In this beta, I found a few problems, bugs and things that I personally wouldn't agree with as a very old BeOS user. I don't use BeOS much anymore, but I do have a good grasp of its excellence in some points and its suckiness in others, so please forgive me for being opinionated.
So, here are a few problems I found in this beta and I wholeheartedly hope they will be fixed for the final version:
1. The fonts are bad. Very bad. I don't know what YTAB changed in the settings of the font rendering engine, but fonts are plain ugly. Dano/EXP also used this new font engine, but it wasn't as bad as in this Zeta beta. Linux's latest font config is worlds better in my opinion.
2. The preference panel "Fonts" is now broken. Changing the font size in the "Plain Font" only changes it to the context menus, but not in the menus of the applications, as it should have.
3. SoundPlay and its plugins and email client Beam crash like there's no tomorrow. No DivX support I could see.
4. There is a problem with the font sensitivity of the UI, and it is especially visible with *some* themes. I don't know if this is a bug of Zeta or a bug of the theme. But on some applications there are problems with widgets colliding with others widgets (they render on top of the other).
5. There is no security/protection for the user, even now with the introduction of BONE, like a personal firewall. Something more advanced like internet connection sharing, would be cool to have too. However YTAB has worked on better ISDN support and that's a plus.
6. The logo of Zeta in the Deskbar is really amateurish and when you click on it you get an even uglier look. But hey, it's a beta, right?
7. No scanner support as of now. Bernd tells me they are working on it though so this is promising.
YellowTAB is a small team and it doesn't have all the resources to work full scale on every aspect of the OS. However, the realities of the marketplace won't be kind to the company, so I feel I am forced to also not soft-pedal my criticism.
Granted, this is a beta. However, the direction the OS is taking is already clear: Add features and more new user-space applications and some more and some more. I will have to ring the bell of danger for YTAB and ask for more fixes rather than more new features. BeOS 5 was by no means perfect. But it worked well for most people. Replacing parts that are known to work well and have served well with new parts that only add unnecessary bloat (e.g. the new Installer) or duplication (e.g. Dockbert), is hardly a step forward for the BeOS paradigm. In fact, this could be considered a step backward. We certainly don't need another Linux, with mind-boggling choice and variety at every turn. Zeta should continue where BeOS has stopped, not transform the OS into bloatware and illogicality like your average Linux distribution. Surely, I love the new drivers. Surely, I like the new Tracker (even if it is still buggy), but instead of filling up the preferences with unneeded panels I would have to ask for things like:
1.Samba. Where is a working samba? We need interoperability!
2. There is _stil_ no support for more than 90 Hz in the monitor panel. In fact, the current screen panel doesn't expose all the abilities of the driver and app_server. Be was thinking of moving to GTF, but Palm bought them before that happens.
3. No spell checking or voice reading in the text views of any Zeta application. No real support for accessibility.
4. No multi-user yet! Yes, this can break a number of older applications (which was why Be didn't go live with it, but Zeta/Dano already breaks apps, so let them breat in one go instead of having to break more apps again in a few years), but someone has to take the big decision and activate the Be implementation (it is just a build flag ;). Zeta is a desktop OS, but multi-user also grows in the minds of people as times goes by. It is not 1998 anymore. Even Microsoft now offers multiuser on all its OSes.
5. Why doesn't Zeta use the _new_ preference panel that was written for Dano/EXP. Why do we still get the old panels that are now filling up the menus? (more than 12 items on any menu is considered bad by usability engineers).
6. No fix for the numlock bug which makes BeOS to not remember if the NumLock was set to ON in the previous booting. Sounds trivial and stupid but really annoys a lot of people.
7. No fix for the 1 GB RAM limit. This is maybe the biggest BeOS/Zeta limitation today. Read here for more explanation and make sure you read the comments too.
8. BeOS can't load more than 32 MB of addons, which is needed for big applications (a problem that almost stalled the Mozilla port back in the day). No more than 192 threads per app. Say you open more than 192 child windows or threads of an application, BeOS goes ca-boom (e.g. ShowImage).
