We have been working to revise the LEDE Project home page.
Draft home page: https://lede-project.org/playground/start2
Draft Showcase Page (detailed list of LEDE features, linked from new home page): https://lede-project.org/playground/why_use_lede_full_list
Please leave comments for this draft here. Thanks!
LEDE has reduced latency/lag and increased network throughput via bufferbloat control algorithms.
Wi-Fi improvements support higher data rates and airtime fairness between stations.
LEDE supports various Quality of Service implementations improving internet quality for connected devices.
Items 1 and 3 (bufferbloat control & QoS implementations) refer pretty much to the same thing, They might be combined.
To my knowledge, item 2 "airtime fairness" only concerns ath9k driver (mostly older routers), but not mvebu, ath10h or mtxxxx drivers (newer routers). So you might specify the wifi driver there.
I like the summary contained in the "Why use LEDE" section. It might be further improved by hyperlinking more of the statements to anchors within the "why_use_lede_full_list" document. It was not immediately apparent to me that there was more to read about security, for example.
The security section in particular should be more strongly backed by the "why_use_lede_full_list" document. I do not think the statement "thanks to its Linux OS which is unaffected by many common attacks" has much merit. The security community continuously finds many vulnerabilities both in the Linux kernel and its user-space software. I would like to see more details. How does LEDE mitigate zero-day vulnerabilities both in the kernel and user space? Does LEDE favor type-safe languages? Is ASLR is on? Stack canaries? Does LEDE prefer software packages whose designs include careful security considerations (e.g., qmail/postifx vs. sendmail)? What does LEDE do better than other Linux firmwares or other firmwares in general? What does LEDE do to encourage the proper configuration of the software it provides, to include strong authentication and encryption?
"LEDE is actively updated" could link to a description of how users can ensure their LEDE installs stay up-to-date.
I think one of the main things that the homepage should have is a link to the status of the OpenWRT/LEDE merge. I'm running LEDE now, and I've seen a discussion about merging back with OpenWRT, but I am confused whether I should now look at the LEDE project or OpenWRT for the merged project (well I guess one will point to the other eventually), and I'm also confused that both projects did/are doing a new release just now. Will the merge happen after these 2 releases are done?
Thanks to all for these comments.
@edwintorok - We've been writing the home page with the upcoming release of LEDE in mind. Although I am deeply hopeful that the two projects can merge cleanly and congenially, the "home page team" has been focused on what we can say about LEDE that's true and current, and will likely be true when we ship the first release.
I would welcome the opportunity to place more explanatory text from the people involved in the merger discussions for the home page.
@flyn - Yes - links from the home page to the sections of the why_use_lede_full_list page are in the wings. You ask good questions about the security section as well. Can you give a few bullet points that we might use (even to replace what we have currently written?)
@hnyman - I suspected that Bufferbloat control and QoS were substantially the same point. I think we'll have to soften the language re: airtime fairness on newer routers. Dang.
@all - Can you help us come up with additional reasons to use LEDE? After all, you're spending hours (days?) to bring it to fruition... Tell us why!
LEDE reduces latency/lag and increased network throughput via bufferbloat/QoS control algorithms. sound OK to you? I think it would make sense to mention QoS as it may be more familiar/translate better than "bufferbloat" to some users.
@richb-hanover -- thanks for removing unneeded QoS entry so quickly.
No hidden backdoors left by hardware vendors. was originally
No hidden backdoors or bugs left by hardware vendors. -- it's no secret that occasionally vendor firmware contain bugs (just google D-Link bugs). Did it not read well enough? I thought the message that there're no vendor-derived bugs in the code was clear.
Sounds better. QoS needs to be mentioned, as it is much more common than bufferbloat.
I would add "optional" as the SQM package needs to be installed. And "increases" instead of "increased"
- LEDE reduces latency/lag and increases network throughput via optional bufferbloat/QoS control algorithms
Yep. You need to do that: the whole airtime fairness things was dropped from the release today by https://git.lede-project.org/?p=source.git;a=commit;h=c296ba834db4ce8c71e0ad7030aab188fe60b27b
> several users reported stability issues. Revert it now so it doesn't cause issues for the upcoming release
Alas. Thanks for the heads up...
Just my two cents: I'm really not comfortable with the style of that page's language. It sounds too much like a Kickstarter steadfastly making loft promises. At the same time it is already hesitant to go full-on advertisement. So now it's neither a good ad nor a good project abstract. IMHO.
