Forum Replies Created
-
Posts
-
I'd welcome any feedback from anyone using Solaris, especially if they have an Expert3D card and can run the OpenGL version.
Reason is that mine is refusing to work correctly under OpenGL, but I'm not sure if my Solaris box is properly configured at the moment.
Also, on Solaris the icon quality is seriously degraded unlesss you have Xrender running. I don't (currently) so am curious to know if any Solaris users are running Xrender, and if the icons looks OK.
If you've a suggestion for an enhancement, please post it. If it's easy enough to implement, it will probably get done – if on the other hand it requires major changes then its less likely, but you never know.
If you find a bug in Glade, please report it – either here or if you wish to support@peardrop.co.uk.
Priority will be given to fixing crashes, please provide a detailed description of what you were doing – ideally with an example, or at least the glade.log file.
Version 4.0.26 is in the works at the moment as I have some spare time :xd:
Some fixes to the pcell code and pcell instanciation, plus (not yet checked in) MOS device extraction. Hopefully the ability to extract and export a SPICE netlist of simple cells.
The extraction is actually up and running, and will recognise 3 terminal MOS devices currently (4 terminal should be a pretty easy extension). Also device L and W can be determined for transistors with straight or bent gates.
Update: the extractMOS command is now working after a bit of coding to get the device width/length working for gates with all angle bends. The geomLabel command now labels shapes according to text names and creates top level pins in the extracted view. And an Export CDL command lets you write a flat CDL netlist of an extracted view (I haven't even started thinking about hierarchical extraction / netlisting yet…)
To do is a general purpose device extraction command suitable for diodes/resistors/capacitors/bipolars etc. This in itself is easy, the work is providing the capability to extract parameter info for devices which means a lot more geometry code in the scanline algorithm.
Also in this version there are some bugfixes for edit in place and some changes for the stretch command – this now keeps 45 degree edges at 45 degrees, and stretching adjacent manhattan edges will keep a 45 degree edge if 'Lock diagonals' is checked in the stretch options dialog, which is the default.
I'll be checking in version 4.0.25 this weekend with a number of fixes for the OASIS reader/writer. A long standing bug has been fixed whereby CBLOCK records over 64K long were not handled correctly. A memory leak in the OASIS reader has been fixed and a case where arrays with negative row/column spacing were getting written with incorrect origins has also been fixed. Another bug where text records were being read as n-strings rather than a-strings (meaning that an ascii space character was considered illegal in a text record) has also been fixed.
A couple of other OASIS related bugs have been fixed today – issues writing 3-deltas and issues with type 0/1 pointlists.
Also the OASIS reader now performs CRC / checksum checks as per the spec now.
Welcome to the Glade user forum. Here you can interact with other Glade users, discuss bugs, enhancements, how to program Glade in Python, pcell scripting… anything within reason!
Please register and login to see all the forum sections. Unregistered users can only see the News and Installation sections.
Please report any bugs to the email address you can find in the Glade Help->About menu. Bug reports are always appreciated; if they can be reproduced they will get fixed!
