This week brings news of the release of Eudora 4.3, along with an exclusive interview with Steve Dorner, Eudora’s primary author. We also ask your opinion of the concept of viewing ads in exchange for getting commercially available features for free. Matt Neuburg weighs in with a review of Deneba’s graphical Swiss Army knife Canvas 7, we announce BoxTop Software’s ProJPEG Photoshop plug-in for saving optimized JPEG files, and we sum up last week’s poll.
ProJPEG 4.0 Adds Compression Control
ProJPEG 4.0 Adds Compression Control — BoxTop Software has released ProJPEG 4.0, a Photoshop plug-in for saving optimized JPEG images. Like BoxTop’s PhotoGIF (see "Crunch GIFs Quickly with PhotoGIF" in TidBITS-479), ProJPEG’s image compression outperforms Photoshop’s built-in filters, creating smaller files for the Web. Version 4.0 adds a Target feature, makes ProJPEG calculate the amount of compression required to meet a file size of your choosing. You can also apply different settings to areas of the image marked as foreground and background (enabling you to apply more compression to a photo’s landscape and less to objects in the foreground, for example). ProJPEG 4.0 is a 580K download and costs $50; owners of any previous version of ProJPEG can upgrade for $20. [JLC]
Poll Results: Macintosh Ownership
Poll Results: Macintosh Ownership — Steve Jobs’s Macworld Expo 2000 keynote trumpeted Apple’s impressive ownership numbers: roughly a quarter of iBook buyers and half of iMac buyers were new to the Mac. But Apple’s repeat customers have traditionally been the company’s ace in the hole, buying Macs even when Apple’s future was murky. So we asked: "Over the years, how many Macs have you personally bought for your individual use at work or at home?" Of more than 2,300 respondents, nearly two thirds indicated they had purchased between two and five Macs over the years. Other responses fit nearly along a bell curve, with the notable exception that about 8 percent of respondents indicated they had purchased ten or more Macs for personal use over the years. These results would seem not only to reflect the loyalty of Apple’s customers, but also to reinforce Apple’s sales figures showing that the majority of Macintosh sales still go to existing Apple customers. Apple definitely needs to grow the size of the Macintosh market by attracting new converts to the platform, but the company has to ensure that initiatives aimed at attracting converts don’t come at the expense of the more valuable long-time customers. [JLC]
Poll Preview: Ad-ing It Up
Poll Preview: Ad-ing It Up — This week’s release of Eudora 4.3 marks the first time a major program has essentially offered to let you use commercially available features for free in exchange for viewing ads. You’ll read below about why Steve Dorner and Qualcomm felt they had to add Sponsored mode to Eudora 4.3’s Paid and Light usage modes, but what do you think? Do you like the option of being able to trade some screen real estate for an ad window in exchange for features that would otherwise cost money? Note that we’re not asking if you think ads in software are inherently good or bad, but instead if you like the option of being able to see ads instead of paying for the software yourself. Also, keep in mind that we’re asking about applications in general; at the moment, we have only the $50 Eudora as an example of this, so also consider how you might feel about programs that cost either $10 or $200 and that you used either occasionally or every day. Register your opinion on our home page (and make sure to scroll down if you can’t see the poll form in your window)! [ACE]
InterviewBITS with Steve Dorner
By the time you read this article, Qualcomm will have released Eudora 4.3, which introduces a new business model for application software by adding an optional mode in which people can use all the features of the commercial version of the popular email program for free in exchange for viewing ads. The upgrade is minor if you’re a Eudora Pro 4.2 user (see our "Eudora Pro 4.2" series of articles for more information), and because of that, Qualcomm has made the upgrade to Eudora 4.3’s "Paid mode" free for Eudora Pro 4.2 owners. Current Eudora Light 3.x users can enjoy a much more significant upgrade, since Eudora 4.3’s "Light mode" offers many of the benefits of the two years of development that’s gone on since Eudora Light was last updated. And "Sponsored mode" provides access to all of Eudora’s commercial features for free. See "Eudora 4.3 Public Beta Adds Free Usage Modes" in TidBITS-509 for details.
