Known Problems
Issues in Keyman
Chillus are created with ZWJ & ZWNJ
For example, n_ creates 0d28 0d4d 200d 200c. This makes wikipedia entries in two different encodings.
Issues in Varamozhi Editor
Hangs on large non-unicode documents
Unnecessary ZWNJ at end of words
If you write 'ath' or 'ath~' both will return with an unnecessary ZWNJ at the end.
Impact: The impact is minimal. The extra ZWNJ does not cause any undesired rendering. It does not cause any issue with linguisting operations like comparison or collation since they would be using collation-key and the weight of ZWNJ is 0. The only issue this would cause is swelling of one extra character. However, making sure chandrakkala at the end by zwnj has an advantage. When space are lost when text is copy pasted or in other occasions, undesired ligatures are not formed. For example the line "അവന് റാണിയെക്കൊണ്ട് ..." will become "അവന്റാണിയെക്കൊണ്ട്..." when space is lost and zwnj is not there after virama. For the output from Varamozhi editor, even if space is lost, the output does not deteriorate considerably: "അവന്റാണിയെക്കൊണ്ട്..."
No ത with Kaveri font
ത does not appear in Kaveri font.
Status: Don't know the root cause.
Analysis: It looks like a font issue. The ത in kaveri is at position 158. But when I copy it to MSWord, it becomes 142. I don't know why. None of the famous fonts have anything in 158th character position. Deshavani has 'vva' there and it appears properly. Cibu 22:45, 26 May 2006 (UTC)
Workaround: None
Advanced rule issues
pate & പറ്റെ
Work Arounds: pate#
Solution: fine tune 'ate' rule
'paint' & പൈന്റ്
'paint' becomes 'പെയിന്റ്’ while the user sometimes expects 'പൈന്റ്’ as per basic Mozhi scheme.
Work arounds: 'peint', 'paint#', 'pai_nt'
Solutions: may need to add 'pint' as ‘പൈന്റ്’
Numerals not converted
Numerals not being converted properly between some ASCII fonts and Manglish while converting (mal)
Conflict with Cygwin installation
Varamozhi Editor installation sometimes conflicts with preexisting Cygwin installation (cygwin.dll).
Status: I don't know anyway to resolve this issue.
Workaround: Download and compile varamozhi source code in cygwin environment
Refresh after paste
Update of the other pane without having to add extra newlines/characters after a paste operation.
Analysis: The pane expects a key press before attempting conversion. Mouse click is not considered as a qualifying event.
Status: I don't know whether this is solvable in Perl/Tk without spoiling default right-click menu.
Work around: Use Cntl-v to paste
'f' and 'tt' in DVMalayalam
The 'f' and 'tt' in DVMalayalam does not get translated properly.
In the new 1.4.1 version, if we type "paNTu~", what appears is പണ്ടു and not പണ്ടു്.
This happens for chandrakkala after all vowels, but only ഉ is a real problem, because people who prefer old lipi prefer to write സംവൃതോകാരം as ഉ followed by chandrakkala.
Workaround: Use _~ instead of ~.
This is a new problem, not present in 1.3.3. - Umesh 20:20, 14 July 2006 (UTC)
Manorama kr^ conjunct is not correct
Revert to disjoint one.
Desired features
More control thru menu
Preferably, switches may be incorporated in the Menu. (Like disable Dictionary Toggle)
Comment: I like the google approach of giving only the essential controls to avoid clutter and confusion. Users always have the choice of editing the configuration file. -Cibu 22:55, 26 May 2006 (UTC)
Issues in AnjaliOldLipi
Font's problem
Conjuncts not present
ദ്മ (should be small മ under ദ)
ഴ്ത്ത - should be small ത്ത under ഴ. Fine in Rachana.
ഴ്വ - fine in Rachana
ത്പ
Missing Essentials
Following characters are in unicode pipeline at stage 4.
സംഖ്യാ ചിഹ്നങ്ങള്
10 - 0D70
100 - 0D71
1000 - 0D72
1/2 - 0D73
1/4 - 0D74
1/8 - 0D75
നു് (date sign) - 0D79
അപൂര്വ്വ അക്ഷരങ്ങള്
ൠ sign - 0D44
ഌ sign - 0D62
പ്രശ്ലേഷം(അവഗ്രഹ) - 0D3D
ചില്ലുകള്
ണ് - 0D7A
ന് - 0D7B
ര് - 0D7C
ല് - 0D7D
ള് - 0D7E
ക് - 0D7F
Incorrect reordering
For any chillu C, a full form X and subscript Y with the reordering vowel V (എ, ഏ, ഒ, ഓ), the position of left part of V goes before C in the sequence CXYV. eg: poorNNE = പൂര്ണ്ണേ. Fine with Rachana and Karthika.
Invalid Conjuncts
gDa = ഗ്ധ - (there is no such conjunct)
Subscript positioning
eyDs = എയ്ഡ്സ് (as we write AIDS) Fine in Rachana
kgko = ക്ഗ്കൊ Fine in Rachana
Dchka = ഡ്ച്ക Fine in Rachana
tnjnjto = റ്റ്ഞ്ഞ്റ്റൊ ( not a font problem : the uniscribe engine doesnt allow three consecutive kooTTaksharams eg: ക്ക്ക്ക്ക്ക; However not a problem in Rachana)
khta = ഖ്റ്റ Fine in Rachana
Dgtr^ = ഡ്ഗ്റ്റൃ Fine in Rachana
phplu = ഫ്പ്ലു
kthho = ക്ഥൊ Fine in Rachana
Typeface issues
Symbol of വ in ന്വ should go down and left a bit. Otherwise it can be confused with മ്പ.
Uneven space after characters. Specifically after കു and മ. See സകുമയരല
ഘ should be like പ, on the line. So ഘ should go down a bit.
Do you know about a പ്പ being written with a tilda over പ ? Personaly I prefer that than small പ below. Cibu 19:14, 25 July 2006 (UTC)
in ല്പ conjuct, the subscript പ should get aligned with ല. (eg: അല്പം)
The size of small ഭ in ബ്ഭു, ബ്ഭൂ etc is very small so that it is difficult to read. I know the smaller glyph of ഭു below ബ can not be too big. But can't the ബ be compressed a bit heightwise?
ഋ symbol part of വൃ and നൃ are too elongated. For me it will look better if it were little more round.
Single quote overlaps the ചന്ദ്രക്കല. Example: 'അവന്'
Other Unicode issues
Implement C2-conjoint ligature draft http://www.unicode.org/review/pr-37.pdf
Uniscribe problems
villu~ = വില്ലു് (present in XP, not in Win2000). For any X and Y = {l, y, v}, XYu~ has the problem. eg: blu~ = ബ്ലു്, yyu~ = യ്യു്, vvu~ = വ്വു്
NThro = ണ്ഠ്രൊ (can be corrected by the font, a conjuct for NNTTHA is needed in the font)
Semi-vowel Signs missing
Lya = ള്യ
ntya = ന്റ്യ
zhya = ഴ്യ
Rya = റ്യ
kutyaaTi = കുറ്റ്യാടി
strong - സ്റ്റ്രോംഗ്