Hi folks
I am trying to install the Husky Project software on my Debian 10 box.
My plan is to run a current copy of HPT and HTick along with whatever
I need to work with these. I understand the components / files I need
to compile are
areafix
fidoconf
hpt
htick
huskybse
huskylib
I have cloned these from a git repo so they sit as subdirs under a
holding folder called huskyproject
My plan is to run a current copy of HPT and HTick along with whatever I need to work with these. I understand the components / files I need to compile are
areafix
fidoconf
hpt
htick
huskybse
huskylib
once all compiled I *think* I need to either place all the complied stuff in one 'husky' working directory or locate complied stuff in sub dirs off a husky dir?
I'm trying to suss how best to start compiling and it looks like for huskybse there's a config file called huskymak.cfg that needs some attention.
I'm a tad lost to know what should/should not be changed in this file, especially in relation to directories and filenames?
Any tips on how to compile and install / shift files to desired locations once installed are appreciated.
Hi folksnormal recomendations would be to create a husky folder, then cd into that folder, and run git clone on the modules you need.. then goto the huskybse folder and cp huskymak.cfg ../
I am trying to install the Husky Project software on my Debian 10 box.
My plan is to run a current copy of HPT and HTick along with whatever
I need to work with these. I understand the components / files I need
to compile are
areafix
fidoconf
hpt
htick
huskybse
huskylib
I have cloned these from a git repo so they sit as subdirs under a
holding folder called huskyproject
once all compiled I *think* I need to either place all the complied
stuff in one 'husky' working directory or locate complied stuff in sub dirs off a husky dir?
I'm trying to suss how best to start compiling and it looks like for huskybse there's a config file called huskymak.cfg that needs some attention.
I'm a tad lost to know what should/should not be changed in this file, especially in relation to directories and filenames?
Any tips on how to compile and install / shift files to desired
locations once installed are appreciated.
My plan is to run a current copy of HPT and HTick along with whatever I need
to work with these. I understand the components / files I need to compile are
I have cloned these from a git repo so they sit as subdirs under a holding folder called huskyproject
once all compiled I *think* I need to either place all the complied stuff in one 'husky' working directory or locate
complied stuff in sub dirs off a
husky dir?
I'm a tad lost to know what should/should not be changed in this file, especially in relation to directories and filenames?
Any tips on how to compile and install / shift files to desired locations once installed are appreciated.
It's important that you build huskylib, smapi and fidoconf before any others, and also, areafix before hpt.
after getting huskymak.cfg in place run "make" in each directory
huskybse there's a config file called huskymak.cfg that needs some attention.
Yes, put that in the root husky dir and make any changes if needed, make will use that each time it is called.
Just the order, you need to start with huskylib, smapi, fidoconfig and areafix. That's all in my case anyway.. but we are here if you need us..
normal recomendations would be to create a husky folder, then cd into
that folder, and run git clone on the modules you need.. then goto the huskybse folder and cp huskymak.cfg ../
then follow the install instructions in huskybse
One of the good things with docker, is application builds are build like
a receipe - here is the receipe I use to build the Hub 3 container.
You'll get an idea of what I do to install husky.
You'll get an idea of what I do to install husky.
It's important that you build huskylib, smapi and fidoconf before
any others, and also, areafix before hpt.
do I need smapi if I plan to to run JAM bases? Or is it best to have
this as an option even if not used?
What of huskybse - when to compile that? Assume I need it to?
What of huskybse - when to compile that? Assume I need it to?
By in place you mean copy it from huskybse to the root dir huskyproject?
Do I need to do the same for each other program as well before compiling?
Deon seems to have set PERL=1 so perl support enabled?
I also see USE_HPTZIP=1 in Deon's script, what's the merits of that one guys?
Can someone comment on the CFGDIR=$(PREFIX)/etc/fido option in huskymak.cfg do I need to alter that or be locked into using that? Or best to go with this default?
huskybse - when to do that one?
