Kelut logo

Login
 
   
In Folder: XML Generation | Up
Sub Folders: None
Issues
Id # Title Description
102 Property file mode when not supplied When a mode is not supplied, the current model rather than omitting mode puts a "". Is this identical from the XML standpoint, i.e. if provided but null, will a parser still interpret it as the default "readwrite" shown on D-21? If not, I think the tool should omit the element instead of including it with an empty string. Ben 5/15: Fixed, changed so that attributes will not be written if nothing is provided.
103 Headings on XML files For all XML files I am used to seeing the and fields, although I no nothing about what they are for and how they are filled out. Is this a standard XML thing, and if so, could the tool fill it in? Ben 5/15: All xml documents now contain xml version and doctype headers.
104 ID and name absent in DCD In the DCD, and presumably the SAD (although I did not check this) the required top level ID and implied name are not present. Some means of supporting these is needed. Ben 5/15: Ooops, I fixed this and several other similar bugs in XML generation. I created the 2bxml example and successfully validated all 50 xml documents.
105 DCD / SAD componentfileid - UUID is a pain For a componentfile, the tool provides a UUID for componentfile. Although technically correct, it makes reading the XML very difficult. In most cases the SPD XML base file name would be sufficient, and it would be really cool if this could be the default (e.g. ModemDevice.spd.xml would use ModemDevice or perhaps ModemDeviceFile as its componentfile id. Ben 5/15: Done.
106 DCD / SAD componentinstantiation UUID is a pain Yup, you guessed it, the componentinstantiation ID is also a big pain as a UUID. It certainly has to be unique to applicaiton, but it would be nice it a kinder name could be given. I haven't thought through this, but one default could be to take the usagename (if present) and append _instance to it, i.e. Audio1 would become Audio1_instance. I know usagename is optional, but the tool could require it neverless, and possibly even require it to be unique. And the SAD file "debuggers" would be ever greatful. Sounds like a good trade to me... Ben 5/16: It appearts that the SCA requires these to be UUIDs so we cannot generate nice names. JVMS now adds a comment below the entry that displays the usage name, as per suggestion by chuck.