Hacking AKAIO 1.2 Released

[lurbi]

Cheat Updating works perfectly here.

Just gotta make sure I'm not whoring my pipe on other things when I do it.

I use it on a 16 MBit/s connection which is IDLE.
(read: I am not "whoring" bandwidth)

I am happy if I manage to download 50 KB before AKAIO 1.2 craps out.
Maybe it would be a good idea to take a look what usercheatup does, because it works veeeeeery reliable here.

 

[dib]

Yeah it was in the news on this site yesterday that Cyclo copied your wifi update idea. That stuff happens all the time with these things, of course. Sure wish Acekard were that fast in adopting popular Cyclo features like save states and slow motion. Or did anything toward improving features, for that matter.

madridi4ever-I don't want to speak for Norm or Smiths, but just going by what's already been considered around these forums, bottom line is they can't support products that they don't own and have no way of testing.

 

[Madridi]

Yeah it was in the news on this site yesterday that Cyclo copied your wifi update idea. That stuff happens all the time with these things, of course.

madridi4ever-I don't want to speak for Norm or Smiths, but just going by what's already been considered around these forums, bottom line is they can't support products that they don't own and have no way of testing.

Yeah I understand and respect that. It was just an idea nothing more

 

[Smiths]

I am happy if I manage to download 50 KB before AKAIO 1.2 craps out.
Maybe it would be a good idea to take a look what usercheatup does, because it works veeeeeery reliable here.

usrcheatup seems to download (from hexediting) a file hosted at Googlepages, gzipped up by the author. the latest changelog is 8-12-08.

So it obviously is working better with google's hosting and the routing googlepages uses. tons of mirrors.

so yeah, so far that's why it works better IMO.

anyone else to contact?

 

[dib]

Possibly use a string in place of the url, loaded from globalsettings where people can enter their own cheat url?

 

[Smiths]

why? cheats.gbatemp.net is the official place and it always has the latest.

i can't help too much on timeouts from overseas or whatever, but on my side it always works. can't fix what isn't broken.

 

[IOwnAndPwnU]

Just curious, why does it stop downloading it halfway? Because it got to 151Kb/1111kb today (CHEAT.XML). I wasn't using the computer at all during all the multiple attempts at updating DB. I didn't have any torrents running and I was like right beside the computer. I noticed it usually stops at 11 or 21Kb but if I'm lucky a bit higher.

 

[dib]

Like Smiths said, could be related to latency. I'm an idiot with networking though, I've never coded anything like that in my life. Even when I used to work on an online game, that wasn't my area of expertise so I understand little about it.

As far as globalsettings, the cheat database currently hosted on Google is compatible, right? Just let people point their own Acekards to that one if they're encountering problems with Gbatemp's host.

 

[lurbi]

usrcheatup seems to download (from hexediting) a file hosted at Googlepages, gzipped up by the author. the latest changelog is 8-12-08.

So it obviously is working better with google's hosting and the routing googlepages uses. tons of mirrors.

so yeah, so far that's why it works better IMO.

anyone else to contact?

hmmm ... never thought of that ... could definitely be that the latency and jitter is too high here in europe.

I just retried 3 times ... and I did make an absolute record.
I got 100KB downloaded!!
Is it normal that there are columns of numbers displayed in the background while downloading.

Is it maybe possible to make the network code more resistant to stalls??
Eg that it is useable outside the US? (or from regions who are geographically far away from the gbatemp server)

sorry ... no contact known

anyways ... I can live without AIO cheat update ... nevetheless it would be good to make it more resistant.

 

[Smiths]

it's already set to be rather good on stalls. i could increase the timeout but that probably won't matter.
the one on googlecodes is out of date, is gzipped, and is frankly the responsibility of the maintainer of usrcheatup to handle and personally i don't want to get another person responsible for this.

my downloads have been working perfectly. so i am sorry it is not working for you but it is fine here.

the columns of numbers are just to "flavor" it up a teeny bit. technically, they're debugging numbers representing the amount of data received in each loop, updated every 8 loops or so. so you can see technically the speed it gets, and that's with a big buffer and everything.

dswifi's tcp stack is not good, and it seems sg is not developing it anymore (real life gets in the way).

 

[Skydancer]

Can confirm that Rumble per Game Setting not work!

Tried different settings, just to be sure:
3in1 Menu Rumble off -> per Game Setting "Low" or "middle" or "high" = no rumble
3in1 Menu Rumble low -> per Game Setting "High"

 

[Smiths]

yeah pergame rumble apparently requires telling the 3in1 beforehand that it's being changed.
that's a missed line by me. since i don't own a 3in1 i was just setting the option... not realizing the 3in1 is told as well to change the thing up. the setting in a window is like half for "show".

that'll be fixed whenever the next one comes out.

 

[IOwnAndPwnU]

We're already discussing the next update of AIO.

Yay!

 

[priv8dan]

any chance for us to use *.zip / *.rar files in our skins directory? Thanks!

 

[lurbi]

any chance for us to use *.zip / *.rar files in our skins directory?

IMHO very, very counterproductive.
The CPU of the DS is weak, and the skin has to be displayed very often.
This would mean that the DS has to decompress the skin very often.

I also do not really see the reason ... we live in a world were 8GB MicroSDHCs are pretty cheap ...
and I'm not sure ... but maybe AKAIO uses JPG or PNG files .... which cannot be compressed further anyways.
If not, and AKAIO uses uncompressed images, than because even the delay to decode JPG files is too high.

 

[Smiths]

it uses bmps and i won't be adding any zip skin support. totally counter productive.
esp since we're stetching memory as it is.

 

[IOwnAndPwnU]

any chance for us to use *.zip / *.rar files in our skins directory? Thanks!

Wouldn't that be kinda useless? Like its only maybe another MB.

 

[f3l1x]

smiths, norm

what about implementing http resume or ftp retr and size with offsets to get the files in little (512KB?) chunks instead? Ive had to do it before with low mem devices with no mmu and it worked for getting larger files without hang-ups. I dunno, im just pulling stuff outta my... uhh, hat... ...What?

also, the boths of ya, anyway to get new dldi features in the next release? I want the ability to chroot homebrew so badly, and I'm sure anyone who knows what I'm talking bout would agree.

Oh and make x the cheat menu and L the switch device key. B does just fine for back/up/cancel and we don't need 3 different buttons to get to the same place. especialy since x does nothing in ak2 now. I know i dont have to tell you, the menus are crowded as it is.

Maybe L could be an action modifier down the road, but thats kinda minor for a pretty big change.

too much stuff in one post? just invite another ass eating and i'll un-willingly comply

 

[GH0ST]

A chroot on Unix operating systems is an operation that changes the apparent disk root directory for the current running process
... yep just fine...

Did someone succefully boot / reset from DSOrganize ?

Update DB works just fine here Thanks ;-)

 

[cory1492]

it uses bmps and i won't be adding any zip skin support. totally counter productive.
esp since we're stetching memory as it is.

Still not sure how you have a lot of stuff set up (are skin files cached into memory on boot or accessed as needed?), but how about an uncompressed package file for skins?

I'd like to reiterate once again, it'd be bloody awesome to see a OS that uses nitro or similar to stow all the loose bits in one single file - seeks should theoretically be faster than per file accesses.