But if you really must, please at least do not deal in absolutes: "LEDE works better", "LEDE is more secure", "LEDE is resistant", "any vulnerabilities are[sic] closed." You do not know that, and you certainly cannot promise it. Sure, it may be like that for you, and may be like that at the moment. But the forums are full of people whose experience has been worse than stock firmware (for example with NAT or Wifi speeds), and we still don't even have any release cycle for people to rely on to get these "vulnerabilities closed." And most importantly, you cannot make promises to the world on behalf of a group of volounteers.
Again, just my two cents.
I pointed that out earlier as well - so maybe I should put it more bluntly: we are not looking to proselytise; we are not peddling any wares; we are not selling anything.
So please, do not communicate like we are.
If i were someone looking for an alternative firmware, I'd skip LEDE if that kind of sales pitch was my first encounter.
Don't make promises you cannot back up. Don't be corporate.
We all have the best interest of this project at heart, and what you do is valued, but don't let your enthusiasm for this project interfere with the neutral and 'dry' approach (for lack of a better word) such a page should have.
I fully agree with metai and Borromini.
@hnyman's note (#8) plus the sincere comments above make me change my opinion about the page.
I had been hoping that we would be able to ship the first release with at least two killer features that are not in stock firmware. (The bufferbloat/QoS fixes are in and solid, but the airtime fairness only applies to ath9k chips, and got pulled from the first release since it seems still to cause problems.)
So as I re-read it, I realize the current wording in start2 overstates our capabilities at this time. It talks about facilities I hope to get, but it would be misleading say it now.
I am willing to step back from the home page discussion to let someone else manage it. My request is that it have the elements I mentioned in LEDE Showcase page, which are (slightly edited):
- An introductory paragraph that tells what LEDE is about
- A note that says LEDE is still in flux (that can go away shortly after the release.)
- A list of "teasers" in a "Why use LEDE" section with...
- A link to the longer "Reasons to Use LEDE" page
- An invitation to have people join the project
With the recent legal changes in UK, highlighting dnscrypt-proxy for UK-based users could be a priority.
I took a break from thinking about the home page. As I come back to it, I can see that my "aspirational" writing was too far advanced (too much "marketing speak") for people's tastes, and for the actual state of the project. That writing needed to be dialed back.
In reviewing all the notes in this topic. I come to the conclusion that @bobafetthotmail had the best basis for the home page. I've revised start2 to incorporate his basic layout and bullets, with a place for links to the corresponding "Reasons to use LEDE" page that expands on it.
I welcome comments on this revision. Thanks.
Overall I like the wording, just some more nit-picking: the page states
Security: LEDE's standard installation is secure by default, with no poor passwords [...]
Given that we ship with no password at all we should probably not mention poor passwords but elaborate on the fact that wifi is disabled by default or something.
Apart from that I do like it!
Ok, I've voiced my opinion, edited the page -- I give up on the fight with terminology like "people believe" and "people find". I strongly believe it does not belong to any of the pages of any technical project. State the facts and let users come to their own conclusions.
PS. Whatever "reasons for LEDE" are included in the start page, they should be duplicated in the "reasons to use LEDE" page. In other words, the start page should include the subset of items listed on "reasons to use LEDE" and maybe the better approach would be to pick 5-6 most important ones rather than trying to come up with something else.
@jow - I tweaked the wording about passwords and wifi off. Is it better?
@stangri - thank you for your concession on "people find" and "people believe." I think it's good balance between saying "LEDE is the Best!" and leaving people to guess about the benefits. I think this text can serve for the first stable release (the branch may occur this weekend.)
You're also right that the Reasons to Use LEDE page should have the same section headings as the bullet points on the home page. I took a shot at harmonizing the bullet items/headings between those two pages.
@all - In order to be ready for an onslaught of new visitors who will come when we publicly announce the first release candidate:
We need to agree on the content of the home page. I think there's broad willingness to use the format of the page at start2. But it would benefit from a final editorial pass.
The "Reasons" page needs a similar editorial pass.
The "Reasons" page also has a bunch of links to non-existent pages. What's the consensus for handling those links?
Thanks all for this hard work!
There's a difference between concession and giving up the fight. I've never advocated "LEDE is the best", what I suggested was "LEDE is superior to most of the factory firmware in the following: ... (shortened list of why use lede)."
I still strongly believe that the way it's phrased right now is weak and reminds me of the herbal supplement sites or more recently a certain political figure (people say... some people...) -- instead of stating the facts you rely on opinion of some unknown people.
I've done everything I could on this, including modifying the wording and yet it's changed back now. So I give up.
There can be only one answer to this: Create those missing pages and fill them with life!
Yes, I know, easier said than done...