-->
Page 2 of 7

Re: Is any one still using the ESP-01 modules?

PostPosted: Sun Sep 04, 2016 9:38 pm
by forlotto
If you really needed to get things working why not just spend $11.20 USD and purchase some 4MB flash chips here http://www.aliexpress.com/item/Free-Shipping-50PCS-W25Q32FVSSIG-W25Q32FVSIG-instead-of-W25Q32BVSSIG-IC-SPI-FLASH-32MBIT-8SOIC-25Q32-W25Q32/1850554689.html?spm=2114.01010208.3.40.0Xq8ie&ws_ab_test=searchweb201556_8,searchweb201602_2_10057_10056_10065_10055_10054_10067_10069_10059_10058_10017_10070_10060_10061_10052_10062_10053_10050_10051,searchweb201603_2&btsid=f0612717-8332-4070-9889-381f8e6d6ca3

And get 50 of them to upgrade sonsoff as well as esp01 etc... These version if my memory serves me right do not have caps on them.

I am curious if anyone has attempted to place an 8MB flash on their esp01 for the kicks of it though I see they sell those as well go ahead and future proof your OTA and grab a ton of space.

I am curious if the new TFT stuff will break OTA?

Re: Is any one still using the ESP-01 modules?

PostPosted: Tue Sep 06, 2016 11:51 am
by bugs
I can see I am in a minority - but my view is that the memory is sufficient for the small jobs I need to use the ESP-01 (and the tiny ESP-11). I can run some DHT22s and an OLED display using just the 2 pins available. Likewise an IR receiver and transmitter. I have not yet needed the extra pins of the ESP-12.

BUT I would like to be able to free up some flash by not having to have all the unwanted baggage of libraries for hardware I never use (TFTs etc). I would much prefer an arduino-style compilation where I can specify the required libraries for the task.

Just my opinion...

:|

Re: Is any one still using the ESP-01 modules?

PostPosted: Wed Sep 07, 2016 12:09 am
by forlotto
Bugs you raise a good point from a personal perspective I wonder how frightitanic builds are done for LUA I am fairly sure it is an automated build process would be great if the build process was open sourced.

Problem is I think it would require an entire rewrite of the code or more module like code in order to remove the code totally possible we have an example of it ...

At the same time I could imagine if I were pouring countless hours into development like mmiscool fun or not I would likely want to keep it simple with the build process and not have to manage several branches and support issues for each branch and each device with the esp07 the esp 01 esp02 the 8285 etc.... The other thing around the corner will be the ESP32 if it is supported will be a lot of additional management, support and code. I have many personal wants with basic but I tend not to mention them all because I don't know that they would be all that useful for everyone with the amount of space they would likely occupy.

Think of it this way if you do not use TFT LCD's version 3.0a43 is likely good enough for most anything you said even the version 2 branch final was likely good enough to handle all the needs you mentioned with bells on. As long as availability is maintained nothing will change and if it is working why not stick with it as there is less bloat in the 2 branch as it stands. Maybe someone will take up dev work for smaller flashes and eliminate some functionality. The only constant in life is change can really pack a wallop at times but eventually we bend or break and heal.

If you did need to upgrade chances are you would use the newer features or be thinking about them and to be honest an upgrade of the device or the flash memory is super cheap to do when you think about it in comparison to off the shelf limited functionality based devices heck one smart device costs you a fortune. These things make cost a lot more bearable to have that end goal of automation. By upgrading you are not only keeping it simple for mmiscool but you are essentially buying products from a company that is in fact going to improve upon this technology and make more things accessible to the average person. I give it less than 20 years and home automation will be the norm.


Anyhow thanks for sharing your thoughts bugs I think more folks should share their vote and opinion on the matter believe it or not your opinion does matter just as much as anyone's and I thank you for sharing it I think we need more thoughts from the other side of the fence so to speak. Don't let what I say here discourage anyone I could totally be off here it is just my train of thought my own overall logic. I make no promises either way but it is likely as with anything else a majority vote will likely be largely the deciding factor while I don't know as I am not mmiscool nor can I speak for him I can only think what I would do if it was me. I likely would not have even went out of my way to have a vote so to speak I would have said if someone wants to take up dev work on this branch we will be moving to 4.x and you are welcome to do so and here is why... So I give him credit for letting the community have some say and a vote.

Re: Is any one still using the ESP-01 modules?

PostPosted: Wed Sep 07, 2016 7:22 am
by Electroguard
Take a poll of any stable version and I suspect the majority would always vote overwhelmingly to retain existing stable functionality rather than make existing functionality redundant for the sake of potential future benefits.
So I haven't voted, because there are no suitable options which are valid for me.


1. I don't have tons of ESP-01's laying around, but I've just needed to order some more 1Mb ESP-01's for creating some real-world working PIR nodes I'm producing as a result of a project I developed on a dev board. The dev board is exactly that, a board for developing ESP projects on - it's even bigger than the less-useful nodemcu (which has non-standard buttons and gpio's without led indicators). The one I'm using contains a 4Mb ESP-12 (it's actually a 202) and leds on the gpios, plus a switchable socket for plugging-in and flashing ESP-01's.
2. So it's not a matter of ESP-01 price versus ESP-12 price, it's about about horses for courses. Because irrespective of the size and features of the development board, the real-world end product needs squeezing into a cramped solar garden light... and the most practical way to achieve that is by using an ESP-01, which has the convenience of only 8 x 0.1" spaced pins/holes to deal with instead of a mixture of 16 + 6 awkward 2mm perforations.
3. I have already used ESP_01's, and will most certainly still need to use them in the future.

So none of those 3 options are true in my case.


As a general rule, 'bigger is usually better' for dev work, and 'smaller is usually better' for the finished product.
But how often does an ESP_Basic dev project have any realistic chance of becoming a real-world working end product?

