newns.sh: added a note-22-dns command, to output text about fix for the newly encountered DNS problem when running openvpn in newns terminal (after update in Ubuntu 22.04.5)
newns.sh: added a startup section, which checks for WezTerm.AppImage (which would be a symlink to a 20.04 AppImage, for Ubuntu versions > 20.04), and runs that instead
sd-my-git-hours.py: encountered problems with cumulatic hours sum calc when using csvA output, fixed by implementing a new algo to calculate split streaks when streaks cross more than one day boundary, and by adding special handling for prepad streaks that cross a day boundary
sd-my-git-hours.py: changed up algo in calcAndPrintHours also for --repos part; just implemented the two-pass technique there too, and it works the same; generates same sum for test.sh test-git-dir repo as latest `--runtestsh`, so looks like this author+commit date timestamp integration is done for now
sd-my-git-hours.py: changed up algo in calcAndPrintHours (for the test part, without --repos) so it uses a two-pass loop to split off author and commit dates if they are different, and to insert labels (AC, AD, CD) on commit msg depending on what timestamp is leading; it seems it works fine, and integrates nicely with existing hours sum calc
sd-my-git-hours.py: changed up algo in calcAndPrintHours so it now uses correct fields in respect to having commit date in addition to author date in the git log entries during parsing; also turns out one command for escaping double quotes stopped (?!) working correctly, but managed to fix that; can confirm that finally `--runtestsh` call works once again ...
sd-my-git-hours.py: also inplemented retrieval of both author and commit date for dtCmtFirstTS/Offset, where it is set to the earlier of the two; finally also made dtSpecdRangeStart/dtSpecdRangeEnd initialize from dtCmtFirst/LastTSOffset, if specified_range_start/end are empty strings - this fixes the first test call in test.sh (which also makes a different commit and author date for first commit now), however now the second one crashes during json.loads (probably because of previous addition...
sd-my-git-hours.py: discovered that --runtestsh does not work anymore, because rest of script uses specified_range_start/specified_range_end regardless; started with inserting also extraction of commit date in addition to author date, added a sim in the generated test.sh script for difference between commit and author date; and implemented retrieval of both author and commit date for dtCmtLastTS/Offset, where it is set to latter of the two (and this will ultimately also be solution for specif...
sd-my-git-hours.py: fixed an issue with quoting; added more file types for csvA output
sd-my-git-hours.py: added a printout of requested range in HTML; fixed bug with passing wrong datetime timestamps from Python to HTML/JS; fixed bug of wrong calc of end of month; fixed zoom button calls in HTML/JS to correct timestamps, and made the HTML/JS show the requested range at start - HTML/JS now works as expected; script should hopefully be done for now
sd-my-git-hours.py: added more extensions in taskAreaMapFext / taskAreaMapProgFext as well as some special file extension checks for csvA output; sorted the list string outputs used in csvA output for easier reading; fix bug where empty strings ended up in taskAreas list, and added any git commit tags as Release under Output Description; csvA output is generally done, but discovered a bug in gitting the timeline view at start in HTML/JS
sd-my-git-hours.py: added another git call, to retrieve any tags associated with commit - if any, they would end up as Release in Task Area; seems to work but not really tested
sd-my-git-hours.py: added total count to Streaks num printout in csvA; added calculation of task description string, based on commit first words with colon; csvA output finally seems complete as all columns are populated, but likely some edge cases are missing
sd-my-git-hours.py: added maps between topics and file extensions changed in commit, taskAreaMapFext and taskAreaMapProgFext, and used them to generate the Task Area column in csvA output; works fine - but the maps have limited choices currently
sd-my-git-hours.py: implemented shell calls to git to retrieve files changed in a commit - works, but csvA output still messy
sd-my-git-hours.py: started extracting first words from commit descriptions, for task descriptions in csv output
sd-my-git-hours.py: added retrieval of commit hashes and repos in the splitStreak processing output for csvA, and used it to output the used repo names under "Project" column, and a sentence "X commits in N repositories" under Output Description; also discovered that CEST was reported as EET for the first UTC offset timezone match, so instead took the second, which fixes the problem (now there is only CET and CEST reported in csv output)
sd-my-git-hours.py: changes, so now ranged commits (that are drawn as ranges in merged commit track/lane) contain a property origcommitid, which refers to the original (gray points) commits; and now in the final step for --repos for csvA, the split streaks contain the origcommitid in their .indices - checked, they select the gray points correctly, looks good
