Seeking an OpenWRT expert to help us build custom OpenWRT firmware and expand support for small devices. Specifically looking to support TP-Link TL-WR840N. The developer will need to flash the firmware to a local device, test functionality, and add the device to the OpenWRT project's supported devices so that it will be available in the menuconfig. Apply on [Moderator Edit: removing obsoltete link]
Just a sidenote
if your end goal is to be able to build images and directly flash them to boards with the original flash chip (4 MB) this is not realistic. Usually this process includes replacing of the flash chip which either requires you or the developer you hire to have physical access and tools to physically swap the chip.
basically either you doing 20% of the work or you sending the device to be physically modified
whether or not it is possible on 4 MB flash to have latest openwrt, it would required massive amounts of change to the code and be a much larger project
otherwise you would be using an outdated kernel / packages
1 Like
This requirement here:
Conflicts with this statement here:
This is not valid in all cases. There are various options to shrink down an image below 4M, in case certain functionality (to be carefully evaluated) is not required. Most notably here to remove LuCI.
However, this probably not valid any more for very recent version of openwrt (20.x), but at least somewhere for the 19.x branch I did it for various devices.
32MB RAM usually sufficient then for basic routing.
The real problem here is, that the author does not specify the version of the WR840N, not even on my explicit question. So this request seems to be a bit dubious.
Even if it is technically possible to build an image for those devices, the challenge is in getting the platform accepted into OpenWrt.
they are looking for someone to make custom builds, not to add to the project
"available in menuconfig" is arbitrary
32 MB RAM is totally fine for normal size builds on 8 MB flash
I just don't see 4 MB flash as reasonable...
In any case, OP has been made aware of this issue (or non-issue)... they will decide what to do with this info.
3 Likes
Within the OpenWRT project there are varying opinions on dropping support for 4/32 device. Notably LEDE developers have voiced alternatives like small builds with limited features. This is the approach we are taking.
Nothing substantially has changed in the last 10 years that justifies ending support for small devices. We are building custom firmware images with a new control plane so we don't need or want LuCI nor do we need many of the management features. Our priorities are different. We are looking to support as many devices as possible, not support as many features as possible.
You are asking " ... to support TP-Link TL-WR840N" . There are lot of various versions of this device, so which one are you looking at ?
" ... add to supported devices ..." questionable to be achieved. Because of minimum required packages, and kernel itself, permanently growing.
BTW, I applied on upwork, having done such downsizing (also) for some WR840N, but no response yet.
Oh, just be careful with freelancers. I once had such a twist in the code that the hardware refused to work completely and there were difficulties in reflashing. My advice to you, better contact the experts, search criteria can be found in net . For example, try researching the team's background, website, social media. Normal developers should have a bunch of examples of work there and, in general, show the attitude towards their responsibilities.So at least you will get some kind of guarantee and you will be able to sue and defend your rights. With freelancers, this may not work ...
In case, you hire the cheapest freelancers, you might be correct in your concerns.
However, you certainly know, that a flashy website, social media etc. can easily be faked.
Otherwise, it is practically impossible, to fake positive feedbacks for previous projects on a freelancer web site.
I agree with you. Although in such a case it is always better to start over, that is, see the problem and the request that needs to be formulated. That is, to understand who to contact, what budget and everything in this spirit. This is the first item in the review that I found at the time when I was [search] [moderator edit: removing obsoltete link] for some of my work questions. Есть конечно в статье и более актуальные советы, но это уж сами решите, когда будете показывать результат работы.
I find it interesting that no mention is made of ongoing support. You aren't planning to embed this firmware in devices you're planning to sell commercially for which no possibility of addressing later discovered CVEs exists, are you?
This is a 2.5 year old thread, so it is likely not relevant anymore. As such, this topic will now be closed.