|
|
#1 (permalink) | |||||
|
Aximsite Rookie
|
Will WM2005 be slow?
Ok, because of the memory issue with the integrated bt/wifi that many of us have had (those of us that have too many today plug-ins
), I had to install wisbar2 and pocket plus onto my BIS in order to get the available memory to a level that will allow bt and wifi to work properly.However, now wisbar and pocket plus run much more slowly (obviously) than they did when they were in the main memory. My question is: If in wm2k5 all apps (installed on the device itself and not on the expansion cards) are in BIS, won't that make wm2k5 run more slowly? Will it somehow enhance the way data is transferred from BIS to ram, or will it be just as slow as it is now? | |||||
|
||||||
|
|
|
| sponsor links |
|
|
#2 (permalink) | ||||||||||
|
Aximsite Rookie
|
Quote:
| ||||||||||
|
Last edited by sniffs; 08-08-05 at 06:48 PM.. |
|||||||||||
|
|
|
|
|
#3 (permalink) | |||
|
Aximsite Minor League
|
There is alot more on this subject. One of the M$ developers has a blog or something that goes into detail explaining the new storage system in WM5.
You can find the links and some other info in this tread: http://www.aximsite.com/boards/showthread.php?t=91558 | |||
|
|
|
|
|
#4 (permalink) | |||||
|
Aximsite Rookie
|
nice links in that thread
Based on what he said about 1/100 of a sec vs 1/10 of a sec and the buffer stuff, that makes sense. But, in the same breath he says that his dictionary takes longer to load now. So, it looks like we'll just have to wait and see, but unless the apps are optimized for PS somehow, I still see wisbar etc... lagging a little | |||||
Last edited by rigg419; 08-09-05 at 05:08 PM.. |
||||||
|
|
|
|
|
#5 (permalink) | |||||
|
Aximsite Minor League
|
It seems likely they'll use the RAM as a cache.
After you do a soft reset all program RAM is clear and programs will have to load from BIS the first time you use each of them after a reset. After a program has been loaded from BIS into RAM it can stay in RAM until the system needs the RAM that program is occupying. In other words, it will stay in RAM even after you quit the program. If you always install programs into the BIS you'll be able to adjust the slider in the Memory settings thing to allocate more RAM to program usage (and less to storage) and should end up with almost all programs staying in RAM. Certainly the more heavily used programs will be end up being in RAM almost all of the time. I'm installing all programs into the BIS and I notice little minor delays here and there but nothing really annoying. | |||||
|
|
|
|
|
#6 (permalink) | |||||
|
Aximsite Minor League
|
I just reread the blog entries of the Microsoft Windows Mobile guy and it sounds like we won't be able to use RAM for storage at all with Windows Mobile 5. According to his blog entry you'll only notice things being slower when there's lots of writing going on, or if you have huge amounts of data your app needs to read. Neither of these things worry me.
| |||||
|
|
|
|
|
#7 (permalink) | |||||
|
Aximsite Minor League
|
Good, that caching system sounds clever. I'm just concerned about what they mean by "staying in memory".
WM2003 doesn't actually close programs, it just hides them. Like many (probably most) people here I know enough about the technical side to understand and dislike this, so I use Wisbar Advance 2 to make the close button actually close the program. Will this style of behaviour make sense with WM05? If we completely close a program in the way WA2 or MagicButton does with WM2003, will it also no longer be cached? Or will the full close concept no longer be relevent? With caching, I presume only content which is already in BiS will be cached? So if I save a document to BiS, it is actually stored on BiS immediately rather than being cached? If it's cached, then battery failure will cause it to be lost. On a more optomistic note, does this mean that programs on an SD card will also see the speed benefits from caching? | |||||
|
|
|
|
|
#9 (permalink) | ||||||
|
Aximsite Minor League
|
Quote:
I have no idea what you're talking about or where you came up with that information. When you delete a file from your hard drive, the file's entry is removed from the FAT (File Allocation Table) under FAT/FAT32 or MFT (Master File Table) under NTFS. The FAT or MFT is essentially a list of every file and where it physically resides on the hard drive. If a file is listed there, the space that file takes up is listed as in use and cannot be written to by the OS (unless you're updating that particular file, of course). When you delete a file, its entry is removed from the FAT/MFT and the space that file was using is now, for all intents and purposes, marked as free space. There is no purposeful staggering of the delete process. The file will phyically remain on the hard drive until it's overwritten (because the OS now sees the space as free); this could happen 2 seconds after you delete the file, or 2 decades. If you delete a file and then never write another bit of data to that drive then that file will physically remain there indefinitely. As for moving files, that also happens immediately . . . sort of. If you move the file from one location to another on the same drive, then the FAT/MFT is simply updated to reflect the new location of the file; the file itself is not physically moved. If you move the file from one drive to another, then the move occurs immediately. There is no "tricking" of the hard drive into thinking anything. The OS doesn't tell the HDD anything in relation to what files are on the drive or where they are. It's the other way around; the drive tells the OS what files are there and where they are located. The used/free space reporting you see in the OS is coming from the OS to you; not from the OS to the drive or anything like that. | ||||||
|
|
|
![]() |
| sponsor links |
| Thread Tools | |
|
|