Beebem Mode 7 glitch

want to talk about MESS/model b/beebem/b-em/electrem/elkulator? do it here!
avengahM
Posts: 6
Joined: Mon May 28, 2018 4:07 am
Contact:

Beebem Mode 7 glitch

Post by avengahM » Sat Jun 23, 2018 2:50 pm

There is a mistake in Mode 7 when using the Hold Graphics function and then changing the colour. On the original BBC, the colour would change from the character AFTER the control code, but on Beebem it changes from the control code itself, one square early. This breaks some of the Mode 7 graphics in games like Granny's Garden. It's not a massive issue but for correctness it ought to be looked at; it threw me off a bit when I saw the mistakes!

EDIT: While I'm at it, what were those things removed due to the wrong licence for the code or something? Speech support and something else? I don't quite understand what the problem is; why would the person who added that support not release the source code as fully open source? Can it be readded? I think it's a shame to have to remove the support.
Last edited by avengahM on Sat Jun 23, 2018 3:03 pm, edited 1 time in total.

User avatar
hoglet
Posts: 7695
Joined: Sat Oct 13, 2012 6:21 pm
Location: Bristol
Contact:

Re: Beebem Mode 7 glitch

Post by hoglet » Sat Jun 23, 2018 3:04 pm

I think a few other people have encountered the same issue:
viewtopic.php?f=4&t=14105

The fix made it into the latest beebem development:
https://github.com/stardot/beebem-windo ... 867ea1b726

Have you tried any of the Beebem 4.15 prereleases?
Last edited by hoglet on Sat Jun 23, 2018 3:06 pm, edited 2 times in total.

User avatar
Richard Russell
Posts: 544
Joined: Sun Feb 27, 2011 10:35 am
Location: Downham Market, Norfolk
Contact:

Re: Beebem Mode 7 glitch

Post by Richard Russell » Sat Jun 23, 2018 3:19 pm

Coincidentally (and spookily) I found this same issue myself today. I've been trying to track down an apparent bug in edit.tf's MODE 7 emulation so I thought I'd compare it with both BeemEm and a real BBC Master (and also BBC BASIC for SDL 2.0). This is a standard test page that everybody should be using to test their emulation so the variation in results is unforgiveable:

BBC Master for reference:
Image

BBC BASIC for SDL 2.0 (ignore the mispositioning of the right end of the bottom row, a mistake at Telstar apparently):
Image

edit.tf (notice colour stripes to right of 'Test Page'):
Image

BeebEm 4.14 (oh dear!):
Image

chrisn
Posts: 401
Joined: Sat Apr 19, 2014 11:31 am
Location: UK
Contact:

Re: Beebem Mode 7 glitch

Post by chrisn » Sat Jun 23, 2018 3:39 pm


User avatar
Richard Russell
Posts: 544
Joined: Sun Feb 27, 2011 10:35 am
Location: Downham Market, Norfolk
Contact:

Re: Beebem Mode 7 glitch

Post by Richard Russell » Sat Jun 23, 2018 4:02 pm

chrisn wrote:
Sat Jun 23, 2018 3:39 pm
BeebEm 4.15b1 … still not perfect.
The Teletext spec says that the 'held graphics' character is defined only when "there has been no intervening change in either the Alphanumerics/Graphics or the Normal/Double Height modes" (3.1.7). BeebEm 4.15b1 seems to be clearing it at the Alpha Red code, but since that is a 'set after' code it should be clearing it one character cell later. However the design of the test page makes me suspect that BeebEm is not alone is its interpretation!

Richard.

User avatar
Richard Russell
Posts: 544
Joined: Sun Feb 27, 2011 10:35 am
Location: Downham Market, Norfolk
Contact:

Re: Beebem Mode 7 glitch

Post by Richard Russell » Sat Jun 23, 2018 4:14 pm

The edit.tf problem is apparently a "known issue" (known to somebody, maybe, but not known to me and not mentioned in its README as far as I can see).

Richard.

User avatar
kieranhj
Posts: 733
Joined: Sat Sep 19, 2015 10:11 pm
Location: Farnham, Surrey, UK
Contact:

Re: Beebem Mode 7 glitch

Post by kieranhj » Sat Jun 23, 2018 4:19 pm

It doesn't help that the Mullard SAA5050 chip interprets Hold Char differently from other Teletext implementations (as observed on different TV's.) So I guess the original spec was open to interpretation somewhere along the line...