sd-my-git-hours.py: found a solution for split prepad ranges - as the earlier split remains empty, simply copy the first of the indices of the later split into indices of the earlier split: this would handle both single and ranged commits
sd-my-git-hours.py: fixed CSS selection of single commits in HTML/JS; can now use either comma or a semicolon to select items in HTML/JS; splitting indices fixed by using "end" as measure for ranged commits, and "start" only for single commits - however, there is still a problem if a padding ends up being split
sd-my-git-hours.py: first attempt to split .indices for csvA output - it generally works, but some edge cases are not covered
sd-my-git-hours.py: splitting of streaks for csvA output started working; had to add a title property to class Streak to make it work; times are correct (final sum of hours is the same as when not splitting), but .numcommits and .indices still need to indicate the split (they are currently repeated/duplicated in split streaks)
sd-my-git-hours.py: now also ranges shorter than --pre_pad_singled get the pre_pad added; seems to work fine
sd-my-git-hours.py: added output of streak .indices into .csv; found an issue that clicking the longest/shortest streak links in HTML/JS selected the wrong commits, when --pre_pad_singles is active - fixed by remapping all streak .indices to new values in that case - now clicking those links select the expected commits
sd-my-git-hours.py: added actual .csv output from dataframe - .csv opens fine in LibreOffice; added debug printout of sum of HoursF - matches the sum in HTML/JS output; added Day and DateH columns; started preparing for splitting of streaks at day boundary
sd-my-git-hours.py: for csvA output: added TZOffsetToAbbrev function, which makes datetimes from timestamps with UTC offset print out CET/3-letter abbreviation with strftime %Z (instead of numeric offset); added HoursF column, which presents a streak duration as floating point hours
sd-my-git-hours.py: started populating DataFrame for csvA output with data from allStreaks, currently just times
sd-my-git-hours.py: started integrating --dumpout csvA option, modded the engine for --repos so it does the same calculations as in html/js case, and added dependency to pandas; for now, --dumpout csvA outputs nothing to stdout
sd-my-git-hours.py: now prepad ranges for single commits are also added as streaks to prepadStreaks, which are then integrated in allStreaks, and the total "work hours" duration calculated during the last iteration of allStreaks is now used as NUMHOURS passed to HTML/JS; finally, --pre_pad_singles functionality is fully done
sd-my-git-hours.py: fixed the insertion algo for --pre_pad_singles loop, now finally expected timeline ranges are seen in the HTML/JS timeline; however those are not counted as streaks yet
sd-my-git-hours.py: --pre_pad_singles loop now finally works without recursion, and prepads are generated correctly, however on end up on different "lines" in the same track in HTML timeline ?!
sd-my-git-hours.py: added --pre_pad_singles command line option, and found eventually where the code for it should sit; however that loop currently ends in a recursion
sd-my-git-hours.py: fixed issue with timeline min/max; zoom buttons in HTML now work
sd-my-git-hours.py: massive rewrite of how range timestamps are calced and passed from Python to JS; added button and handler for Zoom to specified range in HTML, but somehow cannot override the original min/max ranges set for the timeline at start
sd-my-git-hours.py: rearranged code to support calc of specified_range_start and specified_range_end variables, that hold the specification of the date/time range from the command line; added them to HTML output
sd-my-git-hours.py: added OrderedObjectDict class, and made CsvAEntry class inherit it
sd-my-git-hours.py: fixed problem with quoting commit messages; started writing CSV output in Python, but it is maybe better solved in JS?
sd-my-git-hours.py: added --checkrepos command line argument and corresponding function
sd-my-git-hours.py: reintroduced calc of totalDays.ncommits -> finally, numbers for commits per day are correct for merged repos, where ranged and single commits are mixed! Also removed console.logs (and note, green single commit point still remains if at very start of timeline)
sd-my-git-hours.py: fix for isStart calc, so that it does not react if any of startwords appears as a substring in a *word* in a commit message (only reacts on standalone startword); changes to Python code for --repos case in merged track, so that now it is determined programmaticaly if a "point" type commit is a startpoint of streak (in which case it is now blue, className "T"), of if it is "standalone" (remains red; code for startwords green points still exists, but the blue streak startpoi...
sd-my-git-hours.py: added some logs and notes - turns out, need a better condition for singlePointCommits, because it also currently picks up "point" commits that are valid start of range, and we do not want those!
