Hi Eugenia, I'm happy to hear from you, I'm cc-ing this reply to openbeos@xxxxxxxxxxxxx as I get the impression you meant to address this to the whole group.... At 14:38 24-8-2001 -0700, you wrote: >So, I hear you are trying to recreate BeOS by using NewOS (the kernel of >NewOS is new and incomplete but it is certainly capable, fast and clean)... >If I may, I would like to give some advices to speed up the process: Just want to say I feel happy that you (who spend(s) quite some time with people with access to the BeOS kernel internals) agree NewOS looks like a possible base for BeOS-cloning :) >1. One of the most difficult things to recreate it will be the filesystem. I >suggest you port SGI's XFS. It is the one that it is very close to BeOS, as >Dominic who did BFS, has also worked on XFS. It is a ready-made fs, which >means: less hard drive breaks while testing. I was thinking of using an existing journalled filesystem already, is there a reason for prefering XFS over ReiserFS (besides Dominic, which is a good point :))? >2. Instead of trying to re-invent the wheel, use QT. No, I don't mean to use >that horrible and bloated thing called KDE. I am talking about QT itself, >the Toolkit. QT runs on the framebuffer, it does NOT need X or Windows or >MacOS. All you need is to setup the VESA mode, and here you go! This means >that you would be able to have a graphics system in no time where you can on >top of it port the BeOS API. QT is fast, it even runs on PDAs, and >especially QT 3.0 comes with features that you would need years and years of >engineering to implement (like bidirectional writting for example). QT is >C++ and its API is VERY similar to BeOS, but again, please do not assosiate >QT with X. It has nothing to do with X, QT is a very modern toolkit made by >professionals. Personally I was thinking about something like libart in combination with freetype, as it is used in GNOME.... But this is all very theoretical for me, I must say.... >3. Get rid this oh-so-blue background color on your homepage. For God's >sake, I can't read the text. LoL, you're not the only one who felt that way :) BTW, it might be more interesting (and soon linked to http://open-beos.sourceforge.net) to take a look at http://www.obedn.com/ which seems to be the new site for the project... >4. Setup a mailing list... This has been already heavily in use, take a look at //www.freelists.org/archives/openbeos/ for any past issues :)) Would welcome you aboard, if even only for your insightful suggestions and/or OS development experience. >Good luck, >Eugenia Thanks, we need all the luck we can get :))) Regards, Ithamar. PS: Please note that we're talking about a phase by phase (module by module) binary compatible replacement of BeOS, no big: "Here is your BeOS duplicate, just wrote it last month" unbelievable plan.... :) Ithamar R. Adema Information Engineering Associates Ltd http://www.ieadev.com