pixelserv pixelserv - A Better One-pixel Webserver for Adblock

[kvic]

After googling, I think the "Protocol not available" is a problem associated with the Windows WSL. I was hoping to "keep it simple" and have it run on WSL but that seems not to be the case. I'm guessing Arch linux under WSL may have the same issue but I'll try it another time.

Yeah, I would think it doesn't make a difference. What you'll get is a slightly different user space environment.

WSL is not VM that's fully virtualised and emulated. It's one step beyond the early days' OS/2 and POSIX subsystems but all run on top of Windows NT kernel.

For those interested, here is a good blog and an overview of WSL.

 

[preacher65]

Given a domain e.g. snbforum.com, all ads/trackers on that domain will allow to pass through i.e. not blocked.

Sorry to be a bit late responding, but I would really like this. To me, being able to do this is the single largest benefit of a browser-based ad blocker, and the feature I miss most from diversion/pixelserv etc. There are a small number of sites I like to support, like this one, but I don't want to see content from the advertisers they use everywhere else...

 

[cmkelley]

Sorry to be a bit late responding, but I would really like this. To me, being able to do this is the single largest benefit of a browser-based ad blocker, and the feature I miss most from diversion/pixelserv etc. There are a small number of sites I like to support, like this one, but I don't want to see content from the advertisers they use everywhere else...

Hence why I still use Adblock Plus along with diversion/pixelsrv.

 

[kvic]

Hence why I still use Adblock Plus along with diversion/pixelsrv.

The logics doesn't follow through..

Usually people excessively/abusively block domains in their blacklists. Browsers adblock like Adblock Plus provides no additional blocking nor it can whitelist certain domains because blacklists on routers always prevail.

So the whitelisting must happen on router, and I throw out an idea that pixelserv-tls can very smartly get it done.

 

[cmkelley]

The logics doesn't follow through..

Usually people excessively/abusively block domains in their blacklists. Browsers adblock like Adblock Plus provides no additional blocking nor it can whitelist certain domains because blacklists on routers always prevail.

So the whitelisting must happen on router, and I throw out an idea that pixelserv-tls can very smartly get it done.

So, if I want to support, saySNB forums, I whitelist the ad sites it uses on pixelserv/diversion. But Adblock still blocks them at the browser level, so I turn off Adblock for SNBforums. Now when I'm at SNBforums the ads get through and SNBforums is supported, but when I'm at cnet, those same ads are blocked at the browser level.

Unless I'm missing something ...

 

[kvic]

Does anyone notice the "bad" counter is unusually high on the servstats page in the recent 2.2.1 test versions?

I was debugging and fixed a crash issue last night caused by a "bad" request. I thought it was an iOS client issue at the same time. It brings my attention to the "bad" counter which in my case about 10% of total requests, "req". Normally it should be zero or an extremely low value.

I haven't looked into the details yet but I suspect one or more of the reduction in timeout values cause it.

 

[Protik]

Does anyone notice the "bad" counter is unusually high on the servstats page in the recent 2.2.1 test versions?

I was debugging and fixed a crash issue last night caused by a "bad" request. I thought it was an iOS client issue at the same time. It brings my attention to the "bad" counter which in my case about 10% of total requests, "req". Normally it should be zero or an extremely low value.

I haven't looked into the details yet but I suspect one or more of the reduction in timeout values cause it.

You mean the ucb/req? In my case that is 1245/23127. About 5%.

Sent from my Moto G (5) Plus using Tapatalk

 

[kvic]

You mean the ucb/req?

I meant the "bad" request e.g 203 in this screenshot. The ratio will be bad/req

 

[Protik]

I meant the "bad" request e.g 203 in this screenshot. The ratio will be bad/req

That is 21 in my case. So the bad/req is 21/23302.

 

[kvic]

That is 21 in my case. So the bad/req is 21/23302.

Good feedback. Appreciated.

 

[Zastoff]

In my case (bad)1274/31769

 

[cmkelley]

Does anyone notice the "bad" counter is unusually high on the servstats page in the recent 2.2.1 test versions?

I was debugging and fixed a crash issue last night caused by a "bad" request. I thought it was an iOS client issue at the same time. It brings my attention to the "bad" counter which in my case about 10% of total requests, "req". Normally it should be zero or an extremely low value.

I haven't looked into the details yet but I suspect one or more of the reduction in timeout values cause it.