sd-my-git-hours.py: fixed daysWithSingleCommits.ncommits JS calc, so it handles more than one single commit per day (i.e. several commits spread out by more than threshold duration)
sd-my-git-hours.py: changed the JS calc with leftoverRangeCommits; found that moment objects cannot be used to check for "sameness" with Array.include, so switched to using the original .start/.end strings - now the per-day number of commits JS calc seems to be correct for the most part (unless ranged/dual and single-point commits are mixed in a day)
sd-my-git-hours.py: fixed index error; now (merged) commits counts per day are not all 0 anymore, but count is still off
sd-my-git-hours.py: first attempt to generate (merged) commits counts per day; unfortunately it does not work (everything is 0)
sd-my-git-hours.py: fixed shortest streak HTML link; fixed commit ids counted in longest/shortest, so now both shortest and longest streak links work in HTML output, regardless of if we use --repos or not (with indication (but not zoom) on corresponding commit range track)
sd-my-git-hours.py: commented superfluous JS array populating code; added calc of masterGroupName in JS, to distinguish between use of --repos or not, and used it to optimize JS singlePointCommits; restored JS .innerHTML for streak links to previous Python templates - code does not crash, but the streak links still do not work (longest gets generated, but does not zoom; shortest does not get generated?)
sd-my-git-hours.py: additions to JS code, so now Day summaries also include number of commits - however, accurate 1 is printed only for daysWithSingleCommits, while for daysWithStreaks only a placeholder 0 is printed for now
sd-my-git-hours.py: fixed hashes in --repos HTML output; fixed idcount calc for non --repos mode - streaks (and time calc) look fine
sd-my-git-hours.py: ids for merged commits and streaks fixed; visualization and calc seems to work - but commit hashes in merged seem to be wrong
sd-my-git-hours.py: wrote the part of the code which calculates streaks from merged commits in --repos case; python part seems to work (and streaks are identified correctly), but something is still wrong in html output
sd-my-git-hours.py: rewrote code, so when using --repos, each track/lane header shows also number of commits
sd-my-git-hours.py: fixed formatting, so now code works same as before without using --repos; and when using --repos, gray points are rendered for each individual repository
sd-my-git-hours.py: reworked code, so it outputs same visualisation of streaks on the only git track if using without --repos; if using --repos then only dots are shown - however, there is breakage in both cases currently
sd-my-git-hours.py: reworked portion of code to be ready for multiple repos; still works for single one as before
sd-my-git-hours.py: added --repos cmdline arg, unused yet; split off the dicts for defining groups and formatting
sd-my-git-hours.py: fix for index calculation - now also the months get rendered fine
sd-my-git-hours.py: added calc of days with single (no streak) commits, so those days are now also indicated
newns.sh: more attempts to get it working in X11-forwarded desktop, but does not work still
newns: attempt to whitelist DISPLAY, does not help with remote X11
newns.sh: custom history file
added xtra/ssh_tunnel.svg
newns.sh: added note about interference with docker
newns change
added update_wezterm.sh
newns.sh: added changes to script
added single-scripts/newns.sh
added startsynctime.bat
more ubuntu18-postinst.sh
added img_to_gimp_to_pdf.sh
added mousepos_gtk3.py
some notes
some changes to get-OLA-msys2.sh
add get-OLA-msys2.sh
internal-icons-svg: found a procedure for exporting correct bitmaps for Visual Studio 2017
internal-icons-svg: possible fix for visual studio bitmaps
patches: some windows notes, scite-live-patch.notes
patches: some tests of hg rebase in scite-live-patch.notes
changes to ubuntu18-postinst.sh
some more scite-live-patch.notes
freecad: adding testUnion.FCMacro
internal-icons-svg: changed icons background to #C0C0C0
added mingw64-wcalc.md
added msys2_pacman_install.sh
single-scripts: added melt-slideshow.sh
patches: more scite-live-patch.notes
single-scripts: added test-srt-webvtt.js.html
single-scripts: added test-MP4Box.js-v0.3.15.html
source-build-scripts: add get-nbuexplorer.svn.sh
patches: added scite-live-patch.notes
internal-icons-svg: added livefilter
internal-icons-svg: added folder
single-scripts: added example in test.sh for day with multiple splits (visible only in UTC); (temporary) fix to durationToHoursStr for display of whole hours
single-scripts: bugfixes: restored selection outline via CSS; updated remaining unrenamed button id (which broke selection by clicking on streak indices link); fixed disappearing items upon change of TZ (due to a newly made global array not being reinitialized, and containing duplicates that choked vis.js)
single-scripts: fixed problem with objects setup; now copy button works fine, and obtains a plain-text summary in clipboard; that's enough for now
single-scripts: some renames, added a copy function - was about to make a bad mistake, using objects as keys)
single-scripts: implemented per-month timeline calc and visualisation
single-scripts: added numdays printout on page; tighter group/track timeline height in HTML/CSS
single-scripts: day items display fixed; days now show their own cumulative sum - only a second difference from the one in python
single-scripts: per-day timeline drawing started working, but items are duplicated upon changing timezone