Deon seems to have set PERL=1 so perl support enabled?
I also see USE_HPTZIP=1 in Deon's script, what's the merits of that o guys?
Yes, I have both of those options enabled although I don't use them. If
I use hptzip then htick can't get file descriptions from the file_id.diz when tossing files, so I just use zip and unzip.
Can someone comment on the CFGDIR=$(PREFIX)/etc/fido option in huskymak.cfg do I need to alter that or be locked into using that? Or to go with this default?
You can go with the default. If you leave it alone the husky stuff will look for a file called /usr/local/etc/fido/config for it's config file.
I use /usr/local/etc/config but that is a choice you make. You can also call the file something other than config if you like, but it'll work as it is.
huskybse - when to do that one?
Yes, you just need the huskymak.cfg from there, copy it down a directory to your husky root, edit it to your needs and go.
You dont actually compile huskybse, just the ones mentioned in the
message from Al.. I left mine default, I use a different directory for
all my config files and just use a symlink to that dir..
I'm a tad lost to know what should/should not be changed in this file especially in relation to directories and filenames?
I only change one thing IIRC, I change etc/lib to etc/lib64 since I need to do that on slackware.
Yes, I have both of those options enabled although I don't use them.
If I use hptzip then htick can't get file descriptions from the
file_id.diz when tossing files, so I just use zip and unzip.
OK I'll use them then but may need a hand to bypass should I hit the same issues.
it seems a pain to have the config there if I am putting the rest of the files in a dir called husky, I'm wondering if I should change that link. Hmm
2) I think I duffed up a setting in the huskymak.cfg and set CFGDIR=$(PREFIX)/hub/husky
I was wanting the config to go to hub/husky but forgot about the prefix variable which was /usr/local
/usr/local/hub/husky does not seem to be created nor does there seem to be a config file in there, should I have expected to see a default one there?
3) If I were to start over now by making a change to the config file, am I safe to remove the compiled files from /usr/local/bin (do I do that first?) then run the make and make install process again thereafter?
The huskymak.cfg should be in your husky root directory, all your husky builds will use it.
I saw you were using huskyutil, that software was buggy at one time so I switched to using sqpack. Those bugs may have been resolved in the mean time, I don't know but sqpack works with Squish, JAM and *.MSG areas. I have always used sqpack with good results.
I ran the make and sudo make install and found a ton of compiled
programs in /usr/local/bin with root group/owner privs.
1) can I move them to my dir of choice and change them to my user
group and ownership?
2) I think I duffed up a setting in the huskymak.cfg and set CFGDIR=$(PREFIX)/hub/husky
You can take the $(PREFIX) out and put your own path if you like. CFGDIR=/path/to/my/husky/directory
/usr/local/hub/husky does not seem to be created nor does there seem a config file in there, should I have expected to see a default one t
No, you will need to create that directory along with a config of your own.
There are examples to start you off in hpt/config and htick/misc. There are examples in other directories too.. you'll need to work through that.
I can send you my configs in a zip if you'd like to have a look.
3) If I were to start over now by making a change to the config file, safe to remove the compiled files from /usr/local/bin (do I do that first?) then run the make and make install process again thereafter?
Yes, you can just delete them and start fresh. I like the binary files in /usr/local/bin since that directory is on the path. You can put them anywhere but if they are not on the path you have to use ./hpt or
./htick to run them.
When I built it, I ran the executable and it came back with this
hptutil/lnx 1.9.0-cur 2011-03-09
Usage: hptutil [options]
Options: sort - sort unread messages by time and date
link - reply-link messages
purge - purge areas
pack - pack areas
fix - fix base (hptutil fix -? for more help)
-j - link Jam areas by CRC (great speed-up)
-k - keep import.log file
-q - quiet mode (no screen output)
-i <filename> - alternative import.log
So it looks like it's old circa 2011... Hmmm..
I'll grab sqpack and compile it also... but I'm going to wait until I
read your advice about my cfg gaffe and the best way for me to
recompile (delete files already created? etc.)
