 21, 2005 11:58 am

Re: Some comments regarding feature requests

Post by simifilm »

rpcameron wrote:T
FrameMaker used a completely different paradigm and approach to documents: each page was made of floating frames that could be moved and linked to one another. Mellel is a stream–based word processor that uses no frames whatsoever. While free–floating frames may be useful for some situations, I can see absolute need for them in a word processor. (Text wrapping around images is much-requested, but this can be accomplished with a stream.)

If a frame–based workflow is what is required, then there are several options that are available. One is to return to FrameMaker. Just because it is no longer in production does not mean that you cannot continue to use it. Another option is to look at other frame-based word processors, such as KWord or Pages. (I know that Pages is not wholly frame-based, but it can operate that way.) A third option is to switch to a formal desktop publishing (DTP) and/or layout application. QuarkXpress and Adobe InDesign are two excellent applications; Scribus is an open–source DTP solution as well.
I only partly agree with what you say. I think your comment on Framemaker does in some way miss the point. Of course, Framemaker has a different approach, but what you do is to reduce it completely to its frame-based approach. There are are many things about Framemaker which are not directly related to this approach and which any other word processor could adopt or at least learn from. Its ability to bundle several documents into a book is one, the whole styles system and everything related to cross refs has nothing to do with its frame based approach.

In my personal experience, no other word processor comes even close to match Framemaker when it comes to structural writing of long documents and there are many things Mellel could adopt or at least learn from.
TLS
Read the guide!
Posts: 42
Joined: Sat Oct 22, 2005 1:48 pm

Re: Some comments regarding feature requests

Post by TLS »

simifilm wrote: In my personal experience, no other word processor comes even close to match Framemaker when it comes to structural writing of long documents and there are many things Mellel could adopt or at least learn from.
I agree, but (sadly) I find myself agreeing with Patrick's contention that Mellel is something different. Ideally I want a product that is as powerful as Framemaker for long docs, the design functions of InDesign, and the multilingual capabilities of Mellel. I have come to the conclusion that such a product will likely never exist.

Framemaker is the better product but it is dead on the Mac and barely alive on Windows and neither can handle Unicode. InDesign can just barely handle footnotes, but not easily do cross references or right-to-left scripts (unless you buy the very expensive Middle East version). Mellel is great for what it does, but still cannot handle OpenType fonts completely, has a very confusing user interface, is limited in its implementation of styles, and is generally "crippled" in a number of other regards (assigning languages to text for spell checking and hyphenation, poor/difficult implementation of tables, and others); basically I consider it a suped-up version of TextEdit that is wonderful for writing a nice letter or even an article but not something for doing book layout in.

While disappointing, in the end that expectation was perhaps too much on my part. Many of us are in fact looking desperately for a Framemaker replacement and such a product is not on the horizon. Mellel comes closest but is too limited and I really don't see it going in that direction much further.

Anyhow, while this may sound like a criticism of Mellel, it in fact is not. Rather it is the recognition that Mellel is not a software package at the level of Framemaker or InDesign (or even Word), but it does what it is intended to do, limited though that may be, very well. So I will save Mellel for articles and letters, Framemaker for complicated books, and InDesign for less complicated books, using the right tool for the intended result. Unfortunately whatever the choice is, sacrifices to what I would like to do have to be made (no Arabic on Framemaker or InDesign; no cross-references in Mellel or InDesign, etc.), but that is nothing new.
zoul
Knows everything, can prove it
Posts: 120
Joined: Thu Aug 10, 2006 1:48 pm
Location: Boskovice, Czech Republic
Contact:

Re: Some comments regarding feature requests

Post by zoul »

TLS wrote:(…) basically I consider it a suped-up version of TextEdit that is wonderful for writing a nice letter or even an article but not something for doing book layout in.
But that’s exactly the point: Text processor is not something anybody should use to create book layouts! Word tried to follow that path and now it’s a monster that is not a good DTP application nor a good text processor. Mellel is almost perfect for writing books and other structured texts and I hope that it will not attempt anything more than that, because such attempts simply do not seem to work.
ptram
Knows everything, can prove it
Posts: 105
Joined: Tue Nov 14, 2006 9:01 pm
Location: Recanati, I

Re: Some comments regarding feature requests

Post by ptram »

TLS,
TLS wrote:Mellel comes closest but is too limited and I really don't see it going in that direction much further.
I still keep Mellel 1.9.1 in my hard disk, and sometimes open it to see how much Mellel has grown in about one year. It is for sure growing. Not at a fast, nor a slow pace. Just at the right speed, to preserve its impressive reliability.