I'm not an expert on this topic but zxguesser is the best person to chime in.
Last edited by kieranhj on Sat Jun 23, 2018 4:21 pm, edited 1 time in total.
Bitshifters Collective | Retro Code & Demos for BBC Micro & Acorn computers | https://bitshifters.github.io/

User avatar
tricky
Posts: 2952
Joined: Tue Jun 21, 2011 8:25 am
Contact:

Re: Beebem Mode 7 glitch

Post by tricky » Sat Jun 23, 2018 4:46 pm

I believe that the speech was borrowed from MAME.
Both beebem with speech and mame sources are available.
I don't know about licensing, but I think this was why it was removed.
Somewhere arround here I posted some changes to make the speech code more compatible with actual hardware.

User avatar
Richard Russell
Posts: 544
Joined: Sun Feb 27, 2011 10:35 am
Location: Downham Market, Norfolk
Contact:

Re: Beebem Mode 7 glitch

Post by Richard Russell » Sat Jun 23, 2018 5:40 pm

kieranhj wrote:
Sat Jun 23, 2018 4:19 pm
So I guess the original spec was open to interpretation somewhere along the line...
Very much so. A case in point is whether or not a Held Graphics Character can be defined before the Hold Graphics mode is established (i.e. earlier in the row). The SAA5050's interpretation is that it cannot, and therefore a Held Graphics Character will only be stored after the Hold Graphics control code has been seen. Some other chips and emulations will store a Held Graphics Character anywhere on the row ("providing there has been no intervening change in either the Alphanumerics/Graphics or the Normal/Double Height modes"), although of course they won't display it until after the Hold Graphics mode has been established.

The text in the Specification (at least the September 1976 version which is all I have) is not sufficiently unambiguous to decide which of these is correct.

Richard.

guesser
Posts: 240
Joined: Mon Jun 26, 2006 9:21 pm
Contact:

Re: Beebem Mode 7 glitch

Post by guesser » Mon Jun 25, 2018 2:33 pm

That was me on the facebook thread, though I was away from my notes when I made it as I've been away all weekend. By 'known issue' I meant there are hold mode issues open in the bug tracker that rawles has been too busy to get around to fixing.

To the best of my knowledge this his the correct interpretation of that engineering test page at teletext level 1 with and without reveal:
test-page-reveal.png
test-page-normal.png
(during flash on phase)

It doesn't happen to be a test case for the SAA5050 hold mosaic hardware bug unfortunately so a separate test is required to check emulation of that.
Last edited by guesser on Mon Jun 25, 2018 2:54 pm, edited 2 times in total.
A web based teletext editor which can export as Mode 7 screen memory: https://zxnet.co.uk/teletext/editor

guesser
Posts: 240
Joined: Mon Jun 26, 2006 9:21 pm
Contact:

Re: Beebem Mode 7 glitch

Post by guesser » Mon Jun 25, 2018 2:53 pm

Richard Russell wrote:
Sat Jun 23, 2018 5:40 pm
The text in the Specification (at least the September 1976 version which is all I have) is not sufficiently unambiguous to decide which of these is correct.
The current specification can be downloaded from the ETSI website: http://www.etsi.org/deliver/etsi_en/300 ... 10201p.pdf

I believe it removes any ambiguity in the hold mode question as with the SAA5050 issue.
Hold Mosaics ("Set-At")
Generally, all spacing attributes are displayed as spaces, implying at least one space between characters or mosaics with different colours in the same row. In mosaics mode, the "Hold Mosaics" option allows a limited range of attribute changes without intervening spaces. A mosaic character from the G1 set (referred to as the "Held-Mosaic" character) is displayed in place of the character "SPACE" corresponding to a control character.
Substitution only takes place in mosaics mode when Hold Mosaics mode is in force. At a screen location where substitution is permitted, the "Held-Mosaic" character inserted is the most recent mosaics character with bit 6 = '1' in its code on that row. The "Held-Mosaic" character is reset to "SPACE" at the start of each row, on a change of alphanumeric/mosaics mode or on a change of size. It is not reset by reinforcement of the existing size setting. It is not reset by a change in Hold Mosaics mode.
The "Held-Mosaic" character is always displayed in its original contiguous or separated form regardless of the mode prevailing at the time of substitution.
Conceal creates a potential ambiguity or at least opportunity for confusion as there is nothing in the specification explicitly stating that the SPACE displayed in place of a concealed control code should not be replaced in Hold Mosaics mode. i.e. that the conceal rule takes precedence.
Conceal ("Set-At")
The following characters up to the end of the row, or until a Colour Code attribute (codes 0/0 to 0/7 or 1/0 to 1/7) is encountered, are to be displayed as SPACES until revealed by a decoder or user operation.
A web based teletext editor which can export as Mode 7 screen memory: https://zxnet.co.uk/teletext/editor

