PC-BSD Day 30: The verdict
Thirty days with PC-BSD. One month that flew by. In this month I tried to work with PC-BSD every day, sometimes from a more novice viewpoint, sometimes by pushing the limits from the perspective of the more daring user. There are still issues that haven’t been tested yet like mounting network shares at boot time (always easy in a SOHO network) or trying to access the shared printer on my wife’s Windows XP desktop. I also didn’t test PC-BSD on my laptop and try support for wifi in that way. But, overall, I did what I would normally do on a Linux desktop or at work on a Windows desktop, which -for me- indicates I can make a decent judgment about PC-BSD as a day to day desktop. I know I have barely scratched the surface of *BSD, but this would hold true for many novice users.
PBIs: strengths and weaknesses
One of the key selling points of PC-BSD is the PBI system to install and manage new software. You might almost call it the Unique Selling Point for PC-BSD. In essence, the concept is a strong one. You provide novice users with complete, self-contained binary installers which almost have no interference with the core system.
Using the PBIs has been a mixed blessing. A brief overview of the negatives:
– First, not all mirrors work. There were a couple of occasions where clicking on a mirror resulted in an error message that the package did not exist.
– Not all PBIs work. This was most notable with the games.
– Not all PBIs include all the necessary dependencies. A case in point were the GIMP and Abiword PBIs, though I must say that the GIMP PBI was removed shortly afterward and work began on a new version
– The specific PBI pages do include a â€œlatest dateâ€ indication, but don’t say which PC-BSD is supported by the PBI.
– The list of available PBIs is limited.
– Maintaining and updating the PBIs is left to individual developers.
This last point I found to be very risky, especially for a company that uses the PBI system as an important key feature. One might expect a somewhat higher level of central involvement. According to Matt they are working on a way to build more PBIs automatically.
On plus side the PBIs do provide an easy way to install and manage software. The strength of the system was provided by the non-official PBIs with which you could install Windows-based software under PC-BSD/Wine. This was mostly software that is needed in a W2L/BSD strategy. The PBI system also allows the building of ‘creatively’ constructed packages to circumvent present limitations in FreeBSD, like the PBI with Firefox under Wine and Flash 9.
With the PBIs being so important I can only recommend a swift increase in the number of PBIs and to make them easier accessible. The website is not bad, but I would prefer a GUI frontend from within PC-BSD itself where you can select, download and install the PBIs. Such a frontend would select the proper package for your version of PC-BSD. In the mean time a critical look at all the descriptions is definitely needed. I would also like to see some better indicators of the progress of the installation. Especially with larger packages it appears that they don’t work or are stuck somewhere in the middle. This is confusing and might lead to abortions of the install process. Finally, the PBIs should be flawless.
PC-BSD as a FreeBSD system
The Quick Guide describes three methods to install new software. Apart from the PBIs these are the packages and the ports collection. PC-BSD is positioned as a FreeBSD system with full compatibility. This is true as long as you keep the core system as it is and don’t start installing software via packages and ports. At that point I got the impression that PC-BSD is not completely in sync with FreeBSD. I have seen too many second and third digit dependency errors that made installing software via packages or ports a hit-and-miss thing.
One might argue -as some have done- that the PBIs are the method of choice anyway and that if you want to work with ports and packages you’d better use FreeBSD proper. If so -and I tend to agree with both statements at this point- don’t bother the novice user with explanations about FreeBSD, packages or ports. Acknowledge the FreeBSD roots and then explain the best way to use PC-BSD.
Related to this were the update/upgrade problems. I was kind of shocked to find out that upgrading from 1.3 to 1.4 or from the beta to 1.4 final were not recommended or deemed possible. If that is true PC-BSD definitely needs a new partition/slices layout with /home being separate from the rest. Or some good instructions for novice users on how to secure their data from within -for them- the new system.
PC-BSD and hardware issues
In this month I had two hardware issues. First I had to go through a stack of hard drives that have served me well on various occassions, but weren’t acceptible to the PC-BSD installer. I am grateful for the suggestions to fix this issue, but those suggestions are beyond the reach of the novice user. He/she wouldn’t have a clue why a perfectly fine desktop would be unacceptable to PC-BSD and give up on the OS.
The second problem was with the graphics card. The suggestion to change the aperture in the BIOS was good and the solution worked. But PC-BSD is the only OS that needs this change and I have tried quite a few. Again, not something a novice user would understand and hence not something I want to see in an operating system that caters to the desktop user.
I had some issues with mounting USB drives while being logged in. I would get error messages and the drives wouldn’t mount. Rebooting with the USB drives connected corrected the problem.
PC-BSD as an applications platform
I have become an agnostic as to the operating system. The availability of multi-platform open source applications and open source standards have made that possible. Hence, the phrase â€œapplications platformâ€. KDE desktop is well enough. Working with it on a daily basis led to more appreciation for this desktop. What I didn’t like was the lack of consistency on where new applications would wind up in the menu.
When you think about applications I found nothing lacking. Productivity, communications, browsing, multimedia, financial applications, PC-BSD has them all, though mostly via the packages and ports. Multimedia support out of the box is sufficient, but not completely without glitches. The problems with flash 9 based websites or with DVD playback come to mind.
I am used to changing the look and feel of GNOME desktop whenever I want to and use the gnome-look website to find new themes. Installing a new theme is simply a matter of dragging and dropping. Doing the same with the KDE desktop proved to be somewhat more cumbersome and more hit-and-miss.
I was impressed by the DesktopBSD-tools. Though not part of the PC-BSD system it surpasses Synaptic on various points.
Is PC-BSD ready for mainstream use on the desktop?
That’s the real question, isn’t it? But what does it mean to be ready for mainstream use on the desktop? The slide show that PC-BSD provides during the install gives you pictures of everyday users at home, in school and in the office, happily working with the new system. Is it wrong to conclude that PC-BSD wants to be the desktop of choice for your current colleagues, friends and schoolmates that have been shackled to Windows up to now?
The first impression of PC-BSD is good, really good. But once I went further into the system and using it I regularly ran into snags, some bigger than others. The majority of those problems would be encountered by this group of users quite quickly, problems they hardly encounter under Windows today. Yes, I know that *BSD is more stable, more secure and ultimately a much better OS than Windows can hope for, but that argument becomes moot when a simple PBI fails to install properly or when your USB stick appears unusable.
Does PC-BSD have the potential to be a serious contender for the open source desktop? I answered that questions with a yes, because the potential is there. The solid *BSD roots, the very strong and very accessible information, the friendly and mature community and the PBI system provide the foundations for that potential. I don’t think it is ready now and I couldn’t recommend it yet to someone in the early stages of moving away from Windows to an open source desktop. But I do think that the PC-BSD team has the right target audience in mind and is building an system and a support system that adresses it’s needs. PC-BSD 1.4 is a solid step in the right direction. I can recommend it to the more playful and experienced users and encourage them to provide as much feedback as possible.
Next experiment: DesktopBSD
As promised I will continue with another desktop oriented *BSD, DesktopBSD, and write another 30 days series. However, first I have to move house and make sure I’ll get online again in my new study. I want to start this series on November 1st.
To those who have followed this series and all of you who provided feedback and suggestion, I thank you for your patience and help. If anything, you have assured me that there is a part of the open source realm where mature people simply enjoy life and their open source desktop.