I would like OpenWRT to add some special features.
1-Support for the unsupported SOHO ISP routers, and unsupported VOIP and wifi chips.
2-better controllable and stable VPN clients/Servers, Proxy Client/Servers, Obfuscating Proxy clients/Servers.
3-Full control over all types of modems [xDSL, LTE, Cable...etc],with GUI and CLI.(and DLM/DSLAM override
4-LUCI GUI modules for CLI only packages.
5-Per user Bandwidth/Quota monitor and limit.
I dont have the skills nor time nor money to complete these tasks, so i want cost estimation and time table to be able to crowd fund these tasks using Patreon.
Note: if you have any features you want to implement please share your thoughts and ideas and i will probably add them to the list!
Thanks
Update:
*-LUCI based opt in release checks - ( all vs security critical etc.. ) and various alerting triggers. ( android LAN msgs, email, just in LUCI.. etc. )
*-CPU efficacy optimizations ( nss, scheduler etc. )
*-Security inspection agent - ( distributed snort/banip/etc - for all of us - and with lower overhead )
*-Advocacy and partnership facilitation with vendors - including open-sourcing key low level code... in addition to many more mutually beneficial and practical goals + actions.
This would likely cost millions of dollars - if possible whatsoever. After all, there's a reason they're not supported. For CPUs, the common issue is that there is no code for it in the upstream Linux. Also, you made a very broad request.
-pimp my LUCI based policy routing
-Upgrade keep settings, crossversion/incompatible merge magix
-PC bi-directional firmware + config versioning + autobackup
-LUCI based opt in release checks - ( all vs security critical etc.. ) and various alerting triggers. ( android LAN msgs, email, just in LUCI.. etc. )
-CPU efficacy optimizations ( nss, scheduler etc. )
-Security inspection agent - ( distributed snort/banip/etc - for all of us - and with lower overhead )
-Advocacy and partnership facilitation with vendors - including open-sourcing key low level code... in addition to many more mutually beneficial and practical goals + actions.
With the modems, you kinda might have to be realistic and just pick the top 3-5 partially / to-possibly-be chipsets.... Same with the devices... drop the bottom 65% under-resourced or just plain pidgeon-holed models ( broadcom etc. )
Especially on 1 and 3, I would prioritize your list, after understanding the reality of their implementation cost. In many cases the broad list of features will be stymied by one or more of the following:
Proprietary information and/or licensing required (chips and intellectual property both)
Intentionally locked-down interfaces for carrier management
Locked-down boot loaders (such as firmware signing)
No Linux support exists
No Linux support possible
Insufficient resources to support Linux (for example, kernel exists for SoC, but only 16 MB of RAM)
2- not all vpn clients are stable, OpenVPN is working but the gui interface is buggy as hell!!!, obfspoxy is buggy also and hard to configure, iodine, ptunnel ......etc
3-full control over low level config with the ability to override settings enforced by DLM/DSLAM including SNR on upstream and downstream, G.INP depth, SRA, Interleaving depth, upstream and downstream data rate (my isp is enforcing crazy SNR and crazy sync rate which is a lot more my actual speed), Lazy ISPs who don't have fastpath profiles and refuse to manually configure it, for gamers and enthusiasts the latency is a priority.
4-Obviously all network related CLI only software !!!
5-yes couple of solutions exist but they are troublesome to config and they are buggy as hell, i use wrtbmon and every now and then the database goes crazy and show fake numbers.
Wow...I didn't think that was obvious. That's a lot of programming for all CLI syntax available. For example, I find it hard that many people want to to program an app to perform things like ls, cd, cp, rm, etc. - although a file manager app would be a good solution for that (even though mc already exists).
I would NOT suggest that these requests are impossible given “a” budget - although definitely a very, very large one. After all, the hardware manufacturers provide consumer based tools that make these things “point and click” - I even have a Cisco switch in the prosumer category (26 port with level 3 and other enterprise features) that does this.
However, please remember a few things
This is an open source project - absolutely no prospect of a financial return exists.
Hardware support requirements are extremely diverse. Although they’re all Linux based, the underlying chipsets are mix-and-match, large in number. This produces Cartesian like support needs.
Much of the hardware doesn’t have open source drivers, so that has to be reverse engineered just to get the basics working - like WiFi support.
Most people who purchase the underlying devices have no clue what OpenWrt is, much less an understanding of why they should care.
I could go on, and I don’t want to step on your enthusiasm, but I think we need to be realistic about what can be accomplished. OpenWRT often frustrates the heck out of me, but I still use it because I get features and functionality for which I’d otherwise have to pay thousands of dollars - money I don’t have, and even if I did, I still answer to a higher authority- my wife ;-).
So I’ll go on both cursing the time investment while still being thankful for people willing to invest time and treasure to maintaining and enhancing the tool.
As @lleachii and @gschaub mentioned, your suggestions and requirements don't really form a basis for discussion, as they're both far reaching[0, 1] and very vague - while still aiming for the sky[2, 3].
Most of your requests would quite literally require a budget in the millions, an army of full-time developers (well, more than two dozens) and active cooperation (documentation/ mainline involvement) of the relevant chipset vendors (which isn't going to happen, at least not in the 7 figure (EUR/ USD/ GBP/ CHF) range[4]).
--
[0] e.g. your first bullet point, you want 'all unsupported devices to be supported'
[1] e.g. your third bullet point, overriding the ISPs side of the network (and all of them at once, xDSL, cable, LTE, etc.), I hate to be the bearer of bad news - but the ISP own their end of the cable, and whatever restrictions they impose, the CPE can either accept - or get the fsck of their network (either because the handshake fails in the first place (in most cases), or because the ISP takes legal action).
[2] your fourth bullet point is contradicting itself - and going beyond any reason…
[3] furthermore you are expecting features/ performance of your cheap unnamed hardware (again, every hardware from all vendors) which is far beyond the abilities of the provided hardware, at least the most common (even high-end-) OpenWrt supported targets.
[4] not in aggregate, but for each of them alone, in guaranteed additional business income under the stated conditions - you have to compete against the mass produced (batches in the 8-9 figure range) devices from vendors who do accept the current conditions
Hmm, I know the question wasn't for me but... I have donated an unused router to a developer and last time I checked there was no official, organized way, to contribute money to the core developers. Has this changed?
if you just need better support for xDSL chipsets so you can mess with the line config, you might be better off looking at proprietary modem firmwares, they often provide some unsecured shell over uart (or even telnet), so you could poke around for any interesting-looking commands that control the line driver...
However I don't think we need to discuss here whether anyone would recommend you to mess with these settings, especially if your isp already refused to do so.
They refuse because they are lazy.
btw Lantiq xdsl driver is opensource but have no documentation!
Any programmer can browse the code and add mods and features. also the override can be modded directly into the xdsl driver!
Ok so lets take it slowly in easy small steps, lets focus on lantiq XRX xdsl chips, what can be done here? and how much work is needed? and how much money to raise?!
