What is this about: I'm doing a contest with cqrlog. I had started a new log (database) just for this contest. I'm using the contest entry window. I search over the band and hear a station calling CQ. I type the call into cqrlog. I want to know whether I worked this station before, on this band and in this mode, so it is a dupe, or it's a new one I should call.
What I would love: A big "DUPE" displays right in the contest entry window if I have that station before, right in the contest window where I see it immediately. It wouldn't hurt if the window's background color changed to red.
What I actually have to do: Move my eyes attention from the Contest window to the New QSO window. Scan the list of previous QSOs displayed there. Check manually, whether any of the QSOs has the same band and the same mode. I'm much less proficient at this, compared to my computer. In case there are a few, I'd even have to use the scroll bar for that.
Hi,
we have already implemented this feature:
-Add: Contest form has simple dupe check. Call turning red after leaving call edit box if worked at same
band and mode. Needs new log for each contest, or if same log used for all, setting date and
checkbox of "preferences/fldigi/wsjtx interface"-"WB4 chk starts from"-"call"-"contest start date"
-Add: Qso list has now all contest columns (set by preferences/visble columns)
-Add: New columns for contest exchange and freqmem info upgrades database to version 16
All corresponding views (contest, qso list,NewQso,ADIF ex/Import,TRXControl,log filter) updated
to use new columns.
https://github.com/ok2cqr/cqrlog/commit/ddee46edc6df1173d9588c0179217b34...
you can do the following to use this feature
1) compile the official beta version by yourself from Petrs github master https://github.com/ok2cqr/cqrlog
2) or you are using the alpha version 220 from Saku, OH1KH
3) or you are waiting for the delivery of the next official (beta) version of Petr and Martin.
55 & 73 de
Andreas (DL7OAP)
Thank you, Andreas,
that solves the problem for me.
However, just changing the color is by itself not so useful for a color-blind person. So we might want to have an additional visual clue of the dupe.
I took a roundabout way to get this stuff. I am somewhat of a timid person sometimes, so when I found Saku offers stuff via unencrypted http, that made me a bit nervous. When I later found that I'm asked to copy files into places the Debian package manager considers under its control, I got nervous even more. So I tried something else, namely, to build myself. It took me some four hours later, but I succeeded. I'm sure to sleep well now ;-) .
HI!
Dupe check also uses bold font with red color. So I think you can see the difference even without seeing the color change.
Sorry , but my ISP offers only unencrypted http for home pages with their internet connection. And my own server is behind my ADSL that has only 800k out (10M in, as using over 2km copper line) so I do not like to open my own https port to public.
Of course you can/should compile the binary by your self if you want to be sure. Either from Petr's Git, or my cqrlog-devel Git that has the same source as my alphaTest binary.
But can you be sure that there is not troijan inside the code? Cqrlog has quite many lines of code. Do you check it all?
Not mean to be rude, but if we linux users do think that we are safe when we compile our programs by our self from source that is not true. Unless we check EVERY LINE of the code before using it.
--
Saku
OH1KH
Hello, Saku,
you are absolutely right that I don't achieve absolute safety by being careful what and where I fetch. There is no absolute safety. I do not personally check the source code of all software I run. We agree.
On the other hand, even if a mere brick is all the tooling needed to break into my house through a window, I might still choose to lock the door.
In the above, I described my opinion. My opinion is: "It is good current practice to use some kind of tamper-proof cryptography to secure stuff received is the same as sent". But I don't want to force anything on you.
That said: Should you happen to be interested in security improvement suggestions for the alpha version distribution work you do, I'd be happy to help with that.
One option would be HTTP (as presently) + GPG-signatures. That's very very easy to do.
A (possibly more involved) second option would be to check out what Github offers how large binaries can be distributed through their services.
There are other options.
Ok. That's an offer. If it's something you're interested in, and want details from me, you can take me up on it. Just contact me.
If you are not interested, fine. In that case, kindly excuse the noise and ignore this entire post.
In closing: I'm glad that you're working on a logging software that I can use. Thank you very much for that!
Regards,
Andreas DJ3EI
HI!
Yep!
I have been thinking that. You both ideas are good.
Also making md5sum from zip and release the checksum here and binary in own web site could be one solution.
About the dupe check:
Would it help even a bit more if dupe would be marked red, bold + turn to low case letters ?
That would be very small fix to current code.
--
Saku
OH1KH
Hi, Saku,
> I have been thinking that. You both ideas are good.
thank you! I'm happy to help you. Normally, it is the other way round: You help us.
> Also making md5sum from zip and release the checksum here and binary in own web site could be one solution.
Yes. Or almost yes. For some years,
md5sumhas been considered weak for this application. For one statement (admittedly somewhat out-of-context), RFC6151 says: "MD5 must not be used for digital signatures".I recommend you use
sha256sum. This is strong and healthy, it is what people use nowadays for tamper-proofing.(If you went the GPG route, you could also publish the signature itself via insecure http. No need for two different web sites. You could even put the GPG signature of the inner ZIP into the outer ZIP.)
Vy 73, Andreas, DJ3EI
> About the dupe check:
> Would it help even a bit more if dupe would be marked red, bold + turn to low case letters ?
> That would be very small fix to current code.
I'm not color blind myself, but I think it would be an improvement.