So, I'm confident we will see a much more powerful, yet still reliable, application next year, and the year after. It will not become the monster that Word is, because the Redlers are not selling the number of features, but their quality.

Best regards,
Paolo
mkenney
Read the guide!
Posts: 41
Joined: Thu Nov 03, 2005 1:56 pm
Location: Greenville, SC
Contact:

Post by mkenney »

rpcameron:

Recently, I have made a concerted effort to begin using DevonThink (after trying out Journler for a month), which forced me to rethink my use of Mellel. Your original post helped me to crystalize my thoughts. Mellel, I believe, is extremely well-suited for producing text oriented documents that one intends to share with others (handouts for students, letters, articles, books, etc.). When I forced it to do other things (e.g., to function as some type of note-taking application) I became quite frustrated. Now I know why. So, I've stopped. I am much happier with Mellel now. I have a better idea of how it should fit into my scholarship and professional duties, and I call upon it only when its unique features are needed.

I would recommend, as you do (at least implicilty), that users of Mellel recognize that it a word processor and not to expect it to perform functions better suited to other types of applications. One should, as a consequence, fit Mellel into a collection of applications that complement it and not desire that it become mutli-focused cure all. A paraphrase of Plato would perhaps explain my point better: each application should perform that task it is naturally suited for and not meddle in the task of another application. (NB: this is a paraphrase of the definition of justice in the soul and city that is given in the Republic). Or, to speak colloquially, I would rather have a master of one trade, than a jack of all.
DylanMc
Got the styles thing figured out
Posts: 5
Joined: Fri Mar 10, 2006 11:04 am
Location: Portland, OR
Contact:

I'm a bit depressed...

Post by DylanMc »

Even though the posting at the head of this subject says "If a feature is missing from Mellel, then I can fault no one for wishing for its inclusion in a future version," I can't help but think that my very own FrameMaker-related feature request is among those that the topic author considers "out of scope" for a "word processor".

I still miss cross references and elegant figure inclusion, I paid for Mellel basically on the promise that they were coming, I wrote a posting almost a Year Ago, that posting (http://forum.redlers.com/viewtopic.php?t=338) which is now deleted. Maybe because it included a "do you miss Frame" poll?). Things seem to be drifting in a direction away from my personal needs.

Oh well. I currently use Frame on VirtualPC, LaTeX, Word and Pages. Mellel had promise to replace at least two or three of these for me. For now, it replaces none of them.
mkenney
Read the guide!
Posts: 41
Joined: Thu Nov 03, 2005 1:56 pm
Location: Greenville, SC
Contact:

Post by mkenney »

DylanMc:

I would say that cross-references are a necessity for the kind of word processor that Mellel purports to be. In other words, any word processor intended to handle long, academic documents should have a well-executed cross reference feature.

Figures don't show up much in my scholarship, Greek philosophy (in fact I've never used one), but I do understand why some would need a figure placement feature.

Perhaps the great variety of feature requests points to a change in academic publishing. I know that I have had to submit work in PDF format with the expectation that all required formatting would already be completed (for a conference proceedings). Not every publishing house or journal requires that, but some do. So, the private individual has a need for an elegent and easy to use application that can accomplish their formatting needs. This is just a thought, and it may be way off base, but I'm just trying to think of why someone might need Mellel to have features traditionally found in a layout app.
rpcameron
Knows everything, can prove it
Posts: 980
Joined: Wed Oct 26, 2005 12:48 am
Location: IE, CA, USA

Post by rpcameron »

I started this thread about a week ago, and now that there has been some response to it, I’d like to respond to some points that have been raised, as well as clarify a few things.