System requirements for the $50 Eudora 4.3 include a PowerPC-based Mac with 1,800K of RAM running System 7.1.2 or later with the Text Encoding Converter. The download is likely to be about 6 MB. Note that what Qualcomm is releasing Tuesday is the full version of Eudora 4.3; if you’re an existing Eudora 4.x owner, an updater that brings you up to Eudora 4.3 and registers you in Paid mode will be available later this week.
Get the X-Eudora-Settings List via Email — To simplify access for readers of my Eudora Visual QuickStart book, as well as others using Eudora Pro 4.2 or Eudora 4.3 for the Mac, I’ve created an email auto-reply that contains the full list of x-eudora-setting URLs you can use to tweak Eudora’s myriad of hidden settings (see "Eudora Pro 4.2 Continues to Deliver" in TidBITS-489). Qualcomm makes the same information available on their Web site, but since x-eudora-setting URLs are useful only from within Eudora, I felt it would be easier to get the 80K list via email, and Qualcomm granted me permission to make them available in this fashion. I’ve also added information at the top of the file that explains how to use them and pulls out a few especially useful settings.
On to the Interview — This week we have an exclusive interview with Steve Dorner, who created Eudora back in 1988 and who has continued to drive the program forward, though ably assisted by others at Qualcomm these days. Steve managed to squeeze some time free from the hectic final days that surround any software release to chat via email.
[Adam] The major change with Eudora 4.3 is that it adds a new advertising-supported mode to the two previous ways you could get Eudora: the free Eudora Light and the commercial Eudora Pro, all now bundled together into a single program. Is this giving in to commercialism on the Internet?
[Steve] In one sense, it is giving in… to reality. That reality is that many people prefer free stuff with ads to stuff that costs money. How many times do you see people at trade shows saying, "No, thanks, I prefer my t-shirts without your advertising on them. Here’s $10."
In another sense, it’s not giving in at all. People still have the same Eudora choices they had before; pay and get the full features, or don’t pay and get fewer features. The difference is that now there’s a third choice – get all the features but don’t pay, and see ads instead.
[Adam] What was the primary reason you chose to follow this path with Eudora?
[Steve] Let’s face it, people have a lot of choices out there for free mailers. We think Eudora provides unique advantages, but some people just can’t get over the price tag. This is a way we can eliminate the price tag but still afford to produce the software.
[Adam] Are there significant new features in Eudora 4.3?
[Steve] That depends on the audience you’re considering. For Eudora Pro 4.2 users, the feature differences aren’t all that great. There is the link history window, the ability to remember addresses you type or reply to, and a few other small features.
For Eudora Light users, the upgrade to Sponsored mode is huge. People using Sponsored mode get all the Eudora Pro features; spell-checking, styled composition, inline images and movies and sounds, HTML display, toolbar, powerful filtering, etc., etc. And even if they don’t want to see the ads, Eudora 4.3 running in Light mode is much more powerful than the old Light was.
[Adam] Are there any major differences between the Mac and Windows versions of Eudora now, or are they pretty much in parity?
[Steve] They reflect the different emphases of the two platforms and the separate development teams who have worked on them. I very much believe that a Macintosh program should be a Macintosh program, and similarly that it pays to do what Windows users expect on Windows. Whether in Rome or Little Rock, you have to fit in with the natives.
One thing I find amusing is the mail we get from people who say "I wish you guys spent as much time on the Mac version as you do on the Windows version," which is evenly balanced by the mail we get from people who say "I wish you guys spent as much time on the Windows version as on the Macintosh version."
[Adam] My impression is that Eudora is often chosen by individuals and by organizations looking for an Internet email solution, but that it’s losing ground in large organizations where Microsoft Outlook offers scheduling, contact management, and group conferencing features. Do you envision moving Eudora more in that direction in the future?
[Steve] Yes, we are actively looking at ways to do more with schedules and contacts and the like. We think we’re in a position to do some very interesting things, especially in conjunction with the new wave of wireless devices.
[Adam] Was that some of the rationale behind Qualcomm’s purchase of Now Software and the code that turned into the ill-fated Eudora Planner?
[Steve] I’m glad that Now Software’s products are being carried forward by Power On Software. That’s all I have to say on that topic.
[Adam] Fair enough. It’s long been thought that Qualcomm bought Eudora originally because someone was thinking ahead and saw the convergence of email and wireless communication. But that happened in 1992 – was Qualcomm simply too far ahead of the times, or was there something else going on?