9. Still, no Java.
10. Sucky VM in the kernel. Needs fixing. Source of many problems for Zeta, including the 1 GB RAM limit.
11. I'd like to see support for Great Britain's TV cards (some TV cards use a different sound standard). There was an addon for it but never got integrated to BeOS.
12. Second biggest problem: No usable browser. NetPositive just doesn't cut it anymore -- it is a Netscape 2 compliant browser. Useless, at least for me, despite its speed, as it doesn't support SSL (I don't care about javascript, but I need SSL). The Mozilla/Phoenix ports are just _BAD_. BeOS was created to run on computers like P90 and P100. I use BeOS and Zeta on this (fast machine by the BeOS standards) dual Celeron 2x533 (which is the machine most BeOS geeks preffered back in the day) and BeZilla/Phoenix is just _unusable_. I am *not* saying that it is Mozilla's fault, because in this case it is not. I also have Windows XP and Linux on this machine and while Mozilla doesn't fly, it is 100% usable under these OSes. But on BeOS/Zeta, it is not. It crawls like hell. Again, I will have to ask YellowTAB to help fix the port, as a good browser is imperative no matter the platform (and the rest of the Mozilla apps that come with the browser). THIS port should be fixed, regardless. It is a strategic step for YellowTAB, even if YTAB doesn't plan on using it as default.
Conclusion
Even with Mozilla so slow when operating clogging up both my CPUs, mp3 playback did not skip (while it does on Red Hat Linux 9 on the much faster AthlonXP 1600+). The OS still has the BeOS' great UI responsiveness, but overall the system is a bit slower: you will need something like 80-100 Mhz more than the previous low-end (P75-P90) and at least 48 MB (previous low was 32 MB). Still, the OS overall, BeZilla/Phoenix aside, is much faster and responsive on low-end hardware than _any_ of its competition (Linux, Windows XP).
So, what I do I think of this beta? I believe that it is two steps forward and one step backwards. YellowTAB does some hard work to ensure driver support and application support, but at the same time they lose focus and spend time working on things that don't need replacement or fixing and leaving aside other things that do need fixing. The hard problems are still there. YellowTAB must play catch up with Linux and Windows now, as BeOS was paused and not developed for years now. For now, I say "good job," but keep running.
YellowTAB has a great advantage inheriting all this source code from Be, Inc. but also it has the inevitable curse that they will have to live in the shadow of the "legendary days" of R4.5 and R5. To overcome Be's own legacy will take a lot of work. But it is a great help that Zeta is the true and only direct BeOS descentant, so they are currently years ahead in development than the other teams who try to reproduce the BeOS, like OpenBeOS, B.E.O.S, BeFree and Cosmoe. I hope that ex and BeOS developers and users show the support YellowTAB needs in order to survive and continue the development of the authentic and original BeOS code.
However, I would like to see some real engineering from YellowTAB, not just "development". Some real breakthroughs and innovations, like we became accustomed to from Be. Hiring a usability engineer with real BeOS experience would be a good thing in my opinion. I am not talking about a UI artist, I am talking about a usability engineer. YellowTAB needs one. More engineers to join the team would be good too. If they could "hire" Axel, Marcus and 2-3 more "big brains" from the BeOS dev community and OpenBeOS, they could definately get some great results with time.
For version 1.0, I can say that I understand that the product needs to play catch up with the competition because of the lost time Palm created, but from now on, I need to see some achievements in order to draw attention in the market. BeOS was always about impressions and "wows". So, impress me.
Installation: 9/10
Hardware Support: 6.5/10
Ease of use: 10/10
Features: 6.5/10
Credibility: 8/10 (stability, bugs, security)
Speed: 8.5/10 (throughput, UI responsiveness, latency)
Overall: 8.08
- "Installation, Impressions"
- "Applications and Usage"
- "Problems, Conclusion"