I does sound petty to discount feature requests simply because there are better–suited to a text editor or layout application. However, for many of those features, I have not heard any compelling arguments that seem to warrant their inclusion in a word processor instead of the text editor or layout application from which they originated.

Frames: I understand that many users want free–floating frames that can contain a stream of text, tables and/or images. In addition, these frames should either be anchored to the page (meaning they are defined as attached to a particular page, off-set by an x and y value from some origin), or they are to be anchored to a specific point in the text. Furthermore, the frames should allow a continuous stream of text to flow around the frame, regardless of whether it is tied to an absolute position on the page or a position in the stream of text.

As a word processor, I can understand the desire for inclusion of images, to be anchored to a specific point in the text stream. For a more attractive presentation, text should be permitted to flow around the image. However, I do not understand the need of a word processor to include free-floating frames that support arbitrary elements. This is something I feel is part of a layout application; I have yet to hear any compelling arguments for the support of this feature in a word processor. In fact, Word’s inclusion and implementation of this feature is a distinct reason to not include this feature. (Pages includes this feature, but Pages was also engineered from the beginning to be more layout oriented; KWord itself is frame–based from the beginning.)

Long document management: This is indeed something that FrameMaker did exceptionally well. I feel that this feature is something Mellel could benefit from. Also, Mellel’s move to XML and a bundled file format makes it even easier to implement on a conceptual level. Perhaps this is something 2.2 or 2.5 could see ushered in.

Cross–references: I'm not familiar with how FrameMaker implemented cross–references. However, I do know that the next version of Mellel will include them. I am certain that when they make their appearance we will all be pleasantly surprised.

OpenType support: Outside of Adobe, Mellel’s OpenType support is some of the best on the Mac. However, there is still work to be done in this area. We are all aware of its excellent handling of RTL scripts that few, if any, other applications on OS X can match. However, there is still work that can be done. The recent discussions relating to small caps (with mixed case) is one example. Another is the poor Devanagari support. Linking language–specific features to a language meta-class or meta-style would be nice as well.

Those are just some points that I feel Mellel can take from other applications and improve upon. However, I am still missing compelling arguments for other requests. As mkenney stated, I feel that Mellel should strive to become the master of its trade. Layout and plain text manipulation are two things that Mellel may do well to dabble in slightly, but full inclusion of them is something that ought to be carefully considered.

Perhaps its my workflow that stems from an early Unix mentality, but I agree that there should be one application that does one job well. If you need another job, you use another application suited to it. OS X has attempted to embrace this with the Services menu. Also, look at Mail, Address Book and iCal. All three applications are separate and do their respective jobs; however they all interoperate on some level—and that level is increasing with each OS revision.

(Also, I agree with the calls for a revision in the interface, too. I do not like Mellel’s brushed metal interface with an Aqua look to it; it needs to be redone to stay in keeping with the rest of the interface changes going on in OS X … at least, only the good changes. :-)
— Robert Cameron
zoul
Knows everything, can prove it
Posts: 120
Joined: Thu Aug 10, 2006 1:48 pm
Location: Boskovice, Czech Republic
Contact:

Post by zoul »

Maria wrote:Certainly not. There are new applications around that get rid of old paradigms of which kind of software has to do which job. Instead new software is created while listening to users thus serving their needs. When people on this forum offer their experience and tell what they need, no other user should tell them to keep their mouth shut. Good developers will be happy about these contributions and will try to create software for usage scenarios.
Please, let’s not let the discussion drift into personal tone. There is a lot of good, sensible arguments that speak in favor of the “one job, one tool” paradigm. For example, thinking about a document that is being typeset is different from thinking about document that is being written – that is why there is no application that would be great for writing and at the same time great for typesetting. (Sure you can find users that do both in one application and are happy, but that does not change anything.)

