BnetDocs: Redux?

[Barabajagal]

Right-o!

Edit: Just finished formatting all SID packets!

And when are you gonna make it so packets are one line each for send and receive?

[Don Cullen]

Right-o!

Edit: Just finished formatting all SID packets!

And when are you gonna make it so packets are one line each for send and receive?

Your perseverance is mind-boggling and impressive! Thank you.

As for them being on one line-- well, you can see a discussion thread between me and one of the editors on that topic:

http://bnetdocs.dementedminds.net/forums/viewtopic.php?t=11

The issue here is: if I combine it into one line, it'd cause the side navigation menu to take up too much room horizontally. So the problem here is: do we push it horizontally or vertically? Or is there a way to avoid doing both? Especially without resorting to frames?

[Barabajagal]

I still don't see what your problem with frames is. This is a perfect example of when they SHOULD be used, and you're not taking advantage of a feature designed exactly for this purpose. The width of the side is already not enough, and many of the packet names already exceed the size of the space you've alloted them.

[Don Cullen]

I still don't see what your problem with frames is. This is a perfect example of when they SHOULD be used, and you're not taking advantage of a feature designed exactly for this purpose. The width of the side is already not enough, and many of the packet names already exceed the size of the space you've alloted them.

I added one more line to the dark.css file. I see you're using the dark theme, so you should be seeing the changes. Anyone else using different themes will not see the changes unless they use one of the provided pre-set themes. This is an alternative to using iframes. Go to the site, expand one of the packet groups, and you'll see what I'm talking about. I don't like the solution as it's ugly, but it'll have to do for now. Best of both worlds, basically.

As to when I'll modify it so it'll display it in that format, I'll do it after I finish the commenting script. You can see it's already on my to-do list.

[Joe[x86]]

Going through the D2GS packets real quick and de-rawing some.

I flagged these packets as research, since some fields are unknown:
0x1E D2GS_SETWORDATTR (see below)
0x1F D2GS_SETDWORDATTR (see below)
0x26 D2GS_USEBELTITEM
0x38 D2GS_NPCTRADE
0x49 D2GS_WAYPOINT
0x51 D2GS_WORLDOBJECT
0x7A D2GS_LOGONRESPONSE

Various problems/questions:
On 0x49 D2GS_WAYPOINT, what does 'Level' mean?
The 'protocol headers' document is missing.
The 'sizes and types' document is missing.
The D2GS packets are out of order by ID.
On 0x1E D2GS_SETWORDATTR, is the first field supposed to be a WORD? I've marked this as research until someone fixes it.
On 0x1F D2GS_SETDWORDATTR, is the first field supposed to be a DWORD? I've marked this as research until someone fixes it.
When you click the [+] next to W3GS anything, you get a 404 error.

[Don Cullen]

x86] link=topic=17003.msg172529#msg172529 date=1189401766]
Going through the D2GS packets real quick and de-rawing some.

I flagged these packets as research, since some fields are unknown:
0x1E D2GS_SETWORDATTR (see below)
0x1F D2GS_SETDWORDATTR (see below)
0x26 D2GS_USEBELTITEM
0x38 D2GS_NPCTRADE
0x49 D2GS_WAYPOINT
0x51 D2GS_WORLDOBJECT
0x7A D2GS_LOGONRESPONSE

Thanks.

Various problems/questions:
On 0x49 D2GS_WAYPOINT, what does 'Level' mean?

I assume Level refers to which level the player is on? I barely even played Diablo 2, but from what I remember, there would be several levels to specific areas of the game, such as the catacombs -- they were accessed via stairs. Perhaps Level refers to something else? I'm not sure. But I agree more elaboration on that term is needed to avoid obscurity.

The 'protocol headers' document is missing.
The 'sizes and types' document is missing.

Actually, they're not missing. They just haven't been added. You can add them yourself since you also have ability to add documents. I did not have access to the original database for the original BNETDocs, so I was forced to harvest based on the html of the mirrors of the original BNETDocs, and my code only harvested the packet data, not the documents. So as a result the documents have to be re-added manually.

http://bnetdocs.dementedminds.net/old/

You can view the original documents/packets there and compare them against the current ones if you notice missing documents/packets/data.

The D2GS packets are out of order by ID.

Ahh... Thanks for pointing it out. Fixed.

