The LWN.net Weekly Edition for November 21, 2002 is available.

Book review: Linux In The Workplace

We have not managed to write a great many book reviews over the last couple years - they take a lot of time, and were never our most popular feature. Linux In The Workplace, however, is sufficiently interesting that we took the time to read it through. Read on for our impressions and thoughts on why this book is worth noting.

Linux In The Workplace is published by the Linux Journal Press. Interestingly, no authors are named on the cover; instead, it is credited to "SSC, publishers of the Linux Journal." It is, in fact, the result of the Linux Journal's staff's experience with running a Linux-based office over the last few years. As a result, it is well grounded in a lot of real-world, Linux-based office work; it is also deeply tied into the Journal's way of doing things.

The cover does not go out of its way to make it clear, but this book is mostly about KDE. GNOME-based applications are mentioned in spots, but anybody wanting to set up an office around the GNOME desktop will not get what they need from Linux In The Workplace. There is no problem with this - trying to cover both desktops would likely turn the book into a confusing mess - but it's a good thing to be aware of.

Actually, anybody wanting to "set up" an office around any desktop will need to look elsewhere. Linux In The Workplace is very much a user's manual; it expects that somebody else has already gone through the trouble of installing Linux and making it work:

Again, that is appropriate; serious use of Linux in offices is only feasible if most users do not have to deal with the administrative issues - something which is also true of Windows in the office.

So, what's covered in this book? After a quick "what is Linux" chapter, we learn how to log into a KDE-based system, deal with user accounts ("A good password combines upper- and lowercase letters with nonalphanumeric keys. Passwords such as *nCk&Ve or *nG]y$Uds- are good examples."), and deal with the basics of the KDE desktop. The approach is low-level and detailed - we learn about what most of the icons and menus do. Anybody who is used to working with a Linux desktop at all may find the "now click here" pace a bit tiresome, but readers who are entirely new to Linux will likely welcome the detail.

In subsequent chapters, the reader will encounter:

• Chapter 3: a description of Konqueror, but only in its role as a local file manager.

• Chapter 4: "getting organized." Topics like KOrganizer, KPilot, KArm, and KNotes.

• Chapter 5: OpenOffice. OpenOffice is the preeminent free office suite for Linux, and this book recognizes that fact. This chapter provides a whirlwind tour of the OpenOffice applications; it (like much of the book) is more useful for getting an idea of what the application can do than really getting an in-depth understanding. If you want an overview of how the spreadsheet works, this book will help; if you need to learn how to write formulas, you'll need to look somewhere else. (This chapter is available on the net in PDF form).

• Chapter 6: alternative office software. This chapter is a quick overview of KOffice and AbiWord; one gets the sense that the authors expect few readers to go beyond OpenOffice, however.

• Chapter 7: graphics. A quick look at KPaint, Kontour, KView, and the xscanimage tool.

• Chapter 8: the Gimp. There is, of course, no way to do justice to the Gimp in a single chapter; this attempt reads mostly like a quick demo given to somebody who had never seen a serious image editor before.

• Chapter 9: email, netnews, and faxes. KMail is covered in fair detail, though important points are missing. For example, the KMail interface to GNUpg is covered, but GNUpg itself is passed over. There is also an overly scary warning about reading attachments: "Attachments are often the vehicle for transmitting computer viruses that can do great damage to both your computer and any computer to which you are connected. A virus can even attack your address book and send replications of itself to everyone listed." That is a bit strong, given that this scenario has never, to your reviewer's knowledge, happened to a KMail user.

  If you were going to cover a second mail user agent in a book like this, what would it be? The authors chose Netscape mail. Pine, mutt, and elm get passing mentions; evolution does not, for the purposes of this book, seem to exist.

  Quick mention is made of KNode for reading Usenet news and "K Send a Fax" for dealing with faxes. One could certainly imagine other, better established applications in these categories that would have been worth a mention.

• Chapter 10: Konqueror as a web browser. Quite a bit of detail on bookmark management and such. There are passing mentions of Netscape and Opera; nothing about Mozilla or Galeon.

• Chapter 11: customizing the desktop. This chapter will certainly be useful to anybody wanting to tweak how the KDE desktop works.

• Chapter 12: making backups. A quick look at "Ark" and KOnCD.

• Chapter 13: the command line. Only in the very last chapter does this book get around to discussing terminal emulators, shells, and the Linux command line. A quick overview of a number of basic commands is provided. Emacs is covered in three sentences.

There are a number of shortcomings and strange omissions. For example, office workers are likely to want to view PDF files, but there is no discussion of how to do that - even though gv, which reads (most) PDF files happily, is briefly covered. Printing is mostly passed over, as are multimedia applications. And (by design), there is almost no mention of proprietary software packages that can be useful in real office situations.

But, then, few books are perfect. This one is important as proof that you can get a lot of work done on a Linux system without ever having to mess with partitioning menus, shell prompts, mount commands, and so on. The existence of this sort of book is an important prerequisite for widespread adoption of Linux for desktop use. Desktop Linux is, increasingly, being taken seriously; Linux In The Workplace is full of good examples of why that is happening.

Comments (13 posted)

OASIS to create an open office application format

The OASIS Standards Consortium has announced the creation of a "technical committee" which will develop an open, XML-based file format specification for office applications. The goal of the project, of course, is to facilitate interoperability and data exchange between applications. Should they succeed, the days of trying to reverse engineer Word files could come to an end.

It is hard to overemphasize the importance of this effort. Microsoft's office suite monopoly is based on two things: (1) that suite's feature set, and (2) the ability to exchange documents with the rest of the world. There are numerous other office suites which are closing the feature gap (though there is still some ground to cover for the free applications, to say the least). But, without the ability to easily exchange documents with MS Office users (and have them look good when they get there), adoption of alternative office suites will remain limited.

And there, of course, lies the rub. A new, XML-based office suite file format will have a rough life if Microsoft does not play along with it. It is worth pointing out that Microsoft is a member of OASIS; the company has also said that Office will use an XML-based format in the future. But the list of supporting companies in the press release (Arbortext, Boeing, Corel, Drake Certivo, and Sun) does not include Microsoft.

Even without Microsoft, standards for document data can only be a good thing. This particular standard is getting a jump start from Sun, which is contributing the OpenOffice.org format under royalty-free terms (OASIS, in general, is quite happy with RAND or UFO (uniform fee only) terms). Should the committee create a standard based on this format, the existence of a free reference implementation should encourage adoption of the standard in both free and proprietary packages.

Proprietary data formats are a problem for a number of reasons, of which proprietary lockin is only one. Another is the ability of proprietary applications to surprise users by retaining information in documents that those users had thought they had deleted (or never put there in the first place). Future historians will find that much of the documentation of this era is encoded into formats which are no longer readable. An Open format for office information will not, by itself, solve any of these problems. But it sure would be a good start.

Comments (6 posted)

Notes from LWN

There is a relatively small amount of news to report this week. The individual subscriber count stands almost unchanged from last week - which is not entirely a bad thing, since a number of accounts have expired in that time. We will have to figure out a way to bring in the next round of subscribers, however.

We think we have worked out the problem that made it difficult for lynx users to log into the site. Please let us know if it still fails to work for you. (The longer-term task of making the site more lynx-friendly in general remains on the "to do" list).

Next week is the Thanksgiving holiday in the US. We'll be publishing a (perhaps a little smaller than usual) Weekly Edition one day early - on November 27 - because we'll all have eaten too much food to reach the keyboard thereafter. We'll return to our regular schedule for the following week (but there will be no Weekly Edition the week of December 25).

Comments (8 posted)