“Getting rid of old paradigms” could end up as reinventing the wheel. In this case, it could end up by finding out that an application with so many features that tries to be everything at once is hard to maintain, hard to get conceptually right (“what way are the users going to think about this?”) and eventually gets painful for users. For me, a perfect example of this is Word, which is not a good word processor nor a typesetting program.

Thus, if some people want Mellel to be a perfect word processor, they have objections to features that do not belong (for a good reason!) into a word processor. It’s nothing about telling anybody to shut up – it’s just a discussion.
nicka
Knows everything, can prove it
Posts: 677
Joined: Thu Oct 20, 2005 2:55 pm
Location: Oslo
Contact:

Post by nicka »

One tool for the job. Yes. But what is the job? At the moment there is no tool on OS X suitable for producing a cleanly laid out scientific thesis or for technical writing. When Mellel gets cross-references there will be one.

The scope of the job is wider than word processing. At the beginning of the writing process, Mellel offers outlining. This is as it should be. I have tried sketching out ideas in Scrivener, an interesting application, but it isn't suitable for what I need because it can't import text with footnotes and because it doesn't talk to Bookends. My notes for a paper or lecture tend to contain a bunch of references and a lot of footnotes. So I start in Mellel (after scribbling on paper) and I am happy that its outlining works well. Other features for this stage of writing would be welcome.

At the other end of the process, I may or may not need to produce a final layout. More often than not I do. My thesis, working papers, some papers for edited books, lecture notes and (I have heard) text-books all need laying out. Submissions for journals generally don't but they are a small percentage of most academics' output.
It is impractical and unnecessary to move the whole project to another application for layout. Impractical, because you can't just pour the text from Mellel into InDesign or the like. For a start, no layout application supports Mellel's versatile note-streams (or, soon, cross-references). Unnecessary, because the aim is not an art book or a magazine, but a cleanly laid-out version of what I have.

In short, because of Mellel's market and because Mellel does things its own way and does them better, it needs to work for a larger part of the writing process than a conventional word-processor.
(There is a similarity with Pages, also not a conventional word-processor, but a writing tool better suited for the production of the documents most home users actually use Word for -- things like newsletters, wedding invitations, personal and small business letters, all heavy on layout. It does a different job from Mellel, but the reason for integration of writing and layout is similar.)
noumenal
Got the styles thing figured out
Posts: 10
Joined: Tue Nov 21, 2006 7:26 pm
Location: London

Post by noumenal »

nicka wrote:It is impractical and unnecessary to move the whole project to another application for layout. Impractical, because you can't just pour the text from Mellel into InDesign or the like. For a start, no layout application supports Mellel's versatile note-streams (or, soon, cross-references). Unnecessary, because the aim is not an art book or a magazine, but a cleanly laid-out version of what I have.
I'm just curious: what kind of layout do you need, for the kinds of projects you list, that can't already be accomplished in Mellel?
nicka
Knows everything, can prove it
Posts: 677
Joined: Thu Oct 20, 2005 2:55 pm
Location: Oslo
Contact:

Post by nicka »

what kind of layout do you need, for the kinds of projects you list, that can't already be accomplished in Mellel?
The things I am concerned about are mainly to do with Mellel keeping track of structure, and allowing that structure to be laid out flexibly, rather than with the layout capabilities per se, at least for me.

Cross-references are top of the list.
The ability to have endnotes at the ends of sections or chapters is necessary for some projects.
A similar ability to have the bibliography collected per chapter would be handy too.
It would be nice, but not essential for me, to have text flow around diagrams. I know that technical writers really need this, and probably also text boxes and unequal columns.

In practice, when cross-referencing arrives, I intend to move all of my work - thesis, lecture notes, papers, the lot - to Mellel and see if I can make do with the layout features it has.

It's a slightly different topic, but I'd also like to see Mellel become truly multilingual, with proper support for laying out top to bottom, right to left East Asian text, complete with furigana/ruby. (See other threads in this forum.) But for me at least, that's for fun, not for work, luckily.
Post Reply