[LWN Logo]
[LWN.net]

Bringing you the latest news from the Linux World.
Dedicated to keeping Linux users up-to-date, with concise news for all interests


Sections:
 Main page
 Security
 Kernel
 Distributions
 On the Desktop
 Development
 Commerce
 Linux in the news
 Announcements
 Linux History
 Letters
All in one big page

Other LWN stuff:
 Daily Updates
 Calendar
 Linux Stocks Page
 Book reviews
 Penguin Gallery

 Archives/search
 Use LWN headlines
 Advertise here
 Contact us

Recent features:
- RMS Interview
- 2001 Timeline
- O'Reilly Open Source Conference
- OLS 2001
- GaŽl Duval
- Kernel Summit
- Singapore Linux Conference
- djbdns

Here is the permanent site for this page.

See also: last week's LWN.

Leading items and editorials


Red Hat makes a profit - sort of. With some fanfare, Red Hat has announced its results for the first quarter of its 2002 fiscal year. The core point, of course, is that the company claims to have made a $600,000 profit in this quarter, which ended on May 31.

The company made $11.3 million in revenue from subscriptions, and $14 million from the various consulting and development services. There is no separate item for "Linux distribution sales," leading one to believe that they have been folded into "subscriptions." Red Hat also earned almost $4.5 million in interest from its sizeable cash holdings. On the down side, the "cost of revenue" was $11.1 million, sales and marketing ran $10.7 million, research and development $4.3 million, and administrative expenses were $3.4 million.

In other words, Red Hat actually, using these numbers, operated at a $3.9 million loss. It was the interest income that put it nominally into the black.

This is a significant achievement on Red Hat's part. The company had not promised profitability for another year yet, but that state has been reached now. They have gotten there despite a harsh economic downturn that has savaged many other Linux companies. And Red Hat has not drifted from its commitment to free software; the company still employs a great many free software developers, and gives back a great deal to the community. Red Hat shows that it can be done.

Or so we hope. A closer look at the numbers shows a slightly more complicated story than what we have seen so far. The $600K profit comes out of the "adjusted" results. The adjustments that have been made include the omission of three sets of expenses:

  • "Amortization of goodwill and intangibles": $20.8 million.
  • Stock-based compensation: $3.7 million.
  • Merger and acquisition costs: $3.7 million.
Once you figure those in, suddenly Red Hat shows a loss of $27.6 million, or 16 cents per share. The justification for leaving them out is that they are, it is said, one-time charges that do not show the steady state of the company. And one could argue that the goodwill costs, reflecting money that went out the door some time ago, should be left out. But the fact remains that the "total assets" claimed by the company dropped from $505 million at the end of February to $471 million at the end of May. Red Hat may have achieved "operating profit," but it doesn't look 100% sustainable quite yet. At some point the assets need to start increasing.

What about future quarters? During the company's conference call, Red Hat management explicitly refused to provide any sort of guidance regarding future results. They will not even guess, at this point, at whether they can remain profitable or not. These are uncertain times for most companies, but most are sufficiently on top of their situation to offer estimates of how things will go. Red Hat, evidently, has no clue. The company did say, however, that no significant changes in staffing were expected over the next quarter.

So it remains to be seen whether this is the beginning of Red Hat's money-making operations, or whether it is, instead, the company's high point. Red Hat appears to be doing things right in a lot of ways, to the free software community's benefit. With luck, they'll pull it off.

Red Hat to become a database company? One other detail that Red Hat let slip in its conference call was that the company would soon announce a relational database product and associated services. No further information is available from the company at this time; it is making people wait until the marketing people say it's time.

The hope, clearly, is that relational database systems will help drive the next phase of corporate acceptance of free software, and that Red Hat will be able to provide those systems and the services that go with them. It remains to be seen how well that will work; database customers are used to getting their databases and operating systems from different vendors. Red Hat will have to offer something new and compelling to attract customers in this field.

Red Hat will stick with its open source approach for its database offering. Assuming that Red Hat is smart enough not to try to implement a relational database management system from scratch, it will have to adopt one of the existing, free database systems: MySQL, PostgreSQL, InterBase, or SAP DB. Red Hat as been careful, thus far, to not tip its hand regarding its selection.

There are some rumors about, however, that PostgreSQL will be the platform chosen by Red Hat. Certainly it would be a worthy choice; the PostgreSQL team has worked long and hard to produce a top-quality relational database system. Such a move, however, could prove to be a challenge for the companies that are already providing commercial support for PostgreSQL.

The most prominent of those, perhaps, is Great Bridge, the company founded by early Red Hat investor Frank Batten Jr., and the employer of much of the PostgreSQL core team. We talked briefly with Great Bridge CEO Bob Gilbert, who was very upbeat about Red Hat's possible entry into the PostgreSQL market. "What took them so long?" Mr. Gilbert welcomes Red Hat, and looks forward to Red Hat's contribution to PostgreSQL development.

If you believe Mr. Gilbert, the PostgreSQL market is a good one to get into. The company is finding customers in each of several target areas; PostgreSQL is being received well. Larry Ellison and Oracle, he says, should start getting worried "yesterday."

Mr. Gilbert's confidence may well be justified, but Red Hat's entry into the database market still has the potential to shake things up. We'll revisit this topic once the company has made its plans public.

GnuCash and library dependencies, again. Last week's item about the GnuCash 1.6 release and its many library dependencies drew more than the usual amount of mail, including this response from the GnuCash project itself. We seem to have hit a bit of a nerve there. So this week we'll follow up with two more articles; this one looks at the library dependency issue again, and the following one is a quick review of the 1.6 release itself.