On 0x1E D2GS_SETWORDATTR, is the first field supposed to be a WORD? I've marked this as research until someone fixes it.

No, it's supposed to be a byte. Refer to the second field, it's a WORD, as the packet alludes to.

On 0x1F D2GS_SETDWORDATTR, is the first field supposed to be a DWORD? I've marked this as research until someone fixes it.

No, it's supposed to be a byte. Refer to the second field, it's a DWORD, as the packet alludes to.

When you click the [+] next to W3GS anything, you get a 404 error.

Interesting. Do you have javascript disabled?

[Barabajagal]

I don't think you understand. I don't want an IFrame, nor sidescrolling. That's not the point I'm trying to make. The point is reloading that ENTIRE list every time I view a page is a big annoyance.

My bandwidth is now maxed out, so I'm on a slowed-down connection now, at 16 KB/s MAX, and half the time my HTTP requests are lost. Loading one page on your site now takes an average of 15 seconds. As such, I won't be doing any more editing for the next month (that's how long it takes for the restriction to go away). If you used frames, it would only have to load the content, instead of EVERYTHING each time. Not only that, but you could have a vertical scroll on the packet list so you wouldn't have to scroll away from the current content to see other content. Does this sound like I'm repeating what I said earlier? I AM -.-

[Don Cullen]

I don't think you understand. I don't want an IFrame, nor sidescrolling. That's not the point I'm trying to make. The point is reloading that ENTIRE list every time I view a page is a big annoyance.

My bandwidth is now maxed out, so I'm on a slowed-down connection now, at 16 KB/s MAX, and half the time my HTTP requests are lost. Loading one page on your site now takes an average of 15 seconds. As such, I won't be doing any more editing for the next month (that's how long it takes for the restriction to go away). If you used frames, it would only have to load the content, instead of EVERYTHING each time. Not only that, but you could have a vertical scroll on the packet list so you wouldn't have to scroll away from the current content to see other content. Does this sound like I'm repeating what I said earlier? I AM -.-

Your bandwidth is maxed out? What did you do to max it out? What connection are you on (cable, dial up, dsl, satellite, etc)? I'm guessing dial up since you're downloading at 16 KB/s.

Edit: maxing out your bandwidth, on a dial-up, would be very abusive of any ISP. I suggest switching to a different ISP provider that isn't restrictive. When I was on dial up (after first moving to Fremont, cable hadn't switched here yet and Comcast was taking a long-assed time to switch), I was using NetZero's $20/mo unlimited access plan. Never had a problem with bandwidth, and my sole complaint was that it was so damned slow. But then that's what I get for using dial up (56k modem). It was heaven when cable finally switched on.

It takes me about one second to load BNETDocs Redux. The site is served off a virtual dedicated server, so the site isn't local for me. Still takes about one second to load. That's on a 6 Mbps cable connection.

[Joe[x86]]

Ah, you're right. NoScript was blocking JavaScript, so that's fixed now that I've set it to allow. Still, you might want to give a "JavaScript Disabled" error message instead of the 404.

Tangent, but dialup runs at 56kb/s (lowercase kb), which equates to 7KB/s (uppercase KB). Personally, I use HughesNet Satellite since that's all that's offered in my area, and they have a FAP (Fair Access Policy) which limits me to 375MB per day (200MB with the default package), and after exceeding that limit, I'm restricted to about 3KB/s.

Back on topic, could you also hide the Generate Code sections when you hide the packet lists? That'd be useful.

In 0x09 SID_GETADVLISTEX, is it possible for Conditions 1 and 2 to be a DWORD, and 3 and 4 to be a QWORD, hence 2 and 4 being null? Also, what is sent for the three string fields? Since list count is a DWORD, couldn't you specify to list 0xFFFFFFFF games?

[Don Cullen]

x86] link=topic=17003.msg172542#msg172542 date=1189434777]
Ah, you're right. NoScript was blocking JavaScript, so that's fixed now that I've set it to allow. Still, you might want to give a "JavaScript Disabled" error message instead of the 404.

I'm not sure why it's giving you a 404. The hyperlinks are supposed to direct to #, if clicked. In other words, nothing should happen if you click on the [ + ] link since javascript is disabled. I disabled javascript in my browser to test it, and nothing happened, which exactly is what it should be like if javascript is disabled. I'm surprised your browser attempted to actually obtain the # file on the server...

