Latest Bugs & Tasks
All modern operating systems now come bundled with a basic level of accessibility features, namely a screenreader. While these features are out of scope for Haiku R1, these features should be within the scope of Haiku R2.
Screen readers are made up of three parts:
A backend Speech synthesis system
An API for 1stparty & 3rdparty access to the Speech synthesis library
An OS bundled screen reading utility for accessibility
For instance, Orca is the Gnome screen reader that communicates with the Accessibility Toolkit via the AT-SPI API. In Android you have Google Text-to-Speech, which is used by Google Talkback to provide speech feedback on what is being enacted, displayed or selected. In Windows, Microsoft Narrator uses the Microsoft Speech API.
BeOS did have 3rd party applications, TalkBox used the festival synthesis library and SpeakIt which used a database of spoken words. The author of TalkBox has offered to open source it in the past. However before creating or adapting an utility, we need a speech synthesis library for it to use.
Both 'Festival' and CMU's 'Festival Lite' use the FestVox dev tool. These two would be the best choice, with the lightweight flite as default and hevywight Festival swapped in if needed.
Some laungaues are written without any spaces between words, meaning that the implementation as used in BTextView is not sophisticated enough to handle these languages.
This been brought up before in an unsuccessfull GCI task, which suggested a complete replacement of our line breaking code with that of ICU's implementation.
We need to support people whose language writes text Right-to-Left, but who might want to mention/write something in English or another language that writes text Left-to-Right, therefore creating bidirectional text.
This has previously been brought up by Adrien in a past mailing list post.
Ubidi can calculate if a text run (chuck of text to be rendered) is Left-to-Right; Right-to-Left; 'Direction Mixed' or 'Direction Neutral' (i.e. an emoji).
I was reading a thread on the Haiku General Mailing List which triggered a pair of Webpositive crashes. Both debug reports attached.
Sometimes when I run the unit test BApplication::QuitTest4 I get a crash (debug log attached), but this does not happen always. If it does not happen, it says the test passed.
Small excerpt from the debug log:
state: Exception (Segment violation) Frame IP Function Name ----------------------------------------------- 0x72039038 0x5edddc BPrivate::hoardUnsbrk(void*, int32) + 0x48
When running the BClipboard::Clear1 test, the UnitTester application crashes.
The debug log (attached to the ticket) shows the error to be "state: Call (The clipboard must be locked before proceeding.)".
I checked the original BeBook and it says of the Clear call that it should return B_ERROR when the clipboard is not locked. The code of BClipboard however makes a jump into the debugger at this point. This seems incorrect.
When running the NetworkInterfaceTest unit tests of the Haiku source code, Haiku consistently enters KDL.
- jam -q unittests
- Go to generated/tests/haiku/x86/haiku/unittests/
- Run ./UnitTester NetworkInterfaceTest
This happens on hrev50329
Since i upgraded from hrev50141 to hrev50326, I noticed a strange behaviour with checkfs: every time that I run this command I always get "60127 blocks could be freed" (I could be wrong about "60127" but I can recall that was a great number), So i reinitialized my bfs partition and I installed over to the freshly reinitialized partition, my whole backup (my backup disk relies on another BFS partition on the same disk and such partition is healty.)
However, since I restored my system, I got again the same message about [n] blocks could be freed:
checkfs -c /boot
5200 nodes checked,
0 blocks not allocated,
0 blocks already set,
906 blocks could be freed
Every time that I run checkfs (with or without "c" flag) I always get, again, "906 blocks could be freed"
I never see any other error; if I look at the syslog, in facts, I don't see any "bad data" message.
However I attach my syslog (I removed the optical drive due the ticket:12634, because due this bug the syslog was filled with the same message).
I read the VBE 3 specification. Bad news: it does not allow setting more video resolutions than VBE 2 which we are currently using. It only allows changing the refresh rate for already supported display resolutions.
So, this will not solve the original issue, and not bring us much more functionality. Since we now have a native driver for the mentionned Radeon cards, I am closing this ticket.
Resizing the window on an app caused a crash and debug report (attached)
First time seeing this problem, so not sure about component.
Well, this ticket could be very similar to ticket:12717.
Every time that I boot Haiku I am informed about the fact that my network card is ready. But i can assume this thanks to the green dot present in the network status present as replicant in the Deskbar. So I think that this redundant notification could be removed from the startup.
hrev50326 but I see this behaviour since Alpha 4.1, IIRC.
First of all: please notice that the bug tracker lacks the presence of the bluetooth_server under the "Component" list.
Also if in the Bluetooth preflet i selected "discoverable" and if I select "incoming connections policy From all devices", I am however unable to find my computer from any other bluetooth device (eg my phone).
Obviously my bluetooth usb adapter is recognized by Haiku.
due this bug ticket:12634 was very difficult to find the bluetooth log messages, so I used "TextSearch" addon to find every bluetooth related log entries (reported here as BluetoothLogs.png attachment).
in the haiku-depot-web mailing list, in the "Malformed URLs in Packages" thread on Sat, May 21, 2016, Andrew Lindesay wrote:
The HaikuDepotServer (HDS) application server system already provides
the APIs that are required for HaikuDepot (HD) desktop application to
access information and package-data for multiple repositories, but at
this point in time, HD only obtains data for the "HaikuPorts"
repository. If somebody would like to undertake this work with HD, I
would be glad to help explain how the HDS APIs provide the data required
to achieve this.
I'm going to close this one as a duplicate of #12783. Even though this ticket is older, the new one has the syslog more easily available.
The issue should be resolved now.
Works here too. Fixed then, I'd say.
No idea why i've never noticed this before. It looks like the email server's icon is missing from the DeskBar tray. There is a blank space you can click on to bring up the email menu (when the mail server is running)