Some members of the GnuCash development community felt that the project had been unfairly singled out for criticism when they would rather have seen attention paid to the stable release that they had worked so hard to produce. So let us say it here: GnuCash, at the moment, demonstrates the kinds of problems that can come up with massive shared library dependencies, but GnuCash is not, itself, the problem. GnuCash is a high-quality application which fills a pressing need in the free software community, and it has gotten there partly because its developers have taken the greatest possible advantage of work done by others. We never meant to criticize the project itself.

There are pitfalls, however, with a reliance on large numbers of shared libraries. Especially when a number of those libraries are not widely available on common distributions. If nothing else, it makes it very hard for people to use your software.

In the proprietary world, users will expect to be able to install a new "stable" release of a web browser, mail program, file manager, or personal finance program on their current system. Most do not expect to have to massively upgrade parts or all of their system first. (What happens when the application messes with their system anyway is a different, sad story that Linux users, happily, need not experience).

The people who have reported success with GnuCash 1.6 are, for the most part, running distributions like Debian unstable ("sid") or Mandrake's Cooker. The exception appears to be the just-released SuSE 7.2 distribution. Nonetheless, many users of a personal finance application will not be pleased to have to upgrade their operating systems just to install or upgrade it. If you tell them that, not only do they need to upgrade the operating system, but they must use an unstable version of a distribution, they will simply walk away.

There may be no easy solution to this problem. One of the characteristics of free software is rapid development, and few of us would have it any other way. But fast development implies a lot of upgrades if you want to keep up, and, often, the need to run beta versions of software. These requirements may be hard to reconcile with the needs of desktop users, who just want things to work without their needing to mess with them. This will be a continuing challenge for those developing desktop applications.

About GnuCash 1.6. The GnuCash package has long had the features that one really needs to handle personal finance - see LWN's review from back in 1999. It has, however, remained far behind the proprietary packages with regard to the features offered. People who really want to run free software for [gnucash main window] everything have been able to use GnuCash for some time; just about everybody else has been inclined to wait.

With the 1.6 release, the feature gap is closing. The application as a whole has a much more finished look, and the online help is greatly improved. Quite a few important new features have been added. GnuCash still has not caught up with the proprietary packages in a number of ways, but it has gotten a lot closer. Once the distributions catch up and make GnuCash 1.6 easy to install and run, its user base should grow.

The first thing likely to be noticed by GnuCash users who upgrade to 1.6 is the new, XML database file format. The program converts older files to the new format, but must ask a number of questions in the process - especially if the file contains a lot of stock accounts. The XML format is certainly nice for a number of things, but there is a downside as well: the size of the database file grows by almost a factor of ten. Over 1KB is required for each transaction (i.e. a check) (example). GnuCash 1.6 is noticeably slower to load or save data in the new format. That's the sort of price we pay for a transparent file format.

Of course, for those with huge GnuCash files, taking advantage of the new PostgreSQL back-end may well prove the best way to go.

A crucial improvement in GnuCash 1.6 is in the report generator. Reports in version 1.4 were somewhat crude and could not be directly printed - one [GnuCash report window] had to export the report in HTML and feed it to a web browser first. Version 1.6 reports are more tightly integrated into the system, look better, and include the obligatory bar and pie charts.

GnuCash has always had a strongly international approach and supported multiple currencies. The new version has strengthened that approach, and includes detailed support for the Euro.

Other features include: more business-oriented support (things like depreciation reports), tax preparation support, improved QIF importing, internal updating of stock prices (no more need for an external application), a loan calculator, and even a built-in web browser.

GnuCash 1.6 also includes support for the "GnuCash network." The network does not currently provide much in the way of services; the registration window doesn't even work, and will not until version 1.6.1. One can presume, however, that providing useful services through that channel is part of somebody's business plan, and that things should start showing up there soon.

What's still missing? Many users would most like to see support for scheduled transactions; this feature is apparently under intensive development and should be there for the next major release. It's still not possible to directly import information from banks or credit card companies. No budgeting tools are provided. GnuCash still doesn't really understand loans, and will not handle the amortization for you. And several other things, doubtless.

The GnuCash developers have set themselves the goal of blowing the proprietary finance packages (both personal and business) out of the water with a free alternative. It is an ambitious goal, and, as of 1.6, it has not yet been achieved. Things are clearly heading in the right direction, however; GnuCash is more than usable now. If you are still balancing your checkbook with a proprietary package, it may be time to consider making a change. (See also: the GnuCash web site).

Inside this week's Linux Weekly News:

  • Security: The Danger of Posting Images, a look into the world of 13 year old hackers, new vulnerabilities in fetchmail, rxvt and more.
  • Kernel: Linux Device Drivers 2nd Edition; API changes in stable kernels; FOLK.
  • Distributions: Agenda-VR, Sentry Firewall CD, the Mandrake Cooker Weekly News.
  • On the Desktop: Netscape 6.1, XFree86 reviewed, Ogg Vorbis and yet more on pilot-link.
  • Development: GCC 3.0 announced, Ogg Vorbis decode RC1, Stallman on scientific publishing, mod_lisp 2.0, GHC 5.00.2.
  • Commerce: Compaq initiatives, Goldbox, GCC 3.0.
  • History: Software patents, Handhelds.org and "open sores".
  • Letters: GnuCash and shared library hell; non-executable stacks.
...plus the usual array of reports, updates, and announcements.

This Week's LWN was brought to you by:


June 21, 2001

 

Next: Security

 
Eklektix, Inc. Linux powered! Copyright © 2001 Eklektix, Inc., all rights reserved
Linux ® is a registered trademark of Linus Torvalds