Tangent, but dialup runs at 56kb/s (lowercase kb), which equates to 7KB/s (uppercase KB). Personally, I use HughesNet Satellite since that's all that's offered in my area, and they have a FAP (Fair Access Policy) which limits me to 375MB per day (200MB with the default package), and after exceeding that limit, I'm restricted to about 3KB/s.

Damn, pardon my language, but that's bullshit. I'm surprised they'd limit your bandwidth. I've downloaded over 2 gigs in one day, and Comcast didn't even smack my hand for it. Thanks for that lil' info. Good to know.

Back on topic, could you also hide the Generate Code sections when you hide the packet lists? That'd be useful.

Hmm, I don't see why not. Will do.

In 0x09 SID_GETADVLISTEX, is it possible for Conditions 1 and 2 to be a DWORD, and 3 and 4 to be a QWORD, hence 2 and 4 being null? Also, what is sent for the three string fields? Since list count is a DWORD, couldn't you specify to list 0xFFFFFFFF games?

I would look at the packet content now, but the server is down right now for maintenance. When I'm finished updating the server, I'll take a look at it. It'll take me about 20 minutes to finish maintenance.

[Joe[x86]]

The reason my ISP would limit my bandwidth is because Satellite internet is like one big huge pipe that everyone is sending their stuff through. If I'm sending too much of my stuff through the pipe, there isn't enough room for everyone else's stuff, so I need to be stopped.

EDIT -
By the way, the server is still down. What gives?

[Don Cullen]

x86] link=topic=17003.msg172553#msg172553 date=1189443427]
The reason my ISP would limit my bandwidth is because Satellite internet is like one big huge pipe that everyone is sending their stuff through. If I'm sending too much of my stuff through the pipe, there isn't enough room for everyone else's stuff, so I need to be stopped.

EDIT -
By the way, the server is still down. What gives?

Yea, I know how bandwidth works. I'm a web host. I'm just surprised they restrict that much bandwidth, most other broadband ISPs have a much higher cap on bandwidth. Perhaps their bandwidth capacity is much more limited that I had presumed.

As for the site still being down--  I finished most of the maintenance. The details of the maintenance are:

wipe server, reinstall everything (necessary because some files were missing/corrupt, preventing me from properly upgrading to php5/mysql5 -- upgrade goes fine, but plesk breaks). i just finished wiping everything, and am nearly done reinstalling.

One unfortunate consequence of doing this is the ip address of the nameservers that the site uses has changed. The end result is, it'll take about 12 to 24 hours for the nameserver ip addresses to propagate throughout the internet. The maintenance itself however is finished. I'm doing all I can to speed up the process.

[Camel]

In short, no, "This message's official name is not known, and has been invented." isn't necessary.

Why would you remove that message? It's informative.

[Camel]

One unfortunate consequence of doing this is the ip address of the nameservers that the site uses has changed. The end result is, it'll take about 12 to 24 hours for the nameserver ip addresses to propagate throughout the internet. The maintenance itself however is finished. I'm doing all I can to speed up the process.

[offtopic]
Many ISPs run DNS servers that cache addresses for two weeks. There's nothing preventing a DNS from caching for even longer (except that it breaks standard procedure). It's not safe to say that the IP address will propogate throughout the internet within 24 hours, although most users will see it within that time period.
[/offtopic]

[Don Cullen]

Server is back up, maintenance is nearly finished, but everybody can go ahead and use Redux. Thanks for your patience.

If you're unable to access it, go to the command line (Windows, I'm not sure what the Linux or Mac equalivent is) and type without the quotes: "ipconfig /flushdns"

If even after that, you're still unable to access it, give it time. The ip address has to propagate to your area for it to take effect. It usually takes 12 hours or less to propagate nationally, 24 hours or less to propagate internationally.

In short, no, "This message's official name is not known, and has been invented." isn't necessary.

Why would you remove that message? It's informative.

I didn't remove it. He asked if he was required to add that as a footnote. I said it wasn't required. It'd be appreciated if it was used, but it isn't mandatory. The general rule of thumb is: if the official name is already known, use it. Otherwise, invent one for reference use in the meantime until otherwise known.

If you feel it is necessary, feel free to append that footnote to the ones whose name isn't officially known.

As for those rogue servers that feel the need to break standard, nothing I can do.