Originally Posted by Adrynalyne (Post 991622)

212xx were MS non OEM specific builds, pretty sure.
216xx is HTC, if memory serves.

Originally Posted by da_reeseboy (Post 991497)

Does anyone know what the branch-offs are for or who is doing each of them? As in -

212xx were MS non-OEM builds... (thanks Adrynalyne)
215xx were Samsung builds... (thanks blazingwolf)
216xx were HTC builds... (thanks Adrynalyne)
217xx were ???
218xx ???
219xx ???

Who is behind these builds?

Originally Posted by da_reeseboy (Post 991667)

Thanks...updated orginal post.

Anyone else? I'd like to be able to fill in all the blanks if possible...

Originally Posted by joojoobee666 (Post 991845)

Well, according to Da_G, 217xxx was Crossbow COM1 and 218xx was Crossbow COM2. Also, apparently neither 217xx or 218xx are from OEM builds. They are developement builds, but from where? I dunno. 219xx was branched off from the 217xx tree but I do not know what the tree name is.

Originally Posted by joojoobee666 (Post 991287)

What did you use as the donor XIP? If you used the files from say Da_G's kitchen it won't work since XIPport is looking for the text files that correspond to each module (not the imageinfo.txt inside the module). That's one reason. Another is that there or overlaps with o32_realaddr of the modules (the first section of the XIP map.txt). XIPport does not like overlaps here.

Originally Posted by da_reeseboy (Post 991895)

Sorry... I'm sure this is a stupid question, but what is crossbow? Is that a company?

Originally Posted by lennie (Post 991952)

crossbow was the code name for windows mobile 6:

http://www.bing.com/search?q=crossbo...bile&form=QBRE

Originally Posted by lennie (Post 991934)

I dumped the xip myself: I extracted the OS.nb from the rom, I've read that the athena doesn't have any extra data in the os.nb so I renamed it os.nb.payload, I ran the ImgfsFromNb and got the offeset which in the shipped from is at 0x00660000. I didn't know where the xip started in the file so I used Hex Editor Neo to find the beginning of the second xip at the address "FE 03 00 EA 00 00 00 00 00 00 00 00 00 00 00 00 00 00", which is at 0x00310000. then I used RomMaster to extract the xip.

I used xipport and dumped the xip, wright maps, make packages for the xip I just extracted all went fine. with the xip to be ported I used xipport to dump and make packages.......however, when I copy over the msxipkernel, msxipkernelltk and click undo xipport throughs an unhandled exception.......and crashes. this is where I'm stuck.....

Originally Posted by da_reeseboy (Post 992127)

LOL yep... that was a stupid question by me then. So the fact that 217xx and 218xx are referred to as Crossbow only means its windows 6.x - No deep insight there. Should COM1/COM2 mean something to me?

Originally Posted by chump (Post 992212)

anybody interested in a fix for the backlight going dim on lockscreen? this works for KAISER, try on other devices at your own risk!

Originally Posted by blazingwolf (Post 992379)

Can you tell me what was done to the dll? The dll on the Omnia is different so I can't use this one.

Originally Posted by joojoobee666 (Post 992129)

Does that happen when you undo either dump WITHOUT making any changes?

Originally Posted by lennie (Post 992416)

yeah, it still crashes without making any changes......what does this mean?

I checked the map.txt and there are no overlaps going on.....

Originally Posted by joojoobee666 (Post 992481)

Which one? The donor or yours?

Originally Posted by chump (Post 992398)

its the backlight.dll from official kaiser 3.14 rom. nothing else was done to it.

Originally Posted by lennie (Post 992619)

mine.....I'm going to try it on the donor also and see if my xip is the one giving the problem.


EDIT: yep.....it's my xip thats giving the problem.....I thought I followed everything right. now I have no idea where I'm messing up when extracting my xip from the ROM.

question, when I use imgfsfromnb and it shows the start, end and offset which of these values should I use in RomMaster? I used the offset value (which happens to be the same as the start value)....should I be using the end value instead of the offset? I followed this Tutorial http://forum.xda-developers.com/showthread.php?t=453722

Originally Posted by joojoobee666 (Post 992763)

I would try and use Ervius XipporterEX to extract the xip.bin from the payload. Something is clearly wrong with the xip you extracted.

Originally Posted by Wizzard (Post 992716)

Does anyone have an idea when 6.5 will be final???

Or is it going to be updated every day untill they start working on 7??

Thanks

Originally Posted by Da_G (Post 993209)

@Wizzard: 6.5 is already final :) (21234) and we've gone through.. 6.5.0.1, 6.5.0.20, 6.5.0.40 we're working on 6.5.0.50 right now, which is now final with 21280 :) At the very least we'll make it to 6.5.1 before WM7, but I expect it to go beyond that too :)

Originally Posted by Da_G (Post 993221)

I'm not home, or I would have already :) Tethered on my phone in San Francisco for the volunteering conference (michelle obama was there! that was pretty cool)

Originally Posted by iggord (Post 993229)

21820
21915
http://bbs.xbeta.cn/thread-24548-1-1.html

Originally Posted by Da_G (Post 993278)

i might be able to get a build uploaded by morning, if at&t doesn't hate me for uploading it on tether :)

re:21915, her signature shows it's running QFE 5.0.90, which is incorrect, that build is QFE 5.0.20 (older base system than COM2, as i mentioned previously, which is 5.0.40 and is now moving on to 5.0.50)

21915 is also a build number in the future (has not been built yet) - so i call shenanigans :P (todays build should be 21912 if i am keeping track of build numbers properly)

Originally Posted by Da_G (Post 993209)

@Wizzard: 6.5 is already final :) (21234) and we've gone through.. 6.5.0.1, 6.5.0.20, 6.5.0.40 we're working on 6.5.0.50 right now, which is now final with 21280 :) At the very least we'll make it to 6.5.1 before WM7, but I expect it to go beyond that too :)

Let's see if I can recall the build naming convention...