[Steve] I actually think that Qualcomm acquired Eudora because it used Eudora and wanted to see the product continued and improved. Way Back Then, companies actually wrote their own software sometimes.
Of course, some people undoubtedly had some ideas in the back of their minds, but it was really a very practical decision at the time.
[Adam] Assuming that convergence is at some point inevitable, when do you think we’ll see it taking place for a significant number of Eudora users?
[Steve] It’s hard to say. The phone market right now puts heavy pressure on (big shock) cheap or free telephones. Smart phones are a little more expensive and are a hard sell.
One thing to realize is that ordinary folks like you and I are not the "real" customers for cell phones. The "real" customers are the local cell phone carriers; they’re the ones who control which phones people can buy. And they want mass-market phones so they can get lots of users, and they don’t want to "waste time" supporting really smart devices.
So it’s going to take a while for really smart phones to go anywhere, and hence for Eudora on a phone to really go anywhere.
[Adam] Still, you’re obviously making inroads in that direction all the time. What about Qualcomm’s PureVoice plug-in that enables people send and receive voice messages with Eudora?
[Steve] What your readers may or may not know is that PureVoice is the same voice technology used in Qualcomm’s CDMA digital cellular phones. We hope to see a day where PureVoice will make truly efficient telephone to computer communication very easy to do. Imagine being able to record a message on your phone and email it to someone, for example. Or to easily pick up your voice mail with your email program.
[Adam] Moving away from cell phones, what are the major problems you’re seeing in email today that you think email programs can attempt to solve?
[Steve] Volume is obviously one of them. People get more and more mail, and have to deal with it. But rather than worry about solving problems in email, I’m more excited by providing tools so that people can solve problems using email. Eudora 5.0 is going to have some fun things in that direction.
[Adam] Fun is good, especially in a program that people use all day long and despite the sticks-in-the-mud who think email should be serious and professional. I still miss the dialog box that said "You may as well stop typing now because no one is listening." Any other easter eggs you’d like to share? Should someone make a Eudora plug-in that puts a few of the cute little bits back in?
[Steve] We’ve been thinking about such a plug-in. 🙂 As for easter eggs, the "You have no new mail" envelope that appears if you have Eudora set to display alerts after checking for mail actually has words in it that can be read, if you work at it. Note that it went in at about the time a lot of the fun went out.
[Adam] With Mac OS X due to ship sometime this year, what level of support do you plan for Eudora? Classic, Carbon, or Cocoa?
[Steve] That’s up to Apple. Cocoa is a total rewrite for a Macintosh application, and we don’t anticipate doing that initially.
We’ve been working on Eudora under Carbon for quite some time, but the brutal fact of the matter is that Carbon hasn’t been ready for us. There are many things that Eudora does that Carbon just doesn’t do at all, or that don’t work. For example, Eudora can be smart about what to do when you’re on battery, but Carbon doesn’t currently provide access to the power manager. For a while, we kept getting "Tough luck, applications shouldn’t need to do that" responses when we asked about stuff like that, but things seem to have improved lately.
[Adam] You’ve been developing Eudora for over 12 years now, so you’ve seen a wide variety of Internet technologies come and go. What I’d like to do then to finish off is play "Technology Association Test" – I’ll give you a technology and you tell me the first thing that comes to mind.
XML (eXtensible Markup Language), the much-hyped markup language for creating other markup languages that can define meta-information about data):
[Steve] Sound and fury…
HTML mail, which lets people send HTML styled messages:
[Steve] Deal with it.
LDAP (Lightweight Directory Access Protocol), which provides directory lookup services:
[Steve] Vomit. It’s really too bad that XML wasn’t all the rage in time to forestall this idiot OSI protocol from proliferating.
IMAP (Internet Message Access Protocol, at least with version 4 – it used to be Interactive Mail Access Protocol), which changes the mail storage model so all mail remains on the server, rather than being downloaded to the client, as with POP.
POP (Post Office Protocol), the simple and efficient method of retrieving mail that’s still by far the most common on the Internet:
[Steve] Mom and ___.
APOP (Authenticated POP), which encrypts the otherwise clear text passwords used to login to a POP account as a way of increasing security.
