I have been using OpenOffice for the last two years as my primary word processor, even for the sort of short articles for which Word 97 is probably better. I wrote the whole worm book in it without any disasters or even excitements. In fact, one of the things I had meant to do when I started this blog was to record the large number of small things of nerd interest which come up while using it. In order to make OOo useful, I have had to write quite a number of macros. There was the famous word count, which now pleases me a lot; there are a set of small macros to attach styles to keystrokes; and there is transpose characters, the first macro I write in any word processor.
I had help with all of these, though the word count one has now been completely rewritten so that every line is by me. It is a strange and rather good experience to get direct and personal help from some of the people who wrote the program; that’s probably the decisive attraction. Writing and still more printing books in Word 97 is a real pain, but there are plenty of corresponding annoyances in OOo that mean it would be simpler just to upgrade my copy of office.
The first releases towards 2.0 are coming out now, and one of them should have fixes for three separate bugs I have found in the routines to move around by words and sentences.
Out on one of the mailing lists someone has offered a bounty of $1,000 for curing various annoyances with the spreadsheet. Out of the original list of seven or eight, it turns out that the major one, a limit of 32,000 rows, is already being fixed; four or five already have been fixed, but this is cunningly concealed by the documentation, and only three or so are available. What this means in practice is that the money will never be paid, since no one person can claim to have done all the work, and the hardest job is being done by Sun employees who can’t be paid extra for it by individual customers. But it is a sign of vitality. I want people to make money from the software. That way I know it will be improved. There are some more informative bounties on offer here, which suggest that someone is actually using the slideshow bit for more than I do — which is to read completely unavoidable bits of powerpoint.
There are still a number of things that the program won’t do. A proper integration with Python is on the way but hasn’t happened yet. The macro writing environment is a horrible mess compared to Word’s. The spreadsheet is just not as good as Excel, though this isn’t really surprising. I think that Excel is the best thing Microsoft ever wrote, and a real model of power and usability. This may be because I process words.
The word processor lacks the ability to be used as a proper outliner, concealing or hiding paragraphs according to their heading level. Outline numbering is confusing. There are three different methods, and none of them do what you might expect or even want.
One of the big selling points ought to be the XML file format. I haven’t yet derived any practical benefits from it. Once I have discovered how to get at and manipulate the files directly from disk, using python, perhaps I will. There is a perl module that does everything I want, but life is too short to do anything in perl ever. The perl thing comes from the Francophone wing of OOo, which is tremendously vigorous and helpful. One of the nicest things about the program is that it is used in a lot of different nationalities, and some of these communities are adding things which need to be translated into English. It did start out as a German program, but I really like the way in which it is harnessing the efforts and interests of people all round the world who are doing more than simply playing catch-up with an English original.
I’m beginning to think, then, that using the is not a quirky demonstration of my inner beardliness, but a genuine rational choice of the best tool for this kind of job, taking into account all the various hidden costs. One of those costs, for me, is always the tendency to spend time fiddling that could be spent writing. One example would be indexing text. I know that the MS fast find wastes a little time. But I can’t really claim that it wastes more time than my writing a whole python script to read the context of all the files in directory tree into a mySQL database so that I can find things in them quickly.
But I have to say that the one change which really made OOo more usable for me had nothing to do with Sun. I built s new computer, with an Athlon 2400 in place of the P3/400 I had been using, and everything was transformed. This is bad news for poor organisations hoping to run it on ancient hardware. It may run more quickly under minimal window managers in Linux, but i doubt this. Anyway, changing to a new window manager and a new operating system is very expensive, especially when you can’t buy support.