Now that I have two weeks off from work, I can spend more time digging through the code to look into some of the harder-to-track issues.
- So far I have been unable to figure out why some devices do not show up on the initial client scan. However, the issue seem to be a timing-related one, since one tester reported good results so far by having the initial network mapping delayed by 30 seconds at boot time. I also relaxed various other timings inside the networkmap process to improve reliability.
- Looking at the endless-spinning issue, I'm starting to suspect the issue lies at the webui level rather than the networkmap service level, since the scan flag does get cleared in nvram (indicating that networkmap has finished its job), but the webui does not pick up the cleared flag. I have one person able to reproduce that issue that will help me dig through an actual case, since so far I am still unable to reproduce it on my end. I have a feeling this is related to the hostname lookup code.
- With all that digging through the networkmap code, I ended up uncovering and fixing a few other issues that hadn't been reported yet. For example, Asus's DLNA scan code was half broken - and when it was working, it could delay the scan by as long as two minute per device as it was trying to connect to an incorrect IP.
- Next build will introduce a new type of user scripts, called postconf scripts. Essentially, those scripts are run after the router has dynamically generated a config file (for example, dnsmasq.conf), and before the service gets started. This will allow advanced user to do all kind of manipulations on the generated config file through shell utilities such as sed. A basic example would be replacing a line in the generated config file with another line.
- For the time being I am still working with GPL code 339, as I haven't received anything newer from Asus. I will see if I can get my hands on newer code during the Holidays, but I don't know if that will be possible with people being out of the office during that period. We'll see.
The current plan is to have a beta build released in the coming days, mostly to gather feedback on the tweaked networkmap. That will have to wait until I get the chance to spend more time looking at the endless spinning issue.
The focus of that build will be bugfixing. Once out of the door, next release will most likely be focusing on integrating the next GPL code from Asus once it's available.
- So far I have been unable to figure out why some devices do not show up on the initial client scan. However, the issue seem to be a timing-related one, since one tester reported good results so far by having the initial network mapping delayed by 30 seconds at boot time. I also relaxed various other timings inside the networkmap process to improve reliability.
- Looking at the endless-spinning issue, I'm starting to suspect the issue lies at the webui level rather than the networkmap service level, since the scan flag does get cleared in nvram (indicating that networkmap has finished its job), but the webui does not pick up the cleared flag. I have one person able to reproduce that issue that will help me dig through an actual case, since so far I am still unable to reproduce it on my end. I have a feeling this is related to the hostname lookup code.
- With all that digging through the networkmap code, I ended up uncovering and fixing a few other issues that hadn't been reported yet. For example, Asus's DLNA scan code was half broken - and when it was working, it could delay the scan by as long as two minute per device as it was trying to connect to an incorrect IP.
- Next build will introduce a new type of user scripts, called postconf scripts. Essentially, those scripts are run after the router has dynamically generated a config file (for example, dnsmasq.conf), and before the service gets started. This will allow advanced user to do all kind of manipulations on the generated config file through shell utilities such as sed. A basic example would be replacing a line in the generated config file with another line.
- For the time being I am still working with GPL code 339, as I haven't received anything newer from Asus. I will see if I can get my hands on newer code during the Holidays, but I don't know if that will be possible with people being out of the office during that period. We'll see.
The current plan is to have a beta build released in the coming days, mostly to gather feedback on the tweaked networkmap. That will have to wait until I get the chance to spend more time looking at the endless spinning issue.
The focus of that build will be bugfixing. Once out of the door, next release will most likely be focusing on integrating the next GPL code from Asus once it's available.