[Steve] Doomed. Few sites are going to change their authentication databases for APOP, cram-md5, or anything else. They’re going to stick with plain text and just run everything through SSL when RSA’s patent expires in September.
SMTP (Simple Mail Transport Protocol), one of the basic protocols on top of which all Internet email runs.
SMTP AUTH, which is an authentication extension to SMTP that ensures only people with the proper access can send email through an SMTP server as a way of preventing spammers from hijacking servers.
[Adam] Thanks again for taking the time out to talk, and best of luck with Eudora 4.3 and all the fun stuff you have planned for Eudora 5.0.
[Steve] Glad to help – now it’s time to get this new version uploaded.
Canvas 7 – Not Quite Heaven
For the past six years, my relationship with Deneba and its combination draw/paint program Canvas has been something of a roller-coaster ride. With Canvas 3.5, I was contented: it was an excellent replacement for my old favorite SuperPaint, which was starting to show its age; and, Canvas’s splendid import/export capabilities made it useful in many difficult situations. With Canvas 5, I was disgusted: it was slow, it felt like a port, it was unstable, the interface was incomprehensible. With Canvas 6, I was surprised but once again contented: the program had been thoroughly cleaned up and was now a worthy successor to version 3.5. Although it lacked the power of either CorelDRAW or Corel PHOTO-PAINT, which I had reviewed shortly before, Canvas 6 was simple and convenient, satisfying and delightful for general use; and I found myself turning to it often.
Now, with Canvas 7, I’m still contented but also puzzled. Its advances over its predecessor are real, but so miscellaneous that the program seems to have found its feet while losing its way. Bare enumeration of new features would be in vain; you can consult a list at Deneba’s Web site. Instead, I’ll describe their general categories, and try to assess whether the changes are moving this program in the right direction. (It might help, at this point, to read my earlier reviews and Deneba’s feature list, so as to get a sense of where I’m coming from.)
Cleaning Up the Workspace — On startup, Canvas 7 shows no obvious changes; there’s nothing like the dramatic alteration of the workspace that marked the leap from Canvas 3.5 to 5, or even 5 to 6. But you soon discover numerous small interface tweaks. To change the view’s zoom factor numerically, you can now type directly into the window’s zoom value box, without having to bring up a palette. You can rotate the window for a better view. Aligning of objects and the "smart mouse" are now accessible via contextual menus. Guidelines can now be positioned numerically.
Bezier drawing is better on several counts. Colored nodes and anti-aliased drawing make curves far easier to see. Selecting a node now reveals not only its handles but those of the two adjacent nodes as well. The number of nodes can be automatically reduced. It is at last possible to draw all or part of a curve live by fitting it to points.
Breaking Down the Barriers — One of Canvas’s greatest strengths has always been how it combines on a single page three types of objects: vector (draw), text, and bitmap (paint). Previous versions did much to reduce the conceptual barriers between these object types, and Canvas 7 does even more.
For example, in the past, while you could create a bezier curve by drawing freehand, you were then entrenched firmly in the bezier world, forced to work with nodes and handles to reshape the curve. Now a new tool lets you reshape it by continuing to draw freehand, and another lets you pull a point on the curve as if it were a rope.
Similarly, with a bitmap object, you could previously use vector tools to draw a shape, but to select a non-geometric area, if color-based tools couldn’t handle it, you had to draw freehand. Now bezier shapes can be converted to selection shapes (and vice versa).
Deneba has lifted major limitations on transparency modes introduced in Canvas 6. For example, if you added directional vector transparency to an object, you got a simple gradient running from one end-node at 0 percent to the other at 100 percent; you could move the end-nodes, but that was all. Now you can interactively add intermediate nodes and dictate their transparency values.
It was previously possible to turn a vector or text object to a bitmap, but now such objects can receive filters and adjustments previously reserved for bitmaps, such as blurring, posterizing, and color shifts, without turning them to bitmaps; they keep their editability as vector or text objects. In essence, the object is seen through its own private paint-type filter; so it isn’t surprising that Canvas 7 also introduces lenses.
Weaving the Web — Gone is the objectionable Java-based Colada; Canvas now generates HTML documents directly. But this HTML relies on CSS level 2 absolute positioning, so (unlike tables) it can’t be handled by older browsers, nor even by the latest version of Netscape Communicator unless you render text as images.
