In part one, I looked at how the simplicity of the PalmOS has merited into a platform that is not as mature as the users that it wants to attract. Proof of this is quite evident in the Life Drive by Palm (formerly palmOne). The Life Drive is an excellent piece of device in my opinion. For many mobile users, it combines all of the features that one wants in a mobile device (except a camera) and does so in a reasonable design. The hitch in the Life Drive however, comes in the adaptation of the OS to fit the hardware. For example, though the PalmOS can support large capacity drives, it had not ever been used for a hard drive (and a large one at that). This meant that Palm (palmOne) developers had to create a driver that would enable the use of hard drives larger than the 2GB limit that the PalmOS has. The Life Drive is also dual wireless, which had only been done on Sony Clie PalmOS PDAs, and even then only twice. Add to that the use of non-volatile RAM (which changes how the OS governs battery usage), and you have the makings of a lot of patches to mature the PalmOS.
I wanted the opinion of a PalmOS developer before writing this because, in part, I wanted to know if the PalmOS is as bad/good as it can be. I thought about a lot of developers, and found one in particular who isn’t new on the scene, but has done some things that really do make living with the PalmOS much easier. I chose a few questions to ask him, and from there, I got a perspective on the global effect of the PalmOS, and at the same time the fact that more can be done and should be expected from platform.
Here is the text of that interview with Palm OS developer Dmitry Grinberg:
Antoine: Hello Dmitry. If you don’t mind me asking (for our readers), how old are you and where are you from?
Dmitry: I am 17. I was born in Russia and moved to the USA 5 years ago.
Antoine: About how long have you been developing on the PalmOS?
Dmitry: I’ve been developing for PalmOS for about a year and a half now. I did not really develop anything on my first palm – Sony N610C, but then I got a Zire 71 and got a crash trying to edit a hi-res icon, I made IconFX because at the time no other tool could do this.
Antoine: I’ve noticed that you do a great deal of utilities that smooth-out the performance of the PalmOS. Is there a particular performance mark that you are looking for from the PalmOS that isn’t there yet?
Dmitry: As soon as you stop improving, you lose. PalmSource seems to not know this so I try to make up for this with my applications. Many of my applications never get released and stay on my device just for my personal use. This is the case with many of my hacks. I am just too lazy to polish them up to suit everyone. When you release a beta people start complaining and asking for more so with a simple hack for personal use it is simpler to never release it.
Antoine: Being an active member in the PalmOS development community, where do you see the state of the platform, and where can it go?
Dmitry: I am waiting for PalmOS on Linux and hoping for a lot from it, as now that they will have to open up the kernel sources, it will be much easier to do things. As for the current state, it is far ahead of Windows Mobile, but in an interesting way. It is ahead by being behind. Windows Mobile tries to stuff a full desktop OS in the Handheld; and while this is technologically possible, it is hell for users. PalmOS does the simpler thing, and stays simple so in a way they are behind, but in usability (and that is what matters most in a handheld device) they are light-years ahead.
Antoine: If you had to tell one thing to a person who is reading this and did not realize that the PalmOS was as versatile as it is, what would you want them most to know?
Dmitry: The truth – I wouldn’t wish Windows Mobile on my worst enemy. Not even on Bill Gates himself. Try to use both for a day and see how much you get done. I did this experiment a few times and realized that PalmOS is much more efficient when it comes to user interaction. ToDo list, Datebook, AddressBook – they all are as simple as possible and that is a HUGE advantage. Bigger is not always better.
Antoine: Lastly, you are doing a great job. I tried SkinUI on my T5 and I actually had a more stable experience with it than most other programs I had installed. Part of that might have to do with the DBCache program, but your work is very well and the polish in that app really shows your attention to the platform. Hopefully, Palm picks you up, they could use some fresh and solid programmers.
Dmitry: I had my own share of NVFS-related problems in SkinUI, all programmers have problems with badly written OS, and they all hate it.
Antoine: The last question, you mentioned that all developers have been having a problem with the latest releases of PalmOS Garnet (v.5.4.x). Could you be more specific as to what the problem areas are and if, in your opinion, they are fixable (and should be fixed) by PalmSource/palmOne (NOTE: since the asking of this question, Palm has released patches for the Treo 650 and Tungsten T5 which address these identified problem areas)?
Dmitry: Well they keep changing things so so fast for no reason whatsoever. First the connectors: Universal -> minuUSB -> Athena -> Universal Gen 2 Then the 5-way navigator values (key characters used to see what key was pressed) keep changing from Z71->T3->T5 all different. NVFS is another story. [There are] bugs all over the place. It may be cool and useful but it is so damn buggy that if I ever have to get a NVFS device i will rewrite it. (Yes, consider that a promise). The PIM databases also changed from the old (Z71, and m505 format) to the new (Z72, TT5, LD) format.
Antoine: What have you gained from the PalmOS community that you probably wouldn’t have from other development communities (for example; connections with people around the word, firmer career goals, etc)?
Dmitry: Fun really. PPC programming is boring and annoying. Palm is just fun. People are nice (forums, etc.).
The other developer’s perspective that I wanted to draw on was that of a senior developer at PalmSource via an article at All About Palm (www.allaboutpalm.com). Diane Hackborne is quite a well regarded developer and has made many visits to the online PalmOS community to share some clarifying points about the PalmOS. One of the things that I really notice about her end of the approach is that she seems to admit in part that there are things that could have been done differently, and yet, PalmSource is using the acquisition of China MobileSoft as an opportunity to really make the PalmOS “the” mobile platform of choice for venders. For more information about this perspective, read the ongoing question and answer session at All About Palm (http://www.allaboutpalm.com/forum/showthread.php?t=44).
So what am I trying to get at overall; when I started writing this, I wanted to see how much the PalmOS platform has effected mobile computing. More often than not (at least recently) reports have either written off the PalmOS or identified the problem areas and not exactly why they are problem areas. Yes, there could probably be more control over the licensees’ use of specific components, and there could also be a concession to Windows Mobile and Symbian ways of doing things, but the PalmOS platform has never been about copying just for the sake of having. Features added to the OS and hardware are always items of ease of use for the end-user and flexibility for the developer. Things go wrong, and with a smaller company with a large reach like PalmSource; mistakes are not very easy to hide with a patch here, and an upgrade there. But the effect is made clear. The PalmOS platform has changed the way that we view our mobile devices, and whether we want to regard it or not, this effect has made PalmOS a successful platform for all.