@GaelDuval: Please do better - A plea to stay true to yourself and to us

/e/OS Foundation & Community
1

In the recent time, I couldn’t help but gain the impression, that /e/ has lost itself a little. May it be the sloppy implementation of MGM prior to 3.0, the lack of communication when requesting a password reset because of an undefined “temporary security issue” or now and most severely the use of OpenAIs‘ API, feeding our data, even if it‘s anonymized, to BigTech.

I can only speculate that /e/ senses an opportunity for itself in the uncertainty triggered by the current US government. I can understand that they then try to position themselves hastily as an alternative. I can understand that this requires a certain amount of streamlining. And last but not least can I understand that they somehow have to generate revenue. /e/ doesn’t program itself; it is programmed by people who have to make a living.

But why the use of OpenAI?

The only reasonable explanation that comes to my mind is money. Transparent communication about you having a relevant money problem would undermine your attempt to gain attractiveness as a mainstreamed alternative, I guess. And so, you decided to make a profitable implementation that you must have realized would alienate at least your first-hour supporters, which have made a conscious decision by choosing /e/, based on your promises about data protection, open source, privacy and de-googlization, opposing BigTech in general.

But selling your and our ideals for revenue and streamlining isn’t worth it. I, like many others I am sure, would be glad to donate a monthly amount to help /e/ to stick to its (former?) principles and provide a living for its dev team. But to do this it needs transparent communication from your side. And so far, you only offer one-time donations and do not really emphasize the fact that you are a non-profit organisation. What if every convinced /e/-user would donate one or two or more Euro per month? Wouldn’t that be great and make things more calculable? So just talk about your problems and your obstacles with us! If you were to send out a call for help, I am sure it would be heard.

The decision-makers should seriously ask themselves whether they want to continue on this course. Sure, becoming mainstream and generating revenue by selling the ideals that once started /e/ may make the first-hour supporters superfluous. But is this really the path you want to take?

So please: Stay true to the principles you still claim for yourselves. Be transparent about the problems /e/ and its foundation is facing. Communicate with us, find ways to let us talk things out and search for solutions. And stop hiding behind media-effective high-gloss presentations.

18 Likes
2

The feedback related to the OpenAI implementation as communicated by users has already been shared with the team. Will update on the response.

2 Likes
3

I can understand the implementation of OpenAIs VTT usability, alternatives without a data-privacy issue are few and far beyond.

As very reliably working (in terms of VTT) alternatives there’s Google’s approach, well… should be self-explanatory why I wouldn’t want that one integrated in a deGoogled OS.

On the same time it is very understandable, why people would not want MGM / OpenAIs VTT / “Find my device” (even as most of those need active configuration and registration to actually communicate anything) in /e/ OS.

In the future it probably could be an option to allow selection of specific system packages (and corresponding usability scenarios) prior to installation, be it by the web installer or selection of images.

Or, alternatively: to fully remove the questionable applications easily by an assistant which starts on first run (while the applications are inactive before they’re actively selected as “I want to use them” and as such don’t communicate prior).

That is probably the easier to implement way compared to X images / ROMs which have to get maintained, but could be harder to realize due to removal boundaries (necessary system integration which breaks other functions), dunno.

2 Likes
4

Thank you Manoj. You are the face of /e/ for the forum and sometimes have to put up with a lot. I know I’m speculating a lot, but in the meantime, unfortunately, the whole thing now more and more feels like a disappointed love. And I, just like the rest of us, just want the journey with /e/ to go well for us, for you, for the developers and of course for /e/ itself.

6 Likes
5

But they exist, like FUTO and Whisper.

I’m not a die-hard open source advocate nor a paranoid crying wolf every time I see something that is of no use for me personally and looks suspicious at first glance. I understand that people want their phone to work with their companies, their health trackers, their car or whatever. I would welcome it if /e/ would become more widespread. And I see that sometimes concessions have to be made for this.

But the implementation of OpenAI and feeding our data to it is far beyond that.

Apart from OpenAI, one approach, as you say, would certainly be to leave the choice up to the users wherever possible, either through broad opt-ins or, less well, opt-outs. And, equally important, to start communicating transparently alongside this: Why was this or that feature/API/app/whatever chosen, what are the up- and downsides, how does it behave if used and if not used. That would be very much welcomed. And in this context the self-congratulations through media-effective high-gloss presentations feels embarrassing to me.

3 Likes
6

Sure, the easy choice in the hands of users would be a good way - even if on first run, where uninstall of unwanted apps would be possible by a checklist (before they are even run or connected once in any way), but I guess it adds to the technical difficulties of dependencies by system integration or the missing thereof sometimes.