Better advice must come from someone who uses hptutil, or someone who writes/maintains it.
Thanks yes please, I did have some from you ages ago but it wouldn't
hurt to see them now, so thanks.
You make a good point about pathing. I may do the same then. If I do
move I guess I need to change the group:owner to me and find a way to
add the dir I move them to the path
So because I compiled the first attempt of all my binaries with the
wrong pathing to my intended config file, those compiled binaries need
to be deleted and recompiled else they will always look to the wrong location for config info - is that right?
So because I compiled the first attempt of all my binaries with the wrong pathing to my intended config file, those compiled binaries nee to be deleted and recompiled else they will always look to the wrong location for config info - is that right?
Yes, or you can run hpt with the -c option and point it to another location..
I logged into the file manager as root and removed the compiled files
from /usr/local/bin
Then I made the change to the cfg file then went into huskylib folder
and ran make again... this time make just says 'nothing to be done for all' so it's like I can rerun the make process... am I missing something?
That should read 'I can't rerun make'
Do I need to do another git clone of each repo to reset the original filesfirst?
Try running:
make clean
Then retry make.
Questions on a few of the git repos you pull in / use...
htpzip - so this is used to offer zip/unzip within HPT without the need to install infozip or similar? It would also explain
the switch change you make
in your config file for huskybse.
hptutl - this looks like the jam packer, purge etc. tools, I'm picking I should get these as part of this 'standard' build?
htpkill - what do you use this for? Do I need it?
areastat / fidostat / nltools - I take it are all worth getting for the info they offer?
Do I need to do another git clone of each repo to reset the original files first?
Yes, I dont recall why I added hptzip - because my image also has zip/unzip installed anyway. There was something I was doing with husky
and it complained about no hptzip, so I added it.
hptutl - this looks like the jam packer, purge etc. tools, I'm pickin should get these as part of this 'standard' build?
Yes, I believe so.
When you get sick of compiling and configuring and installing, just pull my container and you'll save all that time ;)
You can just delete them an reclone, but you just need to run "make
clean" and run make again.
Do you use this tool?hptutl - this looks like the jam packer, purge etc. tools, I'm pickin should get these as part of this 'standard'Yes, I believe so.
build?
Do you use this tool?
I havent yet - but I probably should soon.
The hub has been running for 9 months or so, so the message bases are
not that big. I'll probably start playing with that util when I get a
free moment.
I haven't even looked at that either... I still need to do something with the log files too... They grow fast!
I actually have logrotate in the image, and cron - so I map in a cron.daily into the image that rotates the logs and stores them in an "old" directory.
Yell out if you want more detail...
do I need smapi if I plan to to run JAM bases? Or is it best to have this as an option even if not used?
Yes, any detail you could send would be great. I've been moving them manually, when I notice they're getting close to 100MB...
3) Inside the container, I already have a /etc/logrotate.d/fido, which assumesthe logs are all in /var/log/fido, and will rotate them to /var/log/fido/old/.My /var/log/fido is a mount with -v ${SRC}/log:/var/log/fido - so on my hostthey are all in ${SRC} (/srv/docker/fidohub)
You should be good to go...
When does this execute? Is it a midnight when systemd will run it?
This is controlled by /etc/crontab (inside the container) - so the cron.dailyruns at 6.25am.
You probably want to set your timezone (/etc/localtime) - which will be Australia/Melbourne, unless you have overridden it.
If you want, you can "extend" the image that you get from me - I can
give yousome tips (if you arent already doing that) - instead of mapping in theoverrides with -v on the command line. Both work...
| Sysop: | altere |
|---|---|
| Location: | Houston, TX |
| Users: | 66 |
| Nodes: | 4 (0 / 4) |
| Uptime: | 11:09:27 |
| Calls: | 728 |
| Calls today: | 1 |
| Files: | 7,667 |
| Messages: | 295,542 |