I like the fact that one of the goals of this tree will be to help users, depending on
what features the user may require, quickly eliminate which builds to avoid.
I realize this is just a draft decision tree and my suggestions are for consideration only,
they are not meant to criticize the great work you’ve done so far.
Suggestions:
After “Q1”, since this decision tree will most likely be read by new users shouldn’t “A1.1:” start with “Yes” instead of “No”?
Q7 is missing
For Q8, use an actual build for an example; i.e.: (dd-wrt.v24-XXXXX_NEWD-2_K2.6_big-nv60.bin )
or (dd-wrt.v24-XXXXX_NEWD-2_K2.4_big-nv64.bin)
R8.1: “Proceed to Q”. Q what?
Q10 seems a little confusing to me as I can’t find a Broadcom .v26 build in the database. Is that “Q” correct then?
You should, in one of your notes, strongly suggest that anyone using this decision tree should make a good effort to determine,
at the least, there router’s chip manufacture, memory size, and cfe.
I’m not sure why you need to know what cfe your router has, as it made no determination
as to which build I decided to use. Maybe I missed that step.
End suggestions.
Anyway I think you’ve started something great here and I’ll be glad to offer any assistance I can, which may not be much. _________________ WNDR3400v1 using dd-wrt.v24-24461_NEWD-2_K3.x_mega-nv64k.bin
WNDR4500v1 OEM Firmware Version V1.0.1.40_1.0.68
“Incredibly handsome and master of all villainy!”
“Ow! My giant blue head!”
I know that this would take a LOT of time and dedication but I think it would eliminate a lot of questions and save a lot of time in the long run for not only beginners, but repeat users on different hardware. I think there are enough supporters to even pay for a host for a separate web page providing it doesn't violate anyone's TOS. As a matter of fact, I know enough language to write the code. It would be rather simple with the provided information.
I'm not sure what sort of code your talking about other than html. I did not want the system to be automated. I rather like the idea of having all the questions on one page, even if the user does not read all questions all the time, it gives him/her an idea what is out there. In addition the tree is likely to need upgrading from time to time. With the spread-out-page, the gurus can get a quick overview, otherwise upgrading the tree is too much like debugging a program. Depending on the size, I could split it into 2 or 3 html pages. If what I understand is correct the older routers have a much simpler tree than the old and might go on a separate page from the newer ones. As far as code is concerned, I was thinking about hacking together a quick python script that renumbers the questions. It is a pain in the ass having to renumber the Q&A's when inserting/removing a question in the middle of the list.
Murrkf wrote:
Another issue is that there has never been a perfect build. All have bugs. Build choices depend on what a person is going to use it for/which bugs they can live with. VPN or No? Do you need loopback? Repeater? All have different optimal build choices.
I never intended the tree to point to the one true build. Instead I was was wanting to re-normalize the info contained in the peacock. The tree could point to threads with relevant information. The main objective is to avoid confusing the new user with information that is not relevant to there situation. An example is, users of k24 routers don't have to worry about nv60k vs nv64k, because it is not relevant to there situation.
Thanks deviouswun for your input. Some of the errors in the example tree number 3 were due to screw-ups when re-numbering the questions. I guess I used step 7 for the warning and step 8 was the question step. Maybe for readability reasons, I should have warning 7 followed by question 7, rather than the way I did. Also after your post, I fixed R8.1.
In the real tree I don't intend to use single builds. The only reason I used in the example tree, is because that is the build I used. I did not know exactly why I choose that one. I had to ask, "Will This build brick my router?", and the answer was no and so I installed that one.
deviouswun wrote:
I'm not sure why you need to know what cfe your router has, as it made no determination as to which build I decided to use. Maybe I missed that step.
So you did not get that one? Q5 and the questions leading up to it seem to be the only thing without errors. As a test of the tree for understandability please re-read Q5 and look at which questions leading up to it (not every user will get to Q5). Then look at the results for Q5. Now, can you answer "why you need to know what cfe your router has?"? Also, who needs to know it?
After looking over the peacock thread and some other threads. I built up this idea that I think is true, but I will need someone to verify. When it comes to selecting a firmware build, there are two main camps of routers based on whether they use kernel 2.4.x or 2.6.x . Those that require k24 builds and those that require k26 builds and beyond. Routers of the k24 variety have a much smaller set of questions and concerns when finding a build that is compatible. That is once you get past the build size and initial vs post-initial flash. If my hypothesis is true, the k24 vs (k24 and 3.x) would be a natural place to create a major split in the decision tree.
The following paragraph is based on my understanding of DD-WRT development. I would like expert opinion to confirm the next paragraph or correct it if my understanding is flawed.
It seems that development of DD-WRT has basically slowed to a crawl or stopped entirely for k24 dependent routers. As a result, when it comes to the "minimum required build" k24 users will look the selection process differently from k26 and beyond users. The way I see it, all BrainSlayer builds from 2010 and beyond exceed the "minimum required build" for all k24 dependent routers. From the users point of view the "minimum required build" is not necessity question to worry about as long as they choose a BrainSlayer build from 2010 and later. The question for users of k24 dependent routers is, "Do I select VINT or NEW-D". On the other hand people not using a k24 dependent router need to look at the actual build dates. The major split would be made by a question very similar to Q10 from my third attempt at a decision tree.
So what is the expert opinion on whether I have an accurate assessment of the situation?
If this true, then I could be ready to create a decision tree that could be in theory workable for people with kernel 2.4 dependent routers.
Another issue is that there has never been a perfect build. All have bugs. Build choices depend on what a person is going to use it for/which bugs they can live with. VPN or No? Do you need loopback? Repeater? All have different optimal build choices.
I never intended the tree to point to the one true build. Instead I was was wanting to re-normalize the info contained in the peacock. The tree could point to threads with relevant information. The main objective is to avoid confusing the new user with information that is not relevant to there situation. An example is, users of k24 routers don't have to worry about nv60k vs nv64k, because it is not relevant to there situation.
So....again, if that is all you are trying to accomplish, the decision tree is simple: Follow the wiki for your router!
There are few modern routers that use k24 builds. All routers from the last few years that I can think of use k26, and most people who come here are now using k26 builds on their more modern routers. Development continues on these routers, but the complexity of a "one" firmware for all has raised numerous bugs. There are k24 builds in the many recent build offerings.
You appear to be now suggesting dealing more with build type than build number. If you keep those concepts separate, your task will be easier and make more sense. _________________ SIG:
I'm trying to teach you to fish, not give you a fish. If you just want a fish, wait for a fisherman who hands them out. I'm more of a fishing instructor.
LOM: "If you show that you have not bothered to read the forum announcements or to follow the advices in them then the level of help available for you will drop substantially, also known as Murrkf's law.."