If brutally honest, ESP_Basic is an interesting experimental novelty for minority enthusiasts of the nodemcu/google chrome club, and the outcome of this poll will decide if ESP_Basic is destined to remain a minority toy, or allowed the stability to become a practical solution for the majority.

But there's not much realistic chance of it being allowed to be a practical solution for the majority, if the decision is taken to make it unsuitable for Sonoff users in preference for various new future novelty features which will probably remain unused by the majority, as is the case with much of the existing display-specific features.

The Sonoff is a cheap, practical, convenient, readilly available off-the-shelf switched mains unit which allows anyone - whether now or in the future - to implement their own practical working ESP_Basic projects.

So the availability of such an item should be considered as provident 'manna from heaven' for the ESP_Basic cause.

Ignoring that practicality for the sake of a future few who may appreciate some proposed new 'bloaty' features is no way for ESP_Basic to achieve any stability or loyalty base. And for what it's worth - which is only my humble opinion - I think ESP_Basic lost much more when it lost cicciocb than anything it may have gained with the release of V3.

Closing the door on 1Mb is going to cause even more avoidable loss, because it is sure to force out many (perhaps even most) of the practical majority.

It will certainly create another apartheid sub-culture, just as V3 did to V2 users.
V3 killed of the series of ESP_Basic Driving Lessons I was doing, including all the planned projects that would have taught users how to create their own modular automation and control systems according to whatever features they wished to include.

That was ESP_Basic users loss, not mine. It took me a long time to pick myself up from that 'gutting', and not before figuring out a way to effectively divorce my future efforts from the nodemcu and google_chrome constraints imposed as a condition of ESP_Basic club membership to avoid the problems which non-club members must just accept and put up with. But I managed to achieve that, and over the last few weeks I've successfully created a selection of projects that provides nearly all of the core functionality of the modular automation and control goal that I am working towards, which already now includes relay node, thermostat node, PIR sensor node, Infra-red extender nodes, Voice Announce node, and even an Infra-red controlled voice menu unit with IR button cloning and multi-language selection features. All of those projects were developed to make use of the advantages of Sonoffs and ESP-01's.

The only reason I haven't already published all of those projects is because a certain person who shall remain nameless threw a big spanner in the works by asking for a demo video - which has opened up a whole new can of worms for me... cos if I do it for one project, I should do it for all, and that will entail setting up my own public website where all the attachments could be publicly hosted from for all of the various project articles.

So I've been planning a series of 'Workshop Video Tutorials' which would include all those projects mentioned above plus all of the future projects which I still have planned, and would effectively be a video successor to the Driving Lessons which was killed off by V2...

... then deja vue, along comes this poll.


Resulting in me feeling compelled to take the time (days) and trouble of doing this response.

I already feel very much part of the unwanted, ignored and abandoned V2 sub-culture anyway, so I don't expect anything I say to influence this outcome in the slightest, but if I don't do my best to point out the full situation, then I have no excuse for avoiding blame for any ensuing consequences and disaster... a la the Brexit and Trump fiasco's.

I don't envisage problems in completing my modular automation and control goal whatever the outcome of this latest crisis (cos make no mistake, this is a crisis even if nobody else has recognised it yet). So the only difference the decision to abandon 1Mb Sonoffs and ESP-01's devices would make to me is that it would officially finally and irrevokably remove any point or reason for me to invest any further time or effort trying to help others with ESP_Basic, cos it would again make everything I've done redundant before people even get a chance to try it. Nothing stopping me from still using 1Mb devices on V2 for my own purposes though, so as with the Driving Lessons it will be other peoples loss not mine, and if they are all considered unimportant and irrellevent to ESP_Basics future, then why should I be the only one that cares!

But of course it matters, which is why all the time and effort has been put into creating this amazing and fantastic ESP_Basic for the benefit of others, else you'd have thrown up your hands and walked away yonks ago.

So there must be a better of achieving progress other than by having to keep damaging the present, else this sort of progress damage will be never-ending, always resulting in ESP_Basic being a painful and transient experience for everyone but a few.


Might it be possible to implement some form of minimal ESP_Basic Core, with an ability to add selectable conditional includes? That may be unwelcome complication, but if it could offer a greatly extended life expectency - possibly even past the arrival of ESP32's - surely that could make it very worth-while effort.

Alternatively for the shorter term, there must be some libraries which are mutually exclusive to others - particulaly regarding different displays etc - so perhaps it might be worth considering maintaining a common core but with different flavours of display library, or perhaps without neo-pixels or some other under-used features which might offer sufficient savings.

Perhaps the simplest thing might be just to keep adding new features to a 4Mb version while still maintaining an occassionally bug-fixed no-frills 1Mb (or possibly even 512K) version which doesn't include anything display-related, but which might still be suitable for the greater majority. Even Microsoft recognise the need to do that for a while.

At the very least, perhaps people could be asked what features they want to use, plus what features they have no interest in, then it may even be possible to reclaim enough from unwanted unused features to be able to include the new features still in 1Mb Sonoffs and ESP-01's to delay their redundancy.


The important point is, this should not be a hasty decision... because it will effectively decide the future of ESP_Basic - by deciding if it is destined to remain a workbench novelty for the few, or allowed to become a working reality for the majority.
I hope I haven't caused any offence by anything I've said, because my intention is simply to give ESP_Basic it's best chance for the future. The only way I have of doing that is by opening eyes to things which may have relevence but which would probably otherwise go unnoticed until it's too late to make any difference. It's all just my opinion though based on my own personal experiences, so it's quite possible that nobody else shares my views, but the Trump and Brexit fiasco's are obvious warnings against quiet complacency.