To export as GIF or JPEG, you no longer need to choose one or the other and then pass through a sequence of confusing dialogs; instead, you make all your choices in a single export dialog. This export dialog isn’t any easier to understand, but it is certainly more convenient, especially since, along with new optimization abilities, it previews up to four different export options simultaneously. Also, it is now possible to rotate a bitmap and retain that rotation when exporting. Thus, one no longer needs another application to post-process a JPEG made at the wrong compression, or from an inverted scan.
A new document type permits creation and editing of animated GIFs, with frames represented by "pages;" you can view several successive frames simultaneously, but amazingly there is no way to play the animation without exporting and switching to a Web browser, and there are no built-in transitions such as wipes and tiles, even though QuickTime is present to help create them.
Reading the Tea Leaves — Canvas 7’s documentation is complete and compact, though I often find the verbiage confusing. An accompanying CD of narrated screen shot animations is very helpful. Online help is good, but balloon help or more copious tooltips would be better: many important options in palettes are represented only by uninformative icons.
Canvas 7 is definitely a better version of what Canvas 6 was, a pleasantly straightforward all-purpose draw/paint program, adequate to the needs of most users. As such, I enjoy it, I turn to it often, and I recommend it.
On the other hand, I feel that Canvas 7 doesn’t introduce new features so much as it cleans up and rationalizes Canvas 6; Canvas 7 feels good mostly just because there were places where Canvas 6 felt clumsy or inconvenient. It’s debatable, therefore, whether Canvas 7 really warrants either the whole-number version increment or the $100 upgrade fee – especially since Deneba has failed once again to attend to some of Canvas’s most basic features. The delay between clicking on a toolbox icon and the popping-up of its sub-palette is still too long. There’s still no keyboard navigation for moving from one bezier curve node to another. The program isn’t scriptable. Too many features are available only through modifier keys (Option-choose this menu, Tab-click this node), making the program hard to learn. And the painting tools are still crude; with Canvas, you could exhaust a whole day and all your patience before you’d achieve the painterly effects you can get with a single stroke in Corel PHOTO-PAINT, with its options for brush and edge texture, bleed and sustain, dab spacing, stroke smoothing, color shifting, and so forth.
This raises the larger question of what an all-in-one program should be. Canvas ranges wide but is spread rather thin. It lags behind the individual competition in every department it has chosen to enter: in painting, page layout, Web images, Web pages, and GIF animations – even in certain aspects of drawing – other specialized applications are ahead, sometimes far ahead. Canvas 7 now seems to be playing copycat, trying to catch up in every area at once, with results that are sometimes inadequate (like the GIF animation) and sometimes inappropriate (like the new FTP capability). Deneba might profitably refocus Canvas on achieving greatness in a few well-defined areas, rather than risk mediocrity in many. Canvas isn’t QuarkXPress, GoLive, WebPainter, or Anarchie, so why pretend? In my view, Deneba should return its attention to achieving its basic mission with Canvas, excellent drawing excellently combined with excellent painting.
Whether Deneba can afford not to do this is really a question of whether users will continue to find that the convenience of a single general application outweighs the inability to accomplish what’s perfectly possible with several individual dedicated ones. This is the common quandary of "Swiss Army knife" applications. If users become so frustrated by Canvas’s inability to play its own GIF animations, or to export text objects as HTML text that browsers can actually render, or if they find themselves repeatedly turning to other paint programs to achieve the fine control that Canvas denies them, they may wonder what real advantage Canvas offers. There is also, to be sure, the matter of price; and Deneba’s new aggressive policy of offering Canvas as a download for just $130 is commendable. But even this policy could backfire, since it could cause long-time Canvas users to feel understandably enraged at being asked to pay almost as much ($100) to upgrade. Also, Corel is fighting back by offering the CorelDRAW and Corel PHOTO-PAINT package, in a slightly crippled form but without time limitations, absolutely free.
Canvas 7 requires a PowerPC-based Mac with Mac OS 8.5 or later and 32 MB of RAM with a monitor running at 16-bit color and with at least 800 by 600 resolution. A standard installation requires about 80 MB of disk space. Canvas 7 costs $375 with CD-ROMs, printed manuals, font and clip art collections, $130 without, and Deneba makes a free 15-day trial version available.