I as well am curious why many of the system packages are not directly uninstallable, as they usually are only started by direct clicks or configuration, like calculator, notes, find my device. Even “clock” seems more embedded than calculator by design (alarms).

They’re still uninstallable via adb, tho. To a degree.

I’ve yet to see a vendor delivered phone aside fairphone, maybe nothing phone (dunno, had it not in my hands) which has not loads of bloatware and most of those are not directly uninstallable as well, because people will uninstall important stuff and then don’t have an idea how to restore it.

Not that problematic in terms of /e/ OS which is installed manually, anyway - but I guess that’s one of the reason why vendors and even Google itself decided to make a lof ot system packages non-uninstallable by default.

On the other hand: if Murena had to communicate clearly about each snippet of non-communicating system package which could be (one example being the Android Assistant snippets) not even used by official builds / releases, but instead came into play by community builds based on a less “scrubbed” LineageOS image, …

… I guess that’s a lot of work which could be better invested in optimizing and developing, integrating even more FOSS solutions instead of using Android / third party packages in coming versions?

7

Yes, this is the path they have chosen, perhaps right from the start: using open source to advance a corporate business model. All the answers you have already given yourself. The only issue is dishonesty. This was to be expected, signs have been enough for some time. That’s it, I’m getting out of here.

3 Likes
8

My only concern is that I lose the option if undesirable software becomes bundled. As long as I have a choice I don’t have a problem with it. I get unhappy when software takes away choice or does things covertly. If I can shut if off then no harm no foul.

9

So you knew in advance that OpenAIs’ API would be implemented? Because covertly is what I would call it when only on closer inspection by users it is revealed that this had happend without any hint at it in advance. Or did I miss something?

In a moment of bewilderment and incomprehension, I may have overshot the mark with my speculations about money, but I can wholeheartedly stand by the rest. I find the lack of understanding in Gael Duvals’ answer to the issue frightening and fear that it means nothing good for the future. My confidence regarding /e/, which I recently expressed elsewhere in this forum, is now close to zero.

Transparent communication looks different and would take place before, not after, such an implementation.

10

If you pay close attention then you notice that transparent communication and managing expectations vs reality as well as communicating in difficult moments was not and probably will not be a big part of murena / e/os. I saw in my couple of years a lot of different situations similar to this and its kind of the same cycle. I had hoped that the server disaster would better the communication, and it somewhat did, but most of the time it just falls back to the usual part of not admitting wrongdoing, not standing up for their own failure, not communicating transparent enough (which is extremely important if you are dealing with users trust).

For me, at least, its soon time to say goodbye and switch to a different ROM in the hopes that one day murena / e/os will be better at communicating and managing users trust. Because right now, most trust is lost for me. Which is sad considering the great project and how many truly wonderful people are working at this foundation / company.

4 Likes
11

I ask: why?

image

It sure looks like trying to shut down any discussion, change my mind.

2 Likes
12

This is not the sort of thing Murena should have to receive as feedback from the users. The fact that this decision was made at all is a very, very bad sign.

I have suspended my financial support for /e/, and I expect complete financial and technical transparency about the terms of the OpenAI deal, as well as (obviously) an immediate end to it, if you expect me to resume my donations.

2 Likes
13

There have been some posts seeking responses from Gaël or team members. While Gaël has already responded to some of the queries on this forum repeatedly pinging is not appreciated and is considered as trolling. It has been observed that some of these folks have been active on the channels as well under different names. While positive feedback is appreciated and accepted misuse of the forum and channels and is not acceptable.
The response regarding OpenAI has already been shared by Gaël . Asking the same queries in multiple places does not change the response. We have already mentioned that a guide on how the STT app would work will be shared and is currently in review. If the responses are not to your satisfaction you are free to withdraw financial support.

4 Likes
14

As the opener of this thread, I would like to say that I have no objection to its closure.

For the time being, the discussion about OpenAI will not get productive anymore. Insensitivity on /e/s’ side and trolls on the forums’ side stand in the way.

I still find it bizarre that this decision was made and that (justified) criticism was apparently not expected. The public relations work is hair-raising and further insight is not to be expected.

In a moment of bewilderment and incomprehension, I have overshot the mark with my speculations about money and motives, but I can wholeheartedly stand by the rest. Not out of opposition to /e/, but out of love and concern for /e/.

But a forum battle between caught innocence on the one hand and haters on the other won’t get anyone anywhere.

So please close with pleasure!

3 Likes