Using -rc.5 bad/req = 8/24312

 

[jrmwvu04]

Still on rc3. Bad = 812 / 207,362

 

[elorimer]

On .rc3 at 100/5400 over the last 10 hours.

 

[elorimer]

Unless I'm missing something ...

Only this. In my target rich environment I have 3 windows laptops, 2 chromebooks, 3 windows desktops, two tivos, two rokus, three android phones and an ipad. I'm just not going to manage those individually.

From time to time something misbehaves. I now routinely think it might be because something is blocked, and it is easy to turn everything off with diversion, check again, and restart diversion. What @kvic is suggesting would allow me to support as well as block with the same simplicity.

 

[cmkelley]

Only this. In my target rich environment I have 3 windows laptops, 2 chromebooks, 3 windows desktops, two tivos, two rokus, three android phones and an ipad. I'm just not going to manage those individually.

From time to time something misbehaves. I now routinely think it might be because something is blocked, and it is easy to turn everything off with diversion, check again, and restart diversion. What @kvic is suggesting would allow me to support as well as block with the same simplicity.

I think the misunderstanding is, I'm talking about what I do now, kvic is suggesting a possible future change to pixelserve. Unless and until he is able to implement that change, I don't see how else I can allow certain ads on SNB while blocking them elsewhere. Currently Diversion/pixelserve are all or nothing.

Certainly, if he is able to roll the functionality of ABP into pixelserve, then ABP becomes redundant. But that's not the situation today.

 

[Clark Griswald]

Up 5days 16hours
rc. 5 bad= 278/556104

 

[Makaveli]

Mine is at 63

pixelserv-tls 2.2.1-rc.5 (compiled: Dec 15 2018 17:40:07 flags: tls1_3) options: 192.168.1.3

uts    7d 14:28    process uptime
log    1    critical (0) error (1) warning (2) notice (3) info (4) debug (5)
kcc    8    number of active service threads
kmx    20    maximum number of service threads
kvg    1.22    average number of requests per service thread
krq    56    max number of requests by one service thread
req    24299    total # of requests (HTTP, HTTPS, success, failure etc)
avg    694 bytes    average size of requests
rmx    2508 bytes    largest size of request(s)
tav    7 ms    average processing time (per request)
tmx    61 ms    longest processing time (per request)
slh    3091    # of accepted HTTPS requests
slm    20    # of rejected HTTPS requests (missing certificate)
sle    0    # of rejected HTTPS requests (certificate available but not usable)
slc    14597    # of dropped HTTPS requests (client disconnect without sending any request)
slu    5902    # of dropped HTTPS requests (other TLS handshake errors)
v13    3039    slh/slc break-down: TLS 1.3
v12    72    slh/slc break-down: TLS 1.2
v10    0    slh/slc break-down: TLS 1.0
uca    0    slu break-down: # of unknown CA reported by clients
ucb    3063    slu break-down: # of bad certificate reported by clients
uce    0    slu break-down: # of unknown cert reported by clients
ush    71    slu break-down: # of shutdown by clients after ServerHello
sct    157    cert cache: # of certs in cache
sch    21740    cert cache: # of reuses of cached certs
scm    15    cert cache: # of misses to find a cert in cache
scp    0    cert cache: # of purges to give room for a new cert
sst    20    sess cache: # of cached TLS sessions (for older non-RFC5077 clients)
ssh    7980    sess cache: # of reuses of cached TLS sessions
ssm    83    sess cache: # of misses to find a TLS session in cache
ssp    0    sess cache: # of purges to give room for a new TLS session
nfe    1426    # of GET requests for server-side scripting
gif    235    # of GET requests for GIF
ico    9    # of GET requests for ICO
txt    1100    # of GET requests for Javascripts
jpg    53    # of GET requests for JPG
png    0    # of GET requests for PNG
swf    0    # of GET requests for SWF
sta    7    # of GET requests for HTML stats
stt    0    # of GET requests for plain text stats
ufe    221    # of GET requests /w unknown file extension
opt    2    # of OPTIONS requests
pst    222    # of POST requests
hed    0    # of HEAD requests (HTTP 501 response)
rdr    367    # of GET requests resulted in REDIRECT response
nou    0    # of GET requests /w empty URL
pth    1    # of GET requests /w malformed URL
204    0    # of GET requests (HTTP 204 response)
bad    63    # of unknown HTTP requests (HTTP 501 response)

 

[Twiglets]

Mine is 1625/48215

 

[vesalius]

Rc5 7 day uptime and bad/req is 11/11661