212xx = AKU1, all builds leading up to and including WM 6.5
213xx = MOT motorola
214xx = forget
215xx = SAM samsung
216xx = HTC htc
217xx = COM1, continuing dev of 6.5.0.1 - 6.5.0.40
218xx = COM2, continuing dev of 6.5.0.50
219xx = MD, feature branch (unstable features are added here, backported to 218xx (this tree is based on COM1, so older base os code, but the UX code is newer [user experience])
22xxx = SEMC sony ericsson

almost all of the oem branches are based off 21234 (iirc SAM was based off an older build than that) - the oem's requested their modifications, and msft did them, incremented build number - some of these trees have stopped receiving modifications already.

For most of you, for the moment, 218xx is the branch to follow (but they might change the number now that COM2 is done as of 21820)

Most of the current focus is on UX as the backend code is mostly at the goal, so we should be seeing some nice changes to the theming api's etc. that will allow us much more modification of the UI (iirc the end goal is complete skinnability and finger friendliness - nothing left over from the old 6.1 interface that used a stylus, essentially)

They might be bandying about the idea of licensing resco keyboard to replace the stock non-finger-friendly keyboard.

Also, an interesting tidbit, any device that has a dpad and IE6 is *required* to have support in the display driver for a mouse pointer, and support in the dpad driver for a mouse pointer, to allow for things like javascript mouseovers that pull drop-down menus :)

Originally Posted by lennie (Post 992884)

I'm searching for this but can't find it...

Originally Posted by Da_G (Post 993209)

@Wizzard: 6.5 is already final :) (21234) and we've gone through.. 6.5.0.1, 6.5.0.20, 6.5.0.40 we're working on 6.5.0.50 right now, which is now final with 21280 :) At the very least we'll make it to 6.5.1 before WM7, but I expect it to go beyond that too :)

Let's see if I can recall the build naming convention...

212xx = AKU1, all builds leading up to and including WM 6.5
213xx = MOT motorola
214xx = forget
215xx = SAM samsung
216xx = HTC htc
217xx = COM1, continuing dev of 6.5.0.1 - 6.5.0.40
218xx = COM2, continuing dev of 6.5.0.50
219xx = MD, feature branch (unstable features are added here, backported to 218xx (this tree is based on COM1, so older base os code, but the UX code is newer [user experience])
22xxx = SEMC sony ericsson

almost all of the oem branches are based off 21234 (iirc SAM was based off an older build than that) - the oem's requested their modifications, and msft did them, incremented build number - some of these trees have stopped receiving modifications already.

For most of you, for the moment, 218xx is the branch to follow (but they might change the number now that COM2 is done as of 21820)

Most of the current focus is on UX as the backend code is mostly at the goal, so we should be seeing some nice changes to the theming api's etc. that will allow us much more modification of the UI (iirc the end goal is complete skinnability and finger friendliness - nothing left over from the old 6.1 interface that used a stylus, essentially)

They might be bandying about the idea of licensing resco keyboard to replace the stock non-finger-friendly keyboard.

Also, an interesting tidbit, any device that has a dpad and IE6 is *required* to have support in the display driver for a mouse pointer, and support in the dpad driver for a mouse pointer, to allow for things like javascript mouseovers that pull drop-down menus :)

Originally Posted by krazy_about_technology (Post 994078)

@Da_G:

This is a little put down kind of thing for users like me with retro devices having 64 MB RAM :( If the base os code is done that means they are not going to work on memory leaks that GWES.exe, Shell32.exe, device.exe and filesys.exe are having. GWES is the top memory eater. In 21815, it increased a bit, people with phones having less ram should have notices it. Memory usage wise, 217xx series was better. Maybe we, old device users, should go back to 6.1 now. But the increased battery life always stops me. I get double the speed and responsiveness in 6.1 but in 6.5, battery life is more than 1.5 times.

Originally Posted by joojoobee666 (Post 994153)

I remember asking this before in this thread a month or so ago, but no one ever replied. I was wondering if any GSM owners with the 6.5 native kernel were seeing these massive memory leaks. Anyone?

Originally Posted by dsi76 (Post 994184)

I'll reply to your question tomorrow after I flash 21815 w/ 6.5 kernel onto my Fuze tonight.

Been using a Diamond the last few days with 6.1 kernel .. and my battery life actually has been decent, most likely because I don't use the memory-hogging Phone Canvas in my 6.5 builds, as I did in 6.1.

Originally Posted by joojoobee666 (Post 994153)

I remember asking this before in this thread a month or so ago, but no one ever replied. I was wondering if any GSM owners with the 6.5 native kernel were seeing these massive memory leaks. Anyone?

Originally Posted by Da_G (Post 994432)

I posted 21820 over at xda from my cell lol :)

btw, gwes and shell32 are both core components of the ui, so they will need to be optimized as more extensibility is added, i think we'll see improvements here :)