mempko / firestr Goto Github PK

Hi, @mempko

Loved firestr! I work on a similar project called Sneer 😄

I was trying it following the vagrant instructions but could not run firestr. Here's a screenshot of the error:

When several apps open in a given session, it becomes impossible to see the chat again. It would be nice to be able to close an app after it was added.

To be able to use firestr on multiple computers, it would be great from a usability perspective if one could import/export one's cryptographic identity via the GUI.

I compiled firestr from source (git hash 88aa7d8) on my thinkpad running ubuntu 14.04.1. When I dragged the transfer.fab app to the window my whole machine seized up for a few seconds and then firestr died with a segfault. Here's the output, fwiw.

QWidget::setWindowModified: The window title does not contain a '[*]' placeholder
QWidget::setWindowModified: The window title does not contain a '[*]' placeholder
...
QWidget::setWindowModified: The window title does not contain a '[*]' placeholder
QWidget::setWindowModified: The window title does not contain a '[*]' placeholder
zsh: segmentation fault  ./src/firestr/firestr
./src/firestr/firestr  22.72s user 4.56s system 1% cpu 35:45.00 total

I have been wanting to implement, pretty close to the idea you have with luakit. Simply for the gui this might have advantages:

• Many people know HTML-related tech, this means they can modify/create things for it.
  • It is.. "evolved far". CSS could make some kinds of customization easy too.
  • It suggests a template-based system where people can change bits of template for customization.
  • Customization can easily be used to make mobile-oriented guis.
  • (edit)Amenable to using web-technologies locally.
• Chrome pages can be turned into server-able pages. (potentially disadvantage when people foolishly use centralized services with this thing)
• It ties it into the browser.
  • Addons immediately operate on any of the pages.
  • If Ethereum API is ffi-ed in, the same javascript bindings can be exposed as Ethereum. A lot could potentially work seamlessly with it.
  • Ties it into URIs. You could make a lot of things into uris, like locations in pdf, video files, identifying files by magnet links.(/checksums, sortah same)

Some disadvantages, like security not 100% sure all the tiem. I am still hesitant to potentially expose sql-queries to it.(which would potentiallyhappen if bookmark queries had URIs) Sqlites inbuild ? thingies presumably protect... Also a disadvantage is that potentially exposing to some injections of nasty javascript, like sending a message with <script>...</script> in it, and a bug allows it through, leading to potential phishing-like stuff or something. (dont know of a tag that is there just to disable html inside? Template would just end the tag itself i guess.. Setting all of them with .innerText would be a bit of a PITA.. and what if i want markdown....)

The way what i want to do is similar; i want to make a F2F network, and people can write lua code to create communication channels. I want it to be a bit agnostic about that, though initially i'd use Tox. Channels should give out some indication of reliability/latency. Based on that, people hold messages for their friends, send messages directly, and keep data from what is inferred to be interesting. It'd also need a way to query the F2F network for information out that. Thinking a bloom filter of stuff already obtained and a friend at a time. For instance, for a "decentralized reddit" you'd query for comments on a URI. (or a URI + "subreddit") Searching friends-of-friends could really help keep out shills. (Implied though is that it'd need a bit of a reputation/trust system) (Except for the bloom filter, had this idea for a while)

Since i wanted to start with Tox, i wanted to make an IM client. I wanted it to be nice, with a good search(in chats/messages/contacts) and stuff, so made a fairly general sql searcher, from that first an alternative bookmarks/history.(including tags and stuff, and site: like stuff) I guess that is very early part of the project, since i have yet to figure out how to incorporate different kinds of communication. (which to choose, largely)

The program should come with a copy of the GNU GPLv3

Any plans to port this to mobile? A messaging app on desktop-only just isn't very practical for me

Hi,

I came across this error while trying to run the application, could you please help?

I'm packaging firestr for Arch (on the AUR) and despite having libopus 1.1 installed I'm getting:

/usr/bin/ld: cannot find -lopus

Is it expecting some other version, or something?

Why no support for botan 1.11? Is there any major holdback from supporting it?

At least on Arch it would be easier to support with an official package, rather than a 1.10 AUR package. Also might be relevant for the new Debian release, and possibly for new versions of Ubuntu as well.

hey there,

when firestr_0.11.dmg was first posted, it had the checksum
7d7d3740b2eb5a78bc8573b24cdc47c15bcd6536e21f0e8e82722b8a7e8e8cf3

now it suddenly has the checksum
01bc1cda3f12b1bf510b668ad437f6cdbe2db3bd249d8840367569becd7df952

since the file was silently updated without explanation and has no codesigning in-place to confirm its authenticity, its hard to confirm whether it is a valid release or whether it has been tampered with.

could you give any confirmation about the validity of the file?

many thanks