User avatar
Richard Russell
Posts: 544
Joined: Sun Feb 27, 2011 10:35 am
Location: Downham Market, Norfolk
Contact:

Re: Beebem Mode 7 glitch

Post by Richard Russell » Mon Jun 25, 2018 4:50 pm

guesser wrote:
Mon Jun 25, 2018 2:53 pm
"It is not reset by a change in Hold Mosaics mode."
This is the key addition, and is different from what the SAA5050 does. Clearly something that is purporting to emulate MODE 7 (such as BeebEm) needs to behave like the SAA5050, but a Videotex client should probably adhere to the later spec. Of course this poses a problem for a Videotex client coded using BBC BASIC (which is what I have been working on) since - without a lot of extra effort - it will not strictly be compliant.
Conceal creates a potential ambiguity or at least opportunity for confusion as there is nothing in the specification explicitly stating that the SPACE displayed in place of a concealed control code should not be replaced in Hold Mosaics mode. i.e. that the conceal rule takes precedence.
Nothing explicit, no, but in my opinion common sense goes a long way to resolve any ambiguity. Consider the special case of alternating 'actual' graphics (mosaic) characters with 'held' graphics characters (such as a contiguous or separated control code); this is not too far fetched. If one were to assume that the 'hold' rule takes precedence over the 'conceal' rule then the result would be that the 'actual' graphics characters would be concealed, but the alternating 'held' graphics would not. This would be an extremely strange and unexpected result. I conclude therefore that the only sensible behaviour is for conceal to take precedence.

I've put it to the test; the following is what is rendered by edit.tf when alternate 'actual' and 'held' graphics are all 'concealed'; hard to justify in my opinion:

Image

I think it should either look like this (when revealed):

Image

or like this (when concealed):

Image

guesser
Posts: 240
Joined: Mon Jun 26, 2006 9:21 pm
Contact:

Re: Beebem Mode 7 glitch

Post by guesser » Mon Jun 25, 2018 9:57 pm

Richard Russell wrote:
Mon Jun 25, 2018 4:50 pm
I conclude therefore that the only sensible behaviour is for conceal to take precedence.
Yup, that's my view too.
A web based teletext editor which can export as Mode 7 screen memory: https://zxnet.co.uk/teletext/editor

Coeus
Posts: 1117
Joined: Mon Jul 25, 2016 11:05 am
Contact:

Re: Beebem Mode 7 glitch

Post by Coeus » Mon Jun 25, 2018 10:47 pm

So where is the source for this version of the test page? There is a very similar one in the other thread but it is not quite the same and no, that's clearly not due to bugs as a picture of a real BBC showing the one in the other thread shows it as different from the reference image in this thread.

guesser
Posts: 240
Joined: Mon Jun 26, 2006 9:21 pm
Contact:

Re: Beebem Mode 7 glitch

Post by guesser » Tue Jun 26, 2018 12:07 am

Coeus wrote:
Mon Jun 25, 2018 10:47 pm
So where is the source for this version of the test page? There is a very similar one in the other thread but it is not quite the same and no, that's clearly not due to bugs as a picture of a real BBC showing the one in the other thread shows it as different from the reference image in this thread.
The source for my images above is this which I have put together from a combination of a mode 7 screen dump from JGH, and a page recovered from VHS by Jason Robertson. In other words the provenance is somewhat shaky as it's a composite of slightly wonky sources (doubled rows in the former and glitched characters in the latter). The best copy would be if someone could dump the vbi data of the British Garden Birds laserdisc. Edit: I somehow managed to introduce a bunch or errors not in my original sources, see below.

There is also this earlier test page which I transcribed character by character from this photo of a modified decoder showing all the control codes https://www.radios-tv.co.uk/Pembers/Tel ... ering2.jpg
Last edited by guesser on Tue Jun 26, 2018 1:04 pm, edited 1 time in total.
A web based teletext editor which can export as Mode 7 screen memory: https://zxnet.co.uk/teletext/editor

User avatar
Richard Russell
Posts: 544
Joined: Sun Feb 27, 2011 10:35 am
Location: Downham Market, Norfolk
Contact:

Re: Beebem Mode 7 glitch

Post by Richard Russell » Tue Jun 26, 2018 8:14 am

guesser wrote:
Tue Jun 26, 2018 12:07 am
I have put together from a combination of a mode 7 screen dump from JGH, and a page recovered from VHS by Jason Robertson. In other words the provenance is somewhat shaky as it's a composite of slightly wonky sources
I'm pretty sure that the version I use (both in my Test Card Generator and supplied with BBC BASIC) comes from a reliable source, and won't have been through any 'analogue' processing or reconstruction, but I still don't know its prevenance in detail. I will keep using it whatever replacement is (in my view unnecessarily) cobbled together . It's here.

Richard.
Last edited by Richard Russell on Tue Jun 26, 2018 8:19 am, edited 1 time in total.

User avatar
Richard Russell
Posts: 544
Joined: Sun Feb 27, 2011 10:35 am
Location: Downham Market, Norfolk
Contact:

Re: Beebem Mode 7 glitch

Post by Richard Russell » Tue Jun 26, 2018 8:51 am

guesser wrote:
Mon Jun 25, 2018 2:33 pm
It doesn't happen to be a test case for the SAA5050 hold mosaic hardware bug unfortunately so a separate test is required to check emulation of that.
I think it's a bit harsh to call it a "hardware bug" because it's a perfectly reasonable interpretation of the specification as it existed at the time. This alternative test page exercises that feature: the difference shows in whether the green stripe to the right of the hash in row 7 renders or not (it doesn't on a SAA5050). I don't know where this version originates, but it's evidently of a later vintage because it incorporates 'double size' control codes which is a level 2.5 feature.

guesser
Posts: 240
Joined: Mon Jun 26, 2006 9:21 pm
Contact:

Re: Beebem Mode 7 glitch

Post by guesser » Tue Jun 26, 2018 1:03 pm

Yep, this is embarrasing, further investigation shows I've got some errors in my page. I've no idea how they got in there because the two sources I used don't have them! :oops:

The corrected version exactly matches Richard's so I'm confident that must be the correct version: https://zxnet.co.uk/teletext/editor/#0: ... ECBAgQIECA

Edit: that's a lie. The packets Jason recovered off VHS have white (7E/7F) clock cracker on row 06 like the 1978 page, Richard's has red (00/01) :?

sorry for causing additional confusion :oops:
Last edited by guesser on Tue Jun 26, 2018 1:42 pm, edited 2 times in total.
A web based teletext editor which can export as Mode 7 screen memory: https://zxnet.co.uk/teletext/editor

guesser
Posts: 240
Joined: Mon Jun 26, 2006 9:21 pm
Contact:

Re: Beebem Mode 7 glitch

Post by guesser » Tue Jun 26, 2018 1:15 pm

Richard Russell wrote:
Tue Jun 26, 2018 8:51 am
I don't know where this version originates, but it's evidently of a later vintage because it incorporates 'double size' control codes which is a level 2.5 feature.
I believe that's actually an earlier version dating from the 70s though an exact date isn't given for any of the photographs https://www.radios-tv.co.uk/Pembers/Tel ... .html#1978

Edit: duh, it was Friday 25th of August 1978 :D
Last edited by guesser on Tue Jun 26, 2018 1:24 pm, edited 1 time in total.
A web based teletext editor which can export as Mode 7 screen memory: https://zxnet.co.uk/teletext/editor

guesser
Posts: 240
Joined: Mon Jun 26, 2006 9:21 pm
Contact:

Re: Beebem Mode 7 glitch

Post by guesser » Tue Jun 26, 2018 1:38 pm

Richard Russell wrote:
Tue Jun 26, 2018 8:14 am
I will keep using it whatever replacement is (in my view unnecessarily) cobbled together
Well when I poorly cobbled it together it was from the sources I could find, I'd have definitely just used yours if I'd seen it :D
A web based teletext editor which can export as Mode 7 screen memory: https://zxnet.co.uk/teletext/editor

User avatar
Richard Russell
Posts: 544
Joined: Sun Feb 27, 2011 10:35 am
Location: Downham Market, Norfolk
Contact:

Re: Beebem Mode 7 glitch

Post by Richard Russell » Tue Jun 26, 2018 1:59 pm

guesser wrote:
Tue Jun 26, 2018 1:38 pm
it was from the sources I could find, I'd have definitely just used yours if I'd seen it
I think (and hope) that the only difference between the version that comes with my Test Card Generator software (in 'teletext.car') and the version that comes with BBC BASIC (in 'mode7dem.car') is that the latter explicitly duplicates the double-height rows - because that's necessary for MODE 7 - whereas the former doesn't - because a teletext decoder is expected to do it in hardware. Both are freely downloadable!

Part of the problem in making them available has been that the files I mentioned contain raw data (one byte per character cell) whereas you seem to want the new-fangled 'edit.tf' URL format. I can convert one to the other, but it needs some thought and care.

Richard.

guesser
Posts: 240
Joined: Mon Jun 26, 2006 9:21 pm
Contact:

Re: Beebem Mode 7 glitch

Post by guesser » Tue Jun 26, 2018 2:10 pm

Richard Russell wrote:
Tue Jun 26, 2018 1:59 pm
Part of the problem in making them available has been that the files I mentioned contain raw data (one byte per character cell) whereas you seem to want the new-fangled 'edit.tf' URL format. I can convert one to the other, but it needs some thought and care.
Well it's just the easiest for me to share. I can trivially generate binary files or mode 7 data but that then means uploading a file each time :)
I've installed TCGEN to play with now anyway :)

What's your take on the different contents of row 06?

Edit: can't put too much faith in my recovered packets from Jason because it has other issues of corruption like row 02 being missing entirely. So the row 06 I have could have been a figment of the recovery algorithms imagination somehow, and of course my other source is from a beeb so has had row 05 duplicated. Funny that they would have changed it from the old test page and it not match the pattern though.

Edit2: my editor can import bin files too where it's just a file of raw 7-bit codes or a mode 7 dump.
Last edited by guesser on Tue Jun 26, 2018 2:40 pm, edited 3 times in total.
A web based teletext editor which can export as Mode 7 screen memory: https://zxnet.co.uk/teletext/editor

User avatar
Richard Russell
Posts: 544
Joined: Sun Feb 27, 2011 10:35 am
Location: Downham Market, Norfolk
Contact:

Re: Beebem Mode 7 glitch

Post by Richard Russell » Tue Jun 26, 2018 4:28 pm

guesser wrote:
Tue Jun 26, 2018 2:10 pm
What's your take on the different contents of row 06?
Obviously I wouldn't choose to use the 'alternative' version on a genuine Beeb or a MODE 7 emulation because the SAA5050's interpretation of the held graphics rules (wrong according to the later spec, but not at the time) makes it look ugly. But then there are other aspects of both test pages which look suspiciously as though the designer expected them to render differently from how they actually do!

Richard.

User avatar
jgharston
Posts: 3296
Joined: Thu Sep 24, 2009 11:22 am
Location: Whitby/Sheffield
Contact:

Re: Beebem Mode 7 glitch

Post by jgharston » Tue Jun 26, 2018 5:08 pm

My engineering test page I downloaded from the broadcast page 199 in about 1985 with the Teletext Adapter, so it's whatever *CH2, *PAGE 199, *TRANSFER 7000, *SAVE P199 7000+400 gave, with bit 7 stripped except for control codes, and bytes 920-1023 cleared, with this code:

Code: Select all

  830 DEFFN_5:CLS:PRINT'"…Translate a file":INPUT"‡What filename? "F$:PROCPage
  840 OSCLI"LOAD "+F$+" "+STR$~st%:PRINTTAB(0,0)"Converting ";F$
  850 FORz%=0TO919:x%=st%?z%:IFINSTR("#`_",CHR$x%):x%=(223AND(x%=35))+(224AND(x%=95))+(163AND(x%=96))
  860   IFx%=255ORx%=223ORx%=224ORx%<164:ELSEx%=x%-128
  870   IFx%<32:x%=x%+128
  880   VDUx%:st%?z%=x%:NEXT:$(st%+920)="-TFS-":FORz%=926TO1023:st%?z%=0:NEXT:=0

Code: Select all

$ bbcbasic
PDP11 BBC BASIC IV Version 0.25
(C) Copyright J.G.Harston 1989,2005-2015
>_

guesser
Posts: 240
Joined: Mon Jun 26, 2006 9:21 pm
Contact:

Re: Beebem Mode 7 glitch

Post by guesser » Tue Jun 26, 2018 6:57 pm

You mean TESTPAGE and/or 9D/9 in your samples? It's exactly the same as Richard's file apart from rows 03 and 06 because of the double height duplication.
I don't know how I managed to make such a mess of things as I was using your file as a source.
Last edited by guesser on Tue Jun 26, 2018 6:59 pm, edited 1 time in total.
A web based teletext editor which can export as Mode 7 screen memory: https://zxnet.co.uk/teletext/editor

avengahM
Posts: 6
Joined: Mon May 28, 2018 4:07 am
Contact:

Re: Beebem Mode 7 glitch

Post by avengahM » Thu Jul 05, 2018 5:00 pm

I found another issue. If there is a double height control code on the very bottom line, the very top line is rendered invisible (or, if you put a double height code there too, it acts as the lower half). This is wrong. On the actual BBC, a double height code on the bottom line has no effect on the top line.

How are we doing with getting Speech support back into Beebem? I played Deathstar on 4.03 with Speech active and it was interesting; it'd be nice to get this support back in.

Does anyone have an actual BBC to test the conceal / hold graphics interplay? Does the conceal code have priority over the hold graphics code?

Finally, here's a MODE 7 editor that I found pretty good; it was on one of the Model B Computing tapes. Useful for messing around in Beebem. The instructions are run by CHAIN "INSTR" but the !BOOT skips the instructions.
Attachments
Mode 7 Editor.ssd
(27.5 KiB) Downloaded 6 times
Last edited by avengahM on Thu Jul 05, 2018 5:07 pm, edited 2 times in total.

User avatar
Richard Russell
Posts: 544
Joined: Sun Feb 27, 2011 10:35 am
Location: Downham Market, Norfolk
Contact:

Re: Beebem Mode 7 glitch

Post by Richard Russell » Thu Jul 05, 2018 7:47 pm

avengahM wrote:
Thu Jul 05, 2018 5:00 pm
Does anyone have an actual BBC to test the conceal / hold graphics interplay? Does the conceal code have priority over the hold graphics code?
I posted a photo of the testpage on a real BBC Master above (it was originally a Master 512 but the internal co-processor has been removed): you can see that 'conceal' has priority over 'hold' (in the coloured stripes to the right of 'Test Page'). I still think that's the only sensible interpretation of the spec.

colonel32
Posts: 57
Joined: Wed Jan 18, 2017 7:59 pm
Location: USA
Contact:

Re: Beebem Mode 7 glitch

Post by colonel32 » Thu Jul 05, 2018 8:52 pm

I've noticed a javascript error in edit.tf when running in Chrome and exporting every format except zxnet editor:

Code: Select all

Not allowed to navigate top frame to data URL: <URL>
The export works fine in Safari.

Possible iframe workaround here: https://github.com/niklasvh/html2canvas/issues/1271
Last edited by colonel32 on Thu Jul 05, 2018 8:54 pm, edited 2 times in total.

guesser
Posts: 240
Joined: Mon Jun 26, 2006 9:21 pm
Contact:

Re: Beebem Mode 7 glitch

Post by guesser » Thu Jul 05, 2018 9:06 pm

colonel32 wrote:
Thu Jul 05, 2018 8:52 pm
Possible iframe workaround here: https://github.com/niklasvh/html2canvas/issues/1271
The real fix is to go through the hassle of converting the code to generate proper file saves instead (which brings its own cross-browser annoyances!). I don't think Simon has the time/inclination to work on edit.tf at the moment though.
A web based teletext editor which can export as Mode 7 screen memory: https://zxnet.co.uk/teletext/editor

colonel32
Posts: 57
Joined: Wed Jan 18, 2017 7:59 pm
Location: USA
Contact:

Re: Beebem Mode 7 glitch

Post by colonel32 » Thu Jul 05, 2018 10:21 pm

guesser wrote:
Thu Jul 05, 2018 9:06 pm
I don't think Simon has the time/inclination to work on edit.tf at the moment though.
Understood. Easily accommodated at my end, and of course ZX Net works a treat ;)

Post Reply