-
-
Notifications
You must be signed in to change notification settings - Fork 26
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Allow sending 50m+ files to telegram channels using MTProto #37
Conversation
I'll check it, but it's unlikely I will change the used library. The existing one already works except for HTML parsing, which I need to fix on this app side and which is unlikely solved in the library you mentioned. |
Got it, thank you for reply. I've added comparison that can help with trade-offs of picking one or other implementation. |
I'll take a closer look into your library now, you got my attention. |
@ernado @tdakkota I see that switch to that library would require complete code rewrite and is not trivial because the only real example is gotd/bot source code. I still need to fix the incomplete HTML parsing logic I have for the currently used library; the message module doesn't seem to solve that problem for me. If you're interested in using that library for this particular use case, I propose you take my code and rewrite it for your library and send an MR to this repo; however, I personally do not want to spend another day on a rewrite of a working code for the sake of rewrite itself. |
Ok, I'd better improve gotd examples and documentation. Anyway it seems like switching from current library to gotd is a matter of time unless someone is going to spend several months fixing issues, introducing missing features and maintaining. |
The trick is, you don't need to fix something which is not broken, and you don't need to introduce missing features if you already have all the required features on the table. I don't see where the necessity to "spend several months fixing issues" in I am running a |
Current library uses MTProto v1 that is deprecated and will be eventually disabled by Telegram in the future, so things can just stop working randomly, not considering that it is like using software that has open CVE. I'm also going to quote xelaj/mtproto comment here:
If you are ok with this, I'm done here. |
let me clarify a little bit from my perspective. Sending 50+ Mb files is a nice to have feature, however, this is not that essential for the particular application. I think the time and efforts @paskal already invested into this task were a little bit too much for the end goal, but I can hardly see anyone "spend several months" to improve it or adopt any other library regardless of the ergonomics. If v1 will be disabled by tg one day - well, so be it. Feed master can always do what it always did with the bot API, i.e. upload under 50m files only and put external links for the bigger file. And, who knows, maybe as tg kill support of v1 they will increase the upload limits. |
Alternatively, you can also host your own local bot api server (can be used instead of https://api.telegram.org), it is official and should be enough to send 50+ Mb files. |
Rebased on top of current master just to show how much more complex it is rather than using the self-hosted bot API server we choose to go with. Closing this PR. |
This PR switches the Telegram connector from Bot API to the Core API, resolves #23 and likely closes #31.
Before and after: