And yet another batch of Comp 2006 reviews

75 views
Skip to first unread message

Richard Bos

unread,
Nov 22, 2006, 7:09:35 PM11/22/06
to
Last year, I couldn't be arsed to post reviews, not even the small ones
I tend to write. With all those games being entered that clearly hadn't
had any effort spent on them, why should I spend effort reviewing them?
One can continue to try and correct, oh, fill in the usual trolls, and
get them to spell-check their games and not take the piss all the time,
but really, one only encourages them.
And yet, I wanted to give _some_ feedback. And I'm vain enough to want
to air my opinion. So here's what I've decided to do. I'm only posting
reviews for those games that I _did_ like, or at least did not find
awful. Perhaps by showing writers what they should do, rather than
telling trolls what they shouldn't, I can have a more positive effect.
I am, arbitrarily, putting the cut-off point at the halfway mark; since
my highest mark in this Comp was an 8 (yes, I'm harsh, but IMO rightly
so), that's 5 and up. There will be one exception, the reason for which
will be obvious.
The usual cautions apply. There will be spoilers. There will be Opinions
and criticism. There will be arbitrariness. Oh, and whatever I write in
these reviews, my final mark is based on one thing, and one thing alone:
how much I enjoyed playing the game. The rest is mere detail.
So, here they come:


Another Goddamn Escape the Locked Room Game: 5
There were one or two odd bugs, all from the looks of it due to less
than thorough implementation. E.g., "look through window" tells you it's
empty. "Remove photo" does separate the frame and the photo, "remove
frame" does not. There's some more guess-the-verb and default messages
that don't fit. It also has no story to speak of and weird puzzles, but
that at least is intentional. An amusing game, but not great.

Pathfinder: 6
A game with several bugs, one involving the game staying in fixed pitch
where it shouldn't, and some with state being remembered for too long
(Steve, whose dead body was lying in his room, opened the door for me
from the outside!) or not long enough (don't leave the company grounds
until you've found everything you needed to do, or the game is stuck).
The story works - it's far from logical, but if you start with the
premise that the PC is the kind of guy who lets his life be run by a
mobile phone or handheld, and such people obviously exist (hence the
nickname Crackberry, e.g.), it does work; the optional points you can
gain work; the technical side lets it down.

The Traveling Swordsman: 6
I like the design of the graphics; it's more tasteful than the usual
banners. They still take up too much disk space, though, which I
disapprove of in a work of IF. Sorry; I know that's a pet hate of mine.
The puzzles are good; tough, but fun to play. The division into separate
scenes doesn't make the game stronger (something I've seen more of in
previous comps, but which was absent this time), but at least the
atmosphere is consistent throughout. Unfortunately, the ending breaks
that consistency down and mars the game for me. FWIW, I hadn't noticed
that the PC was supposed to be deaf, putting the oddness of the world
down to the fantasy atmosphere, but even if I had, it wouldn't have made
a difference to the score.

Mobius: 6
The idea behind this game is good; in fact, it's very good. The writing,
too, is competent, as is the technical execution, with the exception of
a certain sluggishness. Unfortunately, I found Mobius far too fiddly.
I'm sure there's some rule somewhere, but it seems quite inconsistent in
what you can or cannot do, and what you can touch or not. This is not
helped by there being, despite a promise to the contrary in the help
menu, no hints. Some times you can touch, say, the reactor; some times
you cannot; and I couldn't tell which were which until I examined it,
wasting valuable turns. If that had been remedied; if I had been able to
spot that the reactor or the cabinet had started to become "transparent"
without having to examine it each time (it is, after all, something
which would be quite obvious); I probably would have enjoyed Mobius
quite a bit more.

Madam Spider's Web: 7
The writing in this game is well done, very atmospheric. Unusually, the
change to a completely different scene works relatively well, in part
because of the influence your behaviour in the dream(?) can have on the
endgame. The puzzles, too, are nice; not too hard, not too easy, with,
e.g., several ways to discover the solution to the piano puzzle.
Unfortunately, Madam Spider's Web is far too short.

Star City: 5
This game starts well: the implementation is OK, although it misses a
few special commands that I'm used to having (help, full score) and
there's a scrolling bug in the elevator scene, which I think is due to
Inform 7 because I found it elsewhere, too. In this part, the puzzles
are enjoyable, well-judged; and the story is fine. And then... the damn
thing turns into a Lunar Lander game! Sorry, but that just won't do. The
story suffers similarly; the shift from a callous fortune hunter into a
conscientious historian just isn't believable. Good start; lousy finish.

Legion: 6
Weird; in fact, too weird. I never did get the hang of Legion, and only
partly got an idea of what I was supposed to be and do. You do have many
options and quite a bit of choices and possibilities to explore, which
is good. The downside is that it's hard to figure out which of these
options do something, and which of _those_ do something useful - and
this isn't helped by the weirdness being of such a nature that I
couldn't use my normal idea of what would work, much. The writing is
good, though, making your unusual nature - or at least the fact that you
_are_ an unusual nature - quite clear.

Strange Geometries: 7
Apart from some slight spacing and spelling errors, and some odd bugs
with messages, the main problem with this game was that turning on the
detector made it run and react painfully sluggishly. But for that,
though, the puzzles were fun (except at the very end, where there were a
couple that struck me as very unfair); and the story was also good, with
a wonderful twist. Ok, it may be that I was already fond of both of its
sources in their own right, and so was prejudiced in Geometries' favour,
but I did rather like the way in it combined them.

The Primrose Path: 8
A solidly well-written game. The story as a whole makes sense within its
own world, and the text is of similar pleasant quality. The world, odd
as it is, is well designed, and well described. The puzzles, too - more
precisely, the whole string of puzzles, most of them interlocked - are
nicely and intricately balanced. Here and there they were a bit on the
unhinted side, but most were fine. If there's a downside it's that the
game reacts just a bit slowly, but since The Primrose Path does not have
an action-packed thriller plot anyway, that doesn't let it down.

Delightful Wallpaper: 8
As the title says, delightfully quirky. Unfortunately I couldn't solve
the first part without the walkthrough - or at least wouldn't have been
within the time limit - and the second part was doable but very hard
(but at least I could solve it without following the walkthrough by
myself). The pad was a nice invention which at least mitigated this
somewhat. Anyway, this did not stop me enjoying the wonderful oddness
and sheer joy of the language which permeates the game. I see that this
game was written by Andrew Plotkin; I hadn't guessed that while playing
- to be precise, I hadn't tried guessing at all - but I'm not at all
surprised.

The Tower of the Elephant: 6
A well-made game, with a good, even if not very original, story. (I see
now that it was in fact an adaptation; I'd missed that while playing the
game.) The multiple ways through are a good idea; one time I thought I'd
found an oversight when you go through the game in a way which isn't in
the walkthrough at all (to be precise and only a little spoilery, going
through the bottom instead of over the top), but it was catered for
after all. One downside is that there is too much talk for what should
be an action game. The one thing that really lets TTotE down, though, is
that it is much too short, and the replayability only partly helps with
this.

Fight or Flight: 5
This game's strength is at the same time its Achilles' heel. The
independent NPCs are a good idea, and add some flavour to the action,
but in practice they're too fiddly to work with. It's too hard to figure
out who does what (and the hints for this aren't always conclusive), and
you really have too little time to set it all up right. Nor does the
game always gives you the options you should have; the most egregious
example of this being the puzzle with the hood near the start. That
said, having two options out of the game is good; and the story itself
_is_ hackneyed, but that is appropriate for the genre.

Moon-Shaped: 8
A good, original retelling of an old story. Some of the puzzles (in
particular the use of the monocle in several places, and the stick
puzzle) need better clues, and the hints need to follow the game
progress better; but that's the worst of it, really. Overall Moon-Shaped
is an imaginative and enchanting work, with good, enjoyable writing as
its strongest point.

Floatpoint: 4
It's politics. Politics irritate me, especially the conniving,
balancing-all-but-lying-to-everybody kind. Ok, so it's well-implemented
politics; from this author, one wouldn't expect anything less.
Floatpoint still aggravated the stuffing out of me. I spent the whole
game wanting to take both sides in the negotiations by the scruff, shake
them about thoroughly, slap them about for good measure, and tell them
not to be such irritating, sanctimonious, self-important gits. Much, in
fact, as I often would like to do with our own darling politicians and
diplomats. Under such circumstances, the execution can be as perfect as
you like; I'm not going to enjoy a game. And enjoyment is - as I wrote
above - what I score games on. The moral of this review? Probably that
you can't possibly please all the players, all the time. Not even if
you're Emily Short.

Unauthorized Termination: 5
Of course, UT suffers from the usual Adrift bugs: e.g., GIVE PEBBLE (to
Zeta-Theta being implied, and in fact correctly inferred by the 'terp)
does not work, but an explicit GIVE PEBBLE TO ZETA-THETA does. The
puzzles are good enough, though there aren't many of them and most of
the game is just TALK TO <X>; TELEPORT; RINSE; REPEAT, and the last one
was just not right. The story is also not bad at all, but made harder to
follow by too many codes and numbers. These do evoke a suitably robotic
atmosphere (or appropriate lack of atmosphere, if you wish), but
unfortunately also get in the way. The best of the Adrift lot; but it
could have been a lot better.


Finally, a few general thoughts.

Inform 7. Mixed feelings. Clearly it encourages writers, which is a good
thing; but IMO equally clearly, these writers don't always get all the
technical sides right, which is just as important. Some of the Inform 7
games missed a decent hint system (or had a broken one!); they appear
invariably larger than they could've been with Inform 6, which may seem
a minor point, but it adds up; quite a few had scrolling bugs, in
particular near cutscenes; a few were noticeably slower than other ZCode
games.
Some of these problems can be solved by a new, improved edition of I7;
some are simply due to writers underestimating the extent to which
writing a work of IF still is fundamentally _programming_, even if I7
tries to hide that. The result is sometimes rather like driving a
gorgeous sports car, of which the exhaust will keep rattling. Nothing
that will stop you admiring the beauty of the car, you see; but it does
rather spoil the drive occasionally.

"Jar of Tang". Do a web search on it. Then knock it off, already.

Spell-checkers. Yes, they're important, even for good authors.

And finally, I still think that "a ways" does not belong in a work of IF
unless it's from the mouth of a dialect-speaking NPC. But I didn't
reduce any game's score for it.

Richard

Emily Short

unread,
Nov 22, 2006, 9:16:12 PM11/22/06
to

Richard Bos wrote:
> Floatpoint: 4

> Floatpoint still aggravated the stuffing out of me. I spent the whole
> game wanting to take both sides in the negotiations by the scruff, shake
> them about thoroughly, slap them about for good measure, and tell them
> not to be such irritating, sanctimonious, self-important gits. Much, in
> fact, as I often would like to do with our own darling politicians and
> diplomats.

Heh. Sorry you were aggravated, but it's hard not to take that as a
compliment.


> Some of the Inform 7
> games missed a decent hint system (or had a broken one!); they appear
> invariably larger than they could've been with Inform 6, which may seem
> a minor point, but it adds up; quite a few had scrolling bugs, in
> particular near cutscenes;

What were the symptoms, and on what interpreter did it show up, if you
recall? It's possible this is the fault either of my screen effects
extension or of the way it is commonly being used, in which case I
should fix either the code or the documentation.

Andrew Plotkin

unread,
Nov 23, 2006, 12:38:52 AM11/23/06
to
Here, Emily Short <ems...@mindspring.com> wrote:

>
> Richard Bos wrote:
>
> > Some of the Inform 7
> > games missed a decent hint system (or had a broken one!); they appear
> > invariably larger than they could've been with Inform 6, which may seem
> > a minor point, but it adds up; quite a few had scrolling bugs, in
> > particular near cutscenes;
>
> What were the symptoms, and on what interpreter did it show up, if you
> recall? It's possible this is the fault either of my screen effects
> extension or of the way it is commonly being used, in which case I
> should fix either the code or the documentation.

I have now seen a couple of I7 games which (when played in Zoom) left
a long ladder of translucent status lines tracked up the scrollback.
To be extra helpful, I've forgotten which games they were.

My guess is that this is the result of the game (or maybe an
extension) doing something squirrelly in the status line -- something
which seems reasonable when writing I7 code. However, my couple of
experiments in reproducing it have failed.

--Z

--
"And Aholibamah bare Jeush, and Jaalam, and Korah: these were the borogoves..."
*
If the Bush administration hasn't shipped you to Syria for interrogation,
it's for one reason: they don't feel like it. Not because you're patriotic.

JDC

unread,
Nov 23, 2006, 1:00:18 AM11/23/06
to

Andrew Plotkin wrote:
> Here, Emily Short <ems...@mindspring.com> wrote:
> >
> > Richard Bos wrote:
> >
> > > Some of the Inform 7
> > > games missed a decent hint system (or had a broken one!); they appear
> > > invariably larger than they could've been with Inform 6, which may seem
> > > a minor point, but it adds up; quite a few had scrolling bugs, in
> > > particular near cutscenes;
> >
> > What were the symptoms, and on what interpreter did it show up, if you
> > recall? It's possible this is the fault either of my screen effects
> > extension or of the way it is commonly being used, in which case I
> > should fix either the code or the documentation.
>
> I have now seen a couple of I7 games which (when played in Zoom) left
> a long ladder of translucent status lines tracked up the scrollback.
> To be extra helpful, I've forgotten which games they were.
>
> My guess is that this is the result of the game (or maybe an
> extension) doing something squirrelly in the status line -- something
> which seems reasonable when writing I7 code. However, my couple of
> experiments in reproducing it have failed.

I also remember seeing some strange behavior in Zoom with a couple of
games (I6 ones) with a lot of boxed quotes (IIRC I noticed this with
Vespers, and maybe also Christminster). It's been a while, but I think
this happened when the buffer size was set to the smallest setting, and
it would cause translucent versions of old quotes to reappear later.

-JDC

TheR...@gmail.com

unread,
Nov 23, 2006, 3:10:30 AM11/23/06
to
"Jar of Tang". Hmm. I wonder if that sentiment could be applied to
"Sisters" and its horribly contrived tale of Murder With No Freakin'
Motive and subsequent Guilt-Ridden Suicide, all for the sole purpose of
revealing in the end that you're a murderer with a pack of vengeful
spirits chasing you.

That's such a fun phrase, I think I'll keep it. ^_^ "Jar of Tang"...

David Goldfarb

unread,
Nov 23, 2006, 5:37:25 AM11/23/06
to
In article <4564e115...@news.xs4all.nl>,

Richard Bos <rl...@xs4all.nl> wrote:
> Mobius: 6
>The idea behind this game is good; in fact, it's very good. The writing,
>too, is competent, as is the technical execution, with the exception of
>a certain sluggishness. Unfortunately, I found Mobius far too fiddly.
>I'm sure there's some rule somewhere, but it seems quite inconsistent in
>what you can or cannot do, and what you can touch or not. This is not
>helped by there being, despite a promise to the contrary in the help
>menu, no hints. Some times you can touch, say, the reactor; some times
>you cannot; and I couldn't tell which were which until I examined it,
>wasting valuable turns.

Um, I thought it was reasonably simple, and clued by the lab notes
on the workbench. Anything you've affected in one iteration of the
loop gets drawn into the loop with you, and is thus untouchable in
the next. So let's say you've picked up the lab coat and the radiation
dampener in this iteration, and opened the cover of the cabinet.
Then in the next one, you won't be able to touch the lab coat or the
dampener, and the cabinet cover will be intangible -- translucent,
and able to be reached through.

This turns out to be crucial to the ultimate solution, which involves
using this cross-loop effect across *three* iterations. I thought it
was fairly ingenious.

There was one aspect to it that I thought was unfair, or at least
underclued: that the built-up radiation affected only the panel cover,
and not the panel itself. If you try to touch the panel cover directly,
you get knocked for a loop by radiation. But if the panel cover is
intangible, you can safely touch the stuff *behind* the cover. I had
thought that the whole panel would be affected. This kept me from
figuring out the whole solution for myself, although I did work out
just about everything else I needed to know. I wound up knocking one
point off my rating for that, giving the game an 8 instead of a 9.

--
David Goldfarb |"Hello, this is Leslie Down with the daily home
gold...@ocf.berkeley.edu | astrology report.
gold...@csua.berkeley.edu | TAURUS: Contemplate domestic turmoil.
| AQUARIUS: Abandon hope for future plans." -- TMBG

Mike Snyder

unread,
Nov 23, 2006, 11:19:42 AM11/23/06
to
<TheR...@gmail.com> wrote in message
news:1164269430....@j44g2000cwa.googlegroups.com...

I'd say no. It's more that the setting isn't what you were made to believe
(TTS, Madam Spider). In other words, "but wait, see, it wasn't really a
desert with glass walls -- they were in a jar of tang all along." Something
like that. Was this a Twilight Zone episode? I've heard the label before --
and the explanation -- but I forget the context.

---- Mike.


Mike Snyder

unread,
Nov 23, 2006, 9:09:16 AM11/23/06
to
<TheR...@gmail.com> wrote in message
news:1164269430....@j44g2000cwa.googlegroups.com...

I'd say no. It's more that the setting isn't what you were made to believe

TheR...@gmail.com

unread,
Nov 23, 2006, 10:14:14 PM11/23/06
to

"Five Characters In Search Of An Exit". A man in a military dress
uniform awakens to discover that he's trapped along with a ballerina, a
hobo, a clown and a Scottish bagpiper in a mysterious, open-topped
metal chamber with no obvious way out, no memory of their previous
lives and no idea why the hell such an eclectic mess of characters is
in such a place. In the end, the man manages to climb out of the
chamber with the aid of the others, which is where the Jar of Tang
comes in: they're all toys in a charity drive collection bin. Actually
a pretty good episode.

28 IF

unread,
Nov 24, 2006, 10:39:31 AM11/24/06
to
Richard Bos come on down:

> The Traveling Swordsman: 6
>I like the design of the graphics; it's more tasteful than the usual
>banners. They still take up too much disk space, though, which I
>disapprove of in a work of IF. Sorry; I know that's a pet hate of mine.

Of course everyone's opinions are different, but I'm curious: Why? I mean, my
hard drive is 160 gigs. I can download 350K/sec through my Internet
connection. A game that tops in at slightly over one megabyte doesn't really
strike me as bloated, even if it would have been considered as such 20 years
ago.

d...@pobox.com

unread,
Nov 24, 2006, 10:51:08 AM11/24/06
to

Well, some people play on aging hardware, most text adventures are
quite suitable for such hardware. Then there's the PDA / phone thing.
Modern machines, but far from modern desktop performance. On my PDA
it's a bit tiring to play anything from the post Infocom era. The
Inform library is noticeably slower than the one that Infocom used, and
gets increasingly slow over time. The more modern the game, the slower
it runs. Since Richard Bos notes a "certain sluggishness" in a couple
of titles we can assume he's not playing them on a machine with dual 2
GHz cores. I'm curious though, what machine were you using Richard?

drj

Richard Bos

unread,
Nov 24, 2006, 6:18:47 PM11/24/06
to
"Emily Short" <ems...@mindspring.com> wrote:

> Richard Bos wrote:
> > Floatpoint: 4
> > Floatpoint still aggravated the stuffing out of me. I spent the whole
> > game wanting to take both sides in the negotiations by the scruff, shake
> > them about thoroughly, slap them about for good measure, and tell them
> > not to be such irritating, sanctimonious, self-important gits. Much, in
> > fact, as I often would like to do with our own darling politicians and
> > diplomats.
>
> Heh. Sorry you were aggravated, but it's hard not to take that as a
> compliment.

Well, in a way it is. It _was_ well done, and in a way probably even
realistic - it just went straight against my tastes. *Shrug* Such
happens.

> > Some of the Inform 7
> > games missed a decent hint system (or had a broken one!); they appear
> > invariably larger than they could've been with Inform 6, which may seem
> > a minor point, but it adds up; quite a few had scrolling bugs, in
> > particular near cutscenes;
>
> What were the symptoms, and on what interpreter did it show up, if you
> recall? It's possible this is the fault either of my screen effects
> extension or of the way it is commonly being used, in which case I
> should fix either the code or the documentation.

Right... first of all, it happened in several games, but I only remember
which game for the first time I noticed it, and that was Star City. The
'terp was WinFrotz 2.32v5.3. Old, I know, but it never had this problem
before.
What happened was this: if the game would print a lot of text, large
enough to need a [more] prompt, the game will scroll a whole screen of
text. Normally, that will be the whole lower window. In these cases,
though, it looked like it scrolled up the whole lower window _plus one
line_, with the first line being hidden behind the upper window. It's as
if the lower window isn't the size of the 'terp window minus the status
line, but the whole window, status line included.

Richard

Richard Bos

unread,
Nov 24, 2006, 6:18:49 PM11/24/06
to
d...@pobox.com wrote:

> On Nov 24, 3:39 pm, 28 IF <rushoma...@gmail.com> wrote:
> > Richard Bos come on down:
> >
> > > The Traveling Swordsman: 6
> > >I like the design of the graphics; it's more tasteful than the usual
> > >banners. They still take up too much disk space, though, which I
> > >disapprove of in a work of IF. Sorry; I know that's a pet hate of mine.
> > Of course everyone's opinions are different, but I'm curious: Why? I mean, my
> > hard drive is 160 gigs. I can download 350K/sec through my Internet
> > connection. A game that tops in at slightly over one megabyte doesn't really
> > strike me as bloated, even if it would have been considered as such 20 years
> > ago.
>
> Well, some people play on aging hardware, most text adventures are
> quite suitable for such hardware. Then there's the PDA / phone thing.

Quite. I like, for example, to play ZCode games on my Revo. A few years
ago the comp was no problem.

> Modern machines, but far from modern desktop performance. On my PDA
> it's a bit tiring to play anything from the post Infocom era. The
> Inform library is noticeably slower than the one that Infocom used, and
> gets increasingly slow over time. The more modern the game, the slower
> it runs. Since Richard Bos notes a "certain sluggishness" in a couple
> of titles we can assume he's not playing them on a machine with dual 2
> GHz cores. I'm curious though, what machine were you using Richard?

Actually, two. One is a 2GHz, 256 MB WinXP machine, the other I don't
have the details of right now but it's considerably older and still runs
Win98. But the odd thing is, even on the newer machine, the sluggishness
is less, but I can notice the difference, and in particular the reaction
of Strange Geometries when the detector is active is pretty obvious on
both computers.

Richard

Richard Bos

unread,
Nov 24, 2006, 6:18:48 PM11/24/06
to
gold...@OCF.Berkeley.EDU (David Goldfarb) wrote:

> In article <4564e115...@news.xs4all.nl>,
> Richard Bos <rl...@xs4all.nl> wrote:
> > Mobius: 6
> >The idea behind this game is good; in fact, it's very good. The writing,
> >too, is competent, as is the technical execution, with the exception of
> >a certain sluggishness. Unfortunately, I found Mobius far too fiddly.
> >I'm sure there's some rule somewhere, but it seems quite inconsistent in
> >what you can or cannot do, and what you can touch or not. This is not
> >helped by there being, despite a promise to the contrary in the help
> >menu, no hints. Some times you can touch, say, the reactor; some times
> >you cannot; and I couldn't tell which were which until I examined it,
> >wasting valuable turns.
>
> Um, I thought it was reasonably simple, and clued by the lab notes
> on the workbench.

Oh, it's clear enough after the fact what happened, and it's clear
enough in hindsight. But I couldn't keep it all straight within the two
hours.

> This turns out to be crucial to the ultimate solution, which involves
> using this cross-loop effect across *three* iterations. I thought it
> was fairly ingenious.

Oh, definitely. Top marks for interest.

Richard

Andrew Hunter

unread,
Nov 25, 2006, 11:18:06 AM11/25/06
to
On 2006-11-23 05:38:52 +0000, Andrew Plotkin <erky...@eblong.com> said:

> Here, Emily Short <ems...@mindspring.com> wrote:
>>
>> Richard Bos wrote:
>>
>>> Some of the Inform 7
>>> games missed a decent hint system (or had a broken one!); they appear
>>> invariably larger than they could've been with Inform 6, which may seem
>>> a minor point, but it adds up; quite a few had scrolling bugs, in
>>> particular near cutscenes;
>>
>> What were the symptoms, and on what interpreter did it show up, if you
>> recall? It's possible this is the fault either of my screen effects
>> extension or of the way it is commonly being used, in which case I
>> should fix either the code or the documentation.
> I have now seen a couple of I7 games which (when played in Zoom) left
> a long ladder of translucent status lines tracked up the scrollback.
> To be extra helpful, I've forgotten which games they were.

This will occur if the game sets the non-scrolling region of the upper
window to 0, but then draws the status line anyway. It seems to be the
way some menu systems are coded to behave: typically you'll see a game
that displays the status line correctly up until the point where a menu
is shown, and then it gets pasted over the text after that. To
reproduce it manually, just write a status line routine that calls
@split_window 0 at the start.

Zoom does gradually fade out box quotes in the history, so that the
text underneath them becomes readable again.

Andrew.

Reply all
Reply to